aquaculture engineering

Aquaculture Engineering

Odd-Ivar LekangDepartment of Mathematical Sciences and Technology,

Norwegian University of Life Sciences

This Page Intentionally Left Blank

Aquaculture Engineering

Odd-Ivar LekangDepartment of Mathematical Sciences and Technology,

Norwegian University of Life Sciences

© 2007 by Odd-Ivar Lekang

Blackwell Publishing editorial offices:Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK

Tel: +44 (0)1865 776868Blackwell Publishing Professional, 2121 State Avenue, Ames, Iowa 50014-8300, USA

Tel: +1 515 292 0140Blackwell Publishing Asia Pty Ltd, 550 Swanston Street, Carlton, Victoria 3053, Australia

Tel: +61 (0)3 8359 1011

The right of the Author to be identified as the Author of this Work has been asserted in accordance with theCopyright, Designs and Patents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in anyform or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK

Copyright, Designs and Patents Act 1988, without the prior permission of the publisher.

First published 2007 by Blackwell Publishing Ltd

ISBN: 978-1-4051-2610-6

Library of Congress Cataloging-in-Publication DataLekang, Odd-Ivar. Aquaculture engineering / Odd-Ivar Lekang.

p. cm.Includes bibliographical references and index.

ISBN: 978-1-4051-2610-6 (hardback : alk. paper)1. Aquacultural engineering. I. Title.

SH137.L45 2006639.8–dc222006019514

A catalogue record for this title is available from the British Library

Set in 9.5/11.5 pt Timesby SNP Best-set Typesetter Ltd., Hong Kong

Printed and bound in Singaporeby Markono Print Media Pte Ltd

The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy, and which hasbeen manufactured from pulp processed using acid-free and elementary chlorine-free practices. Furthermore, the

publisher ensures that the text paper and cover board used have met acceptable environmental accreditationstandards.

For further information on Blackwell Publishing, visit our website:www.blackwellpublishing.com

Preface xi

1 Introduction 11.1 Aquaculture engineering 11.2 Classification of aquaculture 11.3 The farm: technical components in a system 2

1.3.1 Land-based hatchery and juvenile production farm 21.3.2 On-growing sea cage farm 4

1.4 Future trends: increased importance of aquaculture engineering 61.5 This textbook 6

References 6

2 Water Transport 72.1 Introduction 72.2 Pipe and pipe parts 7

2.2.1 Pipes 72.2.2 Valves 102.2.3 Pipe parts – fittings 122.2.4 Pipe connections – jointing 122.2.5 Mooring of pipes 132.2.6 Ditches for pipes 14

2.3 Water flow and head loss in channels and pipe systems 152.3.1 Water flow 152.3.2 Head loss in pipelines 162.3.3 Head loss in single parts (fittings) 18

2.4 Pumps 182.4.1 Types of pump 192.4.2 Some definitions 212.4.3 Pumping of water requires energy 222.4.4 Centrifugal and propeller pumps 232.4.5 Pump performance curves and working point for centrifugal pumps 252.4.6 Change of water flow or pressure 272.4.7 Regulation of flow from selected pumps 29References 31

3 Water Quality and Water Treatment: an Introduction 323.1 Increased focus on water quality 323.2 Inlet water 32

Contents

iii

iv Contents

3.3 Outlet water 333.4 Water treatment 35

References 36

4 Adjustment of pH 374.1 Introduction 374.2 Definitions 374.3 Problems with low pH 384.4 pH of different water sources 384.5 pH adjustment 394.6 Examples of methods for pH adjustment 39

4.6.1 Lime 394.6.2 Seawater 414.6.3 Lye or hydroxides 41References 42

5 Removal of Particles 445.1 Introduction 445.2 Characterization of the water 455.3 Methods for particle removal in fish farming 45

5.3.1 Mechanical filters and micro screens 455.3.2 Depth filtration – granular medium filters 495.3.3 Settling or gravity filters 525.3.4 Integrated treatment systems 55

5.4 Hydraulic loads on filter units 565.5 Purification efficiency 565.6 Dual drain tank 575.7 Sludge production and utilization 575.8 Local ecological solutions 60

References 61

6 Disinfection 636.1 Introduction 636.2 Basis of disinfection 64

6.2.1 Degree of removal 646.2.2 Chick’s law 646.2.3 Watson’s law 646.2.4 Dose-response curve 65

6.3 Ultraviolet light 656.3.1 Function 656.3.2 Mode of action 656.3.3 Design 656.3.4 Design specification 676.3.5 Dose 686.3.6 Special problems 68

6.4 Ozone 686.4.1 Function 686.4.2 Mode of action 686.4.3 Design specification 706.4.4 Ozone dose 706.4.5 Special problems 716.4.6 Measuring ozone content 71

6.5 Other disinfection methods 726.5.1 Photozone 726.5.2 Heat treatment 726.5.3 Chlorine 736.5.4 Changing the pH 736.5.5 Natural methods: ground filtration or constructed wetland 73References 74

7 Heating and Cooling 757.1 Introduction 757.2 Heating requires energy 757.3 Methods for heating water 767.4 Heaters 77

7.4.1 Immersion heaters 777.4.2 Oil and gas burners 79

7.5 Heat exchangers 797.5.1 Why use heat exchangers? 797.5.2 How is the heat transferred? 807.5.3 Factors affecting heat transfer 807.5.4 Important parameters when calculating the size of heat exchangers 817.5.5 Types of heat exchanger 837.5.6 Flow pattern in heat exchangers 857.5.7 Materials in heat exchangers 867.5.8 Fouling 86

7.6 Heat pumps 877.6.1 Why use heat pumps? 877.6.2 Construction and function of a heat pump 877.6.3 Log pressure–enthalpy (p–H) 897.6.4 Coefficient of performance 897.6.5 Installations of heat pumps 907.6.6 Management and maintenance of heat pumps 91

7.7 Composite heating systems 917.8 Chilling of water 94

References 95

8 Aeration and Oxygenation 978.1 Introduction 978.2 Gases in water 978.3 Gas theory – aeration 99

8.3.1 Equilibrium 998.3.2 Gas transfer 100

8.4 Design and construction of aerators 1018.4.1 Basic principles 1018.4.2 Evaluation criteria 1028.4.3 Example of designs for different types of aerator 103

8.5 Oxygenation of water 1068.6 Theory of oxygenation 108

8.6.1 Increasing the equilibrium concentration 1088.6.2 Gas transfer velocity 1088.6.3 Addition under pressure 108

Contents v

vi Contents

8.7 Design and construction of oxygen injection systems 1098.7.1 Basic principles 1098.7.2 Where to install the injection system 1098.7.3 Evaluation of methods for injecting oxygen gas 1108.7.4 Examples of oxygen injection system designs 111

8.8 Oxygen gas characteristics 1158.9 Sources of oxygen 115

8.9.1 Oxygen gas 1158.9.2 Liquid oxygen 1168.9.3 On-site oxygen production 1178.9.4 Selection of source 119References 120

9 Ammonia Removal 1219.1 Introduction 1219.2 Biological removal of ammonium ion 1219.3 Nitrification 1219.4 Construction of nitrification filters 123

9.4.1 Flow-through system 1239.4.2 The filter medium in the biofilter 1259.4.3 Rotating biofilter (biodrum) 1259.4.4 Fluid bed/active sludge 1269.4.5 Granular filters/bead filters 127

9.5 Management of biological filters 1279.6 Example of biofilter design 1289.7 Denitrification 1289.8 Chemical removal of ammonia 129

9.8.1 Principle 1299.8.2 Construction 129References 130

10 Recirculation and Water Re-use Systems 13310.1 Introduction 13310.2 Advantages and disadvantages of re-use systems 133

10.2.1 Advantages 13310.2.2 Disadvantages of re-use systems 134

10.3 Definitions 13410.3.1 Degree of re-use 13410.3.2 Water exchange in relation to amount of fish 13610.3.3 Degree of purification 136

10.4 Theoretical models for construction of re-use systems 13610.4.1 Mass flow in the system 13610.4.2 Water requirements of the system 13710.4.3 Connection between outlet concentration, degree of re-use and effectiveness

of the water treatment system 13810.5 Components in a re-use system 13910.6 Design of a re-use system 141

References 143

11 Production Units: a Classification 14411.1 Introduction 144

11.2 Classification of production units 14411.2.1 Intensive/extensive 14411.2.2 Fully controlled/semi-controlled 14711.2.3 Land based/tidal based/sea based 14711.2.4 Other 148

11.3 Possibilities for controlling environmental impact 149

12 Egg Storage and Hatching Equipment 15012.1 Introduction 15012.2 Systems where the eggs stay pelagic 151

12.2.1 The incubator 15112.2.2 Water inlet and water flow 15212.2.3 Water outlet 152

12.3 Systems where the eggs lie on the bottom 15312.3.1 Systems where the eggs lie in the same unit from spawning to fry ready for

starting feeding 15312.3.2 Systems where the eggs must be removed before hatching 15512.3.3 System where storing, hatching and first feeding are carried out in the same

unit 157References 157

13 Tanks, Basins and Other Closed Production Units 15813.1 Introduction 15813.2 Types of closed production units 15813.3 How much water should be supplied? 16013.4 Water exchange rate 16113.5 Ideal or non-ideal mixing and water exchange 16213.6 Tank design 16213.7 Flow pattern and self-cleaning 16513.8 Water inlet design 16713.9 Water outlet or drain 16913.10 Dual drain 17113.11 Other installations 172

References 172

14 Ponds 17414.1 Introduction 17414.2 The ecosystem 17414.3 Different production ponds 17414.4 Pond types 176

14.4.1 Construction principles 17614.4.2 Drainable or non-drainable 177

14.5 Size and construction 17814.6 Site selection 17814.7 Water supply 17914.8 The inlet 17914.9 The outlet – drainage 18014.10 Pond layout 182

References 182

15 Sea Cages 18315.1 Introduction 183

Contents vii

viii Contents

15.2 Site selection 18415.3 Environmental factors affecting a floating construction 185

15.3.1 Waves 18515.3.2 Wind 19115.3.3 Current 19115.3.4 Ice 193

15.4 Construction of sea cages 19315.4.1 Cage collar or framework 19415.4.2 Weighting and stretching 19515.4.3 Net bags 19515.4.4 Breakwaters 19715.4.5 Examples of cage constructions 197

15.5 Mooring systems 19815.5.1 Design of the mooring system 19815.5.2 Description of the single components in a pre-stressed mooring system 20115.5.3 Examples of mooring systems in use 204

15.6 Calculation of forces on a sea cage farm 20415.6.1 Types of force 20515.6.2 Calculation of current forces 20615.6.3 Calculation of wave forces 21015.6.4 Calculation of wind forces 210

15.7 Calculation of the size of the mooring system 21015.7.1 Mooring analysis 21015.7.2 Calculation of sizes for mooring lines 211

15.8 Control of mooring systems 213References 213

16 Feeding Systems 21516.1 Introduction 215

16.1.1 Why use automatic feeding systems? 21516.1.2 What can be automated? 21516.1.3 Selection of feeding system 21516.1.4 Feeding system requirements 215

16.2 Types of feeding equipment 21616.2.1 Feed blowers 21616.2.2 Feed dispensers 21616.2.3 Demand feeders 21716.2.4 Automatic feeders 21816.2.5 Feeding systems 222

16.3 Feed control 22416.4 Feed control systems 22416.5 Dynamic feeding systems 225

References 225

17 Internal Transport and Size Grading 22717.1 Introduction 22717.2 The importance of fish handling 227

17.2.1 Why move the fish? 22717.2.2 Why size grade? 228

17.3 Negative effects of handling the fish 23217.4 Methods and equipment for internal transport 233

17.4.1 Moving fish with a supply of external energy 23317.4.2 Methods for moving of fish without the need for external energy 243

17.5 Methods and equipment for size grading of fish 24517.5.1 Equipment for grading that requires an energy supply 24517.5.2 Methods for voluntary grading (self grading) 253References 254

18 Transport of Live Fish 25618.1 Introduction 25618.2 Preparation for transport 25618.3 Land transport 257

18.3.1 Land vehicles 25718.3.2 The tank 25718.3.3 Supply of oxygen 25818.3.4 Changing the water 25918.3.5 Density 25918.3.6 Instrumentation and stopping procedures 259

18.4 Sea transport 26018.4.1 Well boats 26018.4.2 The well 26118.4.3 Density 26118.4.4 Instrumentation 261

18.5 Air transport 26218.6 Other transport methods 26318.7 Cleaning and re-use of water 26318.8 Use of additives 264

References 264

19 Instrumentation and Monitoring 26619.1 Introduction 26619.2 Construction of measuring instruments 26719.3 Instruments for measuring water quality 267

19.3.1 Measuring temperature 26819.3.2 Measuring oxygen content of the water 26819.3.3 Measuring pH 26919.3.4 Measuring conductivity and salinity 26919.3.5 Measuring total gas pressure and nitrogen saturation 26919.3.6 Other 270

19.4 Instruments for measuring physical conditions 27119.4.1 Measuring the water flow 27119.4.2 Measuring water pressure 27319.4.3 Measuring water level 274

19.5 Equipment for counting fish, measuring fish size and estimation of total biomass 27519.5.1 Counting fish 27519.5.2 Measuring fish size and total fish biomass 277

19.6 Monitoring systems 28019.6.1 Sensors and measuring equipment 28119.6.2 Monitoring centre 28119.6.3 Warning equipment 28219.6.4 Regulation equipment 28319.6.5 Maintenance and control 283References 283

Contents ix

x Contents

20 Buildings and Superstructures 28420.1 Why use buildings? 28420.2 Types, shape and roof design 284

20.2.1 Types 28420.2.2 Shape 28420.2.3 Roof design 285

20.3 Load-carrying systems 28520.4 Materials 28720.5 Prefabricate or build on site? 28820.6 Insulated or not? 28820.7 Foundations and ground conditions 28920.8 Design of major parts 289

20.8.1 Floors 28920.8.2 Walls 290

20.9 Ventilation and climatization 291References 293

21 Design and Construction of Aquaculture Facilities 29421.1 Introduction 29421.2 Land-based hatchery, juvenile and on-growing production plant 294

21.2.1 General 29421.2.2 Water intake and transfer 29421.2.3 Water treatment department 30421.2.4 Production rooms 30621.2.5 Feed storage 31021.2.6 Disinfection barrier 31021.2.7 Other rooms 31121.2.8 Outlet water treatment 31121.2.9 Important equipment 311

21.3 On-growing production, sea cage farms 31421.3.1 General 31421.3.2 Site selection 31421.3.3 The cages and the fixed equipment 31421.3.4 The base station 31721.3.5 Net handling 31721.3.6 Boat 319References 320

22 Planning Aquaculture Facilities 32122.1 Introduction 32122.2 The planning process 32122.3 Site selection 32222.4 Production plan 32222.5 Room programme 32422.6 Necessary analyses 32522.7 Drawing up alternative solutions 32822.8 Evaluation of and choosing between the alternative solutions 32822.9 Finishing plans, detailed planning 32822.10 Function test of the plant 32822.11 Project review 329

References 329

Index 330

The aquaculture industry, which has been growingat a very high rate for many years now, is projectedto continue growing at a rate higher than mostother industries for the foreseeable future. Thisgrowth has mainly been driven by static catchesfrom most fisheries and a decline in stocks of manymajor commercially caught fish species, combinedwith the ever increasing need for marine proteindue to continuing population growth. An increasedfocus on the need for fish in the diet, due to mount-ing evidence of the health benefits of eating morefish, will also increase the demand.

There has been rapid development of technologyin the aquaculture industry, particularly as used inintensive aquaculture where there is high produc-tion per m3 farming volume. It is predicted that theexpansion of the aquaculture industry will lead tofurther technical development with more, andcheaper, technology being available for use in theindustry in the future years.

The aim with this book is to give a generaloverview of the technology used in the aquacultureindustry. Individual chapters focus on water trans-fer, water treatment, production units and addi-tional equipment used on aquaculture plants.Chapters where equipment is set into systems, suchas land-based fish farms and cage farms, are alsoincluded. The book ends with a chapter which

includes systematic methodology for planning a fullaquaculture facility.

The book is based on material successfully usedon BSc and MSc courses in intensive aquaculturegiven at the Norwegian University of Life Science(UMB) and refined over many years, the universityhaving included courses in aquaculture since 1973.In 1990 a special Master’s course was developed inaquaculture engineering (given in Norwegian), andfrom 2000 the university has also offered an Englishlanguage international Master’s programme inaquaculture (see details at www.umb.no).

During the author’s compilation of material foruse in this book, and also for earlier books cover-ing similar fields (in Norwegian), many people havegiven useful advice. I would like especially to thankSvein Olav Fjæra and Tore Ensby. Further thanksalso go to my colleagues at UMB: B.F. Eriksen, P.H.Heyerdal, T.K. Stevik, and from earlier, colleaguesand students:V.Tapei. Mott,A. Skar, P.O. Skjervold,G. Skogesal and D. E. Thommassen.

Tore Ensby has drawn all the line illustrationscontained in the book. All the photographsincluded in the book have been taken by theauthor.

O.I. LekangNovember 2006

Preface

xi

1Introduction

production systems. It is therefore a challenge tobring together both technological and biologicalknowledge within the aquaculture field.

1.2 Classification of aquacultureThere are a number of ways to classify aquaculturefacilities and production systems, based on the technology or the production system used.

‘Extensive’, ‘intensive’ and ‘semi-intensive’ aqua-culture are common ways to classify aquaculturebased on production per unit volume (m3) or unitarea (m2) farmed. Extensive aquaculture involvesproduction systems with low production per unitvolume. The species being farmed are kept at a lowdensity and there is minimal input of artificial substances and human intervention. A low level of technology and very low investment per unitvolume farmed characterize this method. Pondfarming without additional feeding, like some carpfarming, is a typical example. Sea ranching andrestocking of natural lakes may also be included inthis type of farming.

In intensive farming, production per unit volumeis much higher and more technology and artificialinputs must be used to achieve this. The investmentcosts per unit volume farmed will of course also bemuch higher. The maintenance of optimal growthconditions is necessary to achieve the growthpotential of the species being farmed. Additionalfeeding, disease control methods and effectivebreeding systems also characterize this type offarming. The risk of disease outbreaks is higherthan in extensive farming because the organism iscontinuously stressed for maximal performance.

1.1 Aquaculture engineeringDuring the past few years there has been consider-able growth in the global aquaculture industry.Many factors have made this growth possible. Oneis development within the field of aquaculture engi-neering, for example improvements in technologyallowing reduced consumption of freshwater anddevelopment of re-use systems. Another is thedevelopment of offshore cages: sites that until a fewyears ago not were viable for aquaculture purposescan be used today with good results. The focus oneconomic efficiency and the fact the salaries areincreasing have also resulted in the increased use oftechnology to reduce staff numbers.

The development of new aquaculture specieswould not have been possible without the contribution of the fisheries technologist. Even if some techniques can be transferred for thefarming of new species, there will always be a need for technology to be developed and optimizedfor each species. An example of this is the devel-opment of production tanks for flatfish with a largerbottom surface area than those used for pelagicfish.

Aquaculture engineering covers a very large areaof knowledge and involves many general engineer-ing specialisms such as mechanical engineering,environmental engineering, materials technology,instrumentation, and monitoring, and buildingdesign and construction. The primary aim of aqua-culture engineering is to utilize technical engineer-ing knowledge and principles in aquaculture andbiological production systems. The production offish has little in common with the production ofnails, but the same technology can be used in both

1

2 Aquaculture Engineering

Salmon farming is a typical example of intensiveaquaculture.

It is also possible to combine the above produc-tion systems – this is called semi-intensive aqua-culture. An example is intensive fry productioncombined with extensive ongrowing.

Another classification of an aquacultural systemcan be according to the life stage of the species produced on the farm, for instance eggs, fry,juvenile or ongrowing. Farms may also cover thecomplete production process, and this is called fullproduction.

According to the type of farming technologyused there are also a number of classificationsbased on the design and function of the productionunit. This will of course be species and life-stagedependent. For fish the following classificationsmay be used: 1. Closed production units where thefish are kept in a enclosed production unit sepa-rated from the outside environment. 2. Open pro-duction units where the unit has permeable walls,such as nets and so the fish are partly affected bythe surrounding environment. It is also possible toclassify the farm based on where it is located: withinthe sea, in a tidal zone or on land.

Land-based farms may be classified by the typeof water supply for the farm: water may be gravity-fed or pumped. In gravity systems the water sourceis at a higher altitude than the farm and the watercan flow by gravity from the source to the farm.When pumping, the source can be at an equal orlower altitude compared to the farm. For tidalthrough-flow farms, water supply and exchange isachieved using the tide.

Farms can also be classified by how the watersupplied to a farm is used. If the water is used once,flowing directly through, it is named a flow-throughfarm. If the water is used several times, with theoutlet water being recycled, it is a water re-use orrecirculating system.

1.3 The farm: technical components ina systemIn a farm the various technical componentsincluded in a system can be roughly separated asfollows:

• Production units• Water transfer and treatment

• Additional equipment (feeding, handling andmonitoring equipment)

To illustrate this, two examples are given: a land-based hatchery and a juvenile farm, and an ongrow-ing sea cage farm.

1.3.1 Land-based hatchery and juvenileproduction farm

Land-based farms normally utilize much moretechnical equipment than sea cage farms, espec-ially intensive production farms with a number of tanks. The major components are as follows (Fig. 1.1):

• Water inlet and transfer• Water treatment facilities• Production units• Feeding equipment• Equipment for internal fish transport and size

grading• Equipment for transport of fish from the farm• Equipment for waste and wastewater treatment• Instrumentation and monitoring systems

Water inlet and transfer

The design of the inlet depends on the watersource: is it seawater or freshwater (lakes, rivers),or is it surface water or groundwater? It is alsoquite common to have several water sources in useon the same farm. Further, it depends whether thewater is fed by gravity or whether it has to bepumped, in which case a pumping station isrequired.Water is normally transferred in pipes, butopen channels may also be used.

Water treatment facilities

Usually water is treated before it is sent in to thefish. Equipment for removal of particles preventsexcessively high concentrations reaching the fish;additionally large micro-organisms may beremoved by the filter.Water may also be disinfectedto reduce the burden of micro-organisms, especiallythat used on eggs and small fry. Aeration may benecessary to increase the concentration of oxygenand to remove possible supersaturation of the gasesnitrogen and carbon dioxide. If there is lack of

water or the pumping height is large pure oxygengas may be added to the water. Another possibilityif water supply is limited is to reuse the water,however, this will involve much water treatment.For optimal development and growth of the fish

heating or cooling of the water may be necessary;in most cases this will involve a heat pump or acold-storage plant. If the pH in a freshwater sourceis too low pH adjustment may be a part of the watertreatment.

Introduction 3

Figure 1.1 Example of major com-ponents in a land-based hatchery andjuvenile production plant.


Production units

The production units necessary and their size anddesign will depend on the species being grown. Inthe hatchery there will either be tanks withupwelling water (fluidized eggs) or units where theeggs lie on the bottom or on a substrate. Afterhatching the fish are moved to some type of pro-duction tank. Usually there are smaller tanks forweening and larger tanks for further on-growinguntil sale. Weening start feeding tanks are normallyunder a roof, while on-growing tanks can also beoutside.

Feeding equipment

Some type of feeding equipment is commonly used,especially for dry feed. Use of automatic feederswill reduce the manual work on the farm. Feedingat intervals throughout the day and night may alsobe possible; the fish will then always have access tofood, which is important at the fry and juvenilestages.

Internal transport and size grading

Because of fish growth it is necessary to divide thegroup to avoid fish densities becoming too high.It is also common to size grade to avoid large size variations in one production unit; for somespecies this will also reduce the possibilities forcannibalism.

Transport of fish

When juvenile fish are to be transferred to an on-growing farm, there is a need for transport.Either a truck with water tanks or a boat with a wellis normally used. The systems for loading may bean integral part of the farm construction.

Equipment for waste handling and wastewater treatment

Precautions must be taken to avoid pollution fromfish farms. These include legal treatment of generalwaste. Dead fish must be treated and stored satis-factorily, for example, put in acid or frozen for lateruse. Dead fish containing trace of antibiotics orother medicine must be destroyed by legal means.

Whether wastewater treatment is necessary willdepend on conditions where the effluent water isdischarged. Normally there will at least be arequirement to remove larger suspended particles.

Instrumentation and monitoring

In land-based fish farms, especially those depen-dent upon pumps, a monitoring system is essentialbecause of the economic consequences if pumpingstops and the water supply to the farm is inter-rupted. The oxygen concentration in the water will fall and may result in total fish mortality.Instruments are being increasingly used to controlwater quality, for instance, to ensure optimal production.

1.3.2 On-growing sea cage farm

Normally a sea cage farm can be run with ratherless equipment than land-based farms, the majorreason being that water transfer and water treat-ment (which is not actually possible) are not nec-essary because the water current ensures watersupply and exchange. The components necessaryare as follows (Fig. 1.2):

• Production units• Feeding equipment• Working boat• Equipment for size grading• Base station

Production units

Sea cages vary greatly in construction and size;the major difference is the ability to withstandwaves, and special cages for offshore farming havebeen developed. It is also possible to have systemcages comprising several cages, or individual cages.The cages may also be fitted with a gangway to theland. Sea cages also include a mooring system. Toimprove fish growth, a sub-surface lighting systemmay be used.

Feeding equipment

It is common to install some type of feeding systemin the cages because of the large amounts of feedthat are typically involved. Manual feeding may

also be used, but this involves hard physical labourfor the operators.

Working boat

All sea cage farms need a boat; a large variety ofboats are used. Major factors for selection are sizeof the farm, whether it is equipped with a gangwayor not, and the distance from the land base to thecages. Faster and larger boats are normally requiredif the cages are far from land or in weather-exposedwater.

Size grading

Equipment for size grading can be necessary if this is included in the production plan. It may, how-

ever, be possible to rent this as a service from subcontractors.

Base station

All cages farms will include a base station; this mayeither be land based, floating on a barge or both.The base station can include storage rooms, messrooms, changing rooms and toilet, and equipmentfor treatment of dead fish. The storage roomincludes rooms and/or space for storage of feed;it may also include rooms for storage of nets andpossibly storage of equipment for washing, main-taining and impregnating them. However, this isalso a service that is commonly rented from subcontractors.

Introduction 5

Figure 1.2 Example of major com-ponents in an on-growing sea cagefarm.


1.4 Future trends: increasedimportance of aquaculture engineering

Growth in the global aquaculture industry will cer-tainly continue, with several factors contributing tothis. The world’s population continues to grow aswill the need for marine protein. Traditional fish-eries have limited opportunities to increase theircatches if sustainable fishing is to be carried out.Eventually, therefore, increase in production mustcome from the aquaculture industry. In addition,the aquaculture industry can deliver aquatic products of good quality all year round, which represents a marketing advantage compared to traditional fishing. The increased focus on optimalhuman diets, including more fish than meat in thediet for large groups of the world’s population, alsorequires more fish to be marketed.

This will give future challenges for aquacultureengineers. Most probably there will be an increasedfocus on intensive aquaculture with higher produc-tion per unit volume. Important challenges tohousing the growth will be availability of fresh-water resources and good sites for cage farming.Limited supplies of freshwater in the world meanthat technology that can reduce water consumptionper kilogram of fish produced will be important;this includes reliable, cost effective re-use technol-ogy. By employing re-use technology it will also bepossible to maintain a continuous supply of highquality water independently of the quality of theincoming water. To have more accurate controlover water quality will also be of major importancewhen establishing aquaculture with new species,especially during the fry production stage.

The trend to use more and more weather-exposed sites for cage farms will continue. Devel-opment of cages that can not only withstandadverse weather conditions but also be operatedeasily in bad weather, and where fish feeding andcontrol can be performed, is important.

Rapid developments in electronics and monitor-ing will gradually become incorporated into theaquaculture industry. Intensive aquaculture willdevelop into a process industry where the controlroom will be the centre of operations and processeswill be monitored by electronic instruments; robotswill probably be used to replace some of today’smanual functions. Nanotechnology will beincluded, for instance by using more and smaller

sensors for more purposes. An example would beto include sensors in mooring lines and net bags tomonitor tension and eventual breakage. Individualtagging of the fish will most probably also be afuture possibility, which makes control of thewelfare of the single individual possible; this can also be important regarding control of escapedfish.

1.5 This textbookThe aim with this book is to give a general basicreview of the total area of aquaculture engineering.Based on the author’s two previously publishedbooks on aquaculture engineering written in Norwegian.1,2 Several of the illustrations are alsobased on illustrations in these books. The textbookis primarily intended for the introductory course inaquaculture engineering for the Bachelor andMaster degrees in aquaculture at the NorwegianUniversity of Life Science (UMB). Several othertextbooks dealing with parts of the syllabus areavailable and referred to in later chapters.The sameis the case with lecture notes from more advancedcourses in aquaculture engineering at UMB.

The focus of the book is on intensive fish farming, where technology is and will becomeincreasingly important. Most of it concerns fishfarming, but several of the subjects are general and will have much interest for molluscan and crustacean shellfish farmers.

Starting with water transport, the book continueswith an overview chapter on water quality and theneed for and use of different water treatment units,which are described in the next few chapters. It continues with a chapter on production unit classi-fication followed by chapters on the different production units. Chapters devoted to additionalequipment such as that for feed handling and fishhandling, instrumentation, monitoring and build-ings follow. Chapters on planning of aquaculturefacilities and their design and construction con-clude the book.

References1. Lekang, O.I., Fjæra, S.O. (1997) Teknologi for akvakul-

tur. Landbruksforlaget (in Norwegian).2. Lekang, O.I., Fjæra, S.O. (2002) Teknisk utstyr til

fiskeoppdrett. Gan forlag (in Norwegian).

2Water Transport

2.2 Pipe and pipe parts

2.2.1 Pipes

Pipe materials

In aquaculture the common way to transport wateris through pipes; open channels are also used insome cases. Channels may be used for transportinto the farm, for distribution inside the farm and for the outlet of water. They are normally built of concrete and are quite large; the water istransported with low velocity. Channels may also be excavated in earth, for example to supply thewater to earth ponds. Advantages of open channelsare their simple construction and the ease withwhich the water flow can be controlled visually;disadvantages are the requirement for a constantslope over the total length and there can be no pressure in an open channel. The greater exteriorsize compared to pipes, and the noise inside the building when water is flowing are other disadvantages.

Plastics, mainly thermoplastics, are the most com-monly used materials for pipes.Thermoplastic pipesare delivered in many different qualities with different characteristics and properties (Table 2.1).A thermoplastic melts when the temperature gethigh enough.10 Thermoplastic pipes can be dividedinto weldable (polyethylene; PE) and glueable(polyvinyl chloride; PVC) depending on the waythe pipes are connected. The opposite of

2.1 IntroductionAll aquaculture facilities require a supply of water.It is important to have a reliable, good-qualitywater source and equipment to transfer water toand within the facility. The volume of water neededdepends on the facility size, the species and the pro-duction system. In some cases can it be very large,up to several hundred m3/min (Fig. 2.1). This isequivalent to the water supply to a quite large villages, considering that in Norway a normal valuefor the water supply per person is up 180 litres perday.

If something fails with the water supply or dis-tribution system it may result in disaster for theaquaculture facilities. This also emphasizes theimportance of good knowledge in this area. Correctdesign and construction of the water inlet system isan absolute requirement to avoid large unnecessaryproblems in the future. For instance, this may be apparent when the inlet system is too small andthe water flow rate to the facility is lower thanexpected.

The science of the movement of water is calledhydrodynamics, and in this chapter the importantfactors of this field are described with emphasisupon aquaculture. In addition, a description of theactual materials and parts for water transport aregiven: pipes, pipe parts (fittings) and pumps. Muchmore specific literature is available in all these fields(basic fluid mechanics,1–3 pipes and pipe parts,4–6

pumps7–9).

7


thermoplastic is hardening plastic, such as fibreglasswhich is made of different materials that are hard-ened; afterwards it is impossible to change its shape,even by heating. Fibreglass can be used in specialcritical pipes and pipe parts, but only in specialcases (see below).

It is also important that materials used for pipesare non-toxic for fish.11 Copper, much used inpiping inside houses, is an example of a commonlyused material that is not recommended for fishfarming because of its toxicity. In the past steel, con-crete or iron pipes were commonly used, but todaythese materials are seldom chosen because of theirprice, duration and laying costs.

PE pipes are of low weight, simple to handle,have high impact resistance and good abrasionresistance. Nevertheless, these pipes may be vul-nerable to water hammer (see later) or vacuumeffects. PE pipes are delivered in a wide variety of

dimensions and pressure classes; they are normallyblack or grey but other colours are also used. Smalldiameter pipes may be delivered in coils, whilelarger sizes are straight, with lengths commonlybetween 3 and 6m. PE may be used for both inletand outlet pipes. PE piping must be fused togetherfor connection; if flanges are fused to the pipe fit-tings, pipes may be screwed together.

PVC is used in pipes and pipe parts inside the fishfarm and also in outlet systems. This material is oflow density and easy to handle. Pipe and parts aresimple to join together with a special solventcementing glue. A cleaning liquid dissolves thesurface and makes gluing possible.There are a largevariety of pipe sizes and pipe parts available. Whenusing this kind of piping, attention must be given tothe temperature: below 0°C this material becomesbrittle and will break easily. PVC is also recom-mended for use at temperatures above 40°C. PVC

Table 2.1 Typical characteristics of actual pipe materials.

Material Temperature range (°C) Common pressure classes (bar) Common size range (mm)

PE −40 to +60 3.2, 4, 6, 10 and 25 20–1600PP 0 to +100 10 and 16 16–400PVDF −40 to +140 16 16–225PVC-U 0 to +60 4, 6, 10, 16 and 25 6–400PVC-C 0 to +80 16 16–225ABS −40 to +60 16 16–225

Figure 2.1 The supply of water to afish farm can be up to severalhundred cubic metres per minute, ashere for a land-based fish farm forgrowing of marked size Atlanticsalmon.

is also vulnerable to water hammer.There are ques-tions concerning the use of PVC materials becausepoisonous gases are emitted during burning of left-over material.

Fibreglass may be used in special cases, forexample in very large pipes (usually over 1m indiameter). The material is built in two or threelayers: a layer of polyester that functions like a glue;a layer with a fibreglass mat that acts as reinforce-ment; and quartz or sand. The ratio between thesecomponents may vary with the pressure and stiff-ness needed for the pipe. A pipe is normally con-structed with several layers of fibreglass andpolyester. Fibreglass has the advantages that it tol-erates low temperatures, is very durable and maybe constructed so thick that it can tolerate waterhammer and vacuum effects. The disadvantage isthe low diversity of pipes and pipe parts available.For joining of parts, the only possibilities are to con-struct sockets on site using layers of polyester andfibreglass, or pipes equipped with flanges by themanufacturer can be screwed together with agasket in between.

At present, materials such as polypropylene(PP), acrylonitrile–butadiene–styrene (ABS) andpolyvinyl difluoride (PVDF) have also been intro-duced for use in the aquaculture industry, but tominor degree and for special purposes. They arealso more expensive than PE and PVC.

Pressure class

Each pipe and pipe part must be thick enough to tolerate the pressure of water flowing throughthe system. To install the correct pipes it is there-fore important to know the pressure of the waterthat will flow through them.The pressure (PN) classindicates the maximum pressure that the pipes and pipe parts can tolerate. The pressure class is given in bar (1 bar = 10m water column (mH2O)= 98100Pa); for instance, a PN4 pipe will tolerate 4 bar or a 40m water column. This means that if the pressure inside the pipe exceeds 4 bar the pipemay split. In fish farming pressure classes PN4, PN6and PN10 are commonly used. Pipes of differentPN classes vary regarding wall thickness: higherpressure requires thicker pipe walls. Pipes of higherPN class will of course cost more, because morematerial is required to make them.

A complete inlet pipe from the source to the

facility may be constructed with pipes of differentPN classes. If, for instance, the water source to a fishfarm is a lake located 100m in height above thefarm, a PN4 pipe can be used for the first 40m drop,then a PN6 pipe for the following 20m drop, and onthe final 40m drop a PN10 pipe is used.

Some problems related to pressure class are asfollows:

Water hammer: Water hammer can occur, forinstance, when a valve in a long pipe filled withmuch water is closed rapidly.This will generate highlocal pressure in the end of the pipe, close to thevalve, from the moving mass of water inside thepipeline that needs some time to stop. The result isthat the pipe can ‘blow’. Rapid closing of valvesmust therefore be avoided.Water hammer may alsooccur with rapid starting and stopping of pumps.This can, however, be difficult to inhibit and it maybe necessary to use special equipment to damp thewater hammer effect. A tank with low pressure airmay be added to the pipe system: if there is waterhammer in the pipes the air in this tank will be com-pressed and this reduces the total hammer effect inthe system.

Vacuum:A vacuum may be generated in a sectionof pipe, for example when it is laid at differentheights (over a crest) and then functions as a siphon(Fig. 2.2). A vacuum may then occur on the topcrest. It is recommended that such conditions beavoided, because the pipeline may becomedeformed and collapse because of the vacuum.Pipes are normally not certified for vacuum effects;however, if vacuum effects are possible, it is rec-ommended that a pipe of higher pressure class is used in the part where the vacuum may occur.By using pipes with thicker walls, higher toleranceto vacuum effects is achieved; alternatively, a fibre-glass pipe which tolerates a higher vacuum could beemployed.

Classification of pipes

Pipe diameters are standardized. There are anumber of sizes available for various applicationsin different industries. In aquaculture, pipes withthe following external diameters (mm) are gener-ally used 20, 25, 32, 40, 50, 63, 75, 90, 110, 125, 160,180, 200, 225, 250, 280, 315, 355, 400, 450, 500, 560and 630. The internal diameter that is used when

Water Transport 9


calculating the water velocity in the pipelines, isfound by subtracting twice the wall thickness.Higher pressure class pipes have thicker walls thanlower pressure class pipes.

All pipes and pipe parts must be marked clearly.For pipes the marking print on the pipe is normallyevery metre, and for pipe parts there is a mark onevery part.The following is included in the standard-ized marking: pipe material, pressure class, externaldiameter, wall thickness, producer and the time when the pipe was produced. It is important to use standardized pipe parts when planning fish farms.

2.2.2 Valves

Valves are used to regulate the water flow rate andthe flow direction. Many types of valve are used inaquaculture (Fig. 2.3). Which type to use must bechosen on the basis of the flow in the system andthe specific needs of the farm. Several materials are used in valves, such as PVC, ABS, PP andPVDF, and the material chosen depends on where the valves will be used. Large valves may also be fabricated in stainless or acid proofsteel.

Ball valves are low cost solutions used in aqua-culture. The disadvantage is that they are not veryprecise when regulating the water flow. They arebest used in an on/off manner, or for approximateregulation of the water flow. The design is simpleand consists of a ball with an opened centre. Whenturning it will gradually open or close.

Figure 2.2 A vacuum may occur inside the pipe on the top crest causingdeformation.

Valves constructed with a membrane pulleddown by a piston for regulation of water flow are called diaphragm or membrane valves. Thesevalves can regulate water flow very accurately.They cost considerably more than a ball valve, andthe head loss through the valve is significantlyhigher.

Angle seat valves have a piston standing in anangled ‘seat’. When the screw handle is turned thepiston moves up or down. The opening is graduallyreduced by pressing the piston down. This type ofvalve is also capable of accurate flow regulation, butis quite expensive and has also a higher head lossthan a ball valve. For accurate flow regulation, forinstance on single tanks, diaphragm valves or angleseat valves are recommended.When selecting thesetypes of valves it is, however, important to be awarethat the head loss can be over five times as high aswith a ball valve.

Butterfly valves are usually located in large pipes(main pipeline or part pipelines) and regulate thewater flow by opening or closing a throttle. Whichis located inside a pipe part; by turning the throttlethe passage for the water inside the pipe is changed.A slide valve or gate valve can be used for the samecases.This consists of a gate or slide that stands ver-tically in the water flow; the water flow is regulatedby lifting or lowering the plate by a spindle. Thisvalve type is also used in large diameter pipelines,but both butterfly valves and sluice valves are quiteexpensive, especially in large sizes. It is, however,better to use too many valves than too few. To havethe facility to turn off the water flow at several

Water Transport 11

A

B

C

Figure 2.3 Valve types used on aquaculture facilities: (A) diagrams showing valve cross-sections; (B) ball valve;(C) angel seat valve; (D) diaphragm or membrane valve; (E) butterfly valve.

places in the farm, for instance for maintenance,will always be an advantage. These types of valvesare, however, not recommended for precise regula-tion of water flow.

The check valve or ‘non-return’ valve is used toavoid the backflow of water; this means that thewater can only flow in one direction in the pipesystem. In many cases it is used in a pump outlet

to avoid backflow of water when the pump stops.Normally the valve is comprises a plate or ball that closes when the water flow tends to go in theopposite direction.

Triple way valves may regulate the flow in twodirections to create a bypass. There are also manyother types of valves, for instance electrically orpneumatically operated valves which make it


possible to regulate water flow functions automati-cally. In new and advanced fish farms such equip-ment is of increasing interest, especially when savingof water is necessary.

It is important to remember, however, that allvalves create a head loss, the size of which dependson the type of valve being used; for example,diaphragm valves have a high head loss. This mustconsidered when planning the farm and decidingwhich valve types to use. It is necessary to haveenough pressure to ensure that the correct flow rateis maintained through the valves; if the head loss istoo high the water flow into or inside the farm willbe decreased.

2.2.3 Pipe parts – fittings

A large variety of pipe parts can be found, espe-cially for PE and PVC pipes (Fig. 2.4). Variousbends or elbows are normally used in aquaculture.T-pipes are also used to connect different pipes.

Different conversion parts allow the connection of pipes or equipment with different diameters.Sockets, flanges or unions are used to connect pipesor pipe parts. Sometimes end-caps are used to closepipes that are out of use. A particularly useful partis the repair socket which allows connection of anadditional pipe (a T-pipe) to a pipeline where thewater in the installation flows continuously, whichmeans that connections can be made to pipelinesthat are in use.

2.2.4 Pipe connections – jointing

The connection or jointing of pipes and pipe partsmay be executed in various ways depending on thematerial used to make the pipe and the pipe part(Fig. 2.5). For PE, fusing (heating) is the only pos-sible jointing method. This process may be carriedout by a blunt heating mirror, or electrofusion maybe used. When using a fusion mirror both the pipesto be joined are heated on the mirror to make them

D

E

Figure 2.3 Continued.

soft and adhesive; then the mirror is removed andthe pipes are pressed together. The materials of thetwo parts are fused together and form a fixed con-nection. Resistance wire is an integral part of anelectrofusion socket. When an electric current ispassed the material around the wire will fuse,including the two pipe parts added to the socket,and then a fixed connection is established.

Pipes that are fused or glued together are per-manently connected and cannot be separated. Ifthere is a need to create non-permanent connec-tions, it is possible to use flanges fused or glued tothe separate pipe parts which are then screwedtogether. To obtain a completely watertight con-nection a gasket is placed between the parts beforethey are screwed together. A union is a very easypipe connection to separate. It is always desirableto have some non-permanent connections becausesometimes it is necessary to separate the pipelinefor maintenance and exchange of equipment. Thepossibility of exchanging pipes and pipe parts in thewater department of the fish farm must also be con-sidered because of the need to allow for possible

increase in farm production and also because of the constant requirement for modernization of theequipment.

It is common to use sliding sockets in the outletpipe. This kind of connection system can only beused on unpressurized or very low pressurepipelines (<0.2mH2O). If this type of connection isused on pressurized pipes they will easily slideapart.

2.2.5 Mooring of pipes

Pipes may carry large amounts of water at highvelocities. This generates large forces that maycause movements of the pipe. In the worst case this

Water Transport 13

Figure 2.4 Cross-sections of fittings used in aquaculture.

Figure 2.5 Connection methods used for different pipematerials and in different places.


can damage the pipeline. For this reason a correctmooring system for the pipeline is of great impor-tance (Fig. 2.6).

When there is a reduction in the pipe size, orwhen using T-pipes or elbows, there is an increasein the forces dependent on the velocity, and therewill normally be a need for mooring to avoid move-ments and breakage of the pipes. Putting the pipesin a ditch will stabilize them and the ditch will func-tion as a mooring for the pipe. In exposed places,however, it may also be necessary to have addi-tional moorings; concrete blocks may, for instance,be used on elbows. This also shows the importanceof having smooth pipe linings. In indoor facilitiesclamps are used to attach the pipes to the ceiling,walls or floor, and in this way moor and stabilize thepipes.

Inlet or outlet pipes placed underwater or sub-surface in the sea or lakes require moorings. Spe-cially designed concrete block weights are normallyused to moor pipes to the ground (Fig. 2.6) toprevent them floating to the surface as a result oftheir buoyancy, especially when they are empty oronly partly filled with water. The distance betweenthe weights depends on the pipe type, diameter,weight size and expected water flow. When placingoutlet pipes sub-surface, it is important to considerthe weight of the pipe both when filled with waterand filled with air, the buoyancy being much greaterin the latter case which will also increase therequirements for weights. Usually pipe suppliershave their own mooring tables with recommendedblock weights and distance between them.

2.2.6 Ditches for pipes

The inlet and outlet pipes may be laid on thesurface or in ditches (Fig. 2.7). It is generally

cheaper to lay the pipes on the surface, but it is thenmore important to moor them, especially in con-nection with obstructions. Pipes on the surface may,however, inhibit transport and it can therefore benecessary to lay them in ditches. If the pipe is putin a ditch, care must be taken to avoid any damagewhen heavy traffic passes over the ditch. Thereforethe ditch must be constructed and overfilled withgravel correctly. A ditch may be constructed in thefollowing way: a layer of compressed crushed rockor gravel is laid as a base for the pipe in the ditch;then the pipe is laid with sufficient slope (>0.05%).Fine gravel is placed around and over the pipe tocreate good protection; this should only be handcompressed. Afterwards the ditch is filled with ordinary ditch material until ground level isreached. It is normal to overfill slightly becauseafter a while the material will compress to normalterrain level.

The use of ditches will also improve the farm aesthetically because there are no pipes on the

Figure 2.6 It is important that pipes aremoored to avoid movement and possiblebreakages. On inlet pipes in the sea or inlakes specially designed block weightsare used.

Figure 2.7 It is important that ditches for pipes are cor-rectly designed.

surface. Ditches also improve the possibilities forpublic traffic to use the area.

2.3 Water flow and head loss inchannels and pipe systems

2.3.1 Water flow

The amount of water that flows through a pipe orin an open channel depends on the water velocityand the cross-sectional area of the pipe or thechannel where the water is flowing. The followingequation may be used for pipes and channels; it isalso called the continuity equation:

Q = VA

where:

Q = water flow (l/min, l/s, m3/s)V = water velocity (m/s)A = cross-sectional area of where the water is flow-

ing. For full pipes the cross-sectional area willbe the interior cross section of the pipe.

The above equation can be used as a basis forconstruction of a chart. If two of the sizes areknown the last can be read from the chart and nocalculation is necessary. Often the head loss is alsoincluded in the chart (see below).

ExampleThe water flow to a farm is 1000 l/min (0.0167m3/s).The acceptable velocity in the pipeline is set at 1.5m/s. Find the necessary pipe dimensions if onepipe is to be used.

A = Q/V

A = 0.0167/1.5= 0.011m2

Now

A = pr2

where:r is the internal radius of the pipeand rearranging gives

rA=p

Therefore

The internal diameter in the pipe must therefore be2 × 59 = 118mm. Standard dimension pipes areavailable with an exterior diameter of 125mm; aPN6 pipe with a wall thickness of 6mm (supplierinformation) therefore has an internal diameter of113mm. This is actually slightly too small, but as thenext stardard exterior dimension is 160mm, it is bestto choose the 125mm pipe. This will result in thewater velocity being slightly higher.

For an open channel the flow velocity depends onthe slope, the hydraulic radius and the Manningcoefficient. The Manning equation is used to calcu-late the flow velocity:

where:

V = average flow velocity in the channelR = hydraulic radiusS = channel slope n = Manning coefficient.

The hydraulic radius is the ratio between the cross-sectional area where the water is flowing and thewetted perimeter, which is the length of the wettedsurface of the channel measured normal to the flow.

To achieve water transport through the channel itmust be inclined. The slope is defined as the ratiobetween the difference in elevation between twopoints in the channel and the horizontal distancebetween the same two points.

ExampleThe horizontal distance between two points A and Bis 500m. Point A is 34m above sea level and point B

R = cross-sectional areawetted perimeter

VR S

n=

23

12

r =

==

0.011

0.059m

59mm

p

Water Transport 15


is 12m above sea level. Calculate the slope (S) of thechannel.

S = (34m − 12m)/500m= 0.044= 4.4cm/m

This means that for each metre of elevation the horizontal distance is 22.7m.

To ensure drainage, it is recommended that theslope is more than 0.0013, while self-cleaning isensured with slopes in the range 0.005–0.010.11

The Manning coefficient is determined by exper-iment, some actual values being about 0.015 forconcrete-lined channels and 0.013 for plastic, whileunlined channels made of straight and uniformearth have a value of 0.023 and those made of rock0.025.11

Based on the flow velocity and the cross-sectionalarea, the flow may rate be calculated with the con-tinuity equation which also can be expressed as:

where:

Q = water flowA = cross-sectional area where the water is flowingV = average flow velocity in the channelR = hydraulic radiusS = slope of the channeln = Manning coefficient.

2.3.2 Head loss in pipelines

All transport of water through a pipe or a channelbetween two points results in an energy loss (headloss). This is caused by friction between the watermolecules and the surroundings. In all pipe partswhere there is a change in the water direction(bends) or narrow passage (valves) additional friction will occur; this will also increase the headloss.

Inside a pipe there is a velocity gradient, with thehighest water velocity in middle of the pipe and thelowest close to the pipe walls because friction ishighest against and close to the wall. In addition tofriction loss against the wall there will be frictionbetween the water molecules because their veloci-ties are not equal.

Q VAR S

nA= =

23

12

The amount of energy in water is constant(Bernoulli equation) if during passage no energy issupplied to or extracted from it. When frictionoccurs, the energy in the water is transformed intoanother form of energy, normally heat. This is verydifficult to perceive with the large amounts of waterthat are common in aquaculture, since much energyis required to heat the water (see Chapter 7).However, in a thin pipe with a large amount ofwater passing through at very high pressure, it ispossible to observe heating of the water.

As a result of frictional losses when flowingthrough a pipeline, the energy of the water must behigher at the beginning (inlet) than at the end(outlet); energy lines can be used to illustrate this.If the water is pumped, the pump pressure mustovercome these frictional energy losses in additionto the pump height.

The energy loss (hm) due to friction through a pipeline may be calculated using the Darcy–Weisbach equation:

hm = fLV 2/2gd

where:

f = friction coefficientL = length of pipelined = diameter of pipeline (wet)V = water velocityg = acceleration due to gravity.

The friction coefficient depends on the pipesurface; this is normally called the roughness of thepipe. Relative roughness (r) is defined as the rela-tion between the absolute roughness (e) and thediameter (d) of the pipe, r = e/d (Fig. 2.8). High

Figure 2.8 Relative roughness describes the relationbetween the absolute roughness (e) and the pipe diam-eter (d ).

relative roughness gives high friction. The amountof friction depends on the pipe material, the con-nection method and the age of the pipe. Forexample, a new plastic pipe will have a lower fric-tion coefficient than an old pipe. The fouling thatoccurs in pipes that have been in use for some timewill increase the roughness of the pipe. The f valuefor the pipe is given by the manufacturer and for PEor PVC pipes normally ranges from 0.025 to 0.035.For new pipes the value is lower, but when doingcalculations values for old pipes must be used.

The friction coefficient also depends on flowtype. The flow pattern can be divided into laminarand turbulent.The frictional losses are much higherwith turbulent flow. This will always be the case inpipes used in aquaculture, because the water veloc-ity is so high. Laminar flow may occur in open chan-nels with low water velocity. The Reynolds numberRe––

is a non-dimensional number used to describethe flow conditions. If Re

––is less than 2000 the flow

is laminar; when it is above 4000 the flow is turbu-lent. Between these Re

––values the flow is unstable

and both turbulent and laminar conditions mayoccur. Re

––can be calculated from the following

equation:

where:

V = average water velocityd = internal pipe diametern = kinematic viscosity.

Kinematic viscosity is the absolute viscosity dividedby the density of the liquid; the unit is m2/s (for-merly the stoke was used: 1St = 16−4 m2/s).The kine-matic viscosity tells us something about how easilythe liquid flows: for instance, oil will flow out slowlywhen drops are allowed to fall onto a horizontalplate, while water will be distributed much faster.The kinematic viscosity of water decreases withtemperature; for example, is it reduced from 1.79 ×10−6 m2/s at 0°C to 1.00 × 10−6 m2/s at 20°C.11 Salin-ity will also increase the kinematic viscosity ofwater: with a salinity of 3.5% it is 1.83m2/s at 0°Cand 1.05m2/s at 20°C.

ExampleThe average velocity of fresh water in a pipe of internal diameter 123.8mm is 1.5m/s (0.1238m).

RVd

e =n

The temperature is 20°C. Calculate the Reynoldsnumber.

This clearly illustrates that the water flow in the pipeis in the turbulent area.

By calculating the Reynolds number and the rela-tive roughness of the pipe, the friction coefficient f,can be found from the Moody diagram (Fig. 2.9).

Computer programs and special diagrams (Fig. 2.10) are available for calculating the head losscaused by friction inside a pipe. It is important tobe aware that these diagrams are specific for givenpipes because the f- value of the actual pipe is usedto construct them.

ExampleCalculate the head loss in an old PE pipe with inter-nal diameter of 110mm (0.11m). The length of thepipe is 500m and the velocity in the pipe is 1.5m/s;the friction coefficient is 0.030.

hm = fLV2/2gd= 0.030 × 500 × 1.52/2 × 9.81 × 0.11= 11.47m

RVd

e =

= ××

=−

n1.5 0.1238

1 10185700

6

Water Transport 17

Figure 2.9 Principle of the Moody diagram showingthe relation between relative roughness and Reynoldsnumber.


This means that the head loss in the water flowingthrough the pipeline is 11.47m. If the water is to flowin the pipe, the intake pressure must be at least 11.47mH2O in addition to atmospheric pressure;if it is less, the flow rate through the pipe will bereduced.

2.3.3 Head loss in single parts (fittings)

In addition to the head loss in the pipe there isenergy loss due to friction in pipe parts (fittings)because any obstructions in the pipe which createextra turbulence will increase the head loss. Addi-tional turbulence occurs in the inlet and outlet ofthe pipe, in valves, bends, reductions, connections,etc. The head loss can be calculated from the equation

Ht = kV2/2g

where:

Ht = head loss in the single partk = resistance coefficient for the pipe partV = water velocityg = acceleration due to gravity.

ExampleThe water must flow through a 90° elbow: either two45° elbows or one 90° elbow with k values of 0.26and 0.9, respectively, can be used to achieve this. Theflow velocity is set to 1.5m/s. Calculate the head lossin the two cases.

For the two 45° elbows

Ht = ΣkV2/2g= 2 × 0.26 × 1.52/2 × 9.81= 0.06m

For the 90° elbow

Ht = kV2/2g= 0.9 × 1.52/2 × 9.81= 0.10m

As this example illustrates, there is a great advantagein using two 45° bends rather than one 90° bend toreduce the head loss. This will apply, for instance,for the outlet pipe from a fish tank.

The k values for different parts may be found fromspecial tables (e.g. Table 2.2). They are also found incatalogues published by suppliers of fittings.

When constructing the pipe system the head loss that results from fittings in the pipeline must be considered in addition to the head loss in thepipe itself. The resistance of every single part mustbe added, so the sum of every single resistance plusthe head loss in the pipeline gives the total headloss.

When designing the inlet pipe to a fish farm, it isimportant to use smooth bends to reduce the totalhead loss in the pipeline.

2.4 PumpsPumps are mechanical devices that add energy tofluids by transforming mechanical energy (nor-mally from electric motors) to potential and/orkinetic energy of the fluid. Increase in potentialenergy is illustrated by the lifting of water to an ele-vated tank, while the increase in velocity and hence

Figure 2.10 Diagram showing the relation betweeninternal diameter, water flow (1000 l = 1m3), water veloc-ity and head loss for a pipe with a known f value. (Repro-duced with permission from Helgeland Holdings.)

the flow rate through a pipeline by pumpingincreases the kinetic energy of the water. Pumps arecommonly used in aquaculture systems, usually toincrease the system pressure and thereby force thewater to move against an energy gradient. In mostaquaculture situations pumps are used to lift waterfrom one level to another. Water will flow onlywhen energy is available to create a flow, i.e. thereis a positive energy gradient. In hydraulic systemspumps are used to create the pressure which is nor-mally high. This allows the fluid to do work, such asturn shafts or extend a hydraulic cylinder against aload. Oil is commonly used in such systems, butthose using water are also available.

Pumps are fairly efficient machines for transfer-ring energy to water, provided they are correctlyselected for the job. The key requirement whenselecting a pump is that there shall be a close cor-relation between the system requirements and themaximum operating efficiency of the pump; sub-optimal pump selection may result in significantlyincreased operating and maintenance costs and/orresult is system failure.

2.4.1 Types of pump

There are several types of pump based on differentprinciples (Fig. 2.11). The type of pump chosen

depends on a number of factors, including theamount of fluid to be pumped and its characteris-tics, and the head.

A major pump type is the displacement pump inwhich liquid is displaced from one area to another.An example is the piston pump: when the pistonmoves up and down it creates, respectively, avacuum and pressure, and in this way the liquid istransported; back-flow valves must be included.Gear pumps and screw pumps are other types ofdisplacement pump. The pumps may break if theoutlet is blocked.

The ejector pump is based on another principle.Here a part flow under high pressure is used todraw a main stream with much higher water flowbut lower pressure. By pumping water into a specially designed narrow passage a vacuum effectwill occur and create a drag on the main stream.The design of the ejector is most important. Thisprinciple is used in pumps for fish transport, forexample.

In air-lift pumps, air is supplied inside an openpipe standing partly below the surface and partlyfilled with water. The air bubbles will then drag thewater towards the surface and in this way a waterflow is created inside the pipe. This principle maybe used to pump water, add air (aeration) and forfish transport.

Water Transport 19

Table 2.2 Typical resistance coefficients, k, for different fittings. Values of k will vary with the producer of the fitting.

Fitting k Comments

Pipe entrance in a basin 0.05–1 Lowest value with rounded inlet pipes; highest with pipes with sharp edges. The or in a river value is increased when the pipe goes into the basin.

Contraction of pipes 0.05–0.5 Lowest value with conical contraction and small alteration in the diameter.

Expansion of pipes 0.05–1 Lowest value with conical expansion and small alteration in the diameter.

Elbow Increasing with reduction in the pipe diameter. Long, smooth bends will have 90 degree 0.5–1 reduced k values. Smaller angles will have lower k values.45 degree 0.1–0.3

T-pipe (divided flow) 0.3–1.8 Depends on the proportion divided from the main flow; increased with increased part flow.

T-pipe (connection flow) 0.1–0.8 Depends on how much water is supplied via part flow in the T-pipe; increases with increased size.

Valves Values are highly dependent on the specification and the producer of the valve.Ball valve 0.1 Values shown are for fully open valves.Angle seat valve 1.3Diaphragm valve 4Check valve 2Gate valve 0.2


An endless screw pump is based on another prin-ciple; among other uses, such pumps are employedfor sludge.

For aquaculture facilities there is a need to pumpa large amount of water with a relatively small

lifting height. Centrifugal pumps or propellerpumps are the most suitable and most commonlyused. Centrifugal and propeller pumps aredescribed in Section 2.4.4.

Figure 2.11 Diagrams to show the prin-ciples of different pump types.

2.4.2 Some definitions

Pump height

When water is lifted from one level to another theheight difference is called the static lifting height.The lift height can either equal the pressure head(in the case of a submerged pump) or it can be acombination of vacuum and pressure head depend-ing on where the pump is placed. In addition to this,the pump has to overcome the head loss caused bythe friction in the pipe on both sides of the pump.If a manometer that measures the pressure is con-nected to the pump outlet the measured pressure isthe sum of the pressure head (static head) and headloss. When the water passes through the pump itneeds to have certain velocity in order to flow; thisis called the velocity head. The total pressure headis obtained by summing the manometric height andvelocity head. The actual pressure head at the endof the pipe, in addition to the difference in levelmust also be considered, for example when thepump is required to deliver water to a pressurizedtank.

To collect water from a lower level, a vacuumhead must be overcome by removing air from theinlet pipe to create a vacuum inside it. The pressureof the atmosphere will force the surrounding waterup the inlet pipe. For this reason the vacuum headmust not exceed atmospheric pressure which is nor-mally 1013mbar (10.3mH2O), although it dependsto some extent on the weather (low pressure, highpressure) and the altitude. When a pipe is com-pletely emptied of air, the water will therefore beforced up the pipe to a height of 10.3m. The actualsuction head of a pump is, however, lower than 10.3mH2O, because of losses such as from the velocityhead in the inlet pipe. To make a centrifugal pumpself-suctioning it must include a mechanism, forinstance a specially designed impeller, to removethe air from the inlet pipe.

Cavitation

If the vacuum into the pump is too high, the watermay boil and vaporize. That the temperature ofvaporization pressure dependent can easily be illus-trated with a pressure boiler, where the boiling tem-perature of the water is increased; similarly, belownormal atmospheric pressure, the boiling tempera-

ture will decrease (see below). When a mixture ofliquid and gas goes through a pump the boilingpoint will increase because the pressure around thewater molecules increases from vacuum upwards.In changing from gas to liquid the bubbles undergoviolent compression (implosion) and collapse cre-ating very high local shock, i.e. a sharp rise and fallin the local pressure; the phenomenon is called cav-itation. If this happens in connections to a pump orin the impeller, small metal parts can be dislodged.Multiple indentations or dimples in the materialcan result. The same may occur on boat propellerswhere worm like holes may be observed in thematerial of the propeller.

Cavitation reduces the effectiveness of pumpsand will also shorten pump life. A characteristic‘hammer’ noise is produced inside the pump whenit cavitates. Cavitation may also occur if there areleakages in the pipe or pipe connection on thesuction side of the pump. If air leaks in here (knownas ‘false air’), it will create air bubbles that enter the pump chamber with the water where theyimplode.

Cavitation can happen if the suction head is toohigh. When the pressure around the water mole-cules drops, the water will boil at a lower tempera-ture (i.e. the boiling point of water is reduced). Forexample, if the pressure drops from atmospheric(10.3mH2O) to 1mH2O, water will boil at 46°C.Thisphenomenon can be observed when boiling waterat high altitude, for instance in the Himalayas. Herethe water will boil below 100°C because the atmos-pheric pressure is less than 10.3mH2O. Atmos-pheric pressure also depends on the weather. Thesafe static suction head will also decrease withsurface water temperature from 10.4mH2O at 10°Cto 7.1mH2O at 21.1°C.12

Net positive suction head (NPSH)

If the pump is not self suctioning, the water levelmust be higher than the level of the pump. Thismeans that the impeller needs a certain pressure tofunction optimally. The net positive suction head(NPSH) gives the absolute lowest pressure thewater must have when flowing into the pumpchamber, or (more easily) the actual height of waterover the impeller. If the water pressure is lowerthan the NPSH the pump will cavitate. NPSH

Water Transport 21


depends on the water flow and increases withincreasing flow; it can be described as follows:

NPSH = hb − hv − hf + hh

where:

hb = barometric pressurehv = vapour pressure of the liquid at the operating

temperaturehf = frictional losses due to fluid moving through

the inlet pipe including bendshh = pressure head on pumping inlet (negative if it

is a static lift on the suction side of the pump).

ExampleA pump is to be chosen for a land-based fish farm.With an actual discharge (Q) and head (H), the nec-essary NPSH can be read from the pump perfor-mance curves to be 4mH2O.The fish farm is situatedclose to the sea and the barometric pressure (hb) ismeasured to be 10.3mH2O. The maximum tempera-ture during summer time is 30°C which correspondsto a vapour pressure (hv) of 4.25N/m2 equal to 0.44mH2O. The friction loss in the inlet pipe includ-ing loss in fittings (hf) with the actual water velocityis 1.5mH2O. The static suction lift (hh) is 2m.

The NPSH can then be calculated as follows:

NPSH = hb − hv − hf + hh

= 10.3 − 0.44 − 1.5 + (−2)= 6.36mH2O

This is higher than the NPSH value of 4mH2O thatthe pump requires, which means that there will notbe any problems regarding NPSH when using thepump.

The NPSH requirements of a specific pump aregiven in the pump diagram (see section 2.4.5). Thisvalue must be higher than the value calculated fromthe above equation. Remember that NPSH is givenin units of pressure (mH2O, bar or pascal).

2.4.3 Pumping of water requires energy

Energy is required to pump water from one level to another. Energy consumption is usuallyexpressed as power (P), which is energy suppliedper unit time. P is measured in joules per second;1J/s = 1 watt (W).

The following equation can be used to calculatethe energy requirement for pumping:

P = rghQ

where:

r = density of water (kg/m3)g = acceleration due to gravity (m/s2)Q = water flow rate (m3/s)h = height that the water is pumped.

ExampleCalculate the energy required to lift 1000 l/minof water by 5m and 15m (including the frictionhead). The density of water is 1025kg/m3, the flowrate is 0.016m3/s and acceleration due to gravity is9.81m/s2.

Case 1: 5m lift

P = rghQ= 1025kg/m3 × 9.81m/s2 × 5m × 0.016m3/s= 804.4J/s= 804.4W.

Case 2: 15m lift

P = ρghQ= 1025kg/m3 × 9.81m/s2 × 15m × 0.016m3/s= 2413.3W= 2.4kW.

This illustrates that by tripling the pump height theenergy requirement is also tripled.

The power supplied from the pump to the wateris called the water effect and is the sum of the veloc-ity head, head loss and static head. As a result ofenergy losses in the pump, more power is suppliedto the pump than the pump supplies to the water.The efficiency of the pump, is given by the equation

h = PD/PS

where:

h = efficiencyPD = energy delivered from the pump to the waterPS = energy supplied to the pump.

ExampleA pump has an efficiency of 75% and consumes 5kW of electric power. How much of this power isused to pump the water?

h = PD/PS

PD = hPS

= 0.75 × 5= 3.75kW

Energy losses occur at several places in thepump. This results in efficiencies below unity (Fig.2.12). Losses occur in the pump motor (m), trans-mission (t) and in the impeller (p), the sum of whichgives the pump efficiency. Low pump efficiencyresults in the creation of heat, because energycannot disappear; for example, when using sub-merged pumps this energy will be transferred to thewater, which is heated. In water re-use systems thisheating can be noticeable. The total efficiency of apump hA, can be calculated as follows:

hA = hp + ht + hm

Efficiency may also be defined as hydraulic effi-ciency hH (loss when the water flows through apump), volumetric efficiency hv (leakage of waterbetween suction and pressure side of the pump, forexample in centrifugal pumps) and mechanical effi-ciency hm (losses in the motor and transmission).

In fish farming the usual efficiency of well-suitedpumps is around 0.7. Efficiency normally variesbetween 0.4 and 0.8.

Pump costs

To calculate the cost of pumping, the followingequation can be used:

Pumping cost = PdEP

where:

P = power (kW)d = duration of pumping (h)EP = electricity price per kWh (kilowatt per hour).

ExampleA centrifugal pump with a 5kW power supply runscontinuously to supply a fish farm with water. Cal-culate the yearly electricity cost of running the pumpwith an electricity price of 0.1€.

Yearly pumping cost = PdEP= 5kW × 24h × 365 days ×

0.1€/kWh= 4380€.

2.4.4 Centrifugal and propeller pumps

Centrifugal pumps

Centrifugal pumps account for the majority ofthose used by aquaculture enterprises. A centrifu-gal pump contains three major units: the powerunit, the pump shaft and the impeller (Fig. 2.13).The power unit, an electric motor, causes the pumpshaft to rotate and with it the attached impeller.Around the rotating shaft where the impeller andmotor are fixed there are seals that preventsleakage via the shaft into the motor.

If the inlet and the pump cavity are filled withwater (primed), the rotation of the impeller causesmovement of water molecules which will be accel-erated outwards towards the periphery of theimpeller. This will reduce the pressure in the centreof the impeller and new water is drawn into the eyeof the impeller via the pump inlet. Vanes on theimpeller may direct the flow of water and help totransfer the energy from the impeller to the water.

Water Transport 23

Figure 2.12 Only part of the electricenergy fed to the pump is used to transportthe water.


The rotation of the impeller imparts high veloc-ity to water molecules at the periphery. When thiswater leaves the impeller, the velocity is rapidlyreduced and the dynamic head is converted to statichead. How the pressure changes from the inlet tothe outlet on a centrifugal pump depends on thecharacteristics of the pump. The outlet pressure is afunction of the inlet pressure and pump character-istics, mainly the design of the impeller.

The impeller may be open, semi-closed orenclosed; all have different characteristics (Fig.2.13). The choice depends on the amount of partic-ulate matter in the water. If there are many parti-cles (sludge like), an open or semi-closed impelleris normally used; if there are few particles or highpressure is required, an enclosed impeller is used.The enclosed impeller has the highest efficiency, buttolerates the lowest amount of suspended solids inthe water. In aquaculture facilities all types ofimpeller are used.

Propeller pump

Propeller pumps are of simple construction with apropeller rotating inside a pipe (Fig. 2.14).The prin-ciple is the same as a propeller on a boat, butinstead of moving the boat, the propeller is fixed sothe water moves instead. Propeller pumps have theadvantage that they can deliver large amounts ofwater at low pressure (normally less than 10mH2O). The reason for the low pressure is leakagethat occurs between the two sides of the propeller(head and suck). On some pumps the flow rate caneasily be varied by adjusting the angles of the pro-peller vanes. The propeller is normally installed ina vertical pipe, but it is also possible to place it in ahorizontal pipe to create a flow.

Dry placed or submerged pumps

Centrifugal pumps can be dry placed, either abovewater level or in a dry well below water level. Theycan also be installed in the water as submergedpumps (Fig. 2.15).

A dry placed pump consists of a pump chamberwith an impeller, a transfer shaft and an electricmotor. On dry placed pumps the motor is cooled bya fan and on submerged pumps by water. Betweenthe pump chamber and the motor there is a seal toprevent water leaking into the motor. Dry placedpumps may also be made self-suctioning by use ofa specially designed impeller. Dry placed pumps arenormally bolted to a rack that is fixed to the floor.

In a submerged pump the motor and the pumpchamber are usually built together and encapsu-lated in one unit which is lowered into the water.As for dry placed pumps, it is important to have agood seal between the pump chamber and themotor. The whole must be watertight, to preventwater from entering the motor.The motor is cooledby the surrounding water and the encapsulatingpart can be equipped with cooling ribs to facilitatethis.

Both dry placed and submerged pumps are com-monly used in aquaculture. The advantage with dryplaced pumps is that maintenance is simple becauseit is easy to access the pump and pump parts. A dis-advantage could be that artificial cooling of themotor is necessary, needing a fan. If there are leak-ages in the inlet pipeline false air may be drawn in,

Figure 2.13 A centrifugal pump contains three majorunits: the power unit (motor), the pump shaft and theimpeller which can be open, closed or semi-closed.

causing cavitation in the pump and supersaturationof nitrogen (see Chapter 8), especially when usingdry placed pumps standing above the water surface.

2.4.5 Pump performance curves and workingpoint for centrifugal pumps

Characteristics curves

The characteristics curves (pump diagram) are usedto describe the performance of centrifugal pumps

(Fig. 2.16). The most important is the head versuscapacity curve which shows the connectionbetween the water flow (discharge) and the head.Depending on its construction the same centrifugalpump can either be used to deliver a large waterflow with a low head or a smaller water flow with alarger head.Adding a valve on the pump outlet andgradually closing it will decrease the water flow butincrease the head, because the cross-sectional areathrough which the water is forced is reduced.As thevalve is closed more pressure is needed to discharge

Water Transport 25

Figure 2.14 A propeller pump com-prises a fixed propeller rotating insidea pipe. This results in transport ofwater through the pipe. The photo-graph shows water delivery from submerged propeller pumps.


the water; again this increases the pump head. Apump characteristics curve can be constructed bymeasuring the water flow that the pump deliverswith different heads: if pump behaviour is ideal, theplot should be a straight line. This seldom happensbecause of the reduction in pump efficiency at thelimits of performance, so the plot is curved. If largeamounts of water are being pumped, the frictionloss through the pump becomes large. In additionto the frictional losses there is so called impact lossresulting from the impact of the water moleculeshitting the impeller and the inlet and outlet parts ofthe pump chamber. The highest losses occur withlarge heads and large flows at the ends of the pumpcharacteristics curve, close to the x- and y-axes.Thisis to be expected, because a pump has the highestefficiency at its construction point in the middle of

the characteristics curve. Normally this point is dis-tinctly marked.

A pump’s characteristics will of course dependon design and size of the chosen impeller, each ofwhich will have individual characteristics. A cen-

A

B

Figure 2.15 Centrifugal pumps are either (A) sub-merged (pump shown ready to lower into the water) or(B) dry placed.

Figure 2.16 Pump performance (characteristics)curves for centrifugal pumps: head, efficiency, powersupplied and net positive suction head are plottedagainst water flow rate (discharge).

trifugal pump can normally be delivered with dif-ferent impellers, and it is quite easy to change them.In many cases several pump characteristics aregiven in the same diagram, one for each diameterof the impeller.

The pump diagram may also show a plot of pumpefficiency versus discharge. A pump that works asclosely as possible to the maximum efficiency (con-struction point) should be chosen; operation awayfrom the construction point will result in decreasedefficiency. Pump efficiency may also be given in theso-called shell or muscle diagrams in which curvesand/or circles represent percentage efficiency.

The power requirements are also given in thepump diagram. Here the necessary power that mustbe supplied to the pump is plotted versus differentdischarges and heads.

The NPSH curve gives information about thenecessary ‘over-pressure’ required by the pumpwhich, as described earlier, increases with increas-ing water flow. It is important to be aware that thecurve showing the NPSH is constructed in relationto atmospheric pressure, i.e. 10.3mH2O. The sup-plied (inlet) pressure is the sum of the air pressureand water pressure to the pump impeller. If this islower than the NPSH requirements shown on thecurve, the pump will cavitate. There will not nor-mally be problems with NPSH if there is positivewater pressure into the pump. Problems are morelikely to occur when there is static suction lift intothe pump.

Pipeline characteristics

When delivering water through a pipeline, the pres-sure and/or head that the pump has to deliverdepends on the flow rate to be pumped through thepipe. If the pump is not running, the only resistanceis the fixed static head; however, when the pumpstarts to work there is in addition a resistance head(manometric head) caused by friction losses in the pipe.

The manometric head increases with the flowrate through the pipe. If the pump does not lift thewater to a higher level, but only overcomes frictionin the pipes, there is no static head, but only a resis-tance head. To find the total friction, the character-istics of the total pipe system, including pipe length,pipe material, number of bends, etc., must beknown. With this information the total head loss

through the system with different water flows canbe calculated. The results may be plotted to showthe connection between the total head and thewater flow. This curve is called the pipeline charac-teristic curve. It starts at zero if there is no starthead and increases exponentially with the waterflow (Fig. 2.17).

Working point

To find the working point for a pump with givenpipeline characteristics, the pipeline and pumpcharacteristic curves can be drawn on the samediagram (Fig. 2.17). The actual working point of thepump will be at the intersection of the two curves.This point shows the water flow that the pump willdeliver through the pipeline. To achieve the bestpossible efficiency it is important that this point isas close to the construction point of the pump aspossible. If not, another pump should be chosen.

2.4.6 Change of water flow or pressure

When choosing a pump it is important that the pump works as closely as possible to its construction point. What may then be done if thechosen pump does not fulfil this condition andcannot be changed?

Change of impeller (pump wheel)

Most of the centrifugal pumps can be supplied withdifferent impellers. By selecting an impeller of dif-ferent shape and diameter it is therefore possible to

Water Transport 27

Figure 2.17 By drawing the pipeline characteristiccurve and the pump characteristic curve on the samediagram the working point of the pump is found at theintersection.


change the pump characteristics (Fig. 2.18). In thisway it is possible to find an impeller that is betteradjusted to the working point of the pump, andbetter efficiency can be achieved.

The following connections can be used to showthe pump performance when reducing or increas-ing the diameter of the impeller:

Q2/Q1 = d2/d1

H2/H1 = (d2/d1)2

P2/P1 = (d2/d1)3

where:

d2 = diameter of smallest impellerd1 = diameter of largest impellerQ = flow rateH = headP = power requirement.

ExampleThe diameter of an impeller to a centrifugal pump isincreased from 515mm (d1) to 555mm (d2).The flowrate in the first case was 700 l/s (Q1) and the head is9m (H1). Calculate the new flow rate (Q2) and thenew head (H2).

Q2 /Q1 = d2/d1

Q2 = 1.08 × 700= 754 l/s

H2/H1 = (d2 /d1)2

H2 = 9 × 1.16= 10.5m

These equations are most reliable with a range of20% increase/decrease in diameter, because largechanges of the diameter of the impeller may alterpump geometry.12

Connection of pumps

Pumps can be connected in series (one afteranother) or in parallel (beside each other). In thisway the water flow and water pressure can bechanged (Fig. 2.19). When pumps are connected inseries, they are placed one after another in the samepipeline, so increasing the head while maintainingthe same water flow. The inlet pressure of thesecond pump is the outlet pressure from the firstpump so the pressure head is increased.

When pumps are connected in parallel, the mainpipeline is divided into two sub-pipelines. Onepump is placed in each sub-pipeline; these are thenconnected to the main pipeline. Parallel connectiondoubles the water flow, while the head remains constant.

High pressure pumps

High pressure pumps are special types of centrifu-gal pumps that utilize a system of pumps connected

Figure 2.18 Changing the impeller alters the charac-teristics of the pump.

Figure 2.19 The flow rate and pressure may bechanged by connecting pumps in series (one afteranother in the same pipeline) or in parallel (beside eachother in separate sub-pipelines).

in series (Fig. 2.20). In a high-pressure (multi-stage)pump the impellers are connected in series, and canshare the same motor and shaft. The motor andshaft are of course larger than if a single impeller is used. When the water leaves the first impeller it is collected and directed into the centre of thenext impeller and so on, and the head graduallyincreases depending on the number of impellers.This is possible through the design of the pumpchamber. High-pressure pumps are used forinstance for pumping groundwater from greatdepths or for other purposes where there is eithera large head to be overcome or if a high water pres-sure is required.

2.4.7 Regulation of flow from selected pumps

In many cases there is a requirement to adjust thewater flow from specific pumps, for instance toreduce it. It is possible to use several methods forthis purpose and these are briefly described below.

Adjustment of RPM

One method for regulating the water flow from apump is to change the speed (RPM) of the motorand hence the impeller. When the speed is dimin-ished both the water flow and the head are reduced.The usual type of electric motor employed forpumps is an asynchronous motor. Its speed may bechanged by coupling in and out extra sets of polesin steps by a switch, which will change the RPM ofthe motor, also in steps. The following connectionis given between the RPM (n), water flow (Q), head(H) and the power (P):

Q2/Q1 = n2/n1

H2/H1 = (n2/n1)2

P2/P1 = (n2/n1)3

From these equations it can be seen that the waterflow increases in direct proportion to the RPM, thehead increases in proportion to the square of theRPM, and the power increases in proportion to the cube of the RPM.

Water Transport 29

Figure 2.20 In a high-pressure (multistage) pump theimpellers are connected in series and have a commonmotor and shaft.

�


Continuous adjustment of the RPM in pumps is,however, difficult in practice. The motor on a cen-trifugal pump normally uses alternating current andthe continuous adjustment of a.c. motors is quitecomplicated and expensive. The usual way to adjustthe motor speed is to transform the frequency ofthe current (Fig. 2.21). If, for instance, the normalfrequency of the electricity is 50Hz, reducing it to40Hz will reduce the speed of the motor, which willreduce the water flow out of the pump but main-tain the head. Reducing the a.c. frequency canaffect the pump motor; if it is reduced too much,the motor will stop and may be destroyed.The samething can happen if the frequency is increased toabove 50Hz.

The disadvantage of this system is that frequencyregulators are quite expensive and some energy

loss occurs during frequency transformation. Largeimprovement have, however, been made in thetechnology in this field during the past few years,and the cost has been considerably reduced.

ExampleA pump delivers 300l/s.The speed of the pump motormust be changed so that the pump only delivers 200 l/s. The original speed of the pump is 2800RPM/min. What is the new speed? The power supply is 30kW at the start. What is the new power supply?The new RPM is going to be:

Q2 /Q1 = n2/n1

n2 = Q2n1/Q1

= 200 × 2800/300= 1867RPM

The new power supply will be:

P2 /P1 = (n2/n1)3

P2 = P1(n2/n1)3

= 30(200/300)3

= 8.9kW

Throttling

By placing a throttle valve on the pressure side, itis possible to close the outlet of the pump (throt-tle) (Fig. 2.22). In this way an artificial head iscreated and hence a reduction in the water flow,according to the pump characteristics. Higher throt-tling results in a move to the left on the pump curve.To throttle on the pressure side is only possible forcentrifugal pumps, because the water is inhibitedfrom going to the peripheries of the impeller.Throt-

Figure 2.21 A frequency transformer changes the fre-quency of the current and may therefore be used to reg-ulate the pump speed.

Figure 2.22 Throttling on the pressure side may beused to reduce the water flow from a centrifugal pump,but is not a good solution.

tling on the pressure side on pump types other thancentrifugal and propeller can result in pump break-age. Throttling centrifugal and propeller pumps isnot very satisfactory as there is significant energyloss because the throttling valve creates an artificialhead.The power consumption is the same when thewater flow is reduced, which easily can be seen fromthe left shift on the characteristics curve. To throt-tle a centrifugal pump on the suction inlet must beavoided as this easily creates cavitation conditionsin the pump.

Several pumps

If there are large water requirements and the dis-charge varies, it is an advantage to use severalpumps of different sizes. In this way it is possible to couple pumps in and out and by doing this vary the total flow rate (parallel connection) and at the same time achieve an overall high efficiencyfor all the running pumps. To use several pumps will also improve the reliability because one pump can stop without halting the total water flowto a farm.

References1. Finnemore, J., Franzini, J.B. (2001) Fluid mechanics

with engineering applications. Mc-Graw Hill.2. Kundu, P.K., Cohen, I.M. (2001) Fluid mechanics.

Academic Press.3. Munson, B.R., Young, D.F., Okiishi, T.H. (2005) Fun-

damentals of fluid mechanics. John Wiley & Sons.4. Nayyar, M.L. (1999) Piping handbook. McGraw-Hill

Professional.5. Frankel, M. (2001) Facility piping systems. McGraw-

Hill Professional.6. Willoughby, D. (2001) Plastic piping handbook.

McGraw-Hill Professional.7. Karassik, I.J., Messina, J.P., Cooper, P., Heald, C.C.

(2000) Pump handbook. Mc-Graw Hill.8. Rishel, J.B. (2002) Water pumps and pumping systems.

McGraw-Hill.9. Sanks, R.L., Tchobanoglous, G., Bosserman,

B.E., Jones, G.M. (1998) Pumping station design.Butterworth-Heinemann.

10. ASM International (1988) Engineering plastics.In: Engineered materials handbook, vol 2. ASMInternational.

11. Huguenin, J.E., Colt, J. (2002) Design and operatingguide for aquaculture seawater systems. ElsevierScience.

12. Lawsons, T.B. (2002) Fundamentals of aquaculturalengineering. Kluwer Academic Publishers.

Water Transport 31

3Water Quality and Water Treatment:

an Introduction

investments costs involved in the running of rearingunits it is of course vital that the production per unitof farming volume is as high as possible. Fish canlive well in the wild even if water quality is sub-optimal. However, the food supply is usuallylimited and the growth rate will be much lower thanthat possible under optimal conditions.

There is often some kind of stress responseinvolved when there is an outbreak of disease.Disease can be latent in the stock but can becomea problem if the fish are exposed to some kind ofstressor, for instance sub-optimal water quality. Infish farms, where fish are grown as quickly as pos-sible, they are already under stress and so diseaseoutbreaks are more likely due to sub-optimal envi-ronmental conditions. It has been shown that bycatching wild fish and holding them under farmingconditions at a high stocking density, it is very easyto cause a disease outbreak, even if the waterquality is equal to that found at the wild site wherethe fish were caught.

Many experiments have been carried out for thespecies farmed today, and there are quite good datafor recommended water quality.8,9 However, as theamount of new water added per kilogram of fish iscontinuously decreasing, research is focused onaccurately documenting lowest acceptable levels ofnutrients, etc. to maintain optimal growth. Norwe-gian salmon smolt production can be taken as anexample. In 1985 the average size of the fish was 40g while today is it over 100g, as the result ofimproved feed and increased individual growthrate. Most of the sites used have limited fresh waterresources and therefore the amount of new watersupplied per kilogram of fish has been reduced,

3.1 Increased focus on water qualityAs the aquaculture industry becomes ever moreintensive, the focus on water quality in the rearingunits will also increase. Higher production densitieswill also increase the requirements for optimalwater quality, because of the degeneration inquality when the water flows through the produc-tion unit.

The importance of water quality is independentof the type of rearing unit and location of the pro-duction facilities. If using open production units inthe sea such as sea cages, it will of course be moredifficult to treat the water to improve quality, eventhough this is starting to happen (see Chapter 8concerning oxygenation in open sea cages). Onland-based fish farms with control of both the inletand outlet, water treatment to improve the waterquality will be much easier to perform.The increasein aquaculture production based on water re-usesystems will focus attention on water treatment toimprove quality.

During the past few years, the focus on the envi-ronmental impacts of the aquaculture industry haveincreased.1–5 In future, more stringent requirementswill be set to reduce these impacts. It is possible toreduce the discharge from the facility by optimalproduction management, and also by treating theoutlet water from closed production facilities. It willalso be important that production is adapted to sitecapacity, and of course that it is sustainable.6,7

3.2 Inlet waterWhile fish will grow in water of sub-optimal quality,their growth rate will not be maximized. With high

32

mainly by adding pure oxygen: in 1985 the licencerequirement for new water was 0.38 l/kg fish/min,while today in dry periods it is down to 0.1–0.2 l/kgfish/min.10

Water quality requirements depend upon thespecies. The same is the case regarding require-ments for the various life stages; the early stagesnormally have the highest requirements for optimalwater quality. Even if the quality requirements vary,it is better to have good quality water than bad, ifthis is possible.

Optimal water temperature is species specificand so general advice is impossible. Species can bedefined generally as warm water species (>20°C)and cold water species (<20°C). Some species tem-peratures prefer below 10°C. If the water tempera-ture falls below 0°C freezing will be a problem. Theoxygen content of the water enclosure (see full saturation, p. 119) is reduced with increasing tem-perature. At 5°C the available oxygen content is12.8mg/l while at 25°C is it reduced to 8.2mg/l.

The water ought to be fully saturated or super-saturated with oxygen gas. It is very important thatthe oxygen content of the rearing water is highenough: for instance, 7mg/l (70% saturation) is thetypical value for the outlet water in salmonidfarming, 30% having been consumed by the fish.

Of the other gases dissolved in the water, the con-centrations of nitrogen (N2) and carbon dioxide(CO2) must not be too high. The nitrogen gas con-centration should be below 100.5% saturation. Forcarbon dioxide, levels are not only dependent uponthe inlet concentration, they also increase in thetank as a result of fish metabolism which releasesCO2 into the water; the outlet concentration mustnot therefore be too high.

Water pH must neither be too low nor too high(the latter is seldom the case). This applies to fresh-water, since seawater has stable pH values of7.5–8.2. Sufficient alkalinity in the water helps tocontrol fluctuations in pH.

Too many particles in the inlet water may havenegative effects on the fish, for instance by cloggingtheir gills. Fish faeces will increase the particulatecontent of the water, the outlet concentration ofwhich must not be too high.

Ammonia may be a problem in production unitoutlet water because of waste products from fishmetabolism, but only with very limited supply andexchange of water.With normal water sources there

are no problems with ammonia concentration inthe inlet water.

The concentration of metal ions in the inlet watermay be of levels that are toxic to fish; low pH mayincrease this toxicity. Problem metals include alu-minium, copper, iron, zinc and cadmium.

Micro-organsims, including parasites, bacteria,viruses and fungi, may be present in the inlet waterat concentrations unfavourable for aquaculture.

Interactions between several water quality para-meters, for instance between pH and metals, mayalso pose quite a challenge. To fully understandwater treatments effects, a good knowledge of basicwater chemistry is an advantage; this topic is notcovered in this book but extensive literature isavailable, for example refs 11-13.

3.3 Outlet waterAll outlet water discharged from aquaculture facil-ities can present environmental problems whichcreate an imbalance in the ecosystem in the recipi-ent water body. This is especially important whenthe outlet water is discharged into freshwater.Freshwater recipient water bodies are of limitedvolume, whereas the sea represents an infinitelylarge water body; therefore the consequences of adischarge into a freshwater recipient, such as asmall lake, will be much greater for both open andclosed production units.

The effluent discharged from a fish farm cancontain three classes of pollutant (Fig. 3.1):

• Nutrients and organic matter• Micro-organisms• Escaped fish

Water Quality and Water Treatment: an Introduction 33

Figure 3.1 Every fish farm, whether it is on land or inthe sea, will discharge effluent.


The quantity of nutrients and organic matter dis-charged depends on the amount of feed used andfarm management practices.The nutrient content inthe feed must be optimal for the fish, and as muchas possible must be available to and taken up by thefish. The amount of feed supplied must be optimalin relation to appetite, so that feed loss is avoided.

Discharge of nutrients to the recipient waterbody will result in increased algal growth leading to eutrophication and imbalance in the recipientecosystem. Discharge of too much organic matterto the recipient water body may result in lack ofoxygen during the night as a result of decomposi-tion. Local accumulation of fish faeces in the recipient water body may cause anaerobic decom-position, possibly accompanied by release of hydro-gen sulphide (H2S) which is toxic for small animalsand fish (Fig. 3.2). This shows the importance ofhaving an adequate water current at the pointwhere the discharge is released.

The larger concentration of biological material ina restricted volume compared to the case in naturalconditions means that possibilities for disease out-breaks in a fish farm are greater than in the wild.The effluent water may therefore also contain a higher concentration of pathogenic micro-organisms such as parasites, bacteria, viruses andfungi, that can cause disease in the fish population.This may again have significant consequences forthe wild fish in the recipient water body. If there arepossibilities for ‘short circuiting’, the fish farm mayfunction as a facility for increasing the concentra-tion of pathogens (Fig. 3.3). For land-based fishfarms this can be illustrated with pathogens that areejected in the effluent water from the farm and are

taken in again with the inlet water. This also showsthe importance of not having these pipes too closeto each other.The best arrangement is to have themin different water bodies, or to treat either the inletor the outlet water. Fish that migrate can also be ahost for transporting micro-organisms up rivers,meaning that short circuiting also can occur here. Migration obstructions in the river may be asolution.

Figure 3.2 Local anaerobic decompo-sition may occur in the recipient waterbody if the point discharge is excessive.

Figure 3.3 Short circuiting between the outlet and inletwater to the fish farm must be avoided to prevent theconcentration of pathogenic micro-organisms increasingin the recipient water body.

Moving aquatic organisms from one farm toanother may, via the water used for transport, bringnew micro-organisms to the recipient farm. Theresult can be an outbreak of disease on the fish inthe recipient farm. Some wild stocks may, throughnatural evolution, have developed natural immu-nity against some pathogenic micro-organismswhile others may represent a large threat. Anexample of this is the salmon parasite Gyrodacty-lus salaris where the Atlantic salmon in a few rivershave developed immunity, whereas in most riversthe stocks have no immunity. When moving fishcontaining such micro-organisms between rivers, orbetween farms with outlets to different recipients,the consequences can therefore be fatal. Treatmentof the effluent water is absolutely necessary in suchcases.

Escape of fish or other aquatic organisms fromfarm conditions may also present environmentalproblems, except where local stock that not hasgone through to a breeding programme is used onthe farm. If the stocks have gone to a centralizedbreeding programme where mixing of local stocksfrom different districts occurs, or much more seriously, if genetic manipulations have been per-formed, there may be significant consequencesattending escapes. The fish farming industry hasbenefitted greatly from national breeding pro-grammes; for example, growth rates and feed uti-lization have been improved.14

What can happen when several fish escape froma farm? One possibility is that they can establish

their own stock in the recipient water body. Thismay result in competition for feed and habitat withthe naturally occurring stocks. Another possibilityis that they can interbreed with the local stocks andcreate unwanted hybrids. Even if no interbreedingoccurs the escaped fish can destroy the breedinggrounds or breeding nests of the wild stocks. Ofcourse, none of this is wanted and therefore is itvery important to try to avoid escape of fish andother aquatic organisms from farm conditions.

3.4 Water treatmentAll treatment of water leads to a change in thewater quality, and it is improvement that is wanted.Regardless of the incoming water quality, it willalways be possible to obtain a quality good enoughfor growing aquaculture products. The problem is,however, the cost; all water treatment operationsinvolve expenditure. A major advantage for a goodfarming site is therefore to have good qualityincoming water with low treatment requirements.

Several processes may be needed to adjust thewater quality. The inlet water to land-based farmsis aerated; pH adjustment and particle removal arealso required. Heating and chilling are normallyused to create optimal growth conditions. In somecases disinfection is needed to reduce the burden of micro-organisms, especially in fry production(Fig. 3.4).

The outlet water from the fish farm may also betreated to avoid affecting the water quality of the

Water Quality and Water Treatment: an Introduction 35

Figure 3.4 Water treatment processes.


recipient water body. If the recipient water body ishighly eutrophicated the outlet water must betreated, and this is also the case if there are valuablewild stocks in the receiving water. Treatment ofoutlet water will, however, increase productioncosts, and is done only when necessary. Often government regulations will dictate the level oftreatment required. Sites that require less watertreatment are therefore favoured.

Treatment of the outlet water is normallyrestricted to removal of part of the suspendedsolids. From a cost perspective, is it normally impos-sible to remove dissolved substances such as nutri-ents and small micro-organisms in a flow-throughfarm with relatively high fish densities, because ofthe size of the water flow; if re-use systems areemployed, such treatments may be included.However, re-use systems have much higher invest-ment costs than flow-through farms.

The treatment requirements of a completesystem are described in Chapters 4–10: the follow-ing six chapters will focus on the methods mostcommonly used in aquaculture for water treatment,concluding with a chapter on water re-use systems.

References1. Midlen,A., Redding,T.A. (1998) Environmental man-

agement for aquaculture. Chapman & Hall.2. Black, K.D. (2001) Environmental impacts of aqua-

culture. Sheffield Academic Press.

3. Read, P., Fernandes, T. (2003) Management of envi-ronmental impacts of marine aquaculture in Europe.Aquaculture, 226: 139–163.

4. Boyd, C.E., McNevin, A.A., Clay, J., Johnson, H.M.(2005) Certification issues for some common aqua-culture species. Reviews of Fisheries Science, 13:231–279.

5. Pillay, T.V.R. (2005) Aquaculture and the environ-ment. Blackwell Publishing.

6. Ervik, A., Hansen, P.A., Aure, J., Stigebrandt, A.,Johannessen, P., Jahnsen, T. (1997) Regulating thelocal environmental impact of intensive marine fishfarming 1. The concept of the MOM system (Model-ling–Ongrowing fish farms–Monitoring). Aquacul-ture, 158: 85–94.

7. NS 9410. (2000) Environmental monitoring of marine fish farms. Norwegian Standardization Asso-ciation.

8. Alabaster, J.S., Lloyd, R. (1980) Water quality criteriafor freshwater fish. Butterworth.

9. Poxton, M. (2003) Water quality. In: Aquaculture,farming aquatic animals and plants (eds J.S., Lucas,P.C. Southgate). Fishing News Books, Blackwell Publishing.

10. Bergheim, A. (1999) Redusert vannforbruk ogpåvirkning av vannkvaliteten ved settefiskanlegg.Kurs i vannkvalitet for settefiskprodusenter. ArrangørHydrogass (in Norwegian).

11. Snoeyink, V.L., Jenkins, D. (1980) Water chemistry.John Wiley & Sons.

12. Stum, W., Morganm, J.J. (1996). Aquatic chemistry.John Wiley & Sons.

13. Benjamin, M.M. (2001) Water chemistry. McGraw-Hill Science.

14. Gjedrem, T. (2005) Selection and breeding programsin aquaculture. Springer-Verlag.

4Adjustment of pH

the alkalinity in aquaculture systems, together withhydroxides (OH−). In the carbonate system com-pounds are related to each other via different equilibria:

CO2 + H2O ↔ H2CO3

H2CO3 ↔ HCO3− + H+

HCO3− ↔ CO3

2− + H+

where:

H2CO3 = carbonic acid; HCO3− = bicarbonate ion;

CO32− = carbonate ion; CO2 = carbon dioxide.

Water with free bicarbonate will take up H+ ions.The amount of each ion in the water is pH-related:in water of low pH there is excess carbondioxide/carbonic acid; in water of pH 7 there isexcess bicarbonate, and in water of high pH excesscarbonate ion. The units for alkalinity are eithermg/l as CaCO3 or milliequivalents per litre (meq/l)where 1meq/l is equal to 50mg/l CaCO3.

A buffer is defined as a substance capable of neutralizing both acids and bases in a solution and thereby maintaining the original acidity oralkalinity of the solution and the resistance to pHchanges when adding moderate amounts of base oracids.

Hardness is sometimes confused with alkalinity,mainly because it can be expressed using the sameunits. Hardness is, however, a term for the sum ofall metal ions in the water. This is dominated by thebivalent cations of calcium (Ca2+) and magnesium(Mg2+), but manganese (Mn2+), iron (Fe2+), sodium(Na+) and potassium (K+) may also be important.Because calcium and magnesium are included, hardwater may also have high alkalinity, but this is notnecessarily always so. If, for instance, sodium and

4.1 IntroductionOn some sites the freshwater pH is too low toachieve optimal growth for fish or shellfish. Atother sites the buffering capacity of the water is low,and it is difficult to avoid pH fluctuation in thewater. This again results in negative effects ongrowth. Sites where acid rain is a problem are par-ticularly exposed to this. Further, in re-use systems(Chapter 10) biological filters are used to removeammonia and this causes a drop in pH that must becorrected to maintain optimal growing conditions.

4.2 DefinitionspH is the measure of acidity or alkalinity in a solu-tion. It is presented on a number scale between 1and 14, where 7 represents neutrality and the lowernumbers indicate increasing acidity and highernumbers increasing alkalinity. Each unit of changerepresents a tenfold change in acidity or alkalinity.What is measured is the negative logarithm of theeffective hydrogen-ion concentration or hydrogen-ion activity in gram equivalents per litre of the solution.

pH = −log[H+]

If substances are added to the water they may act as acids, bases or be neutral. Acids give freehydrogen ions (H+) and bases free (hydroxyl ions)(OH−).

The alkalinity of the water is a measure of itscapacity to neutralize acids, meaning its ability tokeep the pH constant. If the alkalinity of the wateris low fluctuation in pH occurs easily. The carbon-ate system normally represents the major part of

37


potassium are responsible for the alkalinity, thehardness can be low. The following water classifica-tion system can be used: soft water, hardness lessthan 50mg CaCO3/l; moderately hard water 50–150mg CaCO3/l; hard water 150–300mg CaCO3/l;very hard water, above 300mg CaCO3/l.1 Conduc-tivity may be used as a unit to determine hardness,measured as microsiemens per cm (μS/cm).

4.3 Problems with low pHpH that is too high or too low will have negativeeffects on the fish.2–5 Low pH can cause damage tothe gills, skin and eyes. Higher H+ concentrationswill also increase the permeability of the gills,leading to leakage of Na+ and Cl− which createsosmotic problems. First, the effects can be regis-tered as a reduction in growth; too low a pH willkill the fish. In natural populations, the pH may varyfrom 5 to 9, but for aquaculture facilities it is rec-ommended to be in the range 6.5–9.2,3 Problemswith metals in the water are best avoided; tolerancemay vary with fish species and life stage, with newlyhatched fry being especially sensitive. For crayfish,for example, the pH and alkalinity must be highbecause they utilize Ca2+ in the water for shell synthesis.

The solubility of metal ions in the water willincrease with reduction in pH. There have been particular problems with the concentrations of alu-minium (Al+++) in fish farming; this metal leachesfrom the soil or bedrock in the catchment area. The toxicity of the complexes of Al or Al precipitatesvaries. A drop in pH will change the existing Alcomplexes to more toxic ones, meaning that fatali-ties can occur even if the pH itself does not repre-sent any danger. The most stable and non-toxicforms of Al are in the pH range 6.5–6.8. Calciumwill ameliorate problems with aluminium becauseit protects the gills from aluminium and also fromacidity.6 Toxic effects also depend on temperature,because rate of reaction increases with rise in tem-perature. Normally is it therefore not the pH thatis dangerous, but the combination of low pH andmetal ions.

Some of the aluminium or other metal complexesthat are toxic are unstable and will only persist fora short period. This may occur when mixing waterwith different qualities and characteristics, and amixed zone of water with different qualities is

achieved. If this reaches the fish tanks, fatalities mayresult. It is really important to be aware of this inintensive fish farming, where water from sources ofdifferent quality and temperature is mixed justbefore entering the fish tank. For instance, a briefdrop in pH may occur in a water source due to fallof acid rain or ice melting in the catchment area;this may create a mixed zone in the water source(river or lake).7 Where single Al complexes coagu-late and create larger, more toxic Al complexes.

In mixing zones is aluminium entering the gillscauses osmotic stress.6 When it is taken up, it cancause damage to the nervous system and blockenzymatic reactions. As a defence mechanism,mucus may be secreted and oxygen uptake isthereby reduced, which may result in death. If thereis a possibility of problems with metals ions, choiceof the correct adjustment agent is important.

Water alkalinity for intensive fish farming andpond aquaculture is recommended to be above 40mg/l CaCO3 to stabilize the pH and protect healthand physical quality.3 Alkalinities in the range100–200mg/l CaCO3 will, however, give severaladditional advantages, including making a stablewater source for biofilters in a re-use circuit, addingbuffer capacity to avoid pH fluctuation in ponds,and reducing the toxicity of heavy metals.8 In pondsthere will be a fluctuation in pH during the day andnight due to biological processes. During the nightthe algae (phytoplankton) and fish will release CO2

so the pH will drop if the alkalinity is low; duringthe daytime the algae will consume CO2 by photo-synthesis faster than the fish release it; therefore thepH will increase. The pH can vary from 5 to 10.9

4.4 pH of different water sourcesThe pH will of course vary with the water source.Seawater normally has pH value between 7.8 and8.3 and has a good buffering capacity due to theavailable free bicarbonate. There is normally noneed for pH adjustment. The only exception beingwhen re-using water to a great extent.

For fresh surface water, whether river or lake, thepH will be highly dependent on the ground char-acteristics and whether the catchment area isexposed to acid rain. Normal pH values arebetween 4 and 8.5. Groundwater has a more stablepH, but the buffering capacity can be reduced andvalues between 5.5 and 8.5 are common. If there is

limestone rock in the catchment area the pH willbe high, while the pH of water coming from marshyareas can be low. The alkalinity of the differentsources can vary from below 10mg/l CaCO3 in softfreshwater to several hundred of mg/l CaCO3 inseawater and hard freshwater.10

It is important to be aware that there can be fluc-tuations in the pH of freshwater during the year.Floods during spring may result in dramatic reduc-tions in pH, especially when snow is melting in the catchment area where acid has accumulated in the snow. Also dry weather may affect the pH ofthe water.

4.5 pH adjustmentThe need for pH adjustment or neutralization ofthe water depends on the water source, the speciesfarmed and the production system (e.g. water re-use system).

The principle used to adjust pH in acid water isto remove the free H+ ions. Methods must thereforeattract and bind the H+ ions and will thereforeinvolve basic solutions. The pH is usually regulatedto between 6.5 and 7 either by adding hydroxides(OH−) or carbonate compounds. Examples of thehydroxide group include lye, sodium hydroxide(NaOH), calcium hydroxide or slaked lime(Ca(OH)2) and magnesium hydroxide (Mg(OH)2).Carbonate compounds include different forms oflime (calcium carbonate (CaCO3) and quick lime(CaO)), in addition to dolomite (CaMg(CO3)2,magnesium carbonate (MgCO3), sodium carbonate(Na2CO3) and sodium bicarbonate (Na2HCO3).Where there are problems with aluminium silica lyehas also been used with advantage to reduce theacute toxicity.11,12 Silica (SiO2) will attract labile alu-minium and prevent the occurrence of long Alchains and creation of mixed zones.

If carbonate compounds are used there will, inaddition to the increase in pH, be an increase in thebuffering capacity of the water, i.e. an increase inthe alkalinity of the water, which will then be morestable against pH drops. If adding lye complexesthis effect will be minor but the pH will increase; itcan, however, be simple to overdose with lye so thepH gets too high.

The pH must be adjusted before the waterreaches the fish tanks. Normally there will be a needfor some retention time so that the adjustment

agents can function and to prevent unstable toxicmetal complexes reaching the fish tank (mixed zoneproblems). Treatment at the start of the fish farminlet pipe or using a holding tank to retain the waterfor some time after adding the pH treatment cansolve this agent problem. The solubility and rate ofsolution will also depend on the chemical chosen;for example, sodium bicarbonate will react quickly,while dolomite reacts more slowly.

It is also important to achieve good mixing of thepH treatment agent with the incoming water; somekind of mixing equipment is quite normal and can,for instance, be a mixer, use of air bubbles or byaddition of the chemical before pumping. The formof the pH agent (liquid, meal, powder or largergrain or rock) will dictate the requirements for themixing equipment.

In a fish tank there will be an increase in the CO2

concentration resulting from the fish metabolism.If the fish density is high compared to the watersupply and pure oxygen is supplied in addition, thereduction in pH may be noticeable. When addingpH regulating agents to the water, care must betaken because the water will also contain ammoniathat is toxic for the fish, the amount of which is pH dependent (see Chapter 9); decreasing the pH will reduce the amount of dissolved ammonia.If the pH is adjusted without doing anything toreduce the concentration of ammonia, fatalitiesmay result.

4.6 Examples of methods for pH adjustment

4.6.1 Lime

Lime of various forms is a good substance forincreasing the pH of acid water. When adding lime(calcium carbonate (CaCO3)) to water with low pHthe following process will take place, as mentionedearlier:

CaCO3 ⇒ Ca2+ + CO32− (calcium carbonate is

dissolved in water)CO3

2− + H+ ⇔ HCO3− (carbonate ion attracts

H+ from the water)HCO3

− + H+ ⇔ H2CO3 (bicarbonate attracts H+ from the water)

These reactions are part of the carbonate system,which is the most important contributor to the good

Adjustment of pH 39


buffering capacity of the water. What can be seenfrom the chemical equations is that when addingCaCO3 to water, CO3

2− ions will react with free H+

and reduce the amount, so increasing the pH.Lime may be used in different forms, such as

limestone or as limestone powder. Limestone

powder (CaCO3), with a particle size of less than0.005mm, may also be dissolved in water and createlime slurry (approximately 75% dry matter), whichis much used for pH regulation. Lime slurry may beadded in concentrated form to the water or asdiluted (5–10 times in a mixing basin) lime slurry.

A

BFigure 4.1 (A) A lime slurry plant for regulation of pH ina fish farm; (B) lime slurry tank.

Adjustment of pH 41

The container for storing the lime slurry must havean efficient mixing system to avoid settling out ofthe lime particles. Lime slurry is added with adosage pump to a mixing basin where mixing of theslurry with the water where pH is to be regulatedtakes place (Fig. 4.1). Automatic pH control can beachieved by having a pH sensor in the mixing basin,the signals from which can be used to control thedosage pump and the amount of slurry added. Inthis way any variation in the quality of the incom-ing water can be registered and adjusted before thewater reaches the fish tanks.

To ensure good mixing of the slurry with thewater, a diffuser creating air bubbles can be used inthe mixing basin. Use of lime slurry may increasethe particles content (turbidity) of the water. Thismust be considered when using lime slurry on waterfor fry production.

A shell-sand filter or limestone filter representsa simpler system where the water has to passthrough a layer of crushed limestone (particle size1–3mm) or shell sand.13 As the water passesthrough the filter there will be an increase in pH.After using the filter for a period of time there willbe a gradual drop in pH, because the amount ofCO3

2− is reduced.To avoid these drops in pH severalfilters can be connected in parallel. This makes itpossible to refill, clean or maintain a separate filterwithout stopping the whole system. When usingsuch systems, a part flow of the water to be treatedis sent over the limestone filter. The pH increases,and the water is sent back to the main water flowinto the fish tanks. Care must, however, be taken toavoid possible mixing zones. If CO2 gas is added justbefore the limestone filter, the CO2 will increase thedissolution of the limestone. In this way, it is possi-ble to automatically control the pH by controllingthe addition of CO2 gas before the filter. The reac-tion process in this filter can be described with thefollowing equations:

CaCO3 + H2O + CO2 ⇒ Ca2+ + 2HCO3−

HCO3− + H+ ⇔ H2CO3

A shell-sand well is a very simple and low costsystem for regulating the pH (Fig. 4.2). The incom-ing water is forced to flow up through a layer ofshell-sand and in so doing the pH of the water willincrease.This method requires manual refilling withshell sand to keep the pH stable.

Figure 4.2 A shell-sand well for regulation of pH.

4.6.2 Seawater

Seawater has a high buffering capacity and containsfree carbonate ions (CO3

2−) and/or bicarbonate ions(HCO3

−) which, similarly to limestone, will take upH+ ions and increase the pH. Addition of 2–4% seawater to freshwater will increase the pH and thebuffering capacity; the conductivity of the waterwill also increase. Since measuring the conductivityis quite simple, this method can be used for con-trolling the addition of seawater. Also, the amountof seawater may be fixed manually depending onhow much freshwater is used.

When using this method for pH regulation, it isimportant to consider that the seawater containspathogenic micro-organisms, poisonous algae andother substances that may be harmful to the fish. Asolution is to pump the seawater from large depths,where the amount of algae and micro-organisms islower. Seawater pumped from groundwater wellsmay also be used. This water normally has a verylow content of micro-organisms. It is generally recommended that seawater is disinfected beforeusing it for neutralizing freshwater. The operatormust also be aware of mixed zone problems.

4.6.3 Lye or hydroxides

Different types of lye, such as sodium hydroxide(NaOH), may be used for pH regulation, but specialcare must be taken before use in fish farmingbecause it is quite easy to overdose, especially whenthe water quality varies. The result can be waterhaving a pH that is too high, which is also toxic forthe fish. Sodium hydroxide is strongly corrosive on


metals and is not particularly friendly to work with.In solid form it is a white crystalloid non-odoroussubstance that is easily dissolved in water; themixing process produces heat and steam.

Plants for adding of dilute sodium hydroxidesolution (3%) to the inlet water work on the sameprinciple as slurry plants (Fig. 4.3). A dosage pumpis necessary for adding the sodium hydroxide solution into the mixing tank. The mixing tank for sodium hydroxide and water may be slightlysmaller than for lime slurry because it is easier tomix.

The reaction process when using sodium hydrox-ide is:

NaOH ⇒ Na+ + OH−

OH− + H+ ⇔ H2O

In water of low pH, the hydroxyl ions (OH−) willreact with the free hydrogen ions (H+) and increasethe pH of the water. The use of sodium hydroxidewill, as seen from the chemical equation, only neu-tralize the water and not increase the bufferingcapacity; this is also a reason for preferring othermethods. In water of low alkalinity either lye orhydroxides may be used.

References1. Hammer, M.J. (1996) Water and wastewater technol-

ogy. Prentice Hall.2. Randall, D. (1991) The impact of variations in water

pH on fish. In: Aquaculture and water quality (edsD.E. Brune, J.R. Thomasso). World AquacultureSociety, Louisiana State University.

3. Wedemeyer, G.A. (1996) Physiology of fish in inten-sive fish culture systems. Chapman & Hall.

4. Willougby, S. (1999). Manual of salmonid farming.Fishing News Books, Blackwell Science.

5. Poxton, M. (2003) Water quality. In: Aquaculture,farming aquatic animals and plants (eds J.S. Lucas,P.C. Southgate). Fishing News Books, Blackwell Publishing.

6. Rosseland, B.O. (1999) Vannkvalitetens betydning forfiskehelse. In: Fiskehelse og fiskesykdommer (ed. T.Poppe). Universitetsforlaget (in Norwegian).

7. Krogelund, F., Teien, H-C., Rosseland, B.O., Salbu, B.(2001) Time and pH-dependent detoxification of alu-minium in mixing zones between acid and non-acidrivers. Water, Air and Soil Pollution. 130: 905–910.

8. Wedemeyer, G.A. (2000) pH. In: Encyclopedia ofaquaculture (ed. R.R. Stickney). John Wiley & Sons.

9. Boyd, C.E., Tucker, C.S. (1998) Pond aquaculturewater quality management. Kluwer Academic Publishers.

A

B

Figure 4.3 Adjustment of pH using liquid sodiumhydroxide (lye): (A) the tank and dosing pump for addinglye to the inlet water flow (note the small plastic tubegoing up to the pipe); (B) equipment for monitoring thepH of the water flow and for regulating the addition oflye.

10. Wedemeyer, G.A. (2000) Alkalinity. In Encyclopediaof aquaculture (ed. R.R. Stickney). John Wiley &Sons.

11. Åtland, Å., Hektoen, H., Håvardstun, J., Kroglund, F.,Lydersen, F., Rosseland, B.O. (1997) Forsøk meddosering av silikatlut ved syrtveit fiskeanlegg. Nivarapport 3625 (in Norwegian).

12. Camilleri, C., Markich, S.J., Noller, B.N., Turley, C.J.,Parker, G., van Dam, R.A. (2003) Silica reduces

the toxicity of aluminium to a tropical freshwater fish (Mogurda mogurda). Chemosphere, 50: 355–364.

13. Lekang, O.I. Stevik, A.M., Bomo, A.M., Stevik, T.K.,Herland, H. (1999) Bruk av skjellsand til regulering av pH og alkalitet i småskala resirkuleringsanlegg for fiskeoppdrett. ITF rapport 102, Universitet forMiljø og Biovitenskap (in Norwegian).

Adjustment of pH 43

5Removal of Particles

be treated as gently as possible to avoid breakingthe particles and reducing their size, so increasingthe size of the filter necessary for extraction.2

Gentle handling of the particles includes using lowwater velocity and having few bends, valves, etc., inthe system that create extra turbulence where theparticles are flowing. For the same reason, the filterought to be placed as close to the source as possi-ble; for inlet water this means as close to the watersource and for outlet water as close to the produc-tion unit as possible. It is also important to have asufficient flow to prevent particles settling in thepipes, and leakage of nutrients.

The particles in the water are of various differ-ent forms and numbers. Several methods and defi-nitions are used to define the particle content of thewater. Total suspended solids (TSS) is defined asthe amount of particles stopped by a special fibre-glass filter with a pore size of 0.45μm. Total solids(TS) represents the total amount of particles in thewater; this quantity can also be expressed as totaldry matter (DM).

Particles can also be classified according to size.2

Those smaller than 0.001μm are classified assoluble, 0.001–1μm as colloidal, 1–100μm as super-colloidal and larger than 100μm as settleable. Somenutrients may be totally dissolved in the water,which means that they cannot be re-moved with aparticle filter but with other filter types. Exampleshere are biofilters (Chapter 9) that remove dis-solved substances, such as NH4

+ or NO3−.

When removing particles from the water flowthere will also be a reduction in the discharge ofnutrients because some are included in the parti-cles. There will also be a reduction in the number

5.1 Introduction

Removal of particles from a water flow is calledwater treatment or water purification. In aquacul-ture, removal of particles from a water flow is nec-essary in several places: for the inlet water to thefarm; for the outlet water from the farm; or if thewater is re-used. The inlet water is treated to avoidtoo high a concentration of particles reaching thefish. High concentrations will have a negative effecton growth and may increase mortality.1,2 Some par-asites in the water are also of a size that makes itpossible to remove them with a particle filter.3 Theycan therefore be removed from the water before itreaches the farm, or if used on the outlet water afilter could remove them from the water flowbefore it reaches the recipient water body. Anotherreason for removal of particles from the inlet wateris that the function of other water treatment equip-ment can be affected negatively by the particlecontent (see Chapter 6).An example here is the dis-infection plant where a low particle content isrequired. In the outlet water, particle removal isdone to reduce the effect of the outlet water onwater conditions in the recipient body.4,5 For re-usesystems particle removal is particularly importantto avoid accumulation of particles in the system andreduction in fish growth.

The aim of using a filter to remove particles is toextract a certain proportion of particles from thewater flow, not all. How much is removed dependson the design and function of the filter. The biggestparticles are the easiest to remove, regardless of thechosen method. Before the water flow reaches thefilter unit where the particles are removed, it must

44

of micro-organisms because some are attached tothe surface of the particles.

To give some ideas of the different particle sizes,the following can be used for illustration: cocoa andtalcum powder 5–10μm, hair straw 50–70μm, tablesalt 90–110μm. The lower limit for easily identify-ing single particles is around 40μm.

5.2 Characterization of the waterGood characterization of the water to be filtered isnecessary so that the correct filter can be chosen.The characteristics of the inlet water will vary fromsite to site, whether it is lake water, river water,groundwater or seawater. Before choosing a filterit is therefore necessary to take samples to be ableto characterize the water.

The volume of wastewater coming from fishfarms is normally much higher and the concentra-tion of the discharged substances much lower thanthose entering a municipal wastewater treatmentplant; they are, however, and comparable to thosein the water discharged from municipal wastewatertreatment plants, i.e. water that has been purified.6,7

Requirements for the design and construction ofwastewater plants for fish farming are therefore dif-ferent to those used to treat muni-cipal wastewater.Hence, the purification equipment and technologyused in municipal wastewater treatment cannot betransferred directly to fish farming conditions, evenif the basic principles are the same.

The composition of the outlet water from a fishfarm depends upon a number of factors, includingspecies, growth rate, feed composition and utiliza-tion, feed conversion rate and water amount (seefor example, ref. 8). The first step in reducing thedischarge from the fish farm, without using anyfilter at all, is therefore to have an optimal feed thatis fully utilized and consumed by the aquatic organ-isms. This also includes optimal management of thefarm, having correct water quality and quantity, andfeeding in an optimal way.4

Experiments have shown that the predominantparticle size in the outlet water from fish farming isless than 30–40μm.9,10 The large number of smallparticles account for only a limited part of the totalvolume of discharged particles. Since the volume ofparticles is much more important than the numberof particles when talking about the load on therecipient water body, it is of great importance to

remove the few large particles. However, in re-usesystems the small particles will normally dominate,since it is easy to remove the larger particles. Thiscan also been seen with water that goes brown inhigh re-use systems, because the small particlesremain in the water.

The density of faeces from fish farming varies.Reported densities are above 1, from 1.005 to 1.2,which means that the faeces will settle in water.9,11

A study of intact faeces from rainbow trout showedan average sinking velocity of the faeces of 1–2.5m/min depending on fish size.12

5.3 Methods for particle removal infish farmingSeveral principle and methods are used to removeparticles from a water flow.13–18 These can be classi-fied as follows:

• Mechanical filtration, also called straining ormicro screens

• Depth filtration, also called sand filtration or justfiltration

• Settling

It is also possible to use other methods to removeparticles, such as flotation, membrane filtration andozonation. These methods are normally utilized forremoving smaller particles and are usually tooexpensive to use in aquaculture facilities, perhapswith the exception of water supplied to small fry,such as for some marine species.

Regardless of the method chosen, method it isimportant to remember that all filter systems willcause a head loss. This can be quite high, forinstance when using pressurized filters on the inletwater, with the loss depending on the principleemployed. Acceptable head loss is therefore animportant criterion when selecting an appropriatefilter.

5.3.1 Mechanical filters and micro screens

A mechanical filter is an obstruction that is set intothe water flow to collect the particles and largerobjects and allow the water to pass through. Theprinciple of a mechanical filter is to separate parti-cles from water in a straining plane, either a screenor a bar rack. Particles bigger than the aperture in

Removal of Particles 45


the screen or the distance between the bars in therack will be stopped. The simplest type of mechan-ical filters comprise a static screen, a grating or per-forated plate, or a bar rack that is set down into thewater flow. The screen, which has apertures ormeshes of defined size, will stop particles largerthan the aperture/mesh size moving with the waterflow; they are caught on the surface of the screen(Fig. 5.1). After a while the screen will graduallybecome blocked; the head loss will increase untilthe screen is completely blocked with particleswhich prevent any water passing through it. Thisresults in an overflow. The same will be the casewith a bar rack. Typically bar racks are used toremove larger particles and objects (>15mm), whilescreens can also be used on smaller objects (>6μm).16

When a screen is used, the particles have to beremoved from the surface to avoid blockage. Oneway to remove them is to manually take the screenup from the water and clean it. This method is verylabour intensive and is only used in special caseswhere the pore size is very large compared to themajor particle size in the water. Examples areremoval leaves from the water in the autumn, andstopping other large floating objects from enteringthe inlet pipe. A major aim when constructing amechanical filter or screen to be used for removingsmaller particles, is therefore to find ways of pre-venting blockage, which means being self-cleaning.The bar rack can be made self-cleaning by using ascraper mechanism, but this, as mentioned, is adevice for removing larger particles.

It is important that the screen surface is smooth so that the particles are not crushed. Thescreen can be made of perforated plates when theapertures are large (mm or cm). When a screen is

used to filter inlet water or outlet water in fishfarming, a screen cloth of metal or plastic threadswoven to the wanted mesh size is employed. It isimportant that the screen surface is easy to keepclean.

Several methods can be used to make the screenself-cleaning. A basic separation can be done byback-flushing, vacuuming or mechanical vibrationof the filter cloth. If mechanical vibration is used,the filter cloth will shake and the trapped particleswill fall off by gravity. If such equipment is used, thefilter cloth must be installed at an angle to the hor-izontal plane; this method is, however, not com-monly employed in aquaculture. The simplestmethod of cleaning the filter is to back-flush thescreen (Fig. 5.2).

When using back-flushing or vacuuming it isdesirable to have as great an area of the newcleaned screen as possible in the water. The screenis used until it gets blocked, when it is removed forcleaning and new cleaned screen cloth is substi-tuted. This process must not be so rapid that theefficiency is reduced or mechanical breakdownoccurs. One common set-up used in aquaculture isa rotary screen that rotates partly above and partlybelow the water surface. The meshes in the screenare cleaned by back-flushing either with air orwater when the screen is above the water surfaceand where the back-flushing water that contains theparticles removed from the screen can be collected.Straining or micro screening has been shown to bethe most effective cleaning method per unit surfacearea; in aquaculture it is especially effective forremoval of relatively big particles, with the headloss for the water flow through the screen also beingquite low.

Rotary screen construction can be classified onthe basis of how the screen rotates; various systemsare available, of which the most common are (Fig. 5.3):

• Axial rotating screen• Radial rotating screen (drum)• Rotating belt• Horizontally rotating disc.

An axial rotating screen is placed vertically and stands normal to the direction of water flow. One common type based on this principle isthe disc filter. Which can comprise one or severalvertically installed filter plates, with a gradualFigure 5.1 A static screen.

reduction in the mesh size of the screens. Thismeans that the largest particles are removed on the first screen and smaller ones on subsequentscreens.

In a radial rotating screen, the flow is radialtowards or away from the axis of rotation. A rotat-ing drum is a typical filter using this construction.Water to be purified flows into the drum, whichcomprises a straining cloth of appropriate mesh sizefixed on a frame. The water has to pass through thedrum, which means that it must go out normal(radial) to the main flow direction.The particles willbe trapped in the straining cloth when flowingthrough the drum.

Rotary screens, whether axial or radial, are con-structed so that the screen operates when only par-tially submerged in the water flow that is to befiltered. Back-flushing of the screen, which is usedto clear the mesh and remove trapped particles, isdone when the screen is rotating above the watersurface. High-pressure water from nozzles isdirected on to the screen so the particles are dis-lodged in the same direction that they entered thescreen. The back-flush water containing the parti-cles is collected and represents the sludge waterfrom the filter in which the concentration of parti-cles is high. Back-flushing of the straining cloth canbe continuous or step-wise: the latter method is


Figure 5.2 A system where the static screen is back-flushed to avoid blockage. The slide shows a typicalstatic screen while the sketch shows one principle of back-flushing.


most usually used, but choice depends on the loadon the filter and the mesh size. The water level infront of the screen can be used to control the rota-tion and back-flushing. When the screen is clogged,the head loss through the straining cloth willincrease together with the water level in front of thestraining cloth. This increase in water level can beused to start the rotation of the screen and theback-flushing of the straining cloth. It is, however,important to remember that a rotary screen withback-flushing will produce significant quantities ofback-flush water. It is also possible to back-flushwith hot water to remove the layer of fat that canbe created on the screen surface; this is done onlyfrom time to time, not on every back-flush, becauseof the increased costs.

In a rotating belt filter the straining cloth takes theform of a belt stretched out by rollers at both ends.One of the rollers is motorized and causes the beltto rotate partly above and partly below the watersurface, so back-flushing and removal of particles ispossible on the part of the belt that is above thewater surface. Instead of using water for back-flush-ing, air may be used.Air at high pressure is blown onthe straining cloth from nozzles and dislodges theparticles from the mesh in the opposite directiontogether with some water; this device is known as anair knife. When using air instead of water, the parti-cle concentration (TS) in the sludge water isincreased, but is a more cost effective method.

Another type of filter is the horizontally rotatingdisc standing above the water surface the water to

A

C

B

D

Figure 5.3 Different types of rotating filter with automatic back-flushing of the straining cloth. (A) Overhead view ofa disc filter, where axially rotating screens are vertically placed normal to the water flow direction; here two screenswith different mesh sizes are shown. (B) A rotating belt filter where the water passes through the belt while the par-ticles are transported to the surface. (C) Close-up of a drum filter with straining cloth. (D) Nozzles used for back-flushing the straining cloth.

be filtered comes in from above and tricklesthrough the straining cloth. Particles larger than themesh size of the cloth will be trapped on the clothsurface and must be removed, possibly by vacuumto avoid flooding. The concentration of particles inthe water will then be quite high, i.e. the water hashigh TS value.

The flow rate through the screen is determinedby the mesh size of the screen, head loss across thescreen, desired purification efficiency, amount andcharacteristics of particulate material in the inletwater, and the cross-sectional area of the screen.Mesh size selection is based on the maximum par-ticle size that can be allowed in the effluent. Headloss depends on the percentage hole area in thestraining cloth, amount and characteristics of par-ticulate material in the inlet water, efficiency ofback-flushing, screen rotation speed, and flow rate.Typically the volume of back-flushing water isabout 0.2–2% of the bulk flow.7

The choice of mesh size depends on the condi-tions and where the filter is to be used. In inletwater the mesh size may be as small as 20μmbecause some parasites will also be removed. Onoutlet water that is going to a recipient water body,a mesh size between 90 and 100μm is commonlyused. In re-use systems, mesh sizes down to 30μmare used. Reduction of mesh size will increase theneed for new screen cloth exponentially, especiallyif it is reduced below 60μm.19

5.3.2 Depth filtration – granular medium filters

Depth filtration, also called sand filtration, or justfiltration, means removal of particles when water isforced to flow through a layer of a material withparticles (granular filter medium) of various sizesand depths, this filtration layer can be of sand oranother granular material, depending on thepurpose for which the filter is to be used. Becausesand is a commonly used the filter is often called asand filter. The layer is not dense but contains a number of channels and holes created betweenthe particles that constitute the filter medium (Fig. 5.4).

When the water that contains particles goesthrough the filter medium, particles larger than acertain size will be trapped by several mecha-nisms.16 They may be too large to go through chan-nels (straining); they may settle both unaltered and

due to flocculation and adhesion, and may beadsorbed by chemical and physical forces thatattract them to the filter mass.

The first process will occur on the top layer of thefilter medium; particles are trapped because theyare too large to pass into the channels or pores inthe filter media, which will then become blocked. Itis, however, of some concern that a few large parti-cles block surface channels of the medium causinga large reduction in filter capacity. The head lossincreases and the filter becomes blocked morequickly. Therefore it is more important to utilizesettling, adsorption and other effects inside thefilter medium; the total depth of the filter mediumis then utilized, not just the top layer. However, thisrequires a low water velocity through the filtermedium, which again means that the hydraulic loadon the filter surface must be low. When the chan-nels inside the filter are full of particles the headloss will increase.

The maximum particle size that will pass througha depth filter is determined by the grain size of themedium; if the grain size is small, the filter will clogeasily. The flow rate through filter medium and therate of clogging depend not only on the grain sizeof the filter medium, but also on the characteristicsof the particulates in the water to be purified. Theperformance of a depth filter is dependent both on the type of filter and operating procedures, and also significantly on the characteristics of the filter material.


Figure 5.4 Filtration in a depth filter. The black spotsare trapped particles.


Depth filters can be classified depending on thedirection of the water flow through the filtermedium. In an up-flowing filter, the water flows upthrough the filter medium, while in a down-flowingfilter the water flows downwards; the latter type isthe most common. Good filtration can be achieved,but the filter medium becomes clogged after aperiod. In an up-flowing filter the same will occur;here there may also be a break-through of waste-water at one or several places in the filter mediumif there is excessive clogging, provided that thewater pressure is high enough. In this case almostall the water will go through the filter via thesezones and there will be virtually no purification.

A depth filter should be equipped with back-flushing facilities, or regular manual purification ofthe filter medium will be necessary. In a traditionalsand filtration system (on-site system, see later)used for purification of municipal wastewater, thiscan be the case. Such systems can be used formonths or even years without doing anythingbecause the wastewater is only discharged onto thesoil surface. Of course, this depends on thehydraulic load on the surface area and the concen-tration of particles in the water.

If a filter is to maintain capacity, back-flushing isnecessary. Even when this is carried out, thehydraulic capacity of this filter type is lower thanthat of, for instance, rotating microsieves. Whiledoing the back-flushing operation, the filter must bestopped; water is then sent the opposite waythrough the filter media, so removing particles thathave settled. This back-flushing water is sentdirectly to the farm outlet. Most of the runningproblems with depth filters are the result ofimproper back-flushing. To achieve proper purifica-tion of the filter medium it is important that theback-flushing water suspends trapped particles.Typical amounts of back-flushing water are in therange 1–5% of the bulk flow.7

Depth filters can be separated into those operat-ing at normal atmospheric pressure or those havingan over pressure inside the filter, i.e. pressurizedfilters (Fig. 5.5). In a pressurized filter the mediummay be put inside a sealed chamber, with the sameoverflow arrangements as for an unpressurizedfilter. When a pressurized filter starts to clog, thepressure will increase and the particles will bepushed further down in the filter medium; ulti-

Figure 5.5 A pressurized depth filter with back-flushing.

mately such a filter will become totally blocked, butit takes longer than with an unpressurized filter.Theadvantages with pressurized filters are that a largerpart of the filter medium height is utilized and theback-flushing interval is increased. To use a pres-surized filter it is necessary to pressurize the waterto be purified, normally up to 6–8 bar. Usually thewater passes through a pressure increase pump toachieve this, and consideration must be given toavoiding damage to the particles. If such anarrangement is used on the outlet water, the pumpwill break the particles up and the proportion ofsmall particles which are more difficult to removewill increase. Therefore pressurized filters are notrecommended for use on outlet water; however, ifthere is no alternative the outlet water must behandled as carefully as possible.

The normal filter medium is sand or gravel. Thesize of the sand or gravel particles in the filtermedium depends on the characteristics of the par-ticles in the water to be purified. By using small par-ticles in the medium the head losses and cloggingvelocity of the filter are increased whilst the capacity is decreased. In a one size sand filter mostof the filtration occurs in the first few centimetres.Instead of using one type of medium over the wholefilter depth, i.e. a single size medium, media ofseveral sizes (multimedia) may be used to increaseutilization of the filter. In such filters, provided thewater enters from the top of the filter and flowsdown, the largest media are on the top and thesmallest on the bottom. The largest particles arethen removed in the top layer and increasinglysmall particles are removed through the lowerlayers of the filter. This will ensure a more even col-lection of particles distributed over a larger part ofthe filter and not just in the top layer. To avoidmixing of the media during back-flushing, mediawith different relative density can be used. Thelargest media must have the lowest density and thesmallest the highest. When back-flushing frombelow this will ensure that the largest particles stayon the top. In aquaria, a two-media filter compris-ing a layer of crushed anthracite or ilmenite abovea layer of fine sand has been used.20 A third bottomlayer with even finer particles, such as crushedilmenite or garnet, could also be included. Regard-less of the filter media, it is of great importance thatthe back-flushing is done correctly so that all the

media are back-flushed, otherwise there will bezones that are still clogged and that will not be uti-lized when the filter is returned to operation.

Each medium can be characterized by its grainsize, uniformity of grain size, grain shape and rela-tive density, all factors that are important for filter-ing performance.Appropriate grain parameters canconsiderably reduce the resultant head losses.

Several classifications can be used on granularmedia filters, and these can also be classified withrespect to their hydraulic capacity giving (1) slowsand filters, (2) rapid sand filters, and (3) continu-ously back-flushed filters.21 Type 1 is unpressurizedand type 2 is pressurized; type 3 is constructed toenable continuous back-flushing without interrupt-ing the water flow as is necessary for back-flushingtypes 1 and 2. Typical reported loading rates are0.68 l/s/m2 or less for slow sand filters, up to 1.4 l/s/m2

for rapid sand filters and up to 5.4 l/s/m2 for contin-uously back-flushing sand filters.21

In slow sand filters 2–5mm grains are utilized;10

pressurized rapid single sand filters have grain sizesranging from 0.3 to 4mm.22 Reported removal ratesare for 0.3mm sand about 95% of particles downto 6μm, and for 0.5mm sand about 95% of parti-cles down to 15μm.22 A typical pressurized rapidsand filter has a capacity of 136kg sand, a surfacearea of 0.29m2, a maximum pressure of 3.5kg/cm2

and a design flow rate of 238 l/min.Depth filters can also be of the cartridge type.

The cartridge can be made of different materials,such as plastic, ceramic, spun fibre or resinboundfibre,22 often specific to the supplier. The cartridgehas a defined depth, and the waste is collectedeither inside or outside the cartridge, or has to passthrough the cartridge. The typical cartridge filter isused once and then replaced. Cartridge filters canbe used for removal of small particles and are avail-able for sizes below 1μm. They are used on smallwater flows with low particle loads; otherwise thecost of operation becomes too high.

Diatomite (DE) or pre-coated filters may also beclassified as depth filters.23 Diatomaceous earthfilters comprise a filter with a screen coated withDE (fuller’s earth), a granular material of fossildiatoms. A cake of DE is formed and put on thefilter screen that functions as a skeleton. The waterto be purified has to pass through this filter cakeand the particles are trapped. When the filter is



dirty it can be back-flushed, or a new filter cake canbe installed. DE filters can tolerate higher waterflows than cartridge filters, but are best suited tosmall water flows.

Filter bags are another simple type of filter foruse with small water flows. The filter bag is wovenand so forms a surface, the mesh size of whichdepends on the closeness of the weave. Particles aretrapped on the surface of the bag. When no morewater will pass through, the filter bag has to beremoved and replaced with a new or clean bag. Thebags can be cleaned in a washing machine, forinstance.

5.3.3 Settling or gravity filters

Settling basing

Settling is a simple method for removing particlesfrom the water.24 The principle utilized is that par-ticles have a higher relative density than water(1.005–1.2 compared to 1 for freshwater),11,12 sothey will sink. This phenomenon can easily beobserved when water containing suspended parti-cles is allowed to stand for a period. A natural sep-aration process will occur and the particles will sinkto the bottom. The difference in relative densitybetween the particles and the water controls thevelocity of the separation process.

For small particles (0.1–1mm) and unobstructedsettling, the sinking velocity of the particles is givenby Stoke’s law:

Vg dp

sp w= −( )r r

m

2

18

where:

Vs = sinking velocityrp = density of particlerw = density of waterg = acceleration due to gravitydp = diameter of particlem = dynamic viscosity of the water.

This also demonstrates why the particles with thehighest density are the simplest to remove.

The simplest way to utilize gravitational force forseparation is to send the effluent water through abasin with a large surface area where the watervelocity is reduced: separation will then occur in thebasin provided that the sinking velocity of the par-ticle created by the gravitational force does notexceed the horizontal velocity component createdby the water flow through the basin (Fig. 5.6), inwhich case the particle will flow out of the basin inthe water. The following equation can be set up toachieve settling:

Vs > Q/A

Where:

Vs = sinking velocity of the particle (m/h)A = surface area of the basin (m2)Q = water flow through the basin (m3/h)

The relation Q/A is called the surface load or over-flow rate for the settling basin. Because differencein density between water and faeces is small quitea long hydraulic residence time is required. In fishfarming the normal surface load is between 1 and

Figure 5.6 A settling basin.

5m/h (actually m3/m2/h).2 Experiences show that afavorable depth for the settling basin is 1m and thewidth : length ratio is in the range 1 :4 to 1 :8.

ExampleA water flow of 1000 l/min (60m3/h) is to be purifiedby the use of a settling basin. Determine the size anddesign of the basin.

Expect a surface load of 3m/h

Vs > Q/A

A > Q/Vs

> 60m3/h3m/h

> 20m2

This means that the surface area of the pond needsto be more than 20m2 to allow proper settling. If awidth : length ratio of 1 :3 is chosen for the basin, thismeans that the exterior measurements of the basinwill be about 2.6m × 7.7m. Two settling basins areneeded so that one can be running while purificationand sludge removal can be done in the other.

Other designs of settling basin take account ofhydraulic retention time (t) or mean fluid velocity(vm).

t = V/Q

vm = L/t = Q/Acs

where:

V = volume of the settling basinL = length of settling basinAcs = cross-sectional area of the basin.

Reported values for retention times are 15–40min,while recommended mean fluid velocities are in therange 1–4m/min.2

The flow pattern in a settling basin used in fishfarming is normally horizontal. However, it is alsopossible to use settling with a vertical flow pattern.Such filters will have a tower like design. Waterflows slowly upwards and because of the gravita-tional forces the particles will sink with a greatervelocity. Specially designed filters, such as thelamella separation filter are used to improve set-tling conditions, and also the investment require-ments. Sending the water over biofilm can improvethe settling and filtration efficiency. Small particles

will be attracted to and settle on the biofilm.25 Addi-tion of polymers or other chemicals before settlingcan be used to increase the particle size by floccu-lation, and this also removes smaller particles, pos-sibly in the settling basin.16

The great disadvantage with settling basins is thatthe settled particles remain lying in the water flow,so there is a possibility that nutrients will leak fromthe particles, especially phosphorus which is weaklyattached to the particles.2,4 Additionally, resuspen-sion of settled particles into the water may occur,even if particles are regularly removed from thebasin; for this reason there must be adequate depthin the settling basin. Continuous removal of settledparticles from the basin is impossible from a costperspective. The removal of settled particles must,however, be carried out regularly to optimize basinfunction. This can be done by various methods, forinstance by using a vacuum pump. It is importantthat mixing of sludge and water is kept to aminimum to avoid resuspension of particles withthe result that the nutrients go directly to the outlet.It is quite common to have two settling basins, sothat one can be used while the other is purified.

Compared to using a micro strainer, settlingbasins require a much greater area and this can bea disadvantage. However, if the surroundings andground conditions are suitable, settling basins aresimple and cheap to construct.

Results from testing shows reduced removalrates with small particles.26 It is quite difficult toremove particles smaller than 100μm using a set-tling basin.27 Inlet values of less than 10mg TSS perlitre are difficult to treat, and those below 6mg TSSper litre are also difficult to obtain.27 To processvalues in this range special methods must be used,such as addition of polymers or use of biofilm toattract particles.

Swirl separators, hydrocyclones

In a swirl separator or a hydrocyclone the principlethat the particles are more dense than water is alsoused, but here centrifugal forces are used in addi-tion28–30 to increase this difference. To illustrate this,the water inside a cup can be rotated and the par-ticles will be hurled out towards the edges (Fig. 5.7).This is also the reason why this filter is sometimescalled a tea-cup settler. In a swirl separator thewater enters along the periphery of a circular tank,



so the particles follow the periphery and the puri-fied water will be pressed into the centre. The par-ticles will then sink to the bottom, while the purifiedwater is drained out of the centre. Because the cen-trifugal forces are greater than the gravitationalforce, a smaller area is needed than for a traditionalsettling basin to remove the same amount of particles. Re-suspension or leakage is also less thanwith a settling basin, because the area where thisoccurs is smaller. However, in comparison to thesettling basin, the hydraulic load on a swirl separa-tor may be much higher, 20–25m/h.24

An advantage of this type of filter unit is its quitesimple construction with no movable parts; in addi-tion is it cheap to buy and area intensive. The greatdisadvantage with swirl separators is that theyrequire uniform water flow for optimal efficiency. Ifthe flow is higher than the filter is designed for, the

particles will flow out of the unit with the outletwater in the center.

ExampleAn outlet water flow of 10 l/min (0.6m3/h) is to bepurified using a swirl separator. The acceptablehydraulic load is set to 20m/h. Find the necessarysize of the separator.

Vs > Q/A

A > Q/Vs

> 0.03m2

The radius of the separator must be more than 9.8cm.

> 0.60 m /h20 m/h

3

Figure 5.7 A swirl separator.

5.3.4 Integrated treatment systems

Nature-based wastewater treatment technologycovers several methods that can be used alone or incombination.31,32 Three mains systems can be usedfor treatment of effluent water from land-basedaquaculture farms: ground filtration, constructedwetland, and pond systems. Ground filtrationsystems can further be divided into open ponds orsubsurface trenches (Fig. 5.8). In a filtration systemthe soil is used as a filter medium, and is actuallythe same as the depth filter described earlier. Thewater is distributed above the filter bed and trickles through the soil in which filtration, adsorp-tion/precipitation and biological degeneration willbe major processes. Nutrients, organic matter andmicro-organisms will be removed. The local soil,if suitable, is normally used as a filtration mediumand therefore it is also called on-site treatment.However, soil (e.g. sand) or other suitable porousmedia such as Leca may also be trucked to the site.The main problem with ground filtration for treatment of effluent water from aquaculture is thelarge amount of water to be treated and the lowhydraulic capacity of soil-based systems. Even ifsome improvement can be achieved by using suit-able porous media the area needed is still large.Normal values for domestic wastewater when using

subsurface trenches are about 10000 l/m2/day, withsome higher values in open ponds. Effluent waterfrom aquaculture facilities has been less wellstudied and there is a lack of available values.Because the pollutant concentrations are lower, theload may be somewhat higher, but if the system isoverloaded it will function sub-optimally andpurification will be reduced. Results from treat-ment of domestic wastewater in a cold climate lat-itude (69°N) with an annual mean temperature of1.2°C show the following results: 70% nitrogenremoval, 99% phosphorus removal, 70% chemicaloxygen demand (COD) removal and close to 100%removal of faecal coliforms. The normal flow was750m3/day in a 2000m2 open basin, but during snowmelt it can be up to 3500m3/day.33

If the system becomes totally clogged a tractorwith a shovel can remove the upper 5–10cm ofsand, after which the system can be used again. Atypical value for ground filtration systems forwastewater is 100 l/m2/day, but of course this varieswith the soil conditions. If using such values foroutlet water from fish farming, a tremendous areais necessary for water purification. When usingdepth filtration in fish farming there is a need forhigh hydraulic capacity.

Constructed wetlands are commonly used forvarious types of polluted water such as domestic


A

C

B

Figure 5.8 Different types of integrated treatmentsystems. (A, B) Filtration systems where the outlet waterfrom the fish farm is sent to filter through the bottom ofa trench: (A) a newly constructed filtration trench; (B) atrench filled with water. (C) Sketch showing a constructedwetland through which the outlet water from the fish farmhas to pass before it reaches the recipient water body.The plants will take up nutrients from the outlet water.


and industrial wastewater, and have also been uti-lized to treat wastewater from aquaculture.4 Theoutlet water is sent into the wetland that containswetland-adapted plants and a porous earthmedium. The plants take up and utilize the nutri-ents in the outlet water; in addition, the roots of the plants create an environment that increases the purification processes. By sending outlet waterfrom a salmon hatchery to an abatement pondcoupled to a constructed wetland, the followingremoval rates were attained: solids above 98%,ammonia above 84%, biodegradable organics andphosphorus above 90%.34

Both ground filtration systems and constructedwetlands will, in addition to removal of nutrientsand organic matter, inactivate micro-organisms.The great disadvantage with the systems is the lowhydraulic capacity in relation to the amounts ofwater in aquaculture.

5.4 Hydraulic loads on filter unitsIt is important to be aware that a filter system isdesigned for a given flow of water with a given char-acteristic. If either less or more water than the filterunit is designed to treat is used, the filter will notfunctional optimally. The ability of a filter to toler-ate varying water flows, for example when tappingdown a fish tank, depends on its design. Equipmentusing settling as a principle is especially intolerantof variations in the water flow, particularly highflows. For a mechanical filter, variations in load arenormally not so critical. However, if the loads aretoo high the filter cloth may become so clogged thatbreakdown can occur.

A common fault on aquaculture facilities is thatthe tanks and outlet pipes are the incorrect designand size, so settling of particles occurs in the system.Shock drainage of the outlet system is used, oftenonce or twice a day, to remove settled particles andavoid total blockage. If shock drainage is necessaryto keep the outlet pipes open, something is wrongwith the design and construction of the outlet (see Chapter 11). When shock draining the fish production tanks, the water flow in the outlet pipesis increased and so is the particle concentration,because particles that had settled in the outletsystem will now go into suspension as a result of thehigher velocity. If the filter system does not toler-ate variation in water flow, reduced purification

results. This is critical, because it is in these situations that the number of particles is highest,and where good purification is necessary. Here the importance of choosing an appropriate filtersystem, and of doing everything correctly beforethe filter system, is apparent. It is also necessary tobe aware of the interaction between the differentparts in the farming system.

5.5 Purification efficiencyAs shown, there are several methods for removingparticles from a water flow. To evaluate how effectively a filter is functioning and to comparefilter systems, the term purification efficiency iscommonly used (Fig. 5.9). This can be defined asfollows:

Ce = ((Cin − Cout)/Cin) × 100

where:

Ce = efficiency (%)

Cin = concentration of the actual substance enter-ing the filter

Cout = concentration of the actual substance exitingthe filter.

ExampleThe concentration of suspended solids entering thefilter units is measured and found to be 20mg/l; onexiting the filter the concentration is measured as 5mg/l. Find the purification efficiency of the filter.

Ce = ((Cin − Cout)/Cin) × 100= ((20 − 5)/20) × 100= 75%

Figure 5.9 The purification efficiency indicates howmuch of an actual substance is removed by the filter.

In aquaculture the efficiency is normally ex-pressed as the percentage of the incoming TSSremoved by the filter. However, it may also be used for other substances, such as the amount ofremoved nutrients (total phosphorus, TP or totalnitrogen, TN) or as reduction in chemical oxygendemand (COD) or biological oxygen demand(BOD) in the water passing through the filter. Thelast two measurements tell us how the outlet waterwill affect the oxygen concentration in the recipi-ent water body.

When considering removal of nutrients, it isimportant to know the proportion attached to par-ticles and not dissolved, since it is not possible toremove dissolved nutrients with a particle filter.These values depend on the species and the feedcomposition and utilization. Experiments onsalmonids have shown that more than 80% of nitro-gen compounds are dissolved in the water,10

whereas the situation is reversed for phosphoruswith up to 80% being attached to particles.However, phosphorus leach easily from particleslying in the water; this is a major reason to removethe particles from the water flow as quickly as possible.

The efficiency of the particle filter of coursedepends on the particle concentration and characteristics of the water to be purified. For agood comparison of filter systems, they must be tested on exactly the same water. Typical valuesfor removal of TSS from wastewater from fishfarming range from 30 to 80%1,7,24,35 while somelower values have been reported for water re-usesystems.7

5.6 Dual drain tankToday, circular tanks with a dual drain outlet systemare used on some farms (Fig. 5.10).36–38 Here thetank is used as the first purification step. In a dualdrain system gravitational forces, and the fact thatthe waste particles are denser than water, are uti-lized to separate the particles and collect them atone point in the tank where a particle outlet isplaced and a small amount of water can be with-drawn to flush the particles out, while the mainwater flow can be withdrawn elsewhere, normallyhigher up in the water mass. Several methods canbe used to do this, and a number of different dualdrain systems exist.39

By separating the water flow in this way, purifi-cation is achieved inside the tank. The particle con-centration in the particle outlet is considerablyhigher than in the outlet withdrawn higher up in thewater mass. This can be the first purification step,and may be the only one. Since the amount of watercoming through the particle outlet is much lowerthan the total flow, only a small part of the waterflow has to be treated (Fig. 5.11), so a much smallerfilter can be used. As the water flow through theparticle outlet can be stable, a swirl separator isvery suitable.The amount of water sent through theparticle outlet in the dual drain system varies from0.5 to 20% depending on the system used.

5.7 Sludge production and utilizationRemoval of particles from the water creates sludgecomprising water and particles. The water contentand dry matter (DM) or total solids (TS) in thesludge depend on the particle filter used. To give anidea of the amount of sludge (faeces) created by thefish, the following estimate can be used: for eachkilogram of feed eaten by rainbow trout 20% faecesare produced, measured on 100% DM basis.40 Thisvalue, however, depends on a number of factorsincluding feed composition, feed type, fish size andfish species, and will therefore vary. If the feed con-version rate is above unity and traditional dry feedis used, there will be feed loss that goes directly to the outlet in addition to the sludge producedfrom the faeces. This also shows the importance ofcorrect feeding and avoiding feed losses.


Figure 5.10 A tank with a dual drain outlet is used forthe first purification step.


ExampleHow much sludge is produced per kg commercialdry feed supplied to rainbow trout if the purificationefficiency of the filter is set to 50% measured as DM?What happens if the feed conversion rate increasesto 1.2?

Per kg feed supplied, 200g of sludge is produced andthe filter collects 50% of this. This means that theamount of sludge collected per 1kg feed supplied is100g.

If the feed conversion rate is 1.2 only 0.83kg of the1kg feed supplied will be utilized for growth, whilethe rest will be feed loss. Calculating that 20% of thefeed eaten is converted to faeces, this represents 0.17kg. If this is added to the feed loss, the amountis 0.17 + 0.17 = 0.34kg. If collecting 50% of this, theamount collected per kg feed supplied will be 170g.

The actual amount of sludge can be much higher;the percentage DM in the collected sludge depends

on the filter system used. In a mechanical filter it ismainly a consequence of how the straining cloth isback-washed, whether using air or water, and ifwater, the amount. Normally the percentage DM inthe sludge is 0.1–1%, but in special filters it mightbe up to 5%. It is advantageous to have as muchDM in the sludge as possible, which means thatback-flushing with water is disadvantageous. Theonly reason for back-flushing with water is that it isan effective system. A large proportion of DM inthe sludge reduces the amount that must be furthertreated and transported; the sludge is thereforeoften dewatered to increase the percentage DM.Filter presses and special centrifuges can beemployed for dewatering.16,23 The sludge may alsobe sent to a settling system for further separationof particles.26 Vertical sedimentation in a cone hasbeen used and increased the DM in the sludge to7–10%.

Normally the sludge must be stored for a periodto accumulate enough so that it can be collected

Figure 5.11 A dual drain tank can becoupled with a small and cheap purifi-cation system.

economically. A particle removal system will there-fore include storage tanks for the sludge.

To make it possible to store the sludge, it must bestabilized. If the sludge is immediately placed in anopen container with access to air at the surface, anuncontrolled decomposition process (rotting) willtake place. This will smell and have negative effectson the development of bacteria and the content ofnutrients. Subsequent use of the sludge can beinhibited because of this. Correct decompositionmakes the nutrients in the sludge available forplants so that it can be used as fertilizer on agri-culture land. Sludge from fish farms is rich inorganic nitrogen (3–9% of DM) and phosphorus(1–4% of DM). In addition, the concentration ofheavy metals is usually below regulatory limits.41

This makes the sludge useful as a fertilizer.Untreated sludge may contain pathogenic (neg-

ative) micro-organisms such as viruses, bacteria andparasites.42 When infected sludge is spread on agri-cultural land and there is drainage to lakes or rivers,

pathogenic micro-organisms could be transferredto the local fish strains. Birds could also transferpathogenic micro-organisms from the sludge tolakes. Therefore the sludge must be treated to inactivate the negative micro-organisms before it isspread; this is not achieved with uncontrolleddecomposition.

There are several ways to inactivate pathogenicmicro-organisms in the sludge; wet or dry com-posting is commonly used. Another method is toadd lime to raise the pH in the sludge and henceinactivate the micro-organisms. Both thesemethods will also stabilize the sludge so that it canbe stored, and make it suitable for use as fertilizeron agricultural land, a very important use forsludge, normally a good fertilizer because of thehigh nutrient content.

When composting sludge, controlled aerobicdecomposition occurs.42 Due to the low content ofDM, wet or liquid composting is employed for fishfarming sludge (Fig. 5.12). Before composting the


Figure 5.12 A wet compostingreactor for treatment of sludge.44


sludge may be mixed with manure or municipalsludge. In a community there may be a centrallyinstalled reactor. The sludge is poured into a con-tainer where air is added, for instance through aninjector pump. In addition the sludge is circulatedaround in the tank so air comes into contact withall the sludge. This results in controlled bacterialdevelopment in the container which decomposesthe organic matter. The process is thermophilic, sothe temperature increases depending on the sludge.For fish farming the sludge is energy-rich, experi-mentally producing 3.1kWh/kgDM,44 and the tem-perature rises to 60–70°C (Fig. 5.13). This hightemperature is maintained for some time so that thepathogenic micro-organisms are inactivated andthe sludge is stabilized for storage. To get rid of thesmell developed during the composting process it isan advantage to include a smell filter, for instancemade of peat, through which all gases emitted bythe composting process have to pass.43–45 Overall,the composting process will result in biological sta-bilization of the sludge, removal of the major odourcompounds, increased availability of some plantnutrients, nitrification and denitrification, andimproved waste consistency.42

If the process is run without access to oxygen,anaerobic digestion, also known as fermentation,will occur. As for composting, naturally occurringmicro-organisms in the sludge are utilized. It isimportant that no air is supplied, so anaerobicdegeneration of the organic substances will occur.Normally some heating of the sludge will be neces-sary to allow the micro-organisms to develop. The

sludge is stored in a closed digester during thisprocess. For animal manure it takes 10–35 days; ahigh fat content reduces the time necessary. Thisfermentation produces methane gas; the process istherefore also called biogas fermentation. In addi-tion to the production of biogas, there will also bea reduction in offensive odours, a breakdown oforganic mass, a reduction of pathogens and animproved fertilizing value due to easier availabilityof the nutrients.42 The biogas can be used forheating or electricity production, so there is actu-ally a positive output from the process. Sludge thathas gone through this process can be stored forlater use as fertilizer for agricultural land.

Adding lime (CaO) or slaked lime (Ca(OH)2) tothe sludge to increase the pH, will also stabilize anddisinfect the sludge. In experiments with sludgefrom fish farming it was shown that by increasingthe pH to 12 and maintaining this value for 7 days,more than 99.9% of the pathogenic viruses and bacteria were killed.41 The sludge produced is wellsuited for use as organic fertilizer.

When establishing a system for particle removalit is therefore necessary to think not only about theparticle filter itself, but also sludge production andits utilization. This includes tanks for sludge collec-tion (Fig. 5.14).

5.8 Local ecological solutionsIt is also possible to devise total ecological produc-tion methods for treatment of effluent water fromintensive fish farming, not just by using it in more

Figure 5.13 Temperature increase in liquid compostedsludge from aquaculture.

Figure 5.14 A tank for sludge collection for later treat-ment is necessary on farms.

extensive forms of aquaculture such as polyculture.Such a system is shown above using integratedtreatment systems for the water and local use of thesludge (Fig. 5.15).

The fish farm must employ water re-use technol-ogy to reduce the amount of outlet water. A micro-screen is used as the first step to remove the largerparticles.Then the purified outlet water may be sentto a ground filtration unit or constructed wetlandfor further purification.The sludge produced by themicroscreen can be sent to a locally installed liquidcomposting unit together with animal manure andeventually be used as fertilizer on nearby agricul-tural land, so avoiding long distance transport ofthe sludge. A solution with no effluent, either ofnutrients, organic matter or micro-organisms isachieved in this way.

References1. Chen, S., Stechey, D., Malone, R.F. (1994) Suspended

solids control in recirculating aquaculture systems. In:Aquaculture water reuse systems, engineering designand management (eds Timmons, M.B., Losordo,T.M.).Elsevier Science.

2. Summerfeldt, S.T. (1999) Waste-handling systems. In:CIGR handbook of agricultural engineering, part IIaquaculture engineering (ed. Wheaton, F.). AmericanSociety of Agricultural Engineers.

3. Liltvedt, H., Hansen, B.R. (1990) Screening as a method for removal of parasites from inlet water to fish farms. Aquacultural Engineering, 9:209–215.

4. Midlen,A., Redding,T.A. (1998) Environmental man-agement for aquaculture. Chapman & Hall.

5. Bergheim, A., Brinker, A. (2005) Water pollutionfrom fish farms. In: Water encyclopedia 3: Surface andagricultural water (eds Lehr, J.H., Keeley, J.). pp.579–581. Wiley-Interscience.

6. Lekang, O.I., Fjæra, S.O. (2002) Teknisk utstyr tilfiskeoppdrett. Gan forlag. ISBN 82-492-0353-4.

7. Brinker, A., Bergheim, A. (2005) Waste treatment infish farms. In: Water encyclopedia 1: Domestic, munic-ipal and industrial water supply and waste disposal(eds Lehr, J.H., Keeley, J.). pp. 681–684. Wiley-Interscience.

8. Bergheim, A., Aasgaard, T. (1996) Waste productionfrom aquaculture. In: Aquaculture and water resourcemanagement. (eds Baird, D.J., Beveridge, M.C.M.,Kelly, L.A., Muir, J.F.). Blackwell Science.

9. Chen, S., Coffin, D.E., Malone, R.F. (1993) Produc-tion, characteristics, and modeling of aquaculturalsludge from a recirculating aquacultural system usinga granular media filter. In: Techniques for modernaquaculture (ed. Wang, J-K.). pp. 16–25. Proceedingsof Aquacultural Engineering Conference, Spokane,Washington. ASAE Publication 02–93.

10. Cripps, S.J. (1995) Serial particle size fractionationand characterisation of an aquacultural effluent.Aquaculture, 133: 323–339.

11. Wong, K.B., Piedrahita, R.H. (2000) Settling velocitycharacterization of aquacultural solids. AquaculturalEngineering, 21: 233–246.


Figure 5.15 Design of a fish farmingsystem with no direct outlet: a localecological solution.


12. Warrer-Hansen, I. (1982) The right treatment ofwastewater. Fish Farming International, Aug., 36–37.

13. Droste, R.L. (1997) Theory and practice of water andwastewater treatment. John Wiley & Sons.

14. Davis, M.L., Conwell, D.A. (1998) Introduction toenvironmental engineering. McGraw-Hill.

15. Letterman, R.D. (1999) Water quality and treatment.American Water-Works Association McGraw-Hill.

16. Tchobanoglous, G., Burton, F.L., Stensel, D.H. (2002)Wastewater engineering. McGraw-Hill.

17. Salvato, J.A., Nemerow, N.L.,Agardy, F.J. (2003) Envi-ronmental engineering. Wiley-Interscience.

18. Trussel, R., Hand, D.W., Howe, K.J., Tchobanoglous,G., Crittenden, J.C. (2005) Water treatment: principlesand design. Wiley-Interscience.

19. Cripps, S.J., Bergheim, A. (2000) Solids managementand removal for intensive land-based aquacultureproduction systems. Aquacultural Engineering, 22:33–56.

20. Spotte, S. (1979) Fish and invertebrate culture. Watermanagement in closed systems. John Wiley & Sons.


22. Tucker, J.W. (1998) Marine fish culture. Kluwer Aca-demic Publishers.

23. Montgomery, J.M. (1985) Water treatment, principlesand design. John Wiley & Sons.

24. Lekang, O.I., Bomo, A.M. (1999) Alternative sedi-mentation methods for effluent treatment in aquacul-ture. ITF rapport 101, Norwegian University of LifeScience (in Norwegian).

25. Lekang, O.I., Bomo, A.M., Svendsen, I. (2001) Bio-logical lamella sedimentation used for wastewatertreatment. Aquacultural Engineering, 24: 115–127.

26. Bergheim, A., Liltvedt, H.S., Cripps, G., Indrevik,Nygaard Austerheim, L. (1996) Avvanning, stabiliser-ing og utnyttelse av våtslam fra fiskeoppdrett. RapportRogalandsforskning 280 (in Norwegian).

27. Henderson, J.P., Bromage, N.R. (1988) Optimising theremoval of suspended solids from aquacultural efflu-ent in settlement lakes. Aquacultural Engineering, 7:167–181.

28. Wheaton, F.W. (1977) Aquacultural engineering. R.Krieger.

29. Davidson, J., Summerfelt, S.T. (2005) Solids removalfrom coldwater recirculating system – comparison ofa swirl separator and a radial flow settler. Aquacul-tural Engineering, 33: 47–61.

30. Veerapen, J.P., Lowry, B.J., Couturier, M.F. (2005)Design methodology for the swirl separator. Aqua-cultural Engineering, 33: 21–45.

31. Crites, R., Tchobanoglous, G. (1998) Small and decentralized wastewater management systems.McGraw-Hill.

32. Mander, U., Jensen, P.D. (2003) Constructed wetlandsfor wastewater treatment in cold climate. WIT Press.

33. Jenssen, P.D. (1992) Oppfølging av Setermoenrenseanlegg. Jordforsk rapport 7.2400-05 (in Norwegian).

34. Michael, J.H.J. (2004) Nutrients in salmon hatcherywastewater and its removal through the use of awetland constructed to treat off-line settling pondeffluent. Aquaculture, 226: 213–225.

35. Cripps, S.J., Kelly, L.A. (1996) Reductions in wastesfrom aquaculture. In: Aquaculture and water resourcemanagement (eds D.J. Baird, M.C.M. Beveridge, L.A.Kelly, J.F. Muir), pp. 166–201. Proceedings of a Con-ference at the University of Stirling. Blackwell Science.

36. Ulgenes, Y., Eikebrokk, B. (1994) Fish farm effluenttreatment by microstrainers and a particle trap. EIFACWorkshop, University of Stirling. June 1994.

37. Timmons, M.B., Summerfelt, S.T., Vinci, B.J. (1998)Review of circular tank technology and management.Aquacultural Engineering, 18: 51–69.

38. Lekang, O.I., Bergheim, A., Dalen, H. (2000) An inte-grated waste treatment system for land-based fish-farming. Aquacultural Engineering, 22: 199–211.

39. Timmons, M.B., Riley, J., Brune, D., Lekang, O.I.(1999) Facilities design. In: CIGR handbook of agri-cultural engineering, part II aquaculture engineering(ed. F. Wheaton), pp. 245–280. American Society ofAgricultural Engineers.

40. Cho, C.Y., Hynes, J.D., Wood, K.R., Hynes, H.K.(1991) Quantification of fish culture wastes by bio-logical (nutritional) and chemical (limnological)methods; In: Nutritional strategies for aquaculturewaste. University of Guelph.

41. Bergheim, A., Cripps, S.J., Liltvedt, H. (1998) Asystem for the treatment of sludge from land basedfish-farms. Aquatic Living Resources, 11: 279–287.

42. Burton, C.H., Turner, C. (2003) Manure management:treatment strategies for sustainable aquaculture. SilsoeResearch Institute.

43. Donantoni, L., Skjelhaugen O.J., Sæther, T. (1994)Combined aerobic and electrolytic treatment ofcattle slurry. Paper given at CIGR and AgEng Con-gress, Milan, 1994.

44. Skjelhaugen, O.J., Sæther, T. (1994) Local ecologicalwaste water solution for rural areas, based on aerobictreatment and recycling of nutrients into agriculturalland. Paper given at CIGR and AgEng Congress,Milan, 1994. ITF trykk nr 59/1994. Norwegian Uni-versity of Life Science.

45. Lekang, O.I., Jenssen, P.D., Skjellhaugen, O.J. (1995)Local ecological solution for treatment from aqua-culture. In: Technical solutions in the management of environmental effects of aquaculture (ed. J.Makkonen), Seminar 258. Scandinavian Associationof Agricultural Scientists.

6Disinfection

infections. In a water re-use plant, the water mayalso be disinfected before being used again to avoidincreasing the micro-organism burden. The outletwater may also be disinfected to avoid transfer ofmicro-organisms to fish species in the recipientwater body.

There are several methods for disinfecting waterand a number of general textbooks are available(for example, refs 1–4). Disinfectants can be sepa-rated into chemical agents and non-chemicalagents.1 Alternatively, a four-group classificationcan be used: 1, chemical agents; 2, physical agents;3, mechanical agents; 4, radiation.4 The first groupincludes chlorine and its compounds, bromine,iodine, ozone, phenol and phenolic compounds,alcohols, heavy metals and related compounds,soaps and synthetic detergents, quaternary ammo-nium compounds, hydrogen peroxide and variousalkalis and acids.The second group includes heatingand the use of sunlight, especially the ultraviolet(UV) end of the spectrum.The third group includesparticle separation; although particle separation isthe main objective, there will also be a reduction inthe number of micro-organisms because many areattached to particles. Larger parasites such asCostia and Gyrodactylus will also be removed witha particle filter with small (20μm) mesh size. To thefourth group belong different types of radiationincluding electromagnetic, acoustic and particle.For example, gamma rays are used to disinfect andalso sterilize water and food, although this methodis expensive.

Many of the chemical agents employed oxidizethe organic materials, including the micro-organisms. The oxidizing potential indicates how

6.1 Introduction

Disinfection can be described as the reduction ofmicro-organisms such as bacteria, viruses, fungi andparasites to a desired concentration. This is not thesame as sterilization where all micro-organisms areeliminated. The aim of disinfection of water in fishfarming is to reduce to an acceptable level the riskof transfer of infectious disease from the water to thefish. When disinfecting water for fish farming, selec-tive inactivation of fish pathogenic micro-organismsis required in addition to overall reduction in thetotal number of micro-organisms. Pathogenic micro-organisms infect the fish and cause disease. Trans-missions of infectious diseases is possible in twoways, horizontal and vertical. Horizontal transmis-sion includes direct or indirect contact between individuals or populations. Direct contact occursbetween individuals or urine or faeces, while indirectcontact occurs through contact with water, equip-ment and personnel with pathogens. Vertical trans-port includes transmission from one generation tothe next through roe or milt, for example.

Disinfection can be performed in different situa-tions in aquaculture. Water, equipment, buildingsand effluent can all be disinfected. Equipmentincludes tanks, nets, pipes and shoes. Disinfection ofbuildings includes, for instance, disinfection of thehatchery after seasonal use. Effluent may includesludge and by-products. Disinfection of water actu-ally occurs at several places in an aquaculture plant.Usually the inlet water is disinfected, whether it isseawater or freshwater. At the larval stage it is par-ticularly important to reduce the number of micro-organisms because larvaes are more vulnerable to

63


effective the agent is likely to be: ozone has thehighest potential, while bromine and iodine havethe lowest potentials.

Regardless of the method chosen for disinfectionthe quality of the water to be disinfected is of majorimportance. Pure inlet water is much simpler to dis-infect than outlet water because the latter containsmore particles. Turbid water and water with a highcontent of organic substances, such as re-use water,are also more difficult to disinfect and therefore not so commonly disinfected. Before disinfectingcontaminated water, it is essential to carry out some kind of pre-treatment, normally comprisingremoval of particles.

For disinfection of water supplies to aquaculturefacilities, UV light and ozone are most often used.Later in this chapter there is a survey of methodsemployed, with emphasis on UV light and ozone.

6.2 Basis of disinfection

6.2.1 Degree of removal

The term percentage removal of actual micro-organisms is used in environmental engineering. Inmicrobiological terms log10 removal or inactivation(decimal removal) is used to define the disinfectionyield; normally a reduction of between 99 and99.99% of the total number of bacteria is wanted,which corresponds to a log disinfection of 2–4.However, these terms do not give exact values ofthe number of micro-organisms left; they only indi-cate by how much numbers are reduced from thestarting concentration.

ExampleThe normal concentration of bacteria is 107/ml anda reduction of 99.9% is required. Find the concen-tration of bacteria present after disinfection.

Solution

concentration of bacteria = 107(1−0.999)after infection = 10000000− 9990000

= 10000= 104/ml

ExampleThe starting concentration of bacteria is 107/ml. Alog disinfection of 3 is wanted. Calculate the newconcentration of bacteria.

Solution

Let the starting concentration be N1 and the end con-centration N2; log(disinfection) = 3.

log(disinfection) = log(N1/N2)= logN1 − logN2

logN2 = logN1 − log(disinfection)= 7−3= 4

N2 = 104/ml

6.2.2 Chick’s law

Inactivation of micro-organisms in a disinfectionplant depends on the time that the micro-organismare exposed to the disinfectant. This is described byChick’s law:

where:

dn/dt = necessary time to inactivate n micro-organisms

k = time constant depending on disinfectant,type of micro-organism and water quality

N = number of live micro-organismst = time.

This differential equation can be integrated withinlimits to give the following equation:

N1 = N0 e−kt

where:

N0 = number of micro-organisms at the startN1 = number of micro-organisms after time t.

6.2.3 Watson’s law

Based on the results of Chick’s law, Watson’s lawcan be developed:

where:

Λ = coefficient of specific toxicityC = concentration of disinfectantn = exponent (normally around 1)t = time after start-up.

lnNN

C tn1

0

= −Λ

ddnt

kN=

This means that the relation between the numberof active and inactive micro-organisms is a productof the concentration of the disinfectant and theexposure time.

6.2.4 Dose-response curve

Based on Watsons’s law a dose-response relationmay be established for specific types of micro-organism. This gives the proportion of micro-organisms inactivated by fixed doses of disinfectantover various time periods. Exact dose-responserelationships are difficult to determine in practicefor several reasons. It is often difficult to isolate newpathogens and the response to a certain dosedepends, amongst other factors, on the immunestatus of the organism, environmental conditionsand population density.

6.3 Ultraviolet light

6.3.1 Function

Ultraviolet (UV) light is electromagnetic radiationwith a wavelength of 1–400 nanometer (1nm =10−9 m) located at the lower end of the visible spec-trum and beyond (Fig. 6.1) (the spectrum of visiblelight extends down to 380nm). At the opposite endof the visible spectrum is the infrared (IR) region,heat radiation with longer wavelength whichcannot be detected by the human eye.

The ability of the UV light to inactivate anddestroy micro-organisms varies with both wave-length and the micro-organisms to be inactivated.

The most effective wavelength for general disinfec-tion is 250–270nm.4 UV light created by mercuryvapour lamps will have a wavelength of 253.7nm,which is effective for disinfection.

6.3.2 Mode of action

UV light will damage the genetic material (DNAand/or RNA) in the micro-organism by disruptionof the chains which results in inactivation anddeath. Inactivation (D) is proportional to the doseof radiation per unit area (intensity) of the UV light(I) and the exposure time (t):

D = It

The radiation dose is normally given in units ofμWs/cm2 (microwatt second per centimetresquared), i.e. radiation intensity (energy) per unitarea.

The effectiveness of the UV light depends on anumber of factors including lamp intensity, age ofthe lamp, cleanliness of the lamp surface, distancebetween the lamp and the organism to be inacti-vated, type of organism to be inactivated, durationof UV exposure and purity of the water.

UV lamps become less efficient with use, andneed to be replaced regularly, normally at leastonce a year. Normally the lamp is changed when itsoutput has diminished to 60% of the original.5 Theintensity of the lamp should be measured to ensuresufficient exposure to UV radiation.

How well UV light passes through water dependson the characteristics of the water. UV transmissiondepends on the particle content (turbidity) of thewater, for example. Transmission will be lower forre-used water than for new good quality inlet water.When dimensioning UV systems, it is thereforeimportant to be aware of this.

6.3.3 Design

UV lamps can be placed either in the water (Fig.6.2a,b) or above the water surface6 (Fig. 6.2c,d).Usually the lamps are placed in a chamber throughwhich the water flows. UV chambers may beequipped with reflectors or turbulence discs to irra-diate the total water flow more effectively. The UVlamp is normally placed inside a quartz glass pipeto protect it from direct cooling by the water andfouling of the lamp surface.

Disinfection 65

Figure 6.1 Different wavelengths of light.


B D

A C

Fig

ure

6.2

AU

V p

lant

can

be

cons

truc

ted

with

UV

lam

ps p

lace

d in

the

wat

er fl

ow,

whi

ch is

the

usu

al a

rran

gem

ent

(A,

B),

or

abov

eth

e w

ater

flow

(C

, D

).

Fouling may occur on the quartz glass pipeswhich therefore must be cleaned regularly, eithermanually or automatically by washing with brushes,to maintain optimal UV intensity. Fouling willdecrease UV transmission.

The intensity of the UV radiation can be mea-sured continuously and readings linked to a regu-lator, so that sufficient radiation intensity can bemaintained in water of varying turbidity and withvarious extents of fouling on the pipes. If the radi-ation intensity decreases to a limiting value, analarm can be activated so that remedial measurescan be taken.

If the UV lamps are placed above the watersurface, the water flows in a thin layer directlybelow them. In this system the UV rays must passthrough the water surface and therefore must bemore intense than in a submerged system. For thesame reason, the layer of water is thin.As the lampsare above the water surface, fouling will not be socritical for this kind of system, but the problem ofvarying water quality will be the same as in plantswhere the UV lamp is placed inside the water flow.

6.3.4 Design specification

When designing the UV plant, the radiation dose,water retention time in the chamber, and the UVtransmission must be specified. The transmission is given as the percentage of the known transmis-sion for distilled water. If the water is contaminatedwith particles, humus, etc. and is coloured, the transmission will decrease and must therefore bemeasured. This is normally performed for one dis-tance, for instance 5cm. The transmission for otherdistances may then be found from the followingequation:7

TL = T0L/L0

where:

TL = transmission through L cm of waterT0 = transmission through L0 cm of waterL = thickness of the water layer (cm)L0 = thickness of the water layer set at 1cm.

ExampleThe UV transmission for re-used water of a givenquality is found to be 90% through a 1cm thick layer(i.e. 90% of the UV radiation passes through a water

layer 1cm thick). Find the UV transmission for a 5cm thick water layer.

Using the equation

TL = T0L/L0

where:

L = 5cmL0 = 1cmT0 = 90% = 0.90

TL = 0.905/1

= 0.59

Thus 59% of the initial UV radiation passes througha 5cm thick water layer, i.e. the UV transmission is59%.

The UV intensity can be defined as the amountof radiation per unit surface area and the followingequation can be used to find the radiation dose atdistance L from the UV radiation source7:

where:

D = radiation dose (mWs/cm2)P = radiation effect (W)S = area of radiated surface (cm2)T0 = the transmission of the water through 1cm (%)L = thickness of the water layer that is radiated

(cm)t = necessary time for radiation (s).

ExampleA UV tube is mounted in the middle of a cylindricalchamber of radius 5cm. The length of the UV lampis 1m and therefore the largest radiated area will be0.31m2. The UV transmission through a 1cm layerof water is 95%.The time from when the water entersthe chamber to when it goes out is 10s, so it isexposed to UV radition for 10s. The UV radiationeffect of the tube is 16W.

Find the lowest UV radiation dose to which the wateris exposed (this is close to the interior walls of thechamber, almost 5cm from the UV tube).

Using the equation

DPS

T tL= 0

DPS

T tL= 0 P

Disinfection 67


where:

P = 16WS = 0.31m2 = 3100cm2

T0 = 95%L = 5cmt = 10s

= 0.0400Ws/cm2

= 40mWs/cm2

To obtain good disinfection it is necessary to havea low content of particles and humus in the water,so these must be removed before the water entersthe UV chamber, otherwise they will shield themicro-organisms from the UV radiation.

6.3.5 Dose

The dose required to kill pathogenic micro-organisms depends on the organism. In commercialplants a normal UV dose is in the range 30–35mWs/cm2, and this is adequate for a log disinfectionof 3 of most of the common aquaculture bacte-ria.5,6,8,9 However, some viruses, such as IPN,are much more difficult to inactivate and doses of 100–200mWs/cm2 have been suggested.10 Careshould be taken when reading papers where theeffective dose rates are given, as the methods usedin the laboratory are often different from those usedin commercial aquaculture where, for instance, UVtransmission through the water is variable.

6.3.6 Special problems

As previously described, particles can be a problembecause they shade the micro-organisms from theUV light, so it is absolutely necessary to remove thembefore UV irradiation. If the water is very turbid, asmay occur in systems with a high degree of water re-use, UV transmission may be so reduced that it isimpossible to use a UV lamp for disinfection.

6.4 Ozone

6.4.1 Function

Ozone (O3) is a colourless gas with a boiling pointof −112°C, that is sometimes called trioxygen.

D = × ×163100

0 95 105 1.

Ozone gas is unstable and will quickly be brokendown to O2; the half-life of O3 is around 15 minutes.It is therefore necessary to produce the ozone onsite.

Ozone is produced by the corona method; air orpure oxygen gas is passed through a high voltageelectric field11 (Fig. 6.3). This is actually the sameprocess that happens with lightning in a thunder-storm, where ozone gas can also be smelled. Energyis added to the oxygen molecule and ozone iscreated: 3O2 + energy = 2O3. An ozone generator isshown in Fig. 6.4, which can use either pure oxygenor air to produce ozone. If using air, for the highestcost effectiveness, it must be as dry as possiblebefore entering the ozone generator. An air driermust therefore be used. If using air, the water maybecome super-saturated with nitrogen. Using pureoxygen to generate ozone is more expensive. Thebest source of oxygen to use must be decided on acase by case basis. Use of air result in 0.5–3% ozonein the gas stream, whereas the corresponding valuesfor oxygen are 1–6%.4 Energy requirements forproducing ozone are typically in the range 3–30kWh/kg.7,11 Only a small amount of (5–10%) thesupplied energy is used for ozone production.7

Small quantities of ozone may also be produced by radiation of air or oxygen gas with UV light ofspecific wavelength; this is named photozone (seesection 6.5.1).

6.4.2 Mode of action

Ozone is a very strong oxidizing agent, highly toxicto all forms of life. When dissolved in water it startstwo reactions, slow direct oxidation of organic

Figure 6.3 In the corona method, ozone gas is createdby allowing air or oxygen to flow through a high voltageelectric field.

Disinfection 69

A

B

Figure 6.4 (A) An ozone generatorfor production of ozone gas. (B)Ozone generator control panel.


substances by O3 and a chain reaction with forma-tion of different free radicals based on hydroxylradical (OH−). Ozone acts by damaging cell mem-branes and nucleic acids, breaking long chain molecules down into simpler forms which may befurther degraded in the biological filter. This inacti-vates the micro-organisms. Ozone has anothereffect that could be an advantage in aquaculturesystems: by oxidation it reduces the amount of NH3,NO2 and BOD biofilm on surfaces.11,12 This can be seen in water re-use systems where disinfectionwith ozone improves water quality and may therefore be more beneficial than use of other disinfectants. Oxidation by ozone eliminates theyellow/brown coloration of the water that is builtup in a water re-use system with very high degreesof re-use. When using ozone as a disinfectant it isrecommended that particles be removed from thewater before the ozone is added, otherwise much ofthe ozone will be used to oxidize the particles.

6.4.3 Design specification

When adding ozone gas to water a special injectionsystem has to be used to ensure good gas watermixing. The method is similar to those for mixingoxygen gas into water and use of a venturi is quitecommon (Fig. 6.5), 90% transfer of ozone has beenachieved with a properly designed diffuser system.4

An ozone disinfection system therefore consists ofa production system, the generator and an injectionsystem.

Since there is a dose–time relation for ozone dis-infection, the ozone needs to have a certain workingtime in the water to function and oxidize the micro-organisms. A water retention tank is quite com-monly used. Ozone is added to the water which thenenters the retention tank.This must be large enoughto ensure a satisfactory contact time for the ozoneto achieve disinfection. Alternatively, the ozone canbe added at the beginning of the inlet pipe to theaquaculture plant. Because the water takes sometime to reach the fish tanks, this might be enough toachieve sufficient contact between ozone and themicro-organisms to disinfect the water.

When designing an ozone disinfection plant, it isnecessary to include an injection system to get theozone gas into the water and a system that givessufficient retention time between the ozone and thewater. It is important that the residual concentra-tion of ozone is above the value that is needed fordisinfection, this is, of course, less than the inlet concentration.

6.4.4 Ozone dose

To inactivate pathogens with ozone a dose–timerelation applies, as for many other disinfectants.

Figure 6.5 A venturi for addingozone gas to water.

Either a high dose can be used for a short time, orvice versa. Over-dosing must be avoided becausethis may kill the fish.

Example 3Ozone is being used to inactivate the bacteriumVibrio anguillarum. The dose is set to 0.1mg/(lmin)for a log disinfection of 3. This can be achieved byhaving a residual concentration of 0.1mg/l after 1minute working time or 0.5mg/l after 2 minutes.

Most pathogens are killed by an ozone dose of0.1–1.0mg/l and contact time of 1–10 minutes, butthis varies with the organism (for more informationsee refs 7, 9, 13).

The water quality will have large impact on theresidual ozone concentration after a given time;factors such as concentration of dissolved organics,particular organics, inorganic ions, pH and temper-ature will affect the concentration (for example, seerefs 1, 13). An increase in temperature results in areduced lethal dose. Because of these variations it is important to add enough ozone to obtain a satisfactory residual concentration to achieve disinfection.

6.4.5 Special problems

The great problem with the use of ozone is that itis highly toxic for fish and humans. For fish, ozoneis toxic even at relatively small concentrations,because it oxidizes the gill tissue. Recommendedsafe values for fish are generally below 0.002mg/l,but there are large variations in tolerance.14 There-fore after adding ozone to the water and leaving itto react for the necessary time, any residual ozonemust be removed or destroyed. Having an adequateretention time ensures that most of the ozone hasreacted and the product is mainly oxygen gas (O2):this time is normally much longer than that neces-sary to achieve satisfactory disinfection, but ofcourse depends on the decomposition rate.The retention time must therefore be long enoughfor two processes: (1) to get the ozone to destroythe micro-organisms and (2) to remove residualozone toxic to the fish. Non-toxic ozone concentra-tions are normally achieved after 10–20 minutes.15

This can be done either by increasing the size of the retention tank or by increasing the rate ofozone decomposition to oxygen; methods here

include aeration of the retention tank, ozoneremoval by sending the water through a carbonfilter, stripping of the water in a packed column andaddtion of chemicals.11

Ozone is also toxic to humans, even a very lowconcentration of ozone in the air being harmful.In the USA, the maximum during an 8-hour workshift in an enclosed area is set to 0.1mg/l O3; for10 minutes exposure time the maximum level is 0.2mg/l.2 Humans can detect levels of 0.02–0.05mg/l by the smell, so when ozone is smelt the build-ing must be evacuated immediately. The possibili-ties of ozone gas entering the air which humansbreathe must therefore be minimized. Proper ven-tilation in the room where the ozone is producedand added is absolutely essential. In addition theremust be safety equipment to monitor the concen-tration of ozone in the air continuously, and equip-ment that automatically turns off the ozone supplyif the concentration rises above the recommendedvalues; in addition, warning signals must be given(Fig. 6.6).

Another problem with the use of ozone resultsfrom it being a very strong oxidizing agent. It is soeffective that it will oxidize all materials with whichit comes into contact, whether as a gas or dissolvedin water. Plastic and metals are examples of mate-rials that the ozone gas will try to oxidize. Ozonewill destroy most plastics to various extents.Polypropylene pipes are recommended for trans-porting water with a high content of ozone. Ozonewill oxidize metals, causing significant corrosionproblems. Special additives must be used in fibre-glass if it is to be used in retention tanks forozonated water, for instance.

6.4.6 Measuring ozone content

Control of the content of ozone is essential: enoughmust be injected but overdosing avoided to preventharm to the fish and to humans as a result ofdegassing of ozone to the air in an enclosed space.The ozone content of water can be measured eitherby a chemical method employing indigo triosul-fonate or by using a probe for online measurementemploying a potentiometric principle.11 However,online equipment is quite expensive. Measurementof the redox potential which changes with theamount of ozone in the water is therefore quitecommonly used instead.

Disinfection 71


6.5 Other disinfection methods

6.5.1 Photozone

Using photozone gas for water treatment is similarto the use of ozone gas. If air is blown through aUV light chamber photozone will be produced.TheUV wavelength should be less than 200nm. Photo-zone includes the following substances: ozone,atomic oxygen, hydrogen peroxide and hydrogendioxide. All of these substances are oxidants and inwater are strong disinfectants. Concentrations ofthese oxidizing substances must be below toxic

levels when the water reaches the fish. Use of pho-tozone is not common, but it could be employedwhere the water requirements are small, such as fordisinfecting the water supply to hatcheries.

6.5.2 Heat treatment

All micro-organisms can be destroyed if the wateris heated and the high temperature maintained fora certain period of time. The necessary water tem-perature and contact time with hot water dependon the organism and must be determined by exper-iment. Heating of the water is, however, an expen-

Figure 6.6 Warning equipment must be used to avoid toxic ozone concentrations in the production rooms.

sive disinfection method. The high costs of heatingwater (see Chapter 7) mean that the water or theheat must be re-used, for instance by installing heatexchangers.

It is not necessary to boil the water (100°C) toinactivate micro-organisms. The temperature andtime needed vary depending on the specific micro-organisms. A large number of micro-organisms will be inactivated at temperatures of 60–80°C ifthe retention time is correctly chosen. The watermust of course be chilled again before it reaches thefish.

Heat or steam is commonly used for disinfectingboots, nets, tanks, pumps and other equipment.

6.5.3 Chlorine

Chlorine is a very effective disinfectant for waterand the most common method used for disinfectionof municipal drinking water worldwide. It is nor-mally obtained by adding liquid sodium hypochlo-rite (NaOCl) to the water, but solid calciumhypochlorite (Ca(OCl)2) mixed into the water orpure chlorine gas (Cl2) may also be used.1,4 All thesecompounds are strong oxidizing agents and havethe ability to break up organic molecules.

As for ozone, there is a need for a certain contacttime to achieve the necessary effect. This includestime for dissociation in water, time for diffusionthrough cell walls and time to inactivate selectedenzymes. Chlorine concentrations of 0.2–0.5mg/lwith 20–30min contact time or 3–5mg/l for 1–5minutes have been reported.16 To kill parasites inwastewater, typical values of free residual chlorineconcentration of 1–3mg/l with a contact time of10–15 minutes have been reported.17 As for ozone,the residual concentration after a certain retentiontime is also of interest, and overdosing must beavoided. As an example the minimum residual con-centration of chlorine in the drinking water supplyin Norway is 0.02mg/l after 30 minutes contacttime.

Methods for achieving this contact time are as forozone: use of a retention tank, or to add it at thestart of the transfer pipe, so a natural retention timeis achieved when the water is flowing through thepipeline. However, a higher retention time isneeded for chlorine than for ozone.

Water containing free chlorine is very toxic forfish. Concentrations of chlorine should not exceed

3–5μg/l, although for shorter periods of up to 30minutes concentrations up to 0.05mg/l can be tol-erated by most species.18 When disinfecting a tankor other equipment with chlorine, it is importantthat enough clean water is used to wash away thechlorine residues produced. Therefore chlorine isnot normally used for disinfection of inlet water foraquaculture facilities. If chlorine is to be used, amethod for dechlorination must also be includedwhich can, for instance, be use of aeration, UV light,activated carbon, or reducing agents such asNa2S2O3 or Na2SO3.17 However, this will be a sub-optimal method for fish farming purposes, becauseof the costs.

Effluent water could be disinfected with chlorine,but here also it might be necessary to dechlorinatebefore discharge to the recipient water bodybecause it might be toxic for the fish it contains. Ifeffluent with a high content of chlorine were to bedischarged into recipients with high content oforganic substances, chlorinated organic substancesmight be created: trihalomethanes (THMs) arefound in drinking water which has been disinfectedby chlorination; chloroform is the most commonTHM and its presence is well correlated with thedosage of chlorine.

6.5.4 Changing the pH

Increasing or decreasing the pH may also be usedfor disinfections of water. The pH can be increasedby adding lye, or decreased by adding some kind of acid. This is unsatisfactory, however, because thepH has to be normalized again whether the wateris going to be used on the fish farm or whether it isthe outlet water sent to the recipient. This methodis sometimes used for treatment of purified processwater and bleeding water from slaughter houses.

6.5.5 Natural methods: ground filtration orconstructed wetland

On-site methods such as ground filtration and con-structed wetlands may also be used for inactivationof micro-organisms. Put simply, naturally occurringmicro-organisms in the on-site systems destroy thepathogenic micro-organisms in the water.19,20 It isimportant to have adequate residence time in thesystems to ensure that pathogenic micro-organismsare destroyed.

Disinfection 73


References1. Montgomery, J.M. (1985) Water treatment: principles

and design. John Wiley & Sons.2. Langlais, B., Reckhow, D.A., Brink, D.R. (eds) (1991)

Ozone in water treatment: application and engineer-ing. Lewis.

3. Spellman, F.R. (1999) Choosing disinfection alterna-tives for water/wastewater treatment plants. CRCPress.


5. Rodrigues, J., Gregg, T.R. (1993) Consideration forthe use of ultraviolet in fish culture. In: Techniques formodern aquaculture (ed. Wang, J.K.), ProceedingsAquacultural Engineering Conference, Spokane,Washington. ASAE Publication 02-93, AmericanSociety of Agricultural Engineers.

6. Weaton, F.W. (1977) Aquacultural enginering. R.Krieger.

7. Gebauer, R., Eggen, G., Hansen, E., og Eikebrokk, B.(1992) Oppdrettsteknologi – vannkvalitet og vannbe-handling i lukkede oppdrettsanlegg. Tapir Forlag (inNorwegian).

8. Brown, C., Russo, D.J. (1979) Ultraviolet light disin-fection of shellfish hatchery sea water. Aquaculture,17: 17–23.

9. Liltved, H., Hektoen, H., Efrainsen, H. (1995) Inacti-vation of bacterial and viral fish pathogens by ozona-tion or UV radiation in water of different salinity.Aquacultural Engineering, 14: 107–122.

10. Yoshimiza, M., Takizawa, H., Kimura, T. (1986) UVsusceptibility of some fish pathogenic viruses. FishPathology, 21: 47–52.

11. Colt, J., Cryer, E. (2000) Ozone. In: Encyclopedia ofaquaculture (ed. Sickney, R.R.). John Wiley & Sons.

12. Tango, M.S., Gagnon, G.A. (2003) Impact of ozona-tion on water quality in marine recirculation systems.Aquacultural Engineering, 29: 125–137.

13. Lawsons,T.B. (2002) Fundamentals of aquaculturalengineering. Kluwer Academic Publishers.

14. Wedemeyer, G.A., Nelson, N.C., Yasutaka, W.T.(1979) Potentials and limits for the use of ozone as afish disease control agent. Ozone: Science and Engi-neering, 1: 295–318.


16. Johson, S.K. (2000) Disinfection and sterilization. In:Encyclopedia of aquaculture (ed. Sickney, R.R.). JohnWiley & Sons.

17. Wedemeyer, G. (2000) Chlorination/dechlorination.In: Encyclopedia of aquaculture (ed. Sickney, R.R.).John Wiley & Sons.

18. Wedemeyer, G. (1996) Physiology of fish in intensiveculture systems. Chapman and Hall.

19. Stevik, T.K. (1998) Retention and elimination of path-ogenic bacteria percolating through biological filters.PhD thesis, Norwegian University of Life Science.

20. Bomo, A.M. (2004) Application of natural treatmentsystems on fish farm wastewater. PhD thesis,Norwegian University of Life Science.

7Heating and Cooling

able on this subject and also on basic thermody-namics; see, for example, refs 1–6.

7.2 Heating requires energyA supply of energy is needed to heat water. Energycan be transferred in three different ways: by radi-ation, by conduction or by convection. Electromag-netic radiation from the sun or an electric heater,for example, will be partly absorbed when it falls ona substance and become internal energy. With con-duction, the energy is transferred between solids orliquids as vibrational energy of the atoms or mole-cules. Additionally, in materials with a supply of‘free’ electrons, e.g. metals, these electrons share inany energy gain resulting from a temperature riseand their velocities increase more than those ofatoms or molecules, so energy is quickly transferredto other parts. For example, when the end of an ironbar is heated, the energy will be transferred to thewhole bar and after a short while it will be hot alongits entire length. Convection is heat transfer result-ing from mixing of substances with different tem-peratures. Convection can be natural or forced.Natural convection occurs in water as a result ofdensity variation caused by temperature. Water isof maximum density at 4°C. If heated water is sentin below colder water it will move upwards becauseit is less dense and natural convection will takeplace. Forced convection occurs when a mediumsuch as water is exposed to an exterior force, forinstance a pump or mixer.

The power (P) required for heating water is pro-portional to the flow and the temperature, and isgiven by the equation:

7.1 IntroductionIn aquaculture heating of the water may be neces-sary for several reasons, for example to increase thegrowth rate, to get the fish reach a specific size at acertain time, to get them to mature or to spawn. Dif-ferent species have different optimal temperatures;if the ambient water temperature is cooler than theoptimal temperature, it can be useful to heat thewater.

The principles used for heating in aquacultureare normally the same as those used in houses orindustrial facilities; however, systems used in aqua-culture facilities must heat large amounts of waterand therefore be efficient. Important factors whenchoosing a system are the total heating require-ments and the necessary temperature increase. Inthis chapter a survey of methods and equipment isgiven. It starts with some basic physical laws andends with some simple specifications and calculatedexamples.

Instead of purchasing all the heat necessary, itcould be taken from other available sources, such asgeothermic water, or the water could be re-used. Forspecies needing much warmer water than is avail-able from source, both these methods could be used.

In some cases it is necessary to chill the water, forinstance in connection with storing the brood stock,to get the fish to mature, and for storing eggs andfry. Heating and chilling both involve energy transfer.When heating water, energy is added to thesystem, while chilling removes energy from thesystem. In this chapter the focus is on heatingsystems used for aquaculture; much of the basicinformation applies to both heating and cooling. Agreat deal of general engineering literature is avail-

75


P = mcp dt

where:

m = water flow (kg/s)cp = specific heat capacity (kJ/(kg °C))dt = temperature increase for the water (°C).

The specific heat capacity is the amount of energyrequired to heat 1kg water by 1°C: cp = 4.18kJ/(kg °C) for freshwater and 4.0kJ/(kg °C) for seawater. The temperature increase for the water is the difference between inlet and outlet temperatures.

A continuous power supply is required to heatthe flowing water; the unit of power is normally thekilowatt (kW) and 1kW = 1 kilojoule (kJ) persecond. Therefore, it can be seen that power is rateof energy transfer.

ExampleA freshwater flow of 10 l/min (0.17 l/s) is heated fromits original temperature of 2°C to 10°C. What is therate of energy transfer to the water, i.e. the power supplied?

P = mcp dt

The mass of 1 l of water is 1kg.

P = 0.17kg/s × 4.18kJ/(kg °C) × (10°C − 2°C)= 5.7kJ/s= 5.7kW

The total amount of energy that has to suppliedduring a given period or that has been used duringa given period can be calculated from the powersupplied multiplied by the time for which thispower is used.

Q = Pt

where:Q = total amount of energy (kilowatt-hour, kWh)P = power (kW)t = time over which heating takes place (h).

If this is compared to flowing water, the power (kWor kJ/s) corresponds to water flow rate (l/s), whilethe total energy consumed corresponds to the totalamount of water which has flowed past a certainpoint during a given period of time.

It is the energy consumed (Q) that is paid for, beit electricity, oil or another energy source.Electricity is charged per kilowatt-hour, consumed

and oil is sold by volume. The power (P) gives theenergy rating of the heating equipment.

ExampleCalculate the daily cost of heating the water in theprevious example. The price of electricity is 0.1€/kWh.

Q = Pt= 5.7kW × 24h= 136.8kWh

Therefore the cost is

136.8kWh × 0.1€/kWh = 13.68€.

When the water is heated it must always beaerated before it is used on fish because it maybecome supersaturated with gas. When the watertemperature increases the amount of dissolved gasis reduced. If there was equilibrium between thegases in the water and the surrounding air beforethe water was heated, the water will be supersatu-rated with these gases afterwards (Fig. 7.1). Nitro-gen levels will be harmful (see Chapter 8). It is veryimportant to be aware of this, because it means thatwater containing fish must not be heated directly.

7.3 Methods for heating waterSeveral methods are used for heating water, eitherdirectly or indirectly. The latter case requires avail-able sources of hot water that can be used to meetall or a part of the heating requirements. Whenusing direct methods, all necessary energy must be

Figure 7.1 Heating of water that is in gas equilibriumwith the surrounding air (100% saturation) will result insupersaturation with gases that are harmful to the fish.

added to the water. Electricity, oil or gas are theusual sources of energy for direct heating of water.

It is better, however, if other available energysources can be used to meet part or all of the energyrequirement. These can be separated into low andhigh temperature sources. High temperaturesources can be used to meet the entire need forheating, if the amounts are large enough as they aremuch hotter than the required final temperature.Hot industrial effluent water or geothermal waterare examples of such sources.

Low temperature sources are hotter than the rawwater to be heated but their temperature is insuffi-cient to meet total energy needs. The source cantherefore be used as a part supply to the total waterheating requirements. Seawater or groundwater areexamples of sources that can be used to meet partof the energy needs; industrial effluent water canalso be included in this category.

Whether industrial effluent water or geothermalwater can be used directly for the fish, or whetherheat exchange is necessary depends on its quality.Normally it is necessary to use a heat exchanger,because the quality of industrial effluent water isnot continuously good enough for direct use asgrowing water for the fish. Low temperaturesources are also suitable for use with heat pumps(see Section 7.6).

When supplying energy to heat water, it must beutilized as efficiently as possible. Recovery ofenergy from the outlet water is therefore common,either by direct or indirect systems. Re-use systemsare direct, with the water being used again directly,while heat exchangers where only the energy is re-used, represent indirect systems. It is not possible torecover all the energy that is added by using heatexchangers. Even if it were theoretically possible, itis not cost effective because the heat exchangerwould need to be very large and expensive.

7.4 Heaters

7.4.1 Immersion heaters

An immersion heater is an electrical element (heatelement) placed in the water (Fig. 7.2). Electricityis supplied to the element which heats up.The basicprinciple is the same as that is used in a heater in ahouse: the electricity passes through a thin resistantfilament where heat develops. It is important that

the element is sealed to avoid a short circuit; there-fore it is typically placed in a tube. When waterpasses the heated element, energy is transferredfrom the element to the water, the temperature ofwhich increases. Normally a thermostat is attachedto the heater so that heat is transferred until the water reaches a set temperature when the electricity supply is automatically/switched off;when the temperature drops below this value, theelectricity to the element is automatically switchedon. When using an electric heater almost all of theelectric energy supplied is transferred to the water.For greatest efficiency when heating requirementsis vary, it is an advantage to have several elementsin the heater. In addition to a thermostat, the heateris normally equipped with a system to avoid over-heating, for instance when the water flow is reducedor stops. An immersion heater must be correctlyinstalled. There must be no possibility for airpockets inside the heater, because the element mayoverheat in such places. It is important to choosecorrect materials for the heating element to avoidcorrosion resulting in an electrical short circuit.Materials used are acid-proof steel or stainlesssteel, depending on the quality and composition ofthe water to be heated.

The size of heater depends on the energy require-ments.For large water flows or temperature gradients,large heaters are necessary;these consume much elec-tricity as can be inferred from the high cross-sectionalarea of the wires connected to the heater.

ExampleThe ambient water temperature is 3°C while that inthe hatchery is 8°C. The hatchery water flow is 60l/min (1 l/s). Find the size of the heater needed toachieve the required temperature increase.

P = mcpdt= 1 l/s × 4.18kJ/(kg °C) × (8°C − 3°C)= 20.9kW

The necessary size of the heater is 20.9kW. Theheater is ordered from a supplier and adjusted asrequired.

ExampleOn a land-based freshwater production farm thesupply of energy necessary, to increase the watertemperature to at least 12°C throughout the yearmust be calculated. Water consumption is 2m3/min.

Heating and Cooling 77


The incoming freshwater has the following ambientaverage monthly temperatures:

In the months January, February, March, April,May, October, November and December the temperature is so low that it is necessary to heat thewater.Amount of energy required for January is:

P = mcpdt= 33.3 l/s × 4.18kJ/(kg °C) × (12°C − 2°C)= 1392kW

The energy needed for the whole month is

Q = P × number of hours= 1392kW × 24h × 31 days= 1035648kWh

This calculation is repeated for the other monthsgiving the following values:

Figure 7.2 An electric immersionheater.

Month Ambient water temp. Temp. increase (°C) (°C)

Jan 2 10Feb 2 10Mar 3 9Apr 4 8May 8 4Jun 12 0Jul 14 0Aug 14 0Sep 12 0Oct 10 2Nov 6 6Dec 3 9

If the price of the electricity is 0.1€/kWh, the yearlycost for heating the water with an immersion heateris

5859889kWh × 0.1€/kWh = 585989€

As this example shows, the cost of heating water forfish farming is very high because of the large amountof water. This is also the reason for the great interestin alternative hot water sources and recovery ofenergy from the outlet water.

7.4.2 Oil and gas burners

When oil burns a great deal of energy is released,and in an oil burner this energy is used to heat water.The amount of energy released when burning 1 l oildepends on the characteristics and quality of the oil;a typical value is 41800kJ/kg. Letting the oil burn ina combustion chamber around which water flowsensures transfer of energy from the burning oil tothe water (Fig. 7.3).To keep the oil burning, air mustbe supplied to the combustion chamber and achimney is necessary to get rid of the flue gas.Therewill always be energy losses in an oil-fired boilerfrom the flue gas and due to incomplete burning andincomplete transfer of heat to the water. Dependingon the system, oil-fired boilers are usually between60 and 90% efficient. A shunt valve may be used toregulate the temperature of the water flowing outfrom an oil-fired boiler.

Instead of oil, gas can be used as an energysource. The construction of a gas-fired system is

similar to an oil-fired system, with a combustionchamber around which the water circulates.However, the gas-fired boiler is slightly simpler toconstruct, because gas is more flammable than oil.Normally gas-fired appliances are slightly more effi-cient than oil-fired boilers.

7.5 Heat exchangers

7.5.1 Why use heat exchangers?

A heat exchanger is used to transfer energy fromone medium to another and to achieve this a tem-perature gradient is necessary. Energy transfer cantake place via direct contact between the two


Power Total electricity consumptionMonth (kW) per month (kWh)

January 1392 1035648February 1392 935424March 1253 932232April 1114 802080May 557 414241June 0 0July 0 0August 0 0September 0 0October 278 206832November 835 601200December 1253 932232Total 5859889

Figure 7.3 An oil-fired boiler.


media, or indirectly where the media are separatedby a barrier. In indirect exchangers the heat mustbe transferred through this barrier.

An illustration of the direct method is as follows.Water is fed in at the top of a tower where is itspread on a perforated plate; it falls through asdrops which pass through the air into a basin placedbelow. The water drops have a large surface areaexposed to the surrounding air; if this is warm,energy will be transferred from the air into thedrops and the water will be heated. The indirectmethod is illustrated by having water of differenttemperatures separated by a thin metal plate.Energy will be transferred through the plate fromthe warm side to the cold side because of the tem-perature gradient until the temperature on bothsides is the same. This method is the main one usedin fish farming.

Heat exchangers can be used both for heatingand chilling. In heat pumps and cold-storage plants,a refrigerant may be used as one of the media (seeSections 7.6 and 7.8). Heat exchangers may also beair to air as in ventilation systems in houses, or airto water as in air coolers.

Heat exchangers can be used to recover energyfrom the outlet water; for example, to recover theenergy from hot industrial effluent water and use it to heat the water supply to a fish farm.Heat exchangers are also commonly used torecover the energy from the outlet water from the fish farm and transfer it to the inlet water.Seawater may also be used as a heat source on one side of the heat exchanger, to supply part of the heat energy to the inlet water supply to a freshwater fish farm.

7.5.2 How is the heat transferred?

The energy transfer in a heat exchanger havingliquid on both sides of a metal plate occurs in threestages (Fig. 7.4):

(1) Transfer of heat from the liquid with thehighest temperature to the fixed material, theplate

(2) Transfer of heat through the fixed material(3) Transfer of heat from the fixed material to the

liquid of lower temperature

Transfer of heat from the liquid to the fixed mate-rial is normally via convection and conduction.

Close to the fixed material there will be a layer withlaminar flow through which heat is transferred byconduction more slowly than for convection. It isadvantageous to reduce the thickness of thelaminar layer and, to improve convection, have tur-bulent conditions in the flowing liquids. Heat transfer through the fixed material is by conduc-tion; a material with good conductivity (e.g. metal)improves this. Conduction and convection causeheat transfer into the liquid on the other side of thefixed material.

7.5.3 Factors affecting heat transfer

The rate of energy transfer (P) of the heatexchanger depends on the temperature differencebetween the media, the thermal conductivity of thematerial in the heat exchanger and the area overwhich the energy is transferred; it can be calculatedfrom the equation

P = kALMTD

where:

k = heat transfer coefficient (W/(m2 °C))A = heat transfer area (m2)LMTD = log mean temperature difference (°C).

The value of k gives the quantity of energy trans-ferred per square metre surface area and degreetemperature difference. Various factors affect k; inpractice, values of up to 8kW/(m2 °C) are achieved.k can be calculated from the following equation:

Figure 7.4 In a heat exchanger the heat is transferredfrom the hot to the cold side.

Here, a1 and a2 are the heat transfer coefficients oneach side of the material in the heat exchanger.They give the quantity of energy transferred froma liquid or gas to or from unit area of a fixed mate-rial and per degree temperature difference. Thevalues depend on the condition for convection andconduction and can be improved by optimizingoperational conditions.

tp is the thickness of the material that separatesthe two flowing media; increasing tp will decrease kbecause the heat must be transported a greater dis-tance through the fixed material. Use of a thin fixedmaterial results in a low k value.

l is the heat convection factor for the material;steel and other metals have a high factor, whileglass and plastics have a lower factor. This whymetal with good conductivity is used for the heattransfer plates.

Rf is the fouling factor, which gives the amountof fouling on the material of the heat transferplates. Fouling reduces k, and therefore the rate ofenergy transfer will be reduced. High turbulenceclose to the surface of the heat transfer plates willreduce the amount of fouling, in addition toimproving a1 and a2. Cleaning of the exchange sur-faces will also reduce the fouling, Rf. The reductionin k is because the conductivity of the layer offouling is low and the thickness of the transfermaterial is increased.

Manufacturers of heat exchangers normally givekA as a single value. This is because each manufac-turer will have their own design for the heat transfer area to create optimum flow with turbu-lence; to improve conditions for turbulence andincrease the heat transfer area, various patterns areused on the exchange surface, including groovesand corrugations.

The temperature gradient between the warm andthe cold side in the exchanger (Fig. 7.5) ensuresenergy transfer. It is expressed as the LMTD.A log-arithmetic expression is used because temperatureequalization between the media through theexchanger may not be linear. LMTD can be calcu-lated from the following equation:

LMTD =Δ Δ

Δ ΔT T

T T1 2

1

−( )ln /

1 1 1

1 2kt

R= + + +a a l

pf

where:

T1 = tin (hot water) − tout (heated water)T2 = tout (cooled water) − tin (water to be heated).

If the amount of water and the heat exchange areaare the same on both sides of the heat exchanger,ΔT1 = ΔT2 = LMTD. LMTD will vary depending onwhether it is a parallel-flow or counterflowexchanger (see sections 7.5.4 and 7.5.6).

7.5.4 Important parameters when calculating thesize of heat exchangers

Number of transfer units

Number of transfer units (NTU) indicates howmuch energy can be transferred per unit LMTD.Indirectly, it gives an idea of what the exchangerwill look like, its size, etc. The following calculationcan be used to find the NTU of the exchanger:

NTU = (tin − tout)/LMTD

where:NTU = number of transfer unitstin = temperature of water flowing into the

exchangertout = temperature of water flowing out of the

exchangerLMTD = log mean temperature difference (°C).


Figure 7.5 Log mean temperature difference (LMTD)represents the mean temperature difference betweenthe warm and cold side in a heat exchanger. It is thisgradient that ensures heat transfer.


ExampleCalculate the NTU for the warm side in a heatexchanger, i.e. the side where the hot water is chilled.Water flows into the exchanger at 1000 l/min at a tem-perature of 10°C and out at 4°C; the LMTD is 1.5°C.

The NTU can be calculated both for the warm andcold side of the exchanger. Depending on the flowconditions and the transfer area, it could be thesame (if the flow and heat transfer area are thesame on both sides of the exchanger).

In a plate exchanger described in section 7.5.5 the NTU is seldom above 5, or the plates will beunreasonably large. If the LMTD is estimated andthe NTU known, the highest possible out tempera-ture that can be reached can be calculated. If NTUis above 5, heat exchangers can be connected inseries, or several-stroke exchangers can be used.

The design of the heat exchange surface areavaries. A closed design results in a large NTU withgood heat transfer but a high head loss, whereas amore open design results in a lower NTU.

Specific pressure drop

The pressure drop in the liquid flowing through theexchanger is necessary to achieve heat transfer. If ahigh head loss through the exchanger is accepted,the size of the exchanger can be reduced. Higherpressure creates more turbulence and improvedcontact between the exchange plates, but a higherinput pressure to the exchanger is necessary.

The specific pressure drop (J) for a heatexchanger gives the pressure loss for every transferunit, and can be represented by the following equation:

where:

J = specific pressure dropΔP = total head loss through the exchangerNTU = number of transfer units.

The economic optimum head loss through the heatexchanger varies depending on the situation and

JP= Δ

NTU

NTU10 4

1.54

= −

=

has to be calculated in every case. Normally it liesin the range 2–10mH2O per NTU for exchangersused in fish farming. When looking at the differentareas where exchangers are used, the followingapproximate specific pressure drop per transferunit can be used as an approximate base to start asimulation process to optimize the size of a plateheat exchanger:

• Seawater exchanger 5–10mH20• Outlet water exchanger 10–12mH20• Evaporator in a heat pump 3mH20• Condenser in a heat pump 4mH20

Other important measurements

The maximum pressure (design pressure) in heatexchangers used in fish farming is normally in therange 1–2MPa or 10–20mH2O. Usually the plateexchangers have up to 2000m2 of exchange surfacearea and the water flow can be 1000 l/s or more.

Exchanger price of course depends on size: alarge exchanger will cost more than a small one. Ifhigher head loss can be accepted the size of theexchanger can be reduced. Also the size of theLMTD is of great importance. If a small LMTD isneeded the size of the exchanger will increase (Fig. 7.6).

When installing a heat exchanger, the energyprofit must therefore always be evaluated againstthe price of the exchanger. To have a low LMTD isnormally very costly.The necessary transfer area, or

Figure 7.6 If the log mean temperature difference(LMTD) is small, the size of the heat transfer area mustbe increased.

size of the exchanger increases when the LMTD isreduced. Suppliers of equipment normally havetheir own programs to simulate the temperature ofwater discharged (their own kA values) to optimizethe size of their exchangers.

7.5.5 Types of heat exchanger

When designing a heat exchanger the aim is tocreate a large area where exchange of energybetween the two media can take place. Two typesof heat exchangers are common in fish farming:plate, and shell and tube exchangers.

Plate exchangers

A common plate exchanger consists of the follow-ing components: a rack, two end-plates with pipeconnection in one or both sides, the heat transferplates and gaskets that are used between the platesto avoid leakage (Fig. 7.7). In plate exchangers thatcannot be dismantled, it is possible to omit thegaskets and braze or solder the plates togetherinstead. The two media flow into the exchanger atone end; the hot and cold water flow in two sepa-rate circuits divided by the heat transfer plates. Hotand cold water will flow in parallel through thewhole length of the exchanger, assuming the coun-tercurrent principle is used as is usual. The size ofthe energy transfer area can be changed by addingand removing heat transfer plates as long as thereis sufficient space in the rack. The exchanger can beopened to add or remove plates and for cleaningunless the parts are brazed together. The formertype is normally used in fish farming because theyare easy to open for cleaning and removal offouling. The latter type can be used as evaporatorsand condensers in a heat pump, or in places whereboth the flowing liquids are pure and fouling of theheat transfer surfaces does not occur.

Some kind of corrugation on the heat transferplates is normal (Fig. 7.7). This increases the areawere heat transfer occurs and hence the NTU. Inaddition the corrugation will increase the turbu-lence and because of this improve the heat transfer, so increasing the value of k. The gasketbetween the plates which inhibits leakage, is eitherglued or clipped to the plates and made of a resis-tant rubber material.

To increase the heat transfer area, several heatexchangers can be connected in series, one afteranother. The alternative is to use the so called‘several-stroke’ exchanger. Normally an exchangerhas one stroke, but by adding dense plates in themiddle where the direction of the water flow ischanged, a several-stroke exchanger is achieved(Fig. 7.8). The result is the same as adding twoexchangers in series, but only one rack is used.Whether a plate exchanger is a one-stroke orseveral-stroke type can easily be seen from theconnection points. A one-stroke exchanger has allpipe connections on one side to, one of the end-plates, while several-stroke exchangers have pipeconnections on both sides to, both end-plates. To beable to dismantle a two-stroke exchanger, e.g. forcleaning, it must be possible to disconnect the pipeconnections in a simple way.

The value of k for plate exchangers varies signif-icantly depending on the corrugation pattern on thesurface; normal values are from 3.5–8kW/(m2 °C).The plates are very thin, in the region of 0.5mm.

Shell and tube exchangers

Another common type of heat exchanger is theshell and tube.This is widely used in condensers andevaporators in heat pumps and cold storage plantsin fish farming. It is constructed with a shell cover-ing a number of small tubes (Fig. 7.9). The shell isnormally a large tube. One medium flows in thesmall tubes while the other flows around the tubes,in the shell. The small pipes and the large pipe con-stitute the two different circuits where the two dif-ferent media flow and heat can be transferred. Shelland tube exchangers are seldom used for tradi-tional water-to-water heat exchange in fish farmingbecause they are quite difficult to clean manually.They can be opened by removing the cap at one endof the shell, but it is impossible to reach all heattransfer surfaces for simple manual brushing. Theymust be adapted for automatic chemical cleaning.

Special types, pipes in seawater

A pipe that is laid down in water will have a higheror lower surrounding temperature than the waterflowing through the pipe and will function as a heatexchanger. This principle has been used in fishfarming. If the temperature in the surrounding



water is higher, heat will be transferred into thewater flowing inside the pipe.To obtain a noticeabletemperature increase in the water flowing insidethe pipe, the following factors are of importance:length of the pipe, temperature difference betweenthe water inside the pipe and around the pipe, andthe heat transfer coefficient (k) of the pipe mater-ial. This again depends on the thickness of the pipe,material of which the pipe is constructed, contactarea between water and the pipe material on bothsides, and the convection conditions inside andaround the pipe.

In practice, plastic seems to be the material mostoften used. Polyethylene does not, however, trans-fer energy particularly effectively as it has, quite alow k value, but PE pipes are cheap and easy to lay.

A

B

Figure 7.7 A typical plate heat exchanger consists of a rack, two end-plates with pipe connections, heat transferplates and gaskets between the plates to avoid leakage. (A) An exchanger in situ. (B) An open heat exchangershowing the corrugated surfaces of the transfer plates.

Figure 7.8 Multi-pass exchangers include one or more dense plates to change the direction of the waterflow inside. In a multi-pass exchanger there are pipeconnections at both ends of the rack.

The pipes must be moored; normally concreteblocks are used. If in seawater, it is normally to laypipes at depths below 20–30m, too avoid the mostcritical depth that results in much fouling of theexterior surface. In fish farming the system may beused for heating freshwater in the winter. In thiscase the inlet pipeline must be laid into seawaterwhere the temperature is normally higher than infreshwater. A substitute for a traditional on-shoreheat exchanger where seawater is pumped through,is then achieved. What is found in practice is thatthe pipe has to be several hundred metres long to achieve a noticeable temperature increase.Depending on the site conditions and the geo-graphical distance between the site and the watersource, this might be a good solution if the watertransferred from the source to the fish farm is nearthe sea.

7.5.6 Flow pattern in heat exchangers

Two flow principles are used in heat exchangers:with-current or countercurrent, the latter only inplate exchangers of particular construction (Fig.7.10). In the with-current system, the liquids onboth sides of the exchange material flow in thesame direction, or approximately in the same direc-tion as in the shell and tube exchanger. In this casethe temperature gradient between the media is highat the start but gradually decreases. The highestpossible temperature that can be achieved in thecold liquid being heated is the mean temperaturebetween the two flowing liquids.This requires equalflow of both liquids; otherwise the temperaturedepends on the flow ratio of the two liquids. In acountercurrent exchanger the cold media flows in

the opposite direction to the hot media. The tem-perature of the cold media thus gradually increasesand the hot water is correspondingly chilled. Witha countercurrent exchanger, the temperature of thecold media can be raised to almost the temperature


Figure 7.9 The shell and tubeexchanger is constructed with a shellcovering a number of small pipes.

Figure 7.10 The flow pattern in a plate heat exchangercan be either uniflow or countercurrent.


of the hot media, depending of the size of theexchanger.

ExampleFind the LMTD for a countercurrent heat exchangerhaving the following temperatures: hot water into theexchanger, t1 = 11°C; hot water out of the exchanger,t2 = 5°C; cold water into the exchanger, t3 = 3°C; coldwater out of the exchanger, t4 = 7°C.

DT1 = t1 − t4

= 11 − 7 = 4DT2 = t2 − t3

= 5 − 3 = 2

With parallel/linear temperature equalizing theanswer would be slightly higher at 3.0. The way toachieve the conditions shown in the example is tohave different media flow rates on the two sides ofthe transfer plates. If there are no energy losses to thesurroundings and the water flow in both circuits isequal T1 will equal T2 and this value can then be usedfor further calculations.

7.5.7 Materials in heat exchangers

Heat exchangers are built of different materials for different purposes. It is important that the material has good conductivity, such as a metal,and that the heat transfer surfaces are corrosion-resistant. In fish farming, it is also important thatthe materials do not release substances that aretoxic for the fish, e.g. copper. The more contami-nated the flowing media are, either by acids orbases, the more resistant to corrosion the heattransfer surface areas need to be; however, this willincrease the price of the material.

If the media are water and it is uncontaminated,stainless steel could be used. If the water is conta-minated to some extent, for instance with humussubstances, acid-proof steel at least should be used.If the water is salt or brackish, titanium coveredplates ought to be used. For ground water titan-covered plates are also recommended; the price ofthese is, however, much higher than for stainlesssteel plates.

LMTD4 2

ln2.89= −

( )=

4 2/

7.5.8 Fouling

Fouling is a problem with the use of heat exchang-ers in fish farming. This occurs particularly whenusing outlet water on one side in the exchanger, butthere are also problems associated with using sea-water on one side. Fouling of the transfer surfaceswill reduce the value of k and the heat transfer.Fouling will cover the transfer surfaces, and theconduction through the layers of fouling is dra-matically reduced compared to surfaces with nofouling. Since fouling occurs normally and it isimpossible to remove it continuously, this must betaken into consideration when designing heatexchangers for use in fish farming. This is done byincluding the fouling factor (Rf) in the calculationswhich again decreases the value of k (see section7.5.3).

It is important to reduce the amount of foulingas much as possible, but of course within economiclimits. Washing of the surfaces will reduce thefouling, whether by the use of chemicals, manualbrushing, or both. To make chemical washing pos-sible in a simple way, the exchangers should beequipped with a backwashing circuit (Fig. 7.11).For the washing procedure the heating system isstopped and the exchanger backwashed severaltimes with water containing a detergent: caustic

Figure 7.11 To remove fouling, exchangers should beequipped with a washing circuit which is used to back-wash the system using detergents to dissolve fat.

soda or lye may be used as a detergent in fishfarming. Chemical washing, however, is normallynot enough and the exchangers have to be openedand the heat transfer surfaces brushed manually. Itmay not be necessary to do this every time, but onlyfor some of the washings, for example once a week.This of course depends on the degree of foulingwhich varies with the characteristics of the waterflowing through the exchanger, i.e. whether it is newwater or outlet water.

When cleaning the exchangers the heating sys-tem must be switched off. Because of this, it can be advantageous to have at least two exchangers,so at least half the heating capacity is functioningduring the cleaning procedure. It is important tohave enough valves in the pipelines to be able to change the flow direction to permit this (seeChapter 2).

The degree of fouling decreases with increasedwater velocity through the exchanger, becausemore turbulence is created. However, this will alsoincreases the head loss, so there is a balance tofinding the optimal water velocity. What is cer-tain is that a low velocity through the exchangerincreases the degree of fouling; this will occur if the water flow through the exchanger is reducedcompared to what it is designed for; in the worstcases total blockage of the exchanger can result.The normal water velocity through exchangers isaround 2m/s.

Fungus clots and larger particles may block theexchanger totally. To reduce the risk of blockage,for instance when using outlet water, the watermust always be filtered before entering a heatexchanger. This will to some extent, also help toreduce the amount of fouling. A particle removalfilter, for instance a rotating screen filter, is com-monly used for this purpose.

7.6 Heat pumps

7.6.1 Why use heat pumps?

The use of heat pumps is beneficial in many aqua-culture facilities because the temperature increaseis relatively low and the amount of energy that canbe transferred is quite large. Today heat pumps arefinding increasing use in all type of industry, ingreenhouses and housing, as a result of the high costof electricity and oil.

The great advantage with a heat pump is that thelarge amount of low temperature energy availableis transferred to smaller amounts of media withhigher temperature. The energy transfer betweenthe two conditions and the equipment itself mustbe paid for.

A heat pump is basically the same as a refrig-erator or a cold-storage plant. The difference is themethod of utilization. A refrigerator is used toremove energy while the heat pump is used to addenergy. Theoretically a heat pump and a refrigera-tor can be the same unit that is used both forcooling and heating, for instance cooling of thewater supplied to spawning fish and heating ofwater supplied to fry production. It is however dif-ficult to optimize the heat pump for both purposes.

7.6.2 Construction and function of a heat pump

A heat pump consists of four main components(Fig. 7.12):

• Evaporator• Compressor• Condenser• Expansion valve.

Between the four components there is a closedpipeline, the transport circuit, in which the workingmedium or refrigerant circulates. The medium isadapted so that it performs a phase transfer: it willchange phase between liquid and gas when circu-lating between the components in the circuit. Thesystem utilizes the energy needed for evaporationof the working medium, which is released when themedium is condensed.

To explain the construction and function of theheat pump or refrigerator, the working medium canbe followed for one lap round the closed pipelinecircuit. When the medium enters the evaporator, itis a liquid of low temperature and with relativelylow pressure. The boiling point of the medium isquite low. In the evaporator, which is actually a heatexchanger, the temperature is higher than in theworking medium. Energy is therefore transferredfrom the surroundings and into the workingmedium, i.e. heat exchange occurs.The temperatureincreases up to the evaporation point for theworking medium which starts to change phase fromliquid to gas. When a medium changes from liquidto gas much energy is needed which is stored in



the gas. The liquid–gas mix is then sucked into thecompressor where the pressure increases togetherwith the temperature so that all the medium istransformed into gas. The pressure in the workingmedium is now much higher than before.

The compressor is supplied with additionalenergy (normally electricity) to function. Availablecompressors either use a piston principle or thescrew principle, the latter having continuous deliv-ery of gas. Not all the energy supplied to the com-pressor is transferred to the gas, because there aresome energy losses. The efficiency of a compressoris normally about 70–80%, meaning that 70–80% ofthe added electricity is transferred to the gas.

From the compressor the gas is pushed into thecondenser with the help of the created pressure.The condenser is another heat exchanger whereenergy is transferred from the gas, which has thehighest temperature, into the surrounding water.The gas is now chilled and it reaches a temperature

where phase transfer occurs again, the dew point,and changes into liquid. The evaporation energythat has been stored in the gas is now released andtransferred to the cold medium. The gas–liquid mix(or just the liquid, depending on the constructionand conditions) now flows to the expansion valvewhere the pressure is reduced, for instance byincreasing the cross-sectional area of the pipewhere the medium is flowing. The gas–liquid mixnow expands. (A simplification of this is that theliquid is pressed through a small hole and out intoa pipe with larger diameter.) All the medium nowchanges into liquid. The pressure decrease of themedium is accompanied by a drop in temperature.The working medium now enters the evaporatoragain and can then start a new lap round the circuit;thus it circulates continuously.

The working medium or refrigerant must beadapted to the temperature and pressure condi-tions in the evaporator and condenser. It must

Figure 7.12 The main componentsand function of a heat pump.

evaporate and condense at temperatures that fit thetemperature to which the inlet water is to beheated, and of the water from which the heat is collected. Neither must the pressure and pressuredifference be too large. There has been much discussion about refrigerants, because the mostsuitable have negative effects on the environmentby contributing to the greenhouse effect. New moreenvironmentally friendly refrigerants have there-fore been developed during the past few years. Infish farming ammonia is much used, which is a rel-atively environmentally friendly material regardingthe greenhouse effect but highly toxic for the fish,so it is important to avoid leakage of refrigerantinto the inlet water.

7.6.3 Log pressure–enthalpy (p–H)

The heat pump process is often illustrated in a logpressure (p)–enthalpy (H) diagram (log p–Hdiagram) (Fig. 7.13). Enthalpy is a parameter thatis a measure of the energy content of refrigerant;units are kJ/kg. The log p–H diagram illustratesclearly the changes of phase in the refrigerant. Thepressure of the refrigerant is constant through theevaporator, but its energy content (H) increasesbecause it gradually changes phase from liquid togas and in doing so takes up energy from its sur-roundings. When the refrigerant enters the com-pressor it is in the gas phase. Electric energy issupplied to the compressor, most of which is trans-

ferred to the refrigerant, so further increasing itsenthalpy. In the compressor the pressure of the gasincreases and it is then fed into the condenser. Hereenergy is released because the gas changes phase to liquid and the enthalpy drops, but the pressureremains stable: all the energy that was stored whenthe liquid changed phase into gas is now released.The refrigerant exits the condenser as liquid andthen enters the expansion valve where its pressuredrops, but no energy is removed or added (assum-ing ideal conditions). Therefore the enthalpy is thesame, as shown by a vertical line.

The p–H diagram is specific for each medium,and shows the phase of the medium in relation toits pressure and enthalpy content. It is also knownas a nose diagram, depending on the presentation.On one side of the ‘nose’ the phase of the workingmedium is liquid; on the other side it is gas, and in between both gas and liquid. The heat pumpprocesses therefore happen inside the ‘nose’,because it is here that the phase transfers occur.

7.6.4 Coefficient of performance

The coefficient of performance (COP) is an impor-tant parameter when talking about heat pumps andheating systems; it indicates how much energy istransferred to the water in relation to the supply ofenergy (normally electricity). For heat pumps the only electricity supplied is to the compressor.The COP (e) will therefore be the relation between


Figure 7.13 The heat pump processillustrated in a log p–H diagram.


the energy released in the condenser and theenergy added to the compressor.

e = Qreleased in condenser/Qadded to compressor

Normally for commercial heat pumps used in fishfarming, COP can be up to 5.

7.6.5 Installations of heat pumps

When installing a heat pump location of the con-denser is very important, because it is here that theenergy is released. For a refrigeration plant locationof the evaporator will be important because hereremoval of energy occurs. There are a number ofplaces for installing the condenser in the heat pumpin a fish farm; all have advantages and disadvan-tages (Fig. 7.14). One solution is to place it directlyinto the inlet water, which gives very good heattransfer. However, leakage from the condensercould lead to contamination of the inlet water bythe refrigerant, which is dangerous because it couldbe toxic to the fish. To avoid this, an extra closedcircuit is more commonly used between the con-denser and the inlet water consisting of a pump anda heat exchanger through which a non-toxic liquid(water or glycol, for instance) circulates. Glycol has a high thermal capacity so is a good choice of medium. Between 10 and 15% reduction in theCOP is normal when using the extra circuit,because the heat is transferred twice.

Another method is to install the condenser in theoutlet water and afterwards have a heat exchangerbetween the outlet water and the inlet water, alsoa two-step heat transfer process. The disadvantagewith this method is that increased fouling willpresent problems. Cleaning of the condenser, whichis normally of the shell and tube type, is difficult.

The evaporator is normally placed in the outletwater. Here there will also be problems withfouling, and a closed circuit can also be used toadvantage in this situation. A plate exchanger isused in this circuit to transfer the heat, because it iseasier to open for cleaning.

When using the evaporator in the outlet waterand the temperatures are low freezing of the outletwater may be a problem because the temperatureis reduced as much as possible through the evapo-rator to recover the stored energy in the outletwater, meaning that the temperature in the workingmedium is 0°C or less to ensure effective heat trans-

fer. It is therefore possible that the water in theevaporator will freeze, for instance if there is areduction in the water flow or drop in the inletwater temperature.To avoid breakage of a shell andtube evaporator due to ice, it is normal to use glycolwhich functions in the same way here as in the con-denser circuit; glycol is a liquid with a very lowfreezing point.

Figure 7.14 There are several alternatives forinstalling the condenser when using a heat pump. Eitherit can be placed directly in the inlet water, or a circula-tion circuit including a heat exchanger can be used.

In practice, there is not usually enough energy inthe outlet water to get a heat pump to function witha good COP. Therefore the heat pump will usuallybe an integral part of a total energy system whereadditional energy from seawater or another lowtemperature source may be used. Energy may alsobe added directly, for instance by using an immer-sion heater; this will of course require a muchsmaller heater compared to using an immersionheater alone and not in combination with a heatpump.

ExampleInlet water to a fish farm is heated from 2 to 8°C bytransfer of energy from the condenser. The waterflow is 500 l/min (8.33kg/s). Fifty kW of electricenergy is supplied to the compressor. Find the COPfor the heat pump.

Start by calculating the total energy input to theinlet water:

P = mcp dt= 8.33kg/s × 4.2kJ/(kg °C) × 6°C= 210kJ/s= 210kW

The COP is therefore

e = 210/50= 4.2

This means that for every kW of electric energy thatis supplied to the compressor, the water is heated by4.2kW.

7.6.6 Management and maintenance ofheat pumps

Fouling is always a problem when using heat pumpsin fish farming. Plate exchangers with glycol circuits,as mentioned earlier, can therefore be advanta-geous because they are easier to clean and dis-mantle. Chemical washing circuits should also beinstalled.

Normally, heat pumps are thermostatically con-trolled to prevent the water freezing and damagingthe evaporator. If the temperature falls too low,circulation of refrigerant is stopped allowing anyice that has formed to melt. A back-up heatingsystem might be installed to safeguard againstfreezing but is expensive. Alternatively the watercan be pre-warmed by using another heat source,

such as groundwater, before it enters the heatpump.

7.7 Composite heating systemsA composite heating system is normally used to heat water for use in fish farming. The systemcomprises several components that all have someheating effect on the inlet water. Usually there areone or several heat exchangers in addition to eithera heat pump on large farms, or a heater or an oilburner on smaller farms. The COP is calculated forthe entire heating system and are usually in therange 15–25, which means that for each kW of electric energy supplied, the water is heated by15–25kW. Examples given below include heaters,heat pumps and heat exchangers to illustrate the profitability of using a composite heatingsystem.

ExampleHeater and heat exchanger (Fig. 7.15)

Calculate the profitability of adding a heatexchanger compared with using only an electricimmersion heater in a small heating system. A waterflow of 180 l/min (3 l/s) is to be heated from 4 to 8°C.The first calculation is for an electric heater alone.Size of the heater:

P = mcp dt= 3 l/s × 4.18kJ/(kg°C) × (8 − 4°C)= 50.2kJ/s= 50.2kW

The daily cost of using this system with an electric-ity price of 0.1 €/kWh is:

50.2kW × 24h × 0.1 €/kWh = 120.5 €

Now a heat exchanger is included in the circuit torecover the energy in the outlet water; 75% recoveryis quite normal. This value of course depends on thecost of the heater, heat exchanger and electricity; simu-lations should be done to find the most economicalcombination. If 75% of the total heat increase above4°C is provided by the heat exchanger, 3°C of thetemperature rise results from its use. This gives the following temperatures in the heat exchanger asthe same amount of water flows on both sides:

Water entering heat exchanger 4°CWater leaving heat exchanger 7°C



Water entering heater 7°CWater leaving heater 8°C

A smaller heater is therefore needed as the water isonly going to be heated from 7 to 8°C:

P = 3 × 4.18 × (8 − 7)= 12.5kW

This gives the following new daily running costs:

12.5 × 24 × 0.1 = 30.0 €

As can be seen, the daily cost of heating is reducedfrom 120.5 to 30.0 €, a saving of 90.5 € per day, byadding a heat exchanger. This clearly illustrates theadvantage of using a heat exchanger.

The necessary size of the exchanger will now befound.

Since the same amount of water is circulating onboth sides of the heat exchanger, it has the followingtemperature programme:

Water entering heat exchanger 4°CWater leaving heat exchanger 7°CWater entering heater 7°CWater leaving heater 8°C

The energy to be transferred from the warm to thecold side in the heat exchanger given by

P = mcp dt= 3 × 4.18 × 3= 37.6kJ/s

For a heat exchanger the following equation applies:

P = kALMTD

Figure 7.15 A heating system with a heater and a heatexchanger.

The k value for the plates in the exchanger is set to6kW/(m2 °C) and the LMTD (temperature differ-ence that drives the heat transfer) is 1.0°C, whichgives the following area:

A = P/(kLMTD)= 37.6/(6 × 1.0)= 6.3m2

Assuming a plate size of width 0.4m and height 1m,the area is 0.4m2 (this depends on the size of platesupplied). The number of plates required = 6.3/0.4 =15.8; this is an exchanger with 16 plates.

ExampleHeater with heat exchangers in outlet water and inseawater (Fig. 7.16).

The inlet water has a temperature of 2°C and this isincreased to 4°C when the water passes the seawaterexchanger.The water then enters an outlet exchangerwhere the temperature is further increased to 9°C.The last increment up to 10°C, which is the temper-ature in the fish tank, is supplied through an electricheater. Calculate the COP of the system.

e = Qdeliverd/Qsupplied

= (ttank − traw water)/(ttank − tbefore heater)= (10 − 2)/(10 − 9)= 8

The COP is 8 for this system, so that for every kWof electric energy supplied 8kW is supplied to theinlet water.

ExampleHeat pump and heat exchanger (Fig. 7.17)

Find the profitability of installing a heat pump, com-pared with a total energy system including a heat

pump and heat exchangers.A water flow of 300 l/min(50 l/s) is to be heated from 2 to 8°C. The heat pumphas a COP of 5. How much electric energy must besupplied?

First, the total amount of energy that has to be trans-ferred to the water is calculated.

P = mcp dt= 50 l/s × 4.18kJ/(kg°C) × (8 − 2°C)= 1254kJ/s= 1254kW

The COP is 5, meaning that the amount of energytransferred to the compressor is

1254/5 = 250.8kW

The daily cost of using a heat pump with an elec-tricity price of 0.1 €/kWh is therefore

250.8kW × 24h × 0.1€/kWh = 601.9€

Now a heat exchanger that recovers the energy inthe outlet water and transfers it to the inlet water isadded to the circuit. The heat exchanger is assumedto meet 75% of the heating requirement. Of the totaltemperature increase of 6°C, 75% is provided by theheat exchanger, i.e. 4.5°C, which gives the followingtemperatures in the heat exchanger:

Water entering heat exchanger 2.0°CWater leaving heat exchanger 6.5°CWater entering heat pump 6.5°CWater leaving heat pump 8.0°C

The new size of the heat pump can now be calculated:

P = 50 × 4.18 × (8 − 6.5)= 313.5kW


Figure 7.16 A heating system with aheater and heat exchangers both in theoutlet water and to seawater.


With a COP of 5, the amount of energy that must besupplied to the compressor is 313.5/5 = 62.7kW.Therefore new daily costs are:

62.7 × 24 × 0.1 = 150.7€

As can be seen the daily cost of heating is reducedfrom 601.9 € to 150.7 €, a saving of 451.2 € per day,by using a heat exchanger in addition to the heatpump.This illustrates how useful it is to utilize a heatexchanger together with a heat pump. Heat pumpsare nearly always used together with one or severalheat exchangers because of the large reduction inenergy costs compared to the investment costs ofheat exchangers.

The overall COP can now be calculated:

e = total energy transferred to the water/electric energy supplied

= 1254kW/62.7kW= 20.0

This means that of the total energy transferred to thewater, only 1/20 (5%) is supplied as electric energy.

7.8 Chilling of waterWhen water is to be chilled, energy has to beremoved; the amount is the same that has to be

added when heating for a given temperature dif-ference, so of course the same equation can be used.

P = mcpdt

where:

m = water flow (kg/s)cp = specific heat capacity (kJ/(kg°C))dt = temperature decrease for the water (°C),

i.e. the difference between inlet and outlettemperatures.

ExampleA water flow of 60 l/min is to be chilled from 4°C to2°C. Find the rate of energy removal.

P = 1kg/s × 4.18kJ/(kg °C) × (4°C − 2°C)= 8.36kJ/s= 8.4kW

Heat exchange can be used if there is an avail-able water source that could be used as a chillingmedium. For instance, freshwater can be chilledusing bottom water from the sea during the summerwhen freshwater is warmer. Direct mixing of colderwater may also be used to avoid the use of a heat exchanger, but requires satisfactory waterquality.

Figure 7.17 A heating system with aheat pump and a heat exchanger.

Alternatively, ice may be added directly to thewater, or one circuit of a heat exchanger can pass through a basin to which ice is added. If thismethod is used, it is necessary to have an icemachine on the farm; however, this is not very satisfactory because it is a costly method involvingtwo stages: first production of ice and then chillingof the water. Only when small amounts of water are to be chilled at specific times of the year, is it a viable solution, for example in small slaughterhouses, and where ice is available for chilling the fish.

A cooling plant is therefore used to chill thewater.This is basically the same as a heat pump, butis optimized in another way (Fig. 7.18). The evapo-rator is placed in the water or liquid to be chilled.When the working medium or refrigerant evapo-rates, energy is taken from the water or liquid, thetemperature of which falls. This is made possible bychoosing a suitable refrigerant (working medium).After this the working medium goes into the com-pressor and then on to the condenser where it condenses, releasing heat to the surroundings. Acooling plant often contains an air–liquid heatexchanger (the condenser) so the heat is releaseddirectly to the air. This is the same principle that isused in a refrigerator where the energy is alsoreleased to the air. As seen from the set-up, themain component in a cooling plant is the evapora-

tor that ‘removes’ energy. When talking about theefficiency or COP when a heat pump is used as acooler, it is the relation between the energy (Q) thatis added to the compressor and the energy removedfrom the water in the evaporator that is critical.Theequation to use is therefore:

e = Qremoved in the evaporator/Qadded to compressor

The same circulation circuits that are used on heatpump installations to prevent refrigerant from con-taminating the inlet water and avoid freezing arealso used in cooler units (see Section 7.6.5).

If both chilling and heating need to be performedon a fish farm, it might be possible to utilize thesame heat pump for both purposes. Energy can betaken from the inlet water to be chilled and addedto the inlet water to be heated. This means placingthe evaporator in the water to be chilled and thecondenser in the water to be heated. It will,however, normally be a rather unsatisfactory solution because it is difficult to optimize the heat pump with pressure conditions throughout the evaporator and condenser, and to find goodrefrigerant.

References1. Killinger, J., Killinger, L. (2002) Heating and cooling:

essentials. Goodheart-Wilcox.


Figure 7.18 A water cooling plant.


2. Cengel,Y.A. (1998) Heat transfer: a practical approach.McGraw-Hill.

3. Silberstein, E. (2002) Heat pumps. Thomson DelmarLearning.

4. Moran, M.J., Shapiro, H.N. (2003) Fundamentals ofengineering thermodynamics. John Wiley & Sons.

5. Ibarz, A., Barbosa-Cánovas, G.V. (2003) Units opera-tions in food engineering. CRC Press.

6. Incropera, F.P., DeWitt, D.P. (2001) Introduction to heattransfer. John Wiley & Sons.

8Aeration and Oxygenation

at a given temperature, it is less than saturated(<100%); if the water contains more gas than whenfully saturated, it is super-saturated (>100%).

The percentage saturation can be calculated bydividing the measured concentration (Cm) by theconcentration at saturation (Cs):

Percentage saturation = (Cm/Cs) × 100

All the gases in the atmosphere can be dissolvedin water; the sum of the partial pressures of all thedissolved gases is known as the total gas pressure(TGP).The pressure difference ΔP is the differencebetween TGP in the water and the barometric pres-sure (BP) of the air above the water:

ΔP = TGP − BP

If TGP measured in the water is higher than the BPof the air (positive ΔP), the water is super-saturatedand gas will be forced out of the water. If TGP isless than BP, the water is under saturated and gaseswill be forced into the water. TGP may beexpressed as a percentage of the BP:

The water vapour pressure, ΔPH2O, may also beincluded in the equation:

Air and water contains many gases, but in fishfarming oxygen, nitrogen, and in some cases alsocarbon dioxide, are of greatest importance. Sincethe major reason for adding water is to provide

TGPBP +

BPH O2%( ) = −⎛

⎝⎞⎠ ×Δ ΔP P

100

TGP %BP +

BP( ) = ( ) ×ΔP

100

8.1 IntroductionThe purpose of aeration or oxygenation is either toremove gases such as nitrogen (N2) and carbondioxide (CO2) from the water, or to increase theconcentration of gases such as oxygen (O2) in thewater. There are several reasons for aerating andoxygenating the inlet water to a fish farm.The watermay contain insufficient oxygen or too much nitro-gen or carbon dioxide. If the content of oxygen inthe water is increased, the less water need be added.Too much nitrogen (super-saturation) will creategas bubble disease (diving disease) in the fish withhigh possibilities of mortality.1 Too much carbondioxide is also toxic for the fish.2,3 Adding too muchoxygen to the water will also be toxic for the fish.1

Aeration or oxygenation is carried out in almostall production systems. On land-based farms and inponds it is common to use either aeration or a com-bination of aeration and oxygenation. Recentresearch also shows improved production resultingfrom adding oxygen to cages.4 During transport offish, the addition of air or oxygen is also nessesary.

A great deal of literature is available concerningaquaculture, but the literature regarding waste-water treatment, environmental engineering and water chemistry is also a source of useful information.

8.2 Gases in waterWater contains a certain amount of dissolved gases.When the water has taken up the amount possibleunder normal atmospheric pressure it is fully satu-rated (100%) or in equilibrium (see Section 8.3.1).If the water contains less gas than can be taken up

97


oxygen, the added water should have as high anoxygen concentration as possible, close to 100%saturation. If under saturated water is used, morewater must be supplied. Water may also be super-saturated with oxygen, i.e. its concentration in thewater is raised above 100% by the addition of pureoxygen, to reduce the amount of water that must besupplied to the fish. If the other gases in the waterare in equilibrium, water that is super-saturatedwith oxygen will have a positive ΔP, i.e. TGP ishigher than BP. The concentration of nitrogen inthe water must not be above 100%, because this cancause bubble (diving) disease in the fish.

Water treatment and biological processes in thewater source can result in both super-saturationand under saturation of the different gases. Super-saturation can result from naturally occurring orman-made processes.5 Super-saturation can be theresult of rapid heating of water, mixing of water ofdifferent temperatures, freezing of water whensome of the gases remain in the water, mixing airinto the water, for instance under waterfalls or bywaves, photosynthesis that creates oxygen gas, andchanges in the BP.

If the saturation of the gases is 100% or less, theyare completely dissolved in the water. However, ifthe concentration of gas is above 100% saturation,there will be free gas bubbles in the water; these aresmall and difficult to observe. Only the oxygen actu-ally dissolved in the water is available to the fish:when the water is super-saturated with oxygen, thebubbles will gradually be dissolved in the water andthe oxygen made available for the fish as the fishconsume the oxygen that is already dissolved in thewater.

Air and water have different gas contents; in bothnitrogen and oxygen are the major gases. In air therelation between nitrogen and oxygen is approxi-mately 78%–21% by volume (Table 8.1), while in

water the relation is 60% nitrogen and 40% oxygenwhen the gases are in equilibrium. It is importantto be aware of this relationship because if air ispressed into the water, super-saturation of nitrogenmay occur which is harmful to the fish.

The air will exert pressure on the water surface.At sea level this pressure is 1 atmosphere, equal to760 ± 25mmHg; it varies depending on atmosphericconditions.The oxygen partial pressure is related tothe percentage of the total volume that oxygen con-stitutes, which in air is 21%; multiplying this valueby the BP (1atm or 760mmHg) gives the partialpressure of oxygen.

Partial pressure O2 = 0.21 × 760mmHg= 159.6mmHg

This means that the pressure of oxygen on thewater surface is 159.6mmHg.

The total pressure represented by the BP is thesum of the partial pressures of the gases in theatmosphere. This can be described as follows,neglecting the gases that constitute a very smallproportion of the air:

BP = P(N2) + P(O2) + P(Ar) + P(CO2)

The amount of a gas dissolved in water is referredto as the solubility of the gas, and depends onseveral factors, such as water temperature, pressure,salinity and substances in the water.6 At higher tem-perature gases are less soluble in water because themolecules then need more space. The solubility ofoxygen and nitrogen decreases linearly withincreasing temperature (see later) and for thisreason hot water contains less oxygen than coldwater. The solubility for oxygen and nitrogen alsodecreases with increased salt content of the water.Tables for the content of oxygen in water with different temperatures and salinity is given inAppendix 8.1.

The solubility of gases in water is normallyexpressed as mg/l of the actual gas, but it may alsobe expressed as the partial pressure. Equations forconversion are available;1 for oxygen and carbondioxide the following may be used:

where:

partial pressure for the actual gas (i) is in mmHg

Partial pressure =ii

ii

CA

b⎡⎣⎢

⎤⎦⎥

Table 8.1 Characteristics of dry air.1

Gas Weight % Volume %

Nitrogen (N2) 75.54 78.084Oxygen (O2) 23.10 20.946Argon (Ar) 1.29 0.934Carbon dioxide (CO2) 0.05 0.033Other 0.02 0.003

Ci = concentration of the actual gas in mg/lbi = Bunsen coefficient for the actual gas, depend-

ing on temperature and salinity (Appendix8.2)

Ai = constant depending on the actual gas (0.5318for oxygen and 0.3845 for carbon dioxide).7

Solubility may also be expressed on the basis oftension. The oxygen tension may, for instance, bethe necessary partial pressure in the atmosphere tokeep a certain concentration in the water. If theatmosphere is air at normal pressure, the oxygentension is 159mmHg. This creates 100% oxygensaturation in the water. If the pressure is less than

this, the concentration in the water will also bereduced.

8.3 Gas theory – aerationHow much and how fast gases are transferred inand out of water depends on two factors:8

(1) Equilibrium conditions (also known as the sat-uration concentration)

(2) Mass transfer.

Equilibrium is when there is no net transfer of gasin or out of the water. Mass transfer occurs beforeequilibrium is reached when there is transport ofgas into or out of the water.

8.3.1 Equilibrium

When equilibrium has been reached there is no nettransport of gas into the water or from the water tothe air. There is still some transport of gas mole-cules through the water surface, but what goes inequals what comes out; there are no free gasbubbles in the water. In the same way, when salt isadded to freshwater only a certain amount can bedissolved; after reaching this level no more will dis-solve, even if it is added. The excess salt will onlyremain on the bottom as salt crystals. The samehappens with the gas as it stays in bubble formwhen the water is super-saturated.

To find equilibrium conditions, Henry’s law canbe used. Henry’s law can be expressed in severalways;1,9,10 it illustrates that the amount of gas thatcan be dissolved is proportional to the partial pressure:11

Pg = HXg

where:

Pg = partial pressure of gas (atm)H = Henry’s constant (atm/mol fraction)Xg = concentration of gas in water (mol gas/(mol

gas + mol water)).

If SI units are used, Pg is given in pascal (Pa) H inPa m3/mol and Xg in mol/m3.

Henry’s constant depends on temperature andgas type; its value increases with temperature. Thegas obeys Henry’s law if the gas that can be dis-solved in water decreases with temperature.

As explained in section 8.2, partial pressure, Pg,depends on how much of the actual gas is in theatmosphere above the water surface. The pressureit exerts on the surface will press the gas into the water. In air there is about 20% oxygen, so Pg is in this case 0.2; if the atmosphere is pureoxygen, i.e. there is 100% oxygen above the water,Pg = 1.0.

If the gas above the water surface is abovenormal atmospheric pressure, there will be newequilibrium conditions given by Dalton’s law:8

pi = pty

where:

pi = partial pressure of gas ipt = total pressure of a mixture of gasesy = mol fraction of gas i (mol gas i/mol total gases).

Thus if the total pressure of the gas is increased, thepartial pressure will also increase and more gas canbe pressed into the water.

Henry’s law can be combined with Dalton’s lawto give the Henry–Dalton law:

pty = Hxi

where:

y = mol gas ipt = total gas pressureH = Henry’s constantxi = amount of gas i in the liquid.

xi is the key value because, as the equation shows,by increasing the total pressure or by increasing theamount of gas i in the atmosphere, the gas concen-trations in the liquid will be increased.

For instance, when air is under pressure above the water surface, more gas will be pressedinto the water which will be super-saturated.

Aeration and Oxygenation 99


Super-saturation with nitrogen from the air isharmful to fish, so overpressure of air is unwantedand must be avoided.The same may occur if addingair to deep water where the water pressure is higherand the air pressure must be higher than at thesurface, i.e. more than 1 atm. To illustrate this, aircan be added to the water 2m deep, but when doingthis air is compressed by the water before it dissolves.

ExampleA plastic bag containing both air and water islowered to the bottom of a 2m deep basin full ofwater. The air will be compressed by the water pres-sure and the air inside the plastic bag will be pressedinto the water because the pressure on the watersurface increases. We can now see what will happento the nitrogen saturation. The normal pressure onthe surface is 1atm (10mH2O); when lowering thebag to 2m deep the total pressure will increase to 12mH2O. The Henry–Dalton law can be used to findthe concentration of nitrogen:

pty = Hx

y is constant, but the pressure increases from 10 to12mH2O, a 20% increase. Since H is a constant, xwill increase by 20%. Therefore the concentration ofnitrogen increases by 20% which is harmful to fish.Under practical conditions, however, some nitrogenwill go directly to the surface and escape withouthaving time for gas transfer, but an increase in con-centration of 0.5% per 10cm or 5% per metre depthoccurs.

Adding air directly into the water at depths ofmore than 1m is not recommended because super-saturation of nitrogen should be avoided.

8.3.2 Gas transfer

To describe the mechanism of gas transfer intowater, the two-film theory proposed by Lewis andWhitman in 1924 is the simplest and most com-monly used (Fig. 8.1; refs 11–13). The interfacebetween the water and the gas can be divided intotwo films with laminar flow, one gas film and oneliquid film. These films will inhibit the transport ofgas molecules either into the water or out of thewater. The thickness of the two films with laminarflow depends on the amount of turbulence in the

water and in the air; much turbulence results inreduced thickness. To achieve a good transport ofgas molecules it is important to have a thin film toincrease the velocity of the gas transfer through theinterface.

When gas molecules from the air are pressed intothe water, they at first go from the gas phase intothe gas film, also referred to as the surface film.Thisprocess is a combination of diffusion and convec-tion, and is quite fast. The next step is through thegas–liquid film interface where the force is diffu-sion. Diffusion, which results from random molec-ular movements, is a slow process and this step israte limiting. The last step is transfer of gas into theliquid; the main force here is convection.

The gas transfer per unit time through thesurface can be described by the following differen-tial equation:

dc/dt = KL(A/V)(C* − C0)

where:

dc/dt = change in concentration per unit time(mg/(lh))

KL = coefficient for gas transfer (cm/h)A/V = contact area of the gas–liquid interface (cm2)

in relation to the total liquid volume (cm3)C* = saturation concentration for the gas in the

liquid (mg/l)C0 = starting concentration of the gas in the

liquid (mg/l).

The diffusion constant KL is proportional to the dif-fusion rate through the surface film and inversely

Figure 8.1 The two film theory is commonly used todescribe the transfer of gas into water.

proportional to the surface film thickness. Thisshows the importance of a thin surface film. KL isalso dependent on temperature; it will increase withtemperature, because the diffusion velocity willincrease. The following relationship shows this:14

KL = K20qT−20

where:

KL = overall mass transfer coefficientK20 = transfer coefficient at 20°Cq = correction factor (1.024 for freshwater)T = temperature.

It is normal to express the constant KL and thecontact area (A/V) together as the KLA value, whichis known as the overall mass transfer coefficient foran aeration system. It is difficult to calculate KLA foran aeration system; however, if this value could befound it would be easy to calculate the efficiency ofthe aeration system. Mathematical models havebeen developed, but they are difficult to use. There-fore the KLA value for an aeration system is usuallydetermined by experiment.

The gas concentration in a system after time twith mass transfer for a given KLA value can befound by integration of the basic equation to givethe following result:

This may also be expressed as:

This shows that the concentration (C) after usingan aeration system for time t depends on the start-ing concentration (C0), the saturation concentration(C*) and the KLA value:

C = C* + (C0 − C*)eKLAt

When all the oxygen has been removed from thewater and aeration is started, the content of oxygenwill increase rapidly and later level off at the limi-tation value (100% or full saturation); see Fig. 8.2.The value (C − C*)/(C0 − C*) is known as theoxygen deficit. If the natural logarithm (ln) of theoxygen deficit is plotted on the y-axis in a diagramwith time on the x-axis, the KLA value will be givenby the slope of the line.

ln**

C CC C

K t−−

⎛⎝⎜

⎞⎠⎟

=0

LA

C CC C

K t−−

=**0

e LA

8.4 Design and construction of aerators

8.4.1 Basic principles

With aerator the aim is to create conditions as nearequilibrium as possible between the gas in the airand the gas in the water. Eventually super-saturatedgas, especially nitrogen, will be aerated out andoxygen will be supplied if the concentration isbelow saturation. The aim in constructing anaerator is to achieve optimal conditions forexchange of gas between air and water, so that equi-librium can be reached. By creating a large contactarea between the water and the air a goodgas–water exchange will occur; see equations forgas transfer and the A/V relationship. The layers ofair or water ought to be as thin as possible.To createthin layers it is important that both the current ofair and the current of water are turbulent so thateffective gas transfer can occur at the water–airinterface. The aeration process needs time, andeffective aerators will need less time to achieve thesame degree of saturation than ineffective ones.Mass transfer from the air will occur into a tank ofnon-saturated water, almost 100% saturation beingachieved, but this will take a very long time.

Two main methods are utilized for aeration (Fig.8.3). Either air can be supplied into a flow of water,or water may be supplied into a flow of air. Anexample of the first method is bubbling of airthrough a water column; droplets of water passingthrough a layer of air illustrate the second method.Since the droplets are small, a large surface area


Figure 8.2 When all the oxygen has been removedfrom the water and aeration is started, the content ofoxygen will increase rapidly and later level off at the lim-itation value (100% or full saturation).


between the water and the air is created; the smallerthe bubbles or droplets, the larger the surface area.

Aerators can be constructed either for supplyingoxygen to the water (gassing) or removing nitrogenor carbon dioxide from the water (degassing). Anaerator built for gassing is not necessarily good fordegassing; aerators for both O2 and N2 are usual.

8.4.2 Evaluation criteria

Many methods are used for aeration, and severaldesigns of aerators are available either for addingoxygen, or removing nitrogen or carbon dioxide.Objective criteria are needed by which to evaluateaerators; these have been developed for both cleanwater conditions and field conditions.1,11 Most havebeen developed for wastewater applications,however, and may not be the most accurate or validfor use on aerators in aquaculture.13

Performance testing methods depend on theaerator type. For some, such as gravity aerators, thedifference in the gas concentration entering andleaving the aerator can be used. The gas exchangein a tank of a given size is used for testing the per-formance of surface aerators. Standardized set-upsand equations are available for testing the perfor-mance of aerators under laboratory conditions andwith clean water. Normally, the aerator is tested foroxygen, but other tests can be carried out, especiallyfor nitrogen or carbon dioxide. A typical way toperform a standardized aeration test for oxygen isto remove all oxygen from the water with, forexample, cobalt chloride and sodium sulphite,15

after which the aerator is started and the increasein oxygen concentration measured in relation totime. By doing this the KLA value for the aeratormay be found; this value also shows the perfor-mance of the aerator, a high KLA value represent-

ing an effective aerator. Performance tests can alsobe carried out under field conditions, but the resultsare difficult to compare, because the water qualitycan vary from site to site, as can the starting saturation.

The oxygen transfer rate gives the amount ofoxygen transferred into the water through theaerator. If it is possible to measure the gas concen-tration entering and leaving the aerator, as forgravity aerators, the oxygen transfer rate can founddirectly by measuring the water flow (Q) and thedifference in the oxygen concentration enteringand leaving the aerator (Cout − Cin). The oxygentransfer rate (OTR) can then be calculated fromthe following equation:

OTR = Qw(Cout − Cin)

If Qw is given in l/min, and C in mg O2/l, the fol-lowing equation can be used to find the OTR valuein kg oxygen transferred per hour:

OTR = 3.6Qw(Cout − Cin)

As said, this can be measured directly for gravityaerators, and also under field conditions (OTRf).

For basin aerators a standardized test proceduremay be used on clean water; the standardizedoxygen transfer rate (SOTR) in kg/h is given by thefollowing expression:15

SOTR = KLA20C20V × 10−3

where:

KLA20 = gas transfer coefficient determined for theaerating system at 20°C (kg O2/h)

C20 = equilibrium concentration at 20°C (g/m3)V = tank volume (m3)10−3 = factor converting grams to kilograms.

This test is performed with clean water, but com-pensating factors for water quality can be used to

Figure 8.3 Two methods are used foraeration: either air can be supplied into aflow of water or water may be supplied intoa flow of air.

adapt it to field conditions.14,16 The following equa-tion14 can be used to describe this:

where:

OTRf = actual oxygen-transfer rate under field-operating conditions in a respiring system

SOTR = standardized oxygen transfer rate (kg/h)a = field correction factor (varies with type of

aeration device, basin geometry, degree ofmixing and wastewater characteristics)

b = field correction factor for difference inoxygen solubility due to constituents ofsalt, particles and surface active substances

q = temperature correction factor (between1.015 and 1.040; typical value 1.024)

Cs = concentration of oxygen at full saturationand measured temperature

Cs20 = concentration of oxygen at full saturationat 20°C

Cw = concentration of oxygen in the water thatis aerated.

All aerators require a supply of energy; this canbe added directly (electricity), or it may be energythat is stored in the water (potential energy) as ingravity aerators. The potential energy can be uti-lized by sending the water from a high level to alower level through an aerator. When choosing anaerator it is important that it uses the energy sup-plied as efficiently as possible to transfer the gas.The standardized aeration efficiency (SAE) or fieldaeration efficiency (FAE) in units of kg O2/(kWh)is used for this purpose:

Another indicator variable is the oxygen transferefficiency (OTE) in units of kg O2 transferred perhour in relation to the added oxygen gas and themass flow (m), either measured under standardized(OTE) or in field conditions (OTEf). This indicateshow effectively the added gas is transferred to thewater and can be calculated from the followingequations:

FAE =OTR

Power inputf

SAE =SOTR

Power input

OTR SOTRfs w

s20

= −⎛⎝

⎞⎠ ( )−b q aC C

CT 20

For some aerators, such as some gravity aerators, itis impossible to measure the oxygen flow ratebecause they are open to the atmosphere; the OTEvalue can then be measured instead.

The effectiveness (E) of the aerator is anotheruseful indicator that shows how effectively the gasis transferred into the water (in gassing) or super-saturated gas is removed from the water.The valueswill vary, depending on the gas. The effectivenesscan also be used to compare aerators, but they mustbe tested in the same system and under indenticalconditions:

E = (Cout − Cin)/(Csat − Cin) × 100

where:

E = effectivenessCin = concentration of gas in the liquid entering

the aerator or before starting the aeratorCout = concentration of gas in the liquid leaving the

aerator or after using the aerator for a givenperiod

Csat = concentration of gas at full saturation.

The price of getting gas in or out of the water isthe most important factor, and this also makes com-parisons of the methods possible. Included are boththe aerator purchase price and the costs of runningit in relation to the amount of oxygen transferred:

Price for gas transferred in or out of the water = amount of transferred gas/equipment cost

and running cost for the aerator

8.4.3 Example of designs for different types of aerator

Many types of aerators are available, and differentclassifications are possible depending on the designprinciples employed. One classification is surface,subsurface and gravity aerators,1 another ismechanical, gravity and air diffusion systems,17

while a third is gravity, surface, diffusers and turbineaerators.18

In gravity aerators the water falls under gravityand air is mixed into it from the surrounding

OTEOTR

ff=

m

OTESOTR=

m



atmosphere. The simplest type is an artificial waterfall. Gravity aerators require water withnatural pressure (head); otherwise the water mustbe pumped up and into the aerator. In a surfaceaerator the water is sprayed or splashed into the airwith a mechanical device; it functions on the sameprinciple as a water fountain and creates a largesurface area where gas exchange can occur. In sub-surface aerators air is directed under the watersurface, which creates air bubbles that go to thesurface; the air bubbles in the water create a largegas transfer surface.

The design of the aerator will, as mentionedbefore, vary depending on the main purpose of theaerator, whether it is designed for adding oxygen,or removing nitrogen or carbon dioxide. Even if allaerators have some effect on all the gases, the effectis not necessarily optimal. Below is a brief reviewof some major types of aerator.

Gravity aerators

The packed column aerator is one of the most com-monly used in intensive fish farming (Fig. 8.4). Acolumn is filled with a medium with a large specificsurface area. A dispersal plate (perforated plate)installed at the top of the column, over the aerationmedium, ensures proper distribution through thetotal cross-sectional area of the water that entersthe top of the column. There is also a perforatedplate in the bottom of the column to keep themedium stable inside. The water trickles down thecolumn on the surface of the aeration medium in athin film. This arrangement creates a large areabetween the flowing water and the air around it andensures effective gas exchange. It is, however,important to get sufficient air into the aerator. Anopen structured column can therefore be advanta-geous, or a fan that blows air through the columncan be used (see later). The aeration medium iscommonly plastic, and it has a design that creates alarge surface area per unit volume. The relationbetween the surface area (A) and the volume (V)is called the A/V condition for the medium. Foreffective aeration media an A/V value of between100 and 200m2/m3 is normal; if the value is too high, the possibilities for blockage of the columnincrease. To give satisfactory aeration under practi-cal conditions the column ought to be up to 2m inheight. This, however, depends on the medium, the

flow rate and the water temperature.18 The diame-ter of the column cylinder depends on the quantityof water to be aerated, but is normally between 30and 100cm. If the diameter is too small, the pro-portion of the water flow against the wall in thecolumn, where reduced aeration is achieved, willincrease; a design rule of thumb value is 0.5–1lwater/min/cm2 cross-sectional area. The advantagewith a column aerator compared to other aeratorsis the low surface requirement. Column aeratorsare effective with reported SAE values of 1.5–2.0kg O2/kWh.12 This type of aerator also showsquite good results for removal of nitrogen.20

To remove carbon dioxide from the water is quitedifficult because the amounts are so small (less than

Figure 8.4 Packed column aerator.

1% of the total gas volume) and it is rather moresoluble than oxygen.7 Therefore, an aerator that is suitable for removing nitrogen and addingoxygen may not be so suitable for carbon dioxideremoval. What is normally needed is a large flow of air through the water: 3–10 volumes of air forevery 1 volume of water flow treated has been suggested.21 Specially designed packed column aer-ators have proved effective for this purpose.22

Either a long (several metres) packed columnaerator or a short packed column with a fanblowing air through to increase the air flow can beused (Fig. 8.5).

A packed column aerator may also be set underlow pressure. The principle here is that the water is

aerated in a low pressure atmosphere. Thus theamounts of gases that can be dissolved in the waterare reduced (Henry–Dalton law states that byreducing the pressure of the gas the amount of dis-solved gases will be reduced) and it is possible toreduce the saturation in the water to below 100%by natural pressure.This method is especially usefulwhen growing species that have a low toleranceagainst super-saturation, such as marine fry.19

Ejector aerators will also have this advantage.Packed column aerators are normally placed a

certain distance above a level basin to achieve thebest possible air transport through the aerator. Thisis important for the aeration results. Normally thedistance is set to 10cm. Too great a distancebetween column and basin must, however, beavoided because if the water drops from the aeratorfall into the level basin with too high a velocity, airwill be dragged into the water together with thewater drop and some super-saturation of nitrogenmight result. This can be harmful to critical lifestages.

Another much used aerator type is the cascadeaerator (Fig. 8.6). Water is supplied at the top andflows over a series of horizontal perforated platesor trays placed on top of each other. Vertical dis-tances between the trays are typically 10–25cm, andthe number of trays between 4 and 10.18 The effec-tiveness of the aerator increasing with number of


Figure 8.5 Packed column suitable for removal ofcarbon dioxide.

Figure 8.6 Cascade aerator.


trays and distance between them. At least 10 traysare recommended if using 10cm between them.Thewater is distributed throughout the whole perfo-rated tray and drops down from tray to tray until itreaches the level basin located underneath the lastone. When the water flows down through the holesin the perforated trays, drops will be created andthis ensures a large contact area between the airand the water. The space requirement for a cascadeaerator is much higher than for the column aerator;the recommended hydraulic rate is down to one-tweflth of what is recommended for a packedcolumn aerator.19

Typical SAE values for gravity aerators are in therange 0.6 to 2.4kg O2/(kW h).

Subsurface aerators

Diffusers are commonly used as subsurface aer-ators. A diffuser is a construction where smallbubbles of gas are created: the simplest form is atube with holes. If air is pumped in and the tubelowered under the water surface, the air will comeout in the water as bubbles. Tubes with many smallholes may also be specially produced for use as dif-fusers; porous ceramic stones may also be used. Alarge transfer surface is created by air bubblesgoing through the water. Bubble size depends onthe difference in the pressure inside the bubble andaround the bubble.A high pressure difference givessmaller bubbles. Smaller bubbles will give a greatertotal surface area and by this gas transfer area. Therise velocity is also slower for smaller bubbleswhich increases the gas transfer rate.23 SAE valuesvary from 0.6–2.0 depending on the bubble size,1 higher values being achieved with smallerbubbles.11 Bubbles that are too small will, however,be pushed together to create large bubbles, and aretherefore not effective. As mentioned previously,when mixing air under pressure, care should betaken to avoid super-saturation with nitrogen whichmay occur if air is to be added below 1.5–2m depth,or the pressure is over 0.15–0.2bar.

In the Inka aerator the incoming water flows overa perforated plate in a thin layer (Fig. 8.7). An airblower supplies air from the underside. The air willcreate bubbles that flow through the thin waterlayer and aeration is achieved. An Inka aeratorrequires a large surface area but the head loss isvery low.Therefore it may, for example, be installed

under the roof on a fish farm. The aerator type isless used than those described previously, but isuseful for special conditions.

Venturi and air-lift pumps are other principlesthat can be used for aeration where air bubbles aresupplied into a water flow; SAE values are in therange 2–3.3kg O2/(kW h).1

Surface aerators

Surface aerators are commonly use in ponds,24,25 butmay also be used in large tanks, in distributionbasins and in sea cages under special conditions(Fig. 8.7). A great number of different designs ofsurface aerators are available. Normal designs userotating wheels with a type of paddle, or horizon-tally placed propellers. They function by throwingthe water into the air and creating thin films orbubbles. This establishes a large exchange surfacearea between water and air where gas exchange can take place. Surface aerators are driven by electricity, solar power or from tractors through ashaft; SAE values are between 1.2 and 2.9kgO2/(kW h).1

8.5 Oxygenation of waterTo increase the amount of oxygen in the waterabove equilibrium and levels possible with tradi-tional aerators, pure oxygen gas can be added. Theaddition of pure oxygen gas to the water is used inseveral cases. One is to increase fish productionwhen there is not enough water. If the water has tobe pumped to the farm, regardless of whether it issalt water or freshwater, it may be viable to addpure oxygen to reduce the necessary water flow andhence reduce the pumping costs. Normally it ismore economical to add pure oxygen instead ofpumping, but a calculation must be made in everysingle case. In systems with re-use of water it mayalso be worth adding pure oxygen to reduce theamount of new water and amount of water pumpedthrough the system. When transporting fish pureoxygen is usually supplied.

It is important that the water is fully saturatedwith oxygen before starting to add pure oxygen gas.An aerator should therefore be installed before the point where pure oxygen gas is added; if thereis no aerator, pure oxygen is used to saturate thewater up to equilibrium before starting to super-

have a free surface to the atmosphere there wouldbe gas exchange and the oxygen content would fallto 100% saturation. Therefore free surfaces shouldbe avoided after supersaturating the water withoxygen; the best method is to have a short pipe run


B

A

F

E

C

D

Figure 8.7 Various types of surfaceaerator are available. (A) and (B)show a paddle wheel aerator, whilst(C), (D) and (E) show a propelleraerator, (F) shows an Inka aerator.(D) A propellor aerator being used ina closed cage.

saturate. This is, in most cases, unnecessarily ex-pensive, but calculations can be made to evaluatecost-effectiveness.

When the water is supersaturated with oxygenfree gas bubbles are present; if this water were to


from where the oxygen is added to where it reachesthe fish.

An oxygenation plant may be designed either tocover the whole oxygen requirement for the fish oras a top-up oxygenation facility. The design criteriafor the oxygenation plant are determined by the sit-uation where the biomass and water temperatureare highest. A complete oxygenation system on afish farm includes two parts: (1) the injection systemthat brings the gas into the water and (2) the sourceof oxygen gas.

8.6 Theory of oxygenationWhen water is oxygenated pure oxygen gas isadded; by this the saturation of oxygen in the watercan be raised above, the equilibrium level of 100%.The following prossesses are included in oxygena-tion (cf. aeration): (1) increase of the equilibriumconcentration, (2) increase of the gas transfer veloc-ity, (3) addition under higher pressure.

8.6.1 Increasing the equilibrium concentration

Henry’s law can be used to describe what is happening:

pi = Hxi

where:

pi = partial pressure of gas iH = Henry’s constantxi = amount of gas i in the liquid (mg/l).

xi is the value of interest here. The oxygen contentof air is 20% compared to a theoretical value of100% for pure oxygen. The partial pressure (pi)when having pure oxygen atmosphere is thereforealmost five times as high as when using an atmos-phere of air. Since Henry’s constant does not varyalmost five times as much oxygen can be dissolvedin the water by oxygenating in a pure atmospherethan when doing it in air. An example of this is the packed column aerator where the atmosphereinside is pure oxygen.

8.6.2 Gas transfer velocity

The same equation as for aeration can be used todescribe the gas transfer:

dc/dt = KLA(C* − C0)

where:

dc/dt = change in concentration per unit time, i.e.velocity of gas transfer

KLA = diffusion coefficientC* = equilibruim concentration of the gas in the

liquidC0 = concentration of gas in the liquid at the start

point.

The equilibrium concentration C* for oxygen dis-solved in water standing in a pure oxygen atmos-phere will be higher than for water in normalatmosphere (see Section 8.3.1, Henry’s law). There-fore the velocity of gas transfer into the water isgreater, because the difference C* − C is larger.Less time is then needed to increase the concen-tration of oxygen.

8.6.3 Addition under pressure

When using pure oxygen gas it is possible toincrease the pressure and thereby increase theamount dissolved in the water. As distinct fromadding air under pressure, there is no possibility forsupersaturation of nitrogen which is toxic. Here the Dalton and Henry–Dalton laws can be used todescribe what is happening:

pi = pty

where:

pi = partial pressure of gas ipt = total pressure of a mixture of gasesy = mol fraction of gas i (mol gas i/mol total gases).

By increasing the total pressure in the oxygen gasabove the water surface, the partial pressure alsoincreases.

When combining the Henry and Dalton laws thefollowing is obtained:

pty = Hxi

where:

y = mol gas ipt = total gas pressureH = Henry’s constantxi = amount of gas i in the liquid.

The oxygen concentrations in the liquid (xi) is ofinterest and will be increased by increasing the totalpressure (pt) of the oxygen gas.

8.7 Design and construction of oxygeninjection systems

8.7.1 Basic principles

In the injection system the oxygen gas comes fromthe source and is injected or mixed into the water.To get as much gas as possible into the water, aspecial injection system is necessary. Oxygen gas can be mixed into water under normal or high pres-sure. With high-pressure oxygenation, the water is pressurized up to 4 bar (see conversion factor boxbelow) before the oxygen is added, and more than500% supersaturation can be achieved withoutproblems. When using normal pressure or low-pressure oxygenation (up to 1bar), 100–300% satu-ration can be achieved. Normal or low-pressure oxygenation is used on the main inlet pipe to thefarm, while high-pressure oxygenation will normallybe carried out in a part flow divided from the mainflow because it is too costly to pressurize the mainwater inlet. Furthermore, it is easy to get a concen-tration of oxygen higher than either needed or rec-ommended. Too high an oxygen concentration maydamage the gills of the fish. When the water entersthe fish tanks, it is not recommended to be above150–200% oxygen saturation, and even these concentrations have been called into question.

The same design principles as for aerators can be used for the injection system. Either oxygen gas bubbles can be supplied into water or waterdroplets can be supplied into an atmosphere ofpure oxygen.

(3) Directly to the water in the fish tank(4) To an individual circuit where the water is

supersaturated with oxygen.

When adding oxygen to the main flow, the gas isadded directly into the main pipeline to the farm.It is important to be aware that the oxygen gasadded to the water needs space so the water flow through the pipe will be reduced, and this must be taken into consideration when specifying the inlet pipes. With this method it is difficult to


Conversion factors for pressure units are asfollows:

• 1bar = 1 × 105 Pa = 10.19 mH2O = 0.9869atm.• 1Pa = 1 × 10−5 bar = 1.02 × 10−4 mH2O = 9.861

× 10−6 atm• 1 mH2O = 9806.65Pa = 0.098bar = 0.097atm• 1atm = 101325Pa = 1.01325bar = 10.33 mH2O

8.7.2 Where to install the injection system

On a land-based fish farm oxygen gas can be addedto the water in several places (Fig. 8.8):

(1) To the main water inlet pipe to the farm(2) To a part water flow separated from the main

inlet flow

(1)

(2)

(3)

(4)

Figure 8.8 Oxygen can be injected: 1, into the mainwater inlet pipe to the farm; 2, into a part water flow sep-arated from the main inlet flow; 3, directly into the waterin the fish tank; 4, into an individual circuit until the wateris supersaturated with oxygen.


increase the pressure in the water significantlybecause the equipment required to increase thepressure in the entire water supply will be verylarge and costly.

Oxygenation of part of the water flow is carriedout by diverting some of the water from the mainpipeline into a smaller pipe. The pressure in thissmaller pipe can and is normally increased, beforeoxygen gas is added by means of a pump.The pumpis normally of the high-pressure centrifugal multi-stage type. As the pressure in the water is increasedto 2–4bar, a normal centrifugal pump is not power-ful enough. After adding the oxygen gas, the pres-sure is reduced by a reduction valve; the water,which is now supersaturated with oxygen, is pipedinto the main flow.

In the third method oxygen is supplied directlyto the fish tank either by direct addition via the inletpipe to the tank or through diffusers at the bottomof the tank. The advantage with this method is thatindividual oxygenation of the separate fish tanks ispossible; hence it is also possible to improve oxygenutilization because only the amount necessary forthe fish in the tank is added. By having an oxygenmeter in the outlet of the tank an acceptable valuecan be set (for instance 7mg/l) and a simple controlsystem can be used to add the necessary oxygen. Ifthe temperature increases or the fish grow faster,more oxygen will be added to avoid the oxygenlevel in the outlet dropping. Another advantageachieved by using diffusers directly at the bottomof the fish tank is that saturation in the tank willalways be below 100%. It will therefore be easierto get the supplied oxygen into the water, becauseit is not being supplied against a large oxygen gra-dient as is the case when oxygenating super-saturated water (refer to KLA value).

When adding oxygen to a closed circuit, part ofthe inlet water to the farm is taken out and super-saturated with oxygen. This supersaturated water isthen piped directly to each of the fish tanks on thefarm through separate pipelines. The fish tank hastwo inlet pipes, one delivering aerated water at upto 100% oxygen saturation and another deliveringwater supersaturated with oxygen. Because it ispossible to adjust the oxygen supply to the varioustanks on the farm individually, good efficiency can be achieved. The major disadvantage of thismethod is the high cost of the additional pipingneeded.

Oxygenation may be carried out both in fresh-water and seawater. It is simpler to add oxygen toseawater because of its higher ion concentration.Oxygen microbubbles will be smaller in seawaterthan in freshwater because the surface tension isreduced.15 Small bubbles have a reduced rise veloc-ity and so the contact time will increase. In addi-tion, the relative surface area is higher for smallbubbles than for large ones. Hence the efficiency ofmixing, and thus the necessary addition pressure,varies with the salt concentration.

8.7.3 Evaluation of methods for injecting oxygen gas

Many methods are available to inject oxygen intowater, using both low and high pressure. To be ableto evaluate the performance, evaluation methodsare available for aerators. As pure oxygen is muchmore expensive than air, its utilization is veryimportant. This is given by the oxygen transfer effi-ciency (OTE), which is also known as the absorp-tion efficiency6 and is the amount of oxygendissolved in water in relation to the amount ofoxygen supplied:

where:

Cin = concentration of oxygen before injection ofoxygen

Cout = concentration of oxygen after injection ofoxygen

Qw = water flowmO2 = mass flow of oxygen gas.

Performance values ranging from a few percentagepoints up to 100% have been reported.6

Another important parameter is oxygenationefficiency (OE) which is a measure of the amountof oxygen dissolved in relation to the power sup-plied. Normally excess pressure is used to achievehigh oxygen concentrations in the water and for this power is necessary, for example to run a high-pressure pump. This can be described as follows:

Units for OE are mg O2/kW.

Oxygenation efficiency =Power supply

in out wC C Q−( )

Absorption efficiency = in out w

O

C C Qm

−( )2

Most important of course, is the amount ofoxygen dissolved in the water in relation to theinvestment and running costs for the injectionequipment; this is measured in mg O2 per unit cost.

It is important to remember that it is the amountof oxygen available for the fish in the productionunit that is of interest. For example, the absorptionefficiency directly after the injection unit, whichmay be very high, is of no concern; neither arevalues that are achieved in laboratories where allother factors are optimized in a way that is impos-sible to achieve under real farming conditions.Several design factors for the whole farm will havea great influence on the efficiency of the oxygena-tion system; these include the water transfer systemand tank inlet design. No two farms will be exactlythe same and this must be taken into account. Com-parison of systems is therefore not straightforward;tests must be performed under exactly the sameconditions and if possible on the site where theequipment is to be used.

8.7.4 Examples of oxygen injection system designs

A number of systems for injection of oxygen gas are available.6 What distinguishes the methods from aeration is that the gas is injected under pres-sure, so the equipment is pressurized. Injectionsystems can be divided into high- and low-pressuresystems, where the pressure in the former is above1bar.

Low pressure

Packed column: A simple way to add oxygen to thewater is to substitute the air in a packed columnaerator by pure oxygen, and aerate in a pureoxygen atmosphere (Fig. 8.9); there is no excesspressure. Under practical conditions, an oxygen saturation of two to three times the normal value isachievable by changing the atmosphere from air topure oxygen. Theoretically the oxygen contentshould be up to five times as high as in air, but thetime available for the gas to transfer into the wateris normally a limitation.The air in a column aeratorcan easily be exchanged with pure oxygen byplacing the lower end of the column in the waterand feeding pure oxygen gas into the lower part of

the column so that it flows up against the descend-ing water. This results in good mixing of the waterand oxygen. Eventually surplus oxygen that reachesthe top of the column and is not dissolved in thewater can be withdrawn through a special valve and recycled through the column inlet water toprevent any loss of pure oxygen gas. The amount of oxygen that can be transferred into the water andthe efficiency of oxygen supply can be increased if the column is pressurized; here the column is closed at the top and bottom. The pressure in sucha system normally varies from 0.2 to 0.5 bar excess pressure.

Diffuser: A diffuser may also be used to supplyoxygen gas to the water, in the same way that air issupplied for aeration (Fig. 8.9). The efficiencydepends on the depth at which the oxygen is added;a greater depth will result in increased efficiency(see Section 8.3.1, Dalton’s law), a depth of at least2m being recommended. The diffusers can beeither tubes or ceramic. It is important it that theoxygen gas leaves the diffuser as small bubbles,because then better mixing is achieved as a resultof increased contact area and reduced rise veloc-ity.15,23 Usually diffusers are used to add oxygen tothe production units in connection with transportor as security systems. This is normal procedurealmost all fish farms and for all fish transport.Whenusing diffusers, the oxygen is normally suppliedfrom gas bottles because the amounts are quitesmall. Diffusers are also used in normal production;one method is to have some central oxygenation ofthe incoming water and treat each tank individuallyby having a diffuser on the tank bottom.

Diffuser placed directly in the inlet pipe: If oxygenis to be added to the fish tank directly through theinlet pipe, a diffuser is placed inside the inlet pipeafter the regulating valve (Fig. 8.9). Such an oxy-genation system requires a correct design of inletpipe to the tank (Chapter 11). The water depth inthe tank ought to be more than 1.5m to give satis-factory pressure.The highest efficiency is seen whenusing this method in seawater because oxygen ismore soluble in seawater than in freshwater. Witha direct supply via the inlet pipe the greatest advan-tage is that single tanks can easily be supplied withoxygen individually and a better total utilization ofthe added oxygen achieved.



AB

C

D

Fig

ure

8.9

(A–D

)M

etho

ds f

or l

ow p

ress

ure

inje

ctio

n of

pur

e ox

ygen

gas

: ox

ygen

diff

user

in

the

inle

t pi

pe o

f th

e ta

nk (

A),

on

the

botto

m o

f th

eta

nk (

B a

nd D

), in

the

inle

t pi

pe t

o th

e fa

rm (

C).

High pressure

There are various methods utilized for injectingoxygen gas into the water under high pressure.Two of the most commonly used are the oxygencone and deep wells, while others employ differenttypes of injector.

Oxygen cone: In the oxygen cone the oxygen gasand water enter at the top through different pipes(Fig. 8.10). The oxygen will be pushed down in thecone because of the water flow and a large oxygenbubble, actually a layer of oxygen gas, will becreated lower down in the cone. The water will flowin small drops through this layer, creating a situa-tion with water drops flowing through a pureoxygen atmosphere. Since the pressure inside thecone can be up to 5bar, good gas transfer isachieved and because of this quite good efficiency.After flowing through the oxygen layer, the oxy-genated water will flow out, from the bottom of thecone. However, if the cone is ‘overloaded’ withwater, the oxygen layer will be pushed down to thelower end of the cone, perhaps also through thecone outlet and efficiency is reduced. Because ofthe high pressure used inside the cone to transferoxygen gas into the water, the pressure of the watermust be increased before it enters the cone. To dothis, part of the water flow is piped from the main pipeline, and a high-pressure centrifugalpump used to increase the pressure in this part flowbefore it enters the cone. After leaving the cone,having been supersaturated with oxygen, the wateris returned to the main flow. However, the waterpressure must be reduced by passage through apressure reduction valve before entering the mainflow again.

Use of oxygen wells: Oxygenation in wells is some-times used and gives quite good efficiencies(70–90%) (Fig. 8.10). Well design varies, but themain purpose is to mix oxygen and water at quitelarge depths where the high pressure together witheffective mixing ensures good efficiency. Onemethod that utilizes this principle is the U tube.Oxygen gas in injected into the water before it flowsinto the U tube. At the bottom of the U the pres-sure will be increased depending on the depth ofthe U; 10–30m is usual. The construction ensureseffective mixing by creating turbulence between

the oxygen gas and the water and the gas bubblesnow dissolve. The same principle is used whenoxygen is added directly into a deep-water intake.In land-based fish farms in Norway this system hasbeen used with direct addition of oxygen to theinlet pipe at depths of 20–30m. Here the pressureis high and good transfer of the oxygen gas into thewater is achieved.

When using such methods it is important to beaware that the water flow rate through the pipe willbe reduced because the oxygen gas takes up spaceand the real cross-sectional area where the waterflows is reduced. If the water is sent to a pump afterinjection of oxygen, care must be taken because theconditions are now optimal for cavitation with gasbubbles that may implode as the water passesthrough the pump.

Oxygenation in sea cages: Recent research hasshown that lack of oxygen may occur in sea cages,possibly as the result of algal blooms in the areawhere the cages lie. The oxygen concentration maybe reduced to below 70% saturation at the end of dark nights because both fish and algae useoxygen during the night. The conditions can be sobad that the effects of current induced waterexchange cannot meet the reduced oxygen concen-tration caused by the algal bloom.

Lack of oxygen may also be caused by high temperatures and high fish densities in sea cages.The oxygen consumption of the fish increases with temperature and at the same time the oxygencontent of the water is reduced; in such cases the lack of oxygen will be local to the cage.However, if the water current increases, moreoxygen-rich water can be transported to the cage; apropeller may therefore be used to create an artifi-cial current to increase the supply of oxygen-richwater.

It is difficult to monitor the oxygen concentrationin a sea cage in a representative way, because thecage can be large and there may be individual variations within the cage, mainly due to the depth. Several oxygen meters could therefore beadvantageous.

When pure oxygen gas is to be added to the cage,oxygen diffusers can be lowered into the cage. It isvery advantageous to create an equal pressure onthe diffuser, so that an equal amount of oxygen gasis released all around the cage. As it is important to



A

B

C D

Figure 8.10 Oxygen cones and oxygen wells are typical systems for injecting oxygen gas into water under highpressure: (A) oxygen cone cross-section; (B) oxygen well cross-section. (C) cones in situ; (D) locked column.

get an even distribution of the added oxygen, spe-cially designed diffusers are recommended.

8.8 Oxygen gas characteristicsOxygen gas (O2) has a boiling point of −183°C atnormal atmospheric pressure. If the pressure isincreased, the boiling point will also increase. Atnormal temperature O2 is therefore only availablein gas form. Liquid oxygen is light blue withoutodour; the density of the liquid is 1.15 × 103 kg/m3

at boiling point, while the density of the gas is 1.36kg/m3 at 15°C and normal atmospheric pressure. One litre of liquid oxygen will thereforeexpand 820 times when changing phase to gas at15°C.

Liquids with boiling point below −100°C arecommonly called cryogenic liquids. Handling cryo-genic liquids involves certain elements of risk:

• Extremely low temperature can result in frostinjures on the skin when handling; the cold gasmay also cause internal injuries

• Materials that come into contact with the liquideasily become brittle

• There is a danger of explosion when uncontrolledtransformation from liquid to gas is allowed tooccur in closed volumes.

By itself, O2 is inflammable but oxygen supportsa fire and is normally the limiting factor for firedevelopment. If much oxygen is available in theatmosphere, only a small spark can create an explo-sive fire. Care should therefore be taken when handling oxygen gas.

8.9 Sources of oxygenOxygen gas may be delivered from oxygen producers, either as compressed gas in bottles or as liquid oxygen, which the farms transforms to gas under controlled conditions. This oxygen is produced in special factories that produce onlyoxygen. Whether gas or liquid, commercially pro-duced oxygen is characterized by high purity(>99%) and the presence of a small amount ofother gases.

The other alternative is on-site production ofoxygen, which means that the oxygen is producedat the farm with air as the source.

8.9.1 Oxygen gas

Oxygen gas may be delivered as compressed gas inbottles. In Norway the bottles are blue and can bedelivered as single bottles or in batteries, normallyof 12 (Fig. 8.11). The usual pressure inside anoxygen gas bottle is as high as 200bar; the bottlecontains 50 l of compressed gas at this pressure andthis gives 10000 l gas at normal atmospheric pres-sure. On the top of the bottle there is a pressurereduction valve. A manometer, also on the top ofthe bottle shows the internal pressure; this also


Figure 8.11 Compressed oxygen gas in high pressurebottles, either supplied singly or in a battery of 12.


indicates the amount of gas remaining in the bottle,because the pressure inside the bottle will decreaseas oxygen is released. The bottles are normallyrented from the oxygen producers, one reasonbeing that the producer will ensure the bottles areproperly maintained. The price of oxygen gasdepends on the distance from the oxygen produc-tion site amongst other things. The pressurizedbottles represent a real danger, for instance in con-nection with a fire. It is therefore advisable not tohave too many bottles on the farm, and if possible

to place them outside; legal restrictions may limitthe number of gas bottles that can be stored insidebuildings.

8.9.2 Liquid oxygen

Liquid oxygen (LOX) is produced in special facto-ries and transported to the farms by truck or boatin special tanks. On arrival, the liquid oxygen istransferred to the farm’s specially designed tank forstorage (Fig. 8.12). The usual tank size for storing

Figure 8.12 Oxygen can be stored asliquid in special tanks at the farm anda cold-gas evaporator ensures that the liquid oxygen is transformed to gas before entering the fish tanks. Theschematic shows the principle; thephotograph shows a tank in situ.

liquid oxygen is between 2 and 50m3. The oxygen isstored as liquid in the tank with a temperaturebelow the boiling point and of a pressure ofbetween 15 and 25bar.

The pressure in the liquid oxygen tank will stayconstant. An internal evaporator system, normallyplaced under the liquid tank, ensures this. Liquidoxygen is taken from the tank, goes through theevaporator where the liquid is transformed to gas,and back again as gas into the top of the tank. Theexpansion when oxygen goes from liquid to gas isutilized in this process and maintains the pressurein the tank. The tank is fitted with a safety valve toprevent it exploding, because even though the tankis constructed with two shells with space betweenthem it is impossible to achieve 100% effective isolation of the tank. Therefore some liquid willalways evaporate to gas inside the tank. With thelarge expansion in volume when changing phasefrom liquid to gas, the tank will easily blow ifnothing is done, which is where the safety valvecomes into play. If the daily use of oxygen from thetank is too low, oxygen gas will pass through thesafety valve and into the atmosphere. Normally0.1–0.6% of the tank volume must be used daily,depending on tank size, to avoid loss of oxygenthrough the safety valve to the atmosphere. Thisalso shows the importance of choosing a tank of thecorrect size.

When the farm uses oxygen from the tank, it isdrained as liquid from a pipe near the bottom ofthe tank. Because of the pressure in the tank, theliquid oxygen will be pressed out through the outletpipe and into the evaporator when the outlet valveis opened. In the evaporator liquid oxygen changesphase to gas. The gas is transported from the evap-orator through a pipe and into the farm’s oxygeninjection system.

The evaporator is known as a cold gas evap-orator and must be designed according to the farm’s oxygen consumption. It functions as a heatexchanger transferring heat from the air into theoxygen liquid (air–liquid exchange). It is importantto have a large surface area to get a large contactarea between the air and the oxygen. Usually atleast two evaporators are installed, one runs whilethe other is switched off. The evaporator takesenergy from the air and after a while is totallycovered with ice; deicing is therefore necessary andis the reason for at least two evaporators beingrequired.

The liquid oxygen tank should be placed outsideon a concrete base and be fenced in. In addition,there should be a safety zone around the tank.Oxygen tanks are rented from the oxygen suppliersfor safety reasons. With online measurement of theremaining oxygen content in the tank by the sup-plier, it is also possible for the supplier to automat-ically replenish the oxygen before the tank is empty.It is then not necessary for the fish farmer to keepcontrol of this.

8.9.3 On-site oxygen production

The alternative to buying oxygen from a produceris to have on-site production facilities (Fig. 8.13).Air contains 78% nitrogen, 21% oxygen, 0.9%argon and 0.1% of other gases by volume (Table8.1). In a specially designed adsorption unit it ispossible to remove the nitrogen and some of theother gases from the air, leaving mainly oxygen.Theabsorption unit is fed by compressed air from acompressor; different names are used (oxygen gen-erator or xorbox) according to the producer. Putsimply, it is a unit that filters air and lets only oxygenpass. The process is known as pressure swingadsorption (PSA).

The compressed air passes into an absorptionunit that is normally filled with clinoptilolite, anatural zeolite which is a clay mineral. To describethe construction and function in detail is beyondthe scope of this book but a brief and simplifieddescription is as follows. The crystals in the filtermaterial form a latticework where the nitrogenmolecules are trapped. Oxygen molecules, whichare large, do not fit into the latticework and passthrough; almost pure oxygen is produced. Providedthat the filter medium is not contaminated with oilor water from the compressor, it is quite durable.An inlet air filter must be used and cleaned at regularintervals. A normal generator used for aquaculturecan reach a purity of 95% oxygen, the rest beingargon (4%) and nitrogen (1%). Special units mayachieve 99% oxygen purity, but they are veryexpensive and not viable for on-site installation infish farming. The efficiency of the filter graduallydecreases. After a period the filter medium, willbecome saturated with nitrogen. Nitrogen gas willthen follow the oxygen gas out of the generator andto the fish. For this reason two columns are alwaysused in alternation: one is used and the other isregenerated. When a column is regenerated, air



is ‘driven’ through the column in the opposite direction, and the nitrogen is blown out to theatmosphere.

An oxygen generator normally delivers oxygengas with a maximum pressure of 4bar. This must be

taken into consideration because some of theequipment for injecting oxygen gas into the watermay require higher pressure. If so, a speciallydesigned oxygen compressor must be used after theoxygen generator. After the generator there must

A

B

C

Figure 8.13 On-site production of oxygen from air usingan oxygen generator. Principle (A), the air production unitand the two absorption units (B) in a cabinet shown witha bottle of compressed oxygen as security (C).

be a storage tank for the oxygen produced toensure pressure equalization. From this tank theoxygen is tapped and used for injection into thewater on the farm.

When a generator has been in use for some years(depending on type of equipment and runningtime) the purity of the delivered oxygen will fall to 80–85%. The amount of nitrogen gas will alsoincrease, which can be toxic to the fish. If theoxygen gas is supplied under high pressure, therewill be possibilities for supersaturation of nitrogen.Maintenance and replacement or cleaning of thefilter medium inside the columns must then becarried out. It is therefore important to sample theoxygen gas from the generators at fixed intervalsand measure its purity. An easy way to do this is tocollect a sample of the oxygen gas in a plastic bag.The oxygen meter that is used to measure thecontent of oxygen in the water on the farm can thenbe used to measure the approximate oxygen purity.This is done by first adjusting the meter to show21% purity in air; now when the oxygen probe isput into the plastic bag, it will show the purity

of the oxygen gas sample directly, for instance 95%.

8.9.4 Selection of source

The choice of oxygen source depends on severalfactors. If the oxygen consumption is very low or ifoxygen is only used for security purposes, it isnormal to use gas bottles. Otherwise there is achoice between purchasing liquid oxygen and on-site production. If only the price of oxygen is con-sidered, on-site production will in almost every casebe cheaper. The reliability, maintenance costs andnecessary back-up system, however, often causefarmers to buy liquid oxygen from commercial producers because they guarantee the oxygensupply. Other important factors when choosing are the distance to the oxygen production factory,infrastructure for transport of liquid oxygen, andthe price of electricity. The amount of electricenergy used for production of oxygen gas is size-dependent; larger generators are more efficientthan smaller ones.


Appendix 8.1Solubility of oxygen in water with different tempera-ture and salinity with normal atmospheric pressure(1013mbar)

Salinity, parts per thousand (ppt)

Temperature (°C) 0 10 20 30 40

0 14.6 13.6 12.7 11.9 11.12 13.8 12.9 12.1 11.3 10.54 13.1 12.3 11.5 10.7 106 12.4 11.6 10.9 10.2 9.68 11.8 11.1 10.4 9.8 9.1

10 11.3 10.6 9.9 9.3 8.712 10.8 10.1 9.5 8.9 8.414 10.3 9.7 9.1 8.6 8.016 9.9 9.3 8.7 8.2 7.718 9.6 8.9 8.4 7.9 7.420 9.1 8.6 8.1 7.6 7.222 8.7 8.2 7.8 7.3 6.924 8.4 7.9 7.5 7.1 6.726 8.1 7.7 7.2 6.8 6.528 7.8 7.4 7.0 6.6 6.330 7.6 7.2 6.8 6.4 6.1

Appendix 8.2Bunsen’s coefficient for oxygen as a function of temper-ature and salinity

Salinity, parts per thousand (ppt)

Temperature (°C) 0 10 20 30 40

0 0.049 0.046 0.043 0.040 0.0372 0.047 0.043 0.041 0.038 0.0364 0.044 0.041 0.039 0.036 0.0346 0.042 0.039 0.037 0.034 0.0328 0.040 0.037 0.035 0.033 0.031

10 0.038 0.036 0.034 0.032 0.03012 0.036 0.034 0.032 0.030 0.02814 0.035 0.033 0.031 0.029 0.02716 0.034 0.032 0.030 0.028 0.02618 0.032 0.030 0.029 0.027 0.02520 0.031 0.029 0.028 0.026 0.02522 0.030 0.028 0.027 0.025 0.02424 0.029 0.027 0.026 0.024 0.02326 0.028 0.026 0.025 0.024 0.02228 0.027 0.026 0.024 0.023 0.02230 0.026 0.025 0.024 0.022 0.021


Appendices based on:Benson, B.B., Krause, D. (1984) The concentration and

isotopic fraction of oxygen dissolved in freshwater andseawater in equilibrium with the atmosphere. Limnol-ogy and Oceanography, 29: 620–632.

Colt. J. (1984) Computation of dissolved gas concentrationin water as a function of temperature, salinity and pres-sure. American Fishery Society. Special Publication No 14.

Weiss. R.F. (1970) The solubility of nitrogen, oxygen andargon in water and seawater. Deep-sea Research, 17:721–735.

References1. Colt, J., Orwicz, K. (1991) Aeration in intensive aqua-

culture. In: Aquaculture and water quality (eds D.E.Brune & J.R. Thomasso). World Aquaculture Society,Louisiana State University.

2. Fivelstad, S., Haavik, H., Løvik, G., Olsen, A.B. (1997)Sublethal effects and safe levels of carbon dioxide inseawater for Atlantic salmon postsmolts (Salmo salarL.): ion regulation and growth. Aquaculture, 160:305–316.

3. Tucker, J.W. (1998) Marine fish culture. Kluwer Aca-demic Publishers.

4. Bergheim, A., Gausen, M., Næss, A., Krogedal, P.,Hølland, P., Crampton, V. (2006) A newly developedoxygen injection system for cage farms. AquaculturalEngineering, 34: 40–46.

5. Colt, J. (1986) Gas supersaturation – Impact on thedesign and operation of aquatic systems. AquaculturalEngineering, 5: 49–85.

6. Colt, J., Watten, B. (1988) Application of pure oxygenin fish culture. Aquacultural Engineering, 7: 397–441.

7. Colt, J. (1984) Computation of dissolved gas concen-tration in water as a function of temperature, salinityand pressure. American Fisheries Society SpecialPublication 14.

8. Gebauer, R., Eggen, G., Hansen, E., og Eikebrokk, B.(1992) Oppdrettsteknologi – vannkvalitet og vannbe-handling i lukkede oppdrettsanlegg. Tapir forlag (inNorwegian).

9. Mackay, D., Shiu, W.U. (1984) Physical-chemical phe-nomena and molecular properties. In: Gas transfer at water surfaces. (eds W. Brutsaert, G.H. Jirka), D.Reidel.

10. Lincoff, A.H., Gossett, J.M. (1984) The determinationof Henry’s constant for volatile organics by equilib-rium partitioning in closed systems. In: Gas transfer at water surfaces (eds W. Brutsaert, G.H. Jirka),D. Reidel.


12. Boyd, C.E., Watten, B.J. (1989) Aeration systems inaquaculture. Reviews in Aquatic Science, 1: 425–472.

13. Colt, J. (2000) Aeration systems. In: Encyclopedia ofAquaculture (ed. R.R. Stickney). John Wiley & Sons.

14. Stenstrom, M.K., Gilbert, R.G. (1981) Effects ofalpha, beta and theta factor upon the design, specifi-cation and operation of aeration systems. WaterResearch, 15: 643–654.


16. Shelton, J.L., Boyd, C.E. (1983) Correction factors forcalculation of oxygen-transfer rates of pond aerators.Transaction of the American Fishery Society, 112:120–122.


18. Wheaton, F.W. (1977) Aquacultural enginering. R.Krieger Publishing Company.


20. Bouck, G.R., King, R.E., Bouck-Scmidt, G. (1984)Comparative removal of gas supersaturation byplunges, screens and packed columns. AquaculturalEngineering, 3: 159–176.

21. Summerfelt, S.T. (2000) Carbon dioxide. In: Encyclo-pedia of aquaculture (ed. R.R. Stickney). John Wiley& Sons.

22. Summerfelt, S.T., Davidson, J. Waldorp, T. (2003)Evaluation of full-scale carbon dioxide strippingcolumns in coldwater recirculation systems. Aquacul-tural Engineering, 28: 155–169.

23. Lunde, T. (1987) Oksygenering. In: Vannrensing vedakvakulturanlegg. NIF kurs (in Norwegian).

24. Boyd, C.E. (1998) Pond water aeration systems.Aquacultural Engineering, 18: 9–40.

25. Cancino, B., Rothe, P., Reuss, M. (2004) Design of highefficiency surface aerators. Aquacultural Engineering,31: 83–98.

9Ammonia Removal

9.2 Biological removal of ammonium ionWhen using a biological filter, bacteria are used tooxidize ammonium to nitrite and nitrate, andperhaps further to molecular nitrogen.The bacteriaare grown in a biofilm.Three processes are includedin the biological removal of ammonium:

• Transfer of NH4+ (ammonium ion) to NO2

−

(nitrite)• Transfer of NO2

− (nitrite) to NO3− (nitrate)

• Transfer of NO3− to N2 (molecular nitrogen)

The two first processes are carried out simultane-ously and are known as nitrification; the process isperformed in a nitrification filter. The third processis denitrification and is performed in a denitrifica-tion filter. The two first are aerobic, so air must beadded. The last process is anaerobic so air must beremoved from the water. Two different filtersinvolving different bacteria are therefore used. Inmost cases only the nitrification process will beneeded for aquaculture purposes, because fish havea higher tolerance for nitrate than for ammonia.Very high degrees of water re-use and high fish den-sities might require a denitrification filter, butknowledge of denitrifiaction filter function andoptimization for fish farming is scant.

9.3 NitrificationThe nitrification process is carried out in two stepsand is performed by bacteria which oxidizeammonia. These bacteria are autotrophic and useO2 as oxidizing agent and CO2 or HCO3

− as a carbon source for growth. NH4

+ is transformed

9.1 IntroductionIn aquaculture it is often necessary to reduce theconcentration of ammonia in the water, because itis toxic for the fish.1–4 It is particularly importantwhen re-using water (Chapter 10) or when trans-porting fish long distances without changing thewater, since the fish produce ammonium com-pounds as a metabolic waste product.

In water there is an equilibrium between the con-centrations of ammonium ion (NH4

+) and ammonia(NH3) (Fig. 9.1):

NH3 + H+ ⇔ NH4+

This equilibrium depends on pH. The sum of NH3

and NH4+ is known as the total ammonia nitrogen

(TAN). Because of the equilibrium between NH3

and NH4+, reducing one of them automa-

tically reduces the other. NH3 is the more toxic substance for fish, and therefore is the substance ofinterest.

This chapter gives an overview of the mostcommon methods for ammonia nitrogen removal infish farming, including biological and chemicalprocesses. There are also a number of other possi-ble methods for removal of ammonia, such asammonia stripping, break-point chlorination, mem-brane filtration and addition of ozone, but thesemethods are not commonly used in aquaculture.Much information is available about this subject inthe general literature for municipal wastewatertreatment.5,6 Removal of nitrogen from municipalwastewater has become more and more importantduring the past few years. In aquaculture thesubject has also become more important due to theincrease in water re-use systems.7–9

121


to NO2− by Nitrosomonas bacteria and then to NO3

−

by Nitrobacter bacteria. Both of these processesrequire energy that is supplied by the substrate.Thechemical processes involved are as follows:

NH4+ + 3–2 O2 → NO2

− + 2H+ + H2O

The bacteria grow in the biofilm on the filtermedium. Nitrification takes place in this film so thebiofilm must be established for the nitrificationfilter to function. The process creates more biofilmas the bacteria grow and divide, and the cell massincreases. This cell mass can be described asC5H7NO2. The processes including creation of cellmass can then be described as follows.10

Step 1 using Nitrosomonas bacteria:

55NH4+ + 5CO2 + 76O2 → C5H7NO2 + 54NO2

−

+ 109H+ + 52H2O

Step 2 using Nitrobacter bacteria:

400NO2 + 5CO2 + 76O2 → C5H7NO2 +400NO3

− + H+

The effectiveness of a nitrification filter can bedescribed by the nitrification rate, defined as theamount of ammonium oxidized per unit biofilmsurface area and unit time (mg NH4

+/(m2 min)). The

NO O NO212 2 3

− −+ →

efficiency of the nitrification process and the estab-lishment of the biofilm process depend on severalfactors.11 It is important that the bacteria grow asoptimally as possible.

The following important factors regulates thegrowth of the bacterial culture:

• Concentration of ammonia• Temperature• Oxygen concentration• pH• Salinity• Organic substances• Toxic substances.

One of the main factors affecting bacterial growthis the amount of ammonia in the water. In fishfarming this concentration is normally too low formaximal growth, compared to municipal waste-water, for instance, where the same processes isused. Generally values above 3mg ammoniumnitrogen per litre are recommended for maximalgrowth.12 Values that are too high may inhibit bac-terial growth.

Since these bacterial cultures requires oxygen togrow, and in order to have transformation ofammonia to nitrate, it is important that sufficientoxygen is available throughout the entire nitrifica-tion process. Experiments have demonstrated areduction in Nitrosomonas activity with oxygen

Figure 9.1 The equilibrium between theconcentrations of ammonium (NH4

+) andammonia (NH3) is pH dependent.

levels below 4mg/l water, while the correspondingvalue for Nitrobacter is 2mg/l.10

Bacterial growth rate depends on the tempera-ture.13–15 Bacterial activity occurs from 0 to 30°Cand increases with temperature: the optimal rangeis around 30°C. Temperatures that are too high willresult in mortality. However, the bacteria may accli-mate to lower temperatures over time, so high bac-terial activity can also be achieved at lowertemperatures. If the temperature reaches valuesbelow 5°C growth will be slow and it will be diffi-cult and time-consuming to establish new bacterialcultures. It is difficult to give general values for thenitrification rate in relation to temperature, due toboth acclimation and the number of other factorsaffecting bacterial metabolism (Table 9.1).

Nitrification also depends on the pH of the water:optimal values are between 8 and 9.18 It was shownexperimentally that the nitrification rate decreasedby 90% when the pH was reduced from 7 to 6.12 Itis important to remember that hydrogen ions (H+)are produced in the nitrification process and,depending on the buffering capacity of the water,may reduce the pH. If the loads on the biofilter arehigh, systems for adjusting the pH are necessary;addition of lime is an example.

The presence of organic matter can inhibit the functioning of the nitrification filter.19–21 Otherbacterial cultures may start to grow inside the filters in competition with nitrification bacteria.Heterotrophic aerobic bacteria may use the organicmatter as a carbon source and replace the nitrifica-tion bacteria because they have a faster growthrate.22 Nitrification will be reduced by increasingthe carbon/nitrogen ratio;23,24 a 60–70% reductionin nitrification rate was observed when increasingthe chemical oxygen demand/nitrogen (COD/N)

ratio from 0 to 3 for a substrate containing 10mgTAN.l,25

Some substances may be toxic to nitrificationbacteria and may kill or inhibit the bacterial culturein the biofilter. These can include metal ions,organic substances or medicaments such asformaldehyde.26,27

Salinity affects the nitrification rate because thechloride ions inhibit bacterial growth. For thisreason the nitrification is faster in freshwater thanin seawater.28,29

Nitrification filters need to be shielded from light,because it may reduce nitrification.11,30 Predatorsmay also have a negative influence.

The problem about giving reliable values for thenitrification rate is, of course, that all the factorsaffect each other. For this reason it is also normalto incorporate high safeguards when designingnitrification filters.

9.4 Construction of nitrification filtersThe main purpose when constructing a nitrificationfilter is to create a surface for optimal growth of thebiofilm. Depending on the construction and thefilter medium on which the biofilm is established, itis possible to distinguish four types of biologicalfilter:

• Flow-through system• Bioreactor• Fluid bed/active sludge• Granular filters/bead filters.

9.4.1 Flow-through system

The flow-through system may again be divided intothree types depending on how the water flowsthrough the filter medium (Fig. 9.2):

• Trickling filter• Submerged up-flowing system• Submerged down-flowing system.

In a trickling filter, the water trickles through thefilter medium where the biofilm is established.31–33

The filter medium is located above the surface ofthe water and it is very similar to a column aerator.Advantages of this filter type are its simple con-struction, that good natural aeration for the processis achieved and that it is impossible to block. The

Ammonia Removal 123

Table 9.1 Large variations in nitrification rates areobserved at different temperatures and in differentexperiments. The average of values from refs 13, 16 and17 are shown.

Temperature (°C) Nitrification rate (g NH4/(m2 day))

5 0.310 0.515 0.820 1.0


disadvantage is that it has a low nitrification capac-ity compared to other filter types.

In an up-flowing filter the water is supplied underthe filter medium, which is submerged. The incom-ing water is forced through the filter medium by theflow and head of the water. There are severaladvantages with this system: it provides goodmedium structure, good distribution of the water inthe total medium volume is possible, and there isgood contact between the water and the filtermedium where the biofilm grows. The disadvantageis that it is necessary to add oxygen or air with positive pressure, for example by using an air

blower, below the filter medium to ensure that thebacteria in the biofilm get enough oxygen.

A down-flowing filter uses a similar technique to an up-flowing filter; here also the filter mediumis submerged. The water is supplied at the top of the filter and is distributed through it. There is a‘water lock’ at the bottom of the filter and thisensures the filter medium is always submerged. Anadvantage with this filter type is that the water flowsin the opposite direction to the air bubbles, so quitegood mixing of the air/oxygen and water isachieved. The disadvantage is the same as that forthe down-flowing filter: it is necessary to supply

A

B

C

Figure 9.2 Different designs of biofilter with a fixed medium: (A) trickling filter and (B) submerged up-flowing filter;(C) up-flowing filter in use.

oxygen or air; however, in addition, it is difficult to obtain the same good distribution of water in the total volume of the filter medium as in an up-flowing filter.

9.4.2 The filter medium in the biofilter

It is important that the surface of the filter mater-ial on which the biofilm grows is optimal. In fishfarming where the ammonia concentration andconcentration of organic matter are low comparedto levels in municipal wastewater, it is most effec-tive to use systems where a biofilm is established onan artificial surface.17 The following requirementsmust be met by the filter medium:

• Have large specific surface areas where biofilmcan be established (surface area per unit volume,m2/m3)

• Ensure proper contact between the water andthe surface of the medium

• Create low head loss• Be difficult to clog• Ensure an even distribution of water in the entire

filter medium• Be simple to clean.

In practical farming conditions it is difficult to fulfillall these requirements. There is a conflict betweenhaving a nitrification rate as high as possible andthe need for a simply operated and maintainedsystem.

In the past gravel and sand were much used asfilter media in biological filters. Today different

types of plastic filter media have replaced sand andgravel in many applications34 (Fig. 9.3). This isbecause they have a have large specific surfaceareas where biofilm can be established, and they arenot so easily clogged. Leca (lightweight clay aggre-gate) is a material with very large specific surfaceare that has been used to create surface for biofilmwith good results.35 However, there are problemswith clogging when the grain size is small and waterflow large.

9.4.3 Rotating biofilter (biodrum)

A rotating biofilter, also known as a biodrum orrotating biological contactor (RBC), utilizes thesame basic principle as the submerged filter (Fig.9.4). The filter medium where the biofilm is estab-lished rotates at 2–3rpm, partially above the watersurface and partly submerged. The oxygen neces-sary for nitrification is supplied when the mediumis above the water surface. Two different designs ofrotating biofilter are used: (1) a cylinder filled withbiobodies (filter medium in small elements); (2)parallel discs made of plates where the biofilmgrows on the surfaces. In both cases the biofilter ismounted on a rotating shaft.

The advantage of this type of filter is that as a result of the rotation aeration occurs when the filter medium is above the water surface, sonitrification can occur. In such systems the effi-ciency is quite low and it may be necessary to addextra air/oxygen to the tank where the drum isrotating.

Ammonia Removal 125

Figure 9.3 Different filter mediaused in a biofilter.


9.4.4 Fluid bed/active sludge

In an active sludge reactor the bacteria are attachedto suspended sludge material and kept floating inthe water column. The nitrification process occurson the surface of the suspended solids. Because ofthe relatively low content of ammonia, suspendedsolids and organic substances, this is not an appro-

priate method for use in aquaculture. The methodis, however, widely used for treatment of municipalwastewater.

A fairly new method is to utilize a so called flu-idized bed reactor, which has much in common withthe active sludge method (Fig. 9.5). Here an artifi-cial filter medium, normally plastic, floats in an up-flowing current of water and air bubbles (supplied

A

B

C

D

Figure 9.4 Rotating biofilter: (A) rotatingdrum filled with biobodies; (B) rotatingbioplates of drum filled with biobodies; (C)cross-section; (D) filter in use.

by diffusers at the bottom). The medium is main-tained in suspension by the current, like a foun-tain.36 Specially designed filter medium is used forthis purpose, with a density slightly below that ofwater, so it is quite easy to keep fluidized. In addi-tion, the medium is in small elements, so has a largespecific surface area and will float in rotating move-ments in the up-flowing water. A filter grating isnecessary on the outlet to prevent the elements fol-lowing the water out of the reactor.

Biological nitrification takes place on the surfaceof the elements where the biofilm is established. Itis very important to supply enough oxygen/air inthis system to provide optimal growth conditionsfor the biofilm on the elements. Correct construc-tion of the reactor where the elements are floatingis important if high nitrification rates are to beachieved. One advantage with this system is that itis difficult to clog. A quite stable water flow is,however, important to obtain optimum functioning.The efficiency of such filters is also quite goodbecause old biofilm is constantly removed by the airbubbles, the water flow and the rotation of the ele-ments, which all erode to the biofilm. Only new thinfresh biofilm is therefore established on the filtermedium elements. Thin new biofilm has beenshown to have the highest nitrification rate. Thenitrification process is reduced in thick film becauseit is difficult to get enough oxygen and nutrients to the deeper layers. Plastic media have been shownto produce good results in several water re-useplants.

9.4.5 Granular filters/bead filters

A granular filter used for removal of ammonia isthe same as that used for particle removal. Hencetwo operations can be carried out by one filter. Thistype of filter will, however, require frequent back-washing to remove particles.

A bead filter is a type of granular filter, which iscommonly filled with spherical plastic pellets (poly-ethylene (PE)/or polypropylene (PP)) with slightlypositive buoyancy.37,38 The water flows upwardsthrough the layer of plastic pellets and biofilm iscreated on the pellet surface. Back-washing of thefilter medium to avoid blockage can be performedwith air bubbles, circulating water or mechanicallyby paddles or propellers. Experiments have shownreduced nitrification rates in such filters comparedto other flow-through systems, even if on a pervolume basis they are quite effective as a result oftheir the large surface area.39,40

Traditional up-flowing sand filters may also beused as biofilters, but the nitrification rate is lowand requirements for maintenance and back-washing are high.

9.5 Management of biological filtersA biofilter will always need a certain start-up timebefore it becomes functional41,42 (Fig. 9.6) becauseit takes some times to establish the culture of nitri-fying bacteria on the filter medium (the biofilm). Ofcourse there must be nitrificants in the water andthe environment must be suitable, for instance with

Ammonia Removal 127

Figure 9.5 A fluidized bed reactor: biofilmgrows on artificial medium floating in the water.


a water temperature above 7–8°C, dissolved oxygencontent above 4mg/l and biological oxygen demandat 7°C (BOD7) values below 10mg/l.7

If a biofilter is put into service before the bacte-rial culture has been established, there will be no orlow efficiency. Normally, the start-up time for thefilter is between 20 and 40 days, depending upon,among other factors, temperature and pH.28,35 Thestart-up time can be reduced by inoculation orplacing old bacterial cultures in the filter. It isnormal to only clean a part of the biofilter eachtime so that there will always be some biofilm lefton a parts of the filter. Old filter surfaces will alwayshave some film left, so the start-up time for a pre-viously used filter surface will be reduced comparedto completely new ones.

In the start-up period, some nitrite may be sup-plied to the fish with the water that has passedthrough the biofilter because this is the end productof the first process that takes place in the filter. Thisis therefore a critical period, because small concen-trations of nitrite can be toxic to the fish and caremust therefore be taken at this stage. The bestcourse of action is to have as few fish as possible inthe start-up period or to inoculate the filter withbacteria.

9.6 Example of biofilter designIn a fish tank the consumption of feed is 100kg/day.The amount of TAN produced by the fish, dependson growth rate, protein content in the feed and inthe fish, amount of nitrogen in the protein, and theprotein uptake and digestion. A secretion of TANto be between 30 and 40g per kg feed can be usedfor further calculation. The water temperature is

20°C and the pH is 7. The design and size of thebiofilter can be calculated as follows.

The secretion of TAN will be 3–4kg per day. AtpH 7 and a water temperature of 20°C only 0.4%is NH3 (see Fig. 9.1). The TAN concentration istherefore almost equal to the the concentration ofNH4

+.Under normal conditions the nitrification rate

mainly depends on water temperature. From Table9.1 it can be ascertained that at 20°C a value of 1.0g NH4

+ per m2 biofilter surface area per day canbe used. The biofilter medium must therefore havea surface area (Ba) of:

Ba = (3000 to 4000)/1= 3000m2 to 4000m2

It is normal to have a security factor when designing the biofilter: a value of 2 is quite commonly used. The necessary area is therefore8000m2.

The next steep is to choose a filter medium to fitthe necessary volume. If choosing a filter with asurface area of 300m2 per m3 filter medium, thetotal volume required will be 8000/300m3, which isclose to 27m3. Actual designs use submerged up-flowing filters or trickling filters. A combination isalso quite commonly used. The recommended ratioif both are used is from 2 :1 to 1 :1. This means thatthe filter can have an 18m3 submerged filter and a9m3 trickling filter.

9.7 DenitrificationIn a biological filter ammonium (NH4

+) is trans-formed into nitrate (NO3

−). With a high degree ofwater re-use (see Chapter 10) and large loads onthe filter (high fish biomass and rapid growth) theamount of NO3

− may exceed the tolerance of thefish, and result in mortality.To inhibit this, it may benecessary to add a new filter with a denitrificationstep. A denitrification filter transforms nitrate tonitrite (NO2

−) and further on to nitrogen gas (N2),which is degassed from the water. Denitrification isfor instance used in some eel farms that have a veryhigh degree of water re-use.

Biofilm-hosting bacteria may also carry out den-itrification.This is an anaerobic two-step process. Inthe first step nitrate is transformed to nitrite and inthe second step nitrite to nitrogen gas. The bacteriathat carry out this process need a supply of organic

Figure 9.6 A biofilter will require a certain start-up timebefore being fully functional.

carbon for growth, not oxygen. Normally there isnot enough easily available organic carbon in the outlet water so carbon must be supplied.Carbon is commonly added as methanol, ethanol or sugar in liquid form before the denitrificationfilter.When using methanol the following equationsapply:

Step 1

CH3OH + 3NO3− → 3NO2

− + CO2 + 2H2O

Step 2

CH3OH + 2NO2− → N2 + CO2 + H2O + 2OH−

If the created cell mass is described empirically bythe formula C6H7O2N, the total equation, includingsteps 1 and 2, is as follows:4

CH3OH + 0.92NO3− + 0.92H+ → 0.06C6H7O2N +

0.43N2 + 0.7CO2 + 2.25H2O

Since denitrification is an anaerobic process itwill not happen in the presence of oxygen. There-fore it is necessary to remove the oxygen from thewater to get denitrifaction to start. Methanol andethanol may also be used as a means of removingthe oxygen; after addition of methanol or ethanolthere will be no free oxygen left in the water (i.e.the concentration of dissolved oxygen will be closeof 0mg/l).

The design of the denitrification filters is thesame as that of nitrification filters, requiring a largearea where the bacteria can grow. An establishedbiofilm is also necessary. However, the filtermedium must be submerged to avoid ingress ofoxygen. Trickling filters and drum filters are nottherefore used.

9.8 Chemical removal of ammonia

9.8.1 Principle

Ion exchangers have sometimes been used in aqua-culture to remove ammonia.43,44 Ion exchangersutilize the fact that different ions have differentelectrical charges. Some substances have the abilityto attract specific ions in the water because of theircharge and exchange them with ions attached lessstrongly. Ion exchangers may be divided into cationand anion exchangers. For example, cationexchangers are used to remove positively charged

calcium and magnesium ions from hard water,while anion exchangers can be used to remove neg-atively charged nitrate ions. With ammonia, it is theammonium ion (NH4

+) that is removed from thewater, so a cation exchanger is required. Removalof ammonium ions ensures a reduction of ammonialevels because of the equilibrium betweenammonia and ammonium.

9.8.2 Construction

An ion exchanger comprises a column filled withgravel-like ion exchange substance (Fig. 9.7). Thewater to be purified, for example by removal ofNH4

+, flows through this column. The choice of substance to be loaded into the column depends onthe ions to be exchanged. To remove NH4

+ a claymineral (zeolite) called clinoptilolite if often used.This mineral occurs naturally in the ground incertain places. The ion exchange substance is deli-vered as a granulate, normally in sizes up to 5mm.When using clinoptilolite, Na+ is used as theexchange ion and NH4

+ is attracted to the ionexchange substance (R−). The following equationthen applies:

NH4+ + R− Na+ ⇔ Na+ + R− NH4

+

When water with high ammonia content enters theion exchanger, NH4

+ will be attracted to the ionexchange substance and Na+ will be released. After

Ammonia Removal 129

Figure 9.7 Principle of an ion exchanger.


a period of time all the Na+ in the ion exchange sub-stance will have been released and the exchangeprocess will stop; all the NH4

+ in the water will flowdirectly through the column, and the outlet con-centration will be the same as the inlet concentra-tion. The ion exchanger will then have to beregenerated by removing all the NH4

+ bound to theion exchange substance. This can be carried out byallowing a regenerating solution containing a veryhigh concentration of sodium ions, for example asalt solution, to flow through the ion exchanger(Fig. 9.8). Often the regenerating solution is re-used; in this case a cleaning circuit for the salt solu-tion is necessary. A stripping process may be usedto clean the regenerating solution for NH4

+.To function optimally the water entering the

exchanger must be free of particles. A fine particlefilter must therefore be installed before theexchanger. Furthermore, the substance in the ionexchanger must be disinfected at specific intervalsto avoid the exchanger starting to function as a biological filter. This is because biofilm will beestablished on the surface of the exchange sub-stance. The capacity as a biofilter will be very lowcompared to the exchange capacity. Clogging andblockage will also occur rapidly.

Even if the ion exchanger can remove up to 95%of the ammonia from freshwater, the economicyield of the system is normally negative. The system requires several columns with separate

regenerating circuits so that maintenance can becarried out, and to avoid breakdown of the wholesystem if a column is out of order. Because of this,very few fish farms use ion exchanger technologytoday. The advantage of an ion exchanger com-pared to a biofilter for nitrification, is that there isno start-up time and no waste product such asnitrate. There have been examples where this tech-nology has been used in relation to long distancefish transport (Chapter 17).

References1. SECL (1983) Summary of water quality criteria for

salmonid hatcheries. SECL 8067, Report to Depart-ment of Fisheries and Oceans, Vancouver, BC.

2. Mead, J.W. (1985) Allowable ammonia for fishculture. Progress in Fish Culture, 47: 135–147.

3. Knoph, M.B. (1995) Toxicity of ammonia to Atlanticsalmon (Salmo salar L.). PhD thesis. Department ofFisheries and Marine Biology, University of Bergen.

4. Russo, R.C., Thurtston, R.V. (1991) Toxicity ofammonia, nitrite and nitrates to fishes. In: Aquacul-ture and Water Quality. Advances in World Aquacul-ture, vol. 3 (eds D.E. Brune, Thomasso, J.R.). WorldAquaculture Society.

5. Montgomery, J.M. (1985) Water treatment: principlesand design. John Wiley & Sons.


7. Gebauer, R., Eggen, G., Hansen, E. og Eikebrokk, B.(1992) Oppdrettsteknologi – vannkvalitet og vannbe-handling i lukkede oppdrettsanlegg. Tapir Forlag (inNorwegian).

Figure 9.8 Use of an ion exchanger toremove ammonia chemically.

8. Van Rijn, J. (1996) The potential for integrated biological treatment systems in recirculating fishculture – a review. Aquaculture, 3–4: 181–210.

9. Timmons, M.B., Losordo, T.M. (1997) Aquaculturewater reuse systems: engineering design and manage-ment, 2nd ed. Elsevier Science.

10. Haug, R.T., McCarty, P.L. (1971) Nitrification withsubmerged filters. Water Pollution Control Federation,44: 2086–2102.

11. Wheaton, F., Hochheimer, J., Kaiser, G.E. (1991)Fixed film nitrification filters for aquaculture. In:Aquaculture and Water Quality. Advances in WorldAquaculture, vol. 3 (eds D.E. Brune, Thomasso, J.R.).World Aquaculture Society.

12. Odegaard, H. (1992) Wastewater treatment. TapirForlag (in Norwegian).

13. Wortman, B., Wheaton, F. (1991) Temperature effectson biodrum nitrification. Aquacultural Engineering,10: 183–205.

14. Fdez-Polanco, F., Villaverder, S., Garcia, P.A. (1994)Temperature effect on nitrifying bacteria activity inbiofilters – Activation and free ammonia inhibition.Water Science and Technology, 30: 121–130.

15. Zhu, S., Chen, S. (2002) The impact of temperature onnitrification rate in fixed film biofilters. AquaculturalEngineering, 26: 221–237.

16. Speece, R.E. (1973) Trout metabolism characteristicsand rational design of nitrification facilities for waterreuse in hatcheries. Transactions of the American FishSociety, 2: 323–333.

17. Rusten, B. (1986) Ammoniakkfjerning i resirkuler-ingsanlegg for fiskeoppdrett. In: Vannbehandling iakvakultur (ed. H. Odegaard). Tapir Forlag (in Nor-wegian).

18. Henze, M., Harremoës, P. (1990) Chemical–biologicalnutrient removal: the hypro concept. In: Chemicalwater and wastewater treatment (eds H.H. Hann, R.Klute). Springer Verlag.

19. Hem, L.J, Rusten, B., Ødegaard, H (1994) Nitrifica-tion in a moving bed biofilm reactor. Water Research,28: 1425–1433.

20. van Loosdrecht, M.C.M, Tijhuis, L, Wijdieks, A.M.S.,Heijnen, J.J. (1995) Population distribution in aerobicbiofilms on small suspended particles, Water Scienceand Technology, 31: 163–171.

21. Bovendeur, J., Zwaga, B.A., Lobee, B.G.J., Blom, J.H.(1990) Fixed-biofilm reactors in aquacultural waterrecycle system: effect of organic matter eliminationon nitrification kinetics. Water Research, 24: 207–213.

22. Grady, C.P.L., Lime, H.C. (1980) Biological waste-water treatment – Theory and applications. MarcelDekker.

23. Zhu, S., Chen, S. (2001) Effect of organic carbon onnitrification rate in fixed film biofilters. AquaculturalEngineering, 25: 1–11.

24. Carrera, J., Vicent, T., Lafuente, J. (2004) Effect ofinfluent COD/N ratio on biological nitrogen removal(BNR) from high strength ammonia industrial waste-water. Process Biochemical, 39: 2035–2041.

25. Ling, J., Chen, S. (2005) Impact of organic carbon onnitrification performance of different biofilters.Aquacultural Engineering, 33: 150–162.

26. Levine, G., Meade, T.L. (1975) The effects of diseasetreatment on nitrification in closed systems aquacul-ture. University of Rhode Island.

27. Kaiser, G.E.,Wheaton, F.W. (1983) Nitrification filtersfor aquatic culture systems: state of the art. Journal ofthe World Mariculture Society, 14: 302–324.

28. Bovendeur, J. (1989) Fixed-biofilm reactors applied towaste water treatment and aquacultural water recircu-lating system. PhD thesis, Agricultural University,Wageningen.

29. Nijhof, M., Bovendeur, J. (1990) Fixed film nitrifica-tion characteristic in seawater recirculating fishculture systems. Aquaculture, 87: 133–143.

30. Alleman, J.E., Preston, K. (1991) Behaviour and phy-siology of nitrifying bacteria. In: Commercial aqua-culture using water recirculating (ed. L. Swann)Illinois State University.

31. Thorn, M., Mattsson, A., Sorensson, F. (1996) Biofilmdevelopment in a nitrifying trickling filter. WaterScience and Technology, 34: 89–83.

32. Wik, T., Mattsson, A., Hansson, E., Niklasson, C.(1996) Nitrification in a tertiary trickling filter at highhydraulic loads – Pilot plant operation and mathe-matical modeling. Water Science and Technology, 32:185–192.

33. Parker, D.S., Jacobs, T., Bower, E., Stowe, D.W.,Farmer, G. (1997) Maximising trickling filter nitrifica-tion rates through biofilm control: research reviewand full scale application. Water Science and Technol-ogy, 36: 255–262.

34. Kruner, G., Rosenthal, H. (1983) Efficiency of nitrifi-cation in trickling filters using different substrates.Aquacultural Engineering, 2: 49–67.

35. Lekang, O.I., Kleppe, H. (2000) Efficiency of nitrifi-cation in trickling filters using different filter media.Aquacultural Engineering, 21: 181–201.

36. Rusten, B., Hem, L.J., Odegaard, H. (1995) Nitrifica-tion of municipal waste water in moving-bed biofilmreactors. Water Environmental Research, 67: 75–86.


38. Malone, R.F., Beecher, L.E. (2000) Use of float beadfilters to recondition recircualting waters in warmwater aquaculture production systems. AquaculturalEngineering, 22: 57–73.

39. Singh, S., Ebling, J., Wheaton, F. (1999) Water qualitytrials in four recirculating aquaculture systems. Aqua-cultural Engineering, 20: 75–84.

40. Greiner, A.B., Timmons, M.B. (1998.) Evaluation ofthe nitrification rates of microbead and tricklingfilters. Aquacultural Engineering, 18: 189–200.

41. Muir, J.F. (1982) Recirculating water systems in aqua-culture. In: Recent advances in aquaculture (eds. J.F.Muir, R. Roberts). Westview Press.

42. Skjølstrup, J., Nielsen, P.H., Frier, J.O., Mclean, E.(1997) Biofilters in recirculating aquaculture systems

Ammonia Removal 131


– state of the art. In: Technical solutions in the man-agement of environmental effects of aquaculture (ed.J. Makkonen). Scandinavian Association of Agricul-tural Scientists, Seminar 258, pp. 33–42.

43. Dryden, H.T., Weatherly, L.R. (1989) Aquaculturewater treatment by ion exchange with clinoptilolite.Aquaculture Engineering, 8: 109–126.

44. Rosental, H. (1993) The history of recycling technol-ogy. A lesson learned from past experience? In: Fishfarming technology (eds H. Reinertsen, L.A. Dahle,L. Jørgensen, K. Tvinnereim). Proceedings of the firstinternational conference on fish farming technology.A.A. Balkema.

10Recirculation and Water Re-use Systems

10.2 Advantages and disadvantages ofre-use systems

10.2.1 Advantages

Reduction of water flow

Re-use of water will reduce the amount of newwater required for the fish farm. Therefore farmscan be established on sites where the amount ofwater is a limiting factor, or established farms canincrease production without increasing the amountof new water required.

Limited resourses of freshwater are today aserious problem in the world. Water consumptionhas shown exponential growth during the past fewyears. Warning signs are evident, such as loweredgroundwater tables, reduced size of lakes and dis-appearance of marshland.This is indicative of morecompetition for freshwater resources in the future,which will of course affect the fish farming industrybecause of its huge consumption of freshwater. Allmethods that reduce water consumption in fishfarming, such as re-using water, are therefore ofgeneral interest.

Re-use of energy

As shown in previous chapters, the heating of waterrequires much energy and for this reason is expen-sive. By reducing the amount of new water

10.1 IntroductionIn water re-use or recirculation systems, the outletwater from the fish tanks is re-used instead of beingreleased into a recipient water body (Fig. 10.1). Theoutlet water is cleaned and used again, which meansthat the amount of added new water can bereduced. Theoretically, all the outlet water can bere-used, as in aquaria where no new water isrequired except for that which is lost through evap-oration. This is only theoretical, however, becausein most cases the cost of removing all contaminantsfrom the outlet water is very high, but this is ofcourse dependent on the water quality require-ments of the fish species being farmed. Usually, ifmaximum growth is required the water qualityneeds to be high.

A water re-use system includes the fish tank(s)or units for the aquaculture species, an adaptedwater treatment system and a pump to transportthe water around the system. The pump and thewater treatment system are the items that make thesystem distinct from traditional flow-throughsystems. The water treatment system, which is theheart of the re-use system, may include physical,chemical and biological processes to improve thewater quality to acceptable levels.

The aim of this chapter is to describe importantfactors about re-use systems, definitions and equip-ment. Specialized literature is available in this field(ref.1 is recommended for an overview).

133


supplied, the energy requirements for water heatingwill also be reduced; again this will reduce the totalheating costs.

By using a water re-use system, it is possible tofarm fish species that have higher temperaturerequirements than the natural temperature in thearea, for instance to grow warm-water species athigh latitudes in the northern hemisphere.

Simpler cleaning

If there are stringent requirements for cleaning theeffluent water, re-use systems will assist the processbecause the amount of water to be treated isreduced.

Poor water sources

If the water supply to the farm is of poor quality,the requirements for improvement will beincreased. Re-use systems will be of interest in such cases because the amount of new incomingwater, where the quality must be improved, isreduced together with the treatment costs.

If the inlet water has to be pumped to a higherlevel to get it to the farm, the costs can be consider-able; this may favour a re-use system.The same is thecase if new water supplied to the farm is metered.

10.2.2 Disadvantages of re-use systems

Although re-use systems have advantages, they alsohave several disadvantages; these must therefore be

weighed against each other. In most cases the dis-advantages are greater than the advantages. It isbest therefore, to have a site with enough goodquality water of the correct temperature to suit the species grown, and low costs associated withtransferring the inlet water from the source to thefarm.

The two main disadvantages of re-use systemsare the investment and operating costs. Because thenumber and size of the components for water treat-ment is higher than for a flow-through farm, theinvestment costs are also higher. In systems with ahigh degree of re-use (>95%) the investments canbe significantly greater than for traditional flow-through farms, several times as high per unitfarming volume.

In a normal re-use system there is continuoustransport of water performed by some kind ofpump, which results in constant running costs forthe pump(s). Since so much technology is usuallyinvolved in purification systems, a re-use system willalso be more exposed to technical faults. To ensurea functioning system, the requirements for moni-toring water quality and water flow are greater thanin traditional flow-through systems which translatesto larger monitoring systems and more back-upsystems. Furthermore, the time limit/reaction timeis reduced when a fault occurs, e.g. pump failure,filter blockage, which increases the requirementsfor having operators on stand-by.

Some of the equipment used in the re-use systemalso requires a high level of technological and bio-logical knowledge to operate; this imposes extrarequirements for the competence of farm opera-tors. The need for maintenance of the equipment ismuch higher, which also represents a significantcost.

10.3 DefinitionsTo describe re-use or recycling systems a number ofparameters are required.

10.3.1 Degree of re-use

Normally the degree of re-use (R) is used to givethe percentage of the new incoming water (QN) inrelation to the total amount of water (QT) flowinginto the fish tank. This is described by the followingrelationship (Fig. 10.2):

Figure 10.1 Re-use system compared to a traditionallyflow through system, remember the pump.

R = (1 − (QN/QT)) × 100

where:

R = degree of re-useQN = new incoming waterQT = total water supply to the fish tank

= new incoming water + re-used water.

ExampleIf the supply of new water to a farm is 150 l/min andthe total water flow is 1000 l/min calculate the degreeof re-use.

R = (1 − (QN/QT)) × 100= (1 − (150/1000)) × 100= 85%

This, however, requires a continuous supply of newwater, which also represents the normal way tooperate a re-use farm.

Another way to operate a re-use system is tochange the water in batches. This is the same as isdone in aquaria, so information and knowledgefrom the operation of aquaria can be transferred tofish farming. However, in fish farming the emphasisis on optimal growth and this requires optimal envi-ronmental conditions; fish densities are also muchhigher. If the aim is for the fish to survive with somespare capacity, the requirement for water quality is

reduced. For a re-use system that changes the waterin batches, a given amount of water can be changedonce a day or once a week, for instance. In this casethe water quality will gradually decrease, until a new batch is exchanged and the quality returns to top level. In this case the degree of re-use can be expressed either as the percentage of waterexchanged in relation to the total flow in the systemthroughout a day and night, or it can be given asthe amount of water that is exchanged in rela-tion to the total volume of water in the system (Fig. 10.2):

R = (1 − (QB/QT)) × 100

where:

R = degree of re-useQB = size of batch of new incoming waterQT = total amount of water flowing in the system

between the exchanged batches of water.

R = (1 − (QB/QT)) × 100

where:

R = degree of re-useQB = size of batch of new incoming waterQT = total amount of water in the re-use system.

ExampleThe total amount of water in the system includingthe fish tank(s) and re-use circuit is 1000 l. The inter-nal water flow in the system is 10 l/min. The amountof water in the exchange batch is 250 l and it isexchanged once a day. Calculate the degree of recy-cling based on both definitions.

R = (1 − (QB/QT)) × 100= (1 − (250/1000)) × 100= 75%

Total water flow in the system during a day: 10 l/min¥ 60min ¥ 24h = 14400 l.

R = (1 − (QN/QT)) × 100= (1 − (250/14400) × 100= 98.3%

This example shows the importance of knowingwhich definition of re-use is used. Systems with con-tinuous exchange of water and not batch systemsare most commonly used.

Recirculation and Water Re-use Systems 135

Figure 10.2 The degree of re-use can be defined inseveral ways. There can either be a continuous supplyof new water, or there can be a batch exchange of water,for instance once a day.


10.3.2 Water exchange in relation to amount of fish

The degree of re-use does not, however, giveenough information about the performance of thesystem and is not sufficient to describe a recircula-tion system properly. Neither does it take intoaccount the amount of fish in the system. This canbe illustrated in the following way: there are fewproblems in having only one fish in a large re-usesystem, compared to having a high density of fish inthe same system. The degree of re-use for thesystems can, however, be equal. To describe the re-use system it will also be necessary to know theamount of new water added per kg fish (litres newwater per kg fish).

ExampleA re-use system with a total tank volume of 100m3

has a total circulating water flow of 2000 l/min. There-use degree is 95%, meaning that the amount ofadded new water is 100 l/min. The fish density is incase 1, 10kg/m3 and in case 2, 100kg/m3. This repre-sents a total amount of fish of respectively, 1000kgand 10000kg, so the amount of new water in thesecases is 0.1 l/(min kg fish) and 0.01 l/(min kg fish).

Even with this information it is difficult tocompare re-use systems, because factors such asspecies, size and growth rate will have effects. Foreasily evaluating a separate re-use system, thegrowth rate of the fish is the best indicator. If thegrowth rate is optimal it can either be compared togrowth tables or to growth in a flow-throughsystem.

10.3.3 Degree of purification

Another important factor in re-use systems is thedegree of cleaning CP of the water treatmentsystem. This factor indicates how effectively thecleaning plant in the re-use circuit removesunwanted substances. It can be described as follows(Fig. 10.3):

CP = (Cin − Cout)/Cin × 100

where:

CP = degree of purificationCin = concentration of actual substance entering

the cleaning unit

Cout = concentration of actual substance leaving thecleaning unit.

ExampleThe concentration of suspended solids entering theparticle filter is 25mg/l; after the filter the concentra-tion is measured as 10mg/l. Find the effectiveness ofthe filter.

CP = (Cin − Cout)/Cin × 100= (25mg/l − 10mg/l)/25mg/l × 100= 60%

10.4 Theoretical models forconstruction of re-use systems

10.4.1 Mass flow in the system

To understand more about what is happening in are-use system, a theoretical approach can betaken.1–4 In re-use systems is it important to havecontrol over the different water flows (Q), amountof substances added and removed from the system(M) and, based on this, the concentration of sub-stances at different stages in the system (C):

C = M/Q

where:

Q = water flow (l/min)M = mass flow of substances (mg/min)C = concentration of substances (mg/l).

Based on this, a total picture of the re-use system,including the fish tanks, the water treatment systemand water transport system can be drawn up asshown in (Fig. 10.4). Here the following indices areused:

Figure 10.3 The filters in the re-use circuit removessubstances according to their efficiency.

i = new water into the tanko = out from the tank to outlet (= Qi + Qri)ri = out from the tank and into the re-use circuitro = out of the re-use circuit and into the tank (Qri)f = amount of substances produced or consumed

by the fishr = amount of substances removed from the re-use

system.

10.4.2 Water requirements of the system

The necessary amount of flowing water in thesystem, and the separation of new and re-usedwater depends on a number of factors including:

• Amount of water to satisfy the oxygen require-ments of the fish

• Amount of water to dilute and remove wasteproducts to acceptable levels

• Amount of water to ensure self-cleaning of thetanks (see Chapter 13)

• Degree of re-use• Effectiveness of the water treatment system.

Water flow to satisfy oxygen requirements of the fish

The addition of water to a fish tank to satisfy theoxygen requirements depends on the oxygen con-sumption of the fish, the oxygen concentration inthe inlet water and the lowest acceptable concen-tration in the outlet water to achieve optimalgrowth for the fish species. The specific waterrequirements can be calculated from:

Qin = Mf/(Cin − Co)

where:

Qin = specific water flow per kg fishMf = specific oxygen consumption of the fish (mg

O2/(min kg fish))Ci = concentration of oxygen in the inlet water to

the tank (mg/l)Co = concentration of oxygen in the outlet water

from the tank (mg/l).

ExampleThe fish size is 2000g, the water temperature 12°Cand the specific oxygen consumption is 3.63mgO2/(min kg fish). The oxygen concentration in fullysaturated water is 10.8mg/l (from tables). Theacceptable concentration in the outlet is set to 7mg/l.Calculate Qin.

Qin = Mf /(Cin − Co)= 3.63/(10.8 − 7)= 0.96 l/(min kg fish)

Here is it also important to remember the require-ments to dilute for other substances and those forself-cleaning.

By supersaturating the inlet water with pureoxygen the water requirements can, of course, bereduced. This means that the Cin is increased.

ExampleThe same numbers as in the previous example areused, but the inlet water has a supersaturation ofoxygen of 150%, meaning that the concentration is16.2mg/l. Calculate the new Qin.

Qin = Mf /(Cin − Co)= 3.63/(16.2 − 7)= 0.39 l/(min kg fish)

Water flow to dilute waste products to acceptable concentrations

The amount of water required to dilute and removesubstances produced by the fish (suspended solids(SS), CO2 and total ammonia nitrogen (TAN)) toacceptable concentrations can be calculated basedon mass balance equations for the single substances:

Min + Mro + Mf = Mo

where:

Min = mass of substances in new incoming water


Figure 10.4 Sketch showing the water flow, concen-tration of substances and mass flow of substances in are-use system.


Mro = mass of substances from water entering fromthe re-use circuit

Mf = mass of substances produced by the fish inthe tank

Mo = Mass of substances in the outlet from the tank.

If Mo and the water flow out of the tank Qo areknown, the concentration of a substance in the outletfrom the tank (Co) can be calculated; this must notexceed a value that is acceptable for the fish. Basedon this, can the lowest acceptable outlet water flowcan be calculated from the following equation:

Qo ≥ Mo/Co-acc

where:

Co-acc = acceptable concentration of the substancein the outlet to avoid reduction in growth.

ExampleA fish of size 50g has a specific growth during oneday of 31g resulting in waste production measuredas SS = 6.2g. The acceptable level of SS in the outletis set to 25mg/l. Calculate the necessary water flowout (Qo).

SS produced per minute = 6200mg/(24h × 60)= 4.3mg/min

Qo = 4.3mg/min/25mg/l = Mo/Co-acc

= 0.17 l/min

This means that the water flow into the tank must behigher than 0.17 l/min to dilute the concentration ofSS to acceptable levels or less.

10.4.3 Connection between outlet concentration,degree of re-use and effectiveness of the watertreatment system

When starting up a re-use system the concentrationof substances in the system will gradually increaseuntil it is stabilized at a given level.

ExampleA simple re-use system uses a degree of re-use of50%; a filter with 50% efficiency is installed in there-use circuit (Fig. 10.5). Show how many times thewater must circulate in the re-use system before thesystem is in balance (this condition is assumed asideal) regarding the concentration level of metabolicproducts, presented as parts of M.

As shown, the system will be stabilized when thewater has completed four circuits.

An equation to determine the concentration inthe tank outlet in a re-use system (C) compared tothe outlet concentration in a flow-through tank hasbeen developed.2 This is based on the degree of re-use (R) and the removal efficiency (re) of thefilter system in the circuit and is as follows:

C = 1/(1 − R + (R re))

ExampleA system has a degree of re-use of 96%, while theremoval efficiency is 50%. Calculate the concentra-tion in the outlet of the system compared to a tradi-tional flow-through system.

C = 1/(1 − 0.96 + (0.96 × 0.5))= 1.92

This means that the concentration of substances is1.92 times those from a flow-through system. Forinstance, if the flow-through system had an SS con-centration of 20mg/l in the outlet, the concentration

1 (all water No. circuits is new water) 2 3 4 5

Concentration M 5–4 M (M is 21––16M 4–3M 4–3Mof metabolic added toproducts (M) the incomingout of the concentrationproduction unit 1–4M)

Concentration 1–2 M (half of 5–8 M 21––32M 2–3M 2–3Minto re-use the water circuit flow)

Concentration 1–4M (50% is 5–16M 21––64 M 1–3 M 1–3 Mout of re-use removed) bycircuit the filter

Figure 10.5 The system used in the previous example.

in the re-use system will be 20 × 1.92 = 38.4mg/l. Ifthe maximum concentration that the fish toleratewithout growth reduction is 25mg/l, the re-usesystem is not useful because the SS concentration istoo high. Either a better filter must be installed or thedegree of re-use must be lowered, which meansgreater dilution by adding more new water.

Based on this, it is possible to calculate how muchnew water has to be added to have a system thatfunctions. First the maximum allowed SS concen-tration Cmax, is found:

Cmax = 25/20 = 1.25mg/l

Then this value substituted in the formula and in theequation solved for the degree of re-use (R):

C = 1/(1 − R + R re)

1.25 = 1/(1 − R + 0.5R)

R = 0.4

This means that the maximum degree of re-use thatcan be used is 40% and 60% of new water must beadded. In practice, however, a better filter unit will beinstalled instead of adding so much new water.

The general formula is as follows:

C = (1/(1 − R + R re))Mf/Qout

where Mf/Qout represents the outlet concentrationin a flow-through system. By rearranging this equa-tion it can be used to find the necessary efficiencyof a filter system, based on acceptable outlet con-centrations and degree of re-use. The equation mayalso be rearranged and solved to find the accept-able degree of re-use when a filter system has beenchosen.

If the inlet water contains the substances in ques-tion, their concentrations must also be added to theequation. The mass of substances in the new inletwater (Mi) must be calculated using the massbalance equation and be added to Mf. This can be expressed as CinQi, which gives the followingequation:

C = (1/(1 − R + R re))(Mf + (CinQi))/Qout

10.5 Components in a re-use systemFish require oxygen for respiration and producefaecal waste, urine and dissolved substances

released over the gills (Fig. 10.6). The reasons fortreating the water are therefore either to add newsubstances (oxygen) or to dilute waste products.The components required for water treatment,either for addition or removal of substances in there-use circuit, can be calculated based on the equa-tion given earlier in the chapter, and depend onspecies, size and growth rate. The usual order ofnecessary water treatment efforts is shown below,and compared to what happens in a tank were theinlet water flow is reduced (Fig. 10.7).To reduce theamount of incoming water to the tank is actuallythe same as using a re-use system without anypurification units such as particle or ammoniaremoval filters.

If the water is re-used without any treatment, theeffect is the same as reducing the water inletvolume. The first problem is that the concentrationof oxygen in the outlet water will be below recom-mended values; the main reason for adding waterto the tank is to have an outlet oxygen level highenough to achieve maximum growth of the fish. Ifthis water is re-used directly, the oxygen concen-tration will be too low and will result in growthreduction. If the reduction is excessive, mortalitywill occur; values are species-dependent. There arenumerous ways to increase the concentration ofoxygen in the water (see Chapter 8), including theusual method of adding an aerator before the inletto the tank system. In addition, after the water hasbeen aerated up to near 100% saturation pureoxygen can be added to increase the concentrationfurther (supersaturation). Both aeration and oxygenation can also be done directly into the


Figure 10.6 Fish in a system will produce faecal wasteand soluble waste.


tank which will also reduce the requirement foradditional water.

If aeration/oxygenation is carried out, theamount of water added to the system can bereduced. This new value is normally set by theacceptable SS concentration, but may vary withorganism and species. The water flow-through thetanks must now be sufficiently high to dilute the SS concentration to avoid reduction in growth.Methods of achieving this are given in Chapter 5,but usually the outlet water is sent over some kindof micro-strainer where particles are removedbefore the water is recycled. Another way to dealwith this is to use a dual drain outlet with a sepa-rate particle outlet from which the SS are drawn offin 1–5% of the total water flow (see Chapter 5)enabling the rest of the water to be re-used withoutany other treatment.

If the amount of new water is reduced signifi-cantly, the next problem that may occur is accumu-lation of CO2 in the tank to critical levels, becausethe fish release CO2 through the gills as a wasteproduct of metabolism and the concentration in thewater increases. Tank aeration or piping the waterthrough an aerator are two possible solutions to thisproblem; however, normally this has already beendone to increase the concentration of oxygen in thewater. If continuously high concentrations of CO2

are experienced a vacuum aerator may be usedbecause this is more effective for removing CO2

than a traditional aerator designed to increase theoxygen concentration (see Chapter 8).

Reduction of the water flow into the tank canlead to excessive concentrations of NH3 (actuallyTAN, because there is a connection between NH3

and NH4+ levels which depends on the pH). The

normal method of reducing the NH3 concentrationin the system is to use a biofilter in the re-use circuitthat transforms ammonia to less harmful nitrate(NO3).

The normal units included in a re-use system aredescribed above. However, other problems mayoccur when adding a small amount of new water orwith a high degree of re-use. This problem mainlyconcerns the components in the re-use circuit: thecomplexity of the system means that a good knowl-edge of water quality and the way in which differ-ent water quality parameters influence each otheris very important.

There will be a drop in the pH in the system withhigh degrees of re-use. The reasons for this are thatthe biofilter process release H+ ions to the water,and that CO2 is released by the fish. Therefore pHregulation must be included in the water treatmentfor the system, for instance addition of lime.

When having high percentages of water re-use(>99%), high fish densities and a biofilter in thesystem for ammonia removal, the concentration ofnitrate (NO3) can reach values that can be toxic tothe fish if it is not removed. Usually a denitrifica-tion filter is added to the re-use circuit. Here nitrateis transformed to nitrogen gas (N2) that can beremoved by aeration.

High degrees of water re-use will normally alsoincrease the total number of bacteria, some ofwhich might be pathogenic. It is therefore normalto include disinfection in the circuit, for instanceUV irradiation.

Figure 10.7 Diagrams of plants suitable for differentdegrees of water re-use.

10.6 Design of a re-use systemA re-use plant may be established either with con-tinuous addition of new water or batch exchange;the former is most common because it maintainsstable water quality.

Two different principles are used for construction ofre-use plants (Figs.10.8 and 10.9) regardless of whetherthere is continuous or batch exchange of water:

• Centralized re-use system for handling waterfrom several fish tanks


Figure 10.8 A centralized re-use systemserving several fish tanks.

Figure 10.9 Two designs of tank inter-nal re-use system serving only one tank.


• Re-use system placed in a single fish tank, alsoknown as a tank internal re-use system.

Most current re-use systems are based on thecentralized principle. Outlet water from all the fishtanks is colleted in a common pipeline that leadsdirectly to one centrally placed water treatmentsystem which includes all the necessary water treat-ment components. After treatment the water isreturned to the tanks via a common inlet pipe (Fig.10.10). Addition of the new water and removal ofold water is also performed here, either on a con-tinuous or batch basis.

The advantage of a centralized system is thatmore investment can be put in the water treatmentcomponents because they handle more tanks andgreater weight of fish. However, one disadvantage

with this system is that if infection occurs in onetank it will be transferred via the water to all othertanks in the re-use circuit, although this can be elimi-nated or reduced by installing a disinfection plantin the circuit. Another disadvantage with a central-ized water treatment system is that it is more difficult to gradually increase the size of thefarm/system.

In a single tank re-use system, the outlet waterfrom the tank is lead directly into a water treatmentsystem before it is returned to the same tank. Thusevery tank has its own water treatment system andthere is no mixing of water from different tanks.The water treatment system can either be an inte-gral part of the fish tank, partly inside the tankvolume, or it can be a separate external unitattached to the fish tank. Great flexibility is the

Figure 10.10 The re-use plant at theNorwegian University of Life Sciences iscentralized and the water treatmentsystem consists of particle removal facil-ities (swirl separators and drum filter),submerged biofilter filled with bioblocks,dry-placed pumps and addition of ozonefor combined disinfection and oxygena-tion of the water.

major advantage with this system. It is easier suc-cessively to expand a fish farm. In addition, thereare possibilities for better adaptation to individualloads, meaning that the degree of re-use can varyfrom tank to tank, and the single systems can in this way be operated optimally. The risk of spread-ing pathogens between tanks is also eliminatedbecause there is no water connection between thetanks.

The disadvantage is, however, the price, whichinhibits development of the model. Each tankneeds a separate water treatment system; this onlyallows use of low cost simple systems, but even thenit is difficult to compete with centralized systems.The management cost of such systems is alsoincreased because there are several units that haveto be maintained and controlled; for this a largermonitoring system is needed. Such systems are notfavoured when having high fish densities or veryhigh percentage of re-use (>99.5%) because thisnormally includes several high cost steps, such aspH regulation, denitrification and disinfection.What is generally important when constructing awater re-use plant is that the components are com-patible with each other and of the correct size.

In addition to the water treatment components,it is necessary to establish a water flow in the re-usecircuit; this is done by some kind of pump. Typesused include airlift pumps, propeller pumps or cen-trifugal pumps. When using airlift pumps it is pos-

sible to combine aeration with the transport ofwater and by this eliminate the need for traditionalpumps. In tank internal systems with a low degreeof re-use this might be done to create a low costsystem. In larger systems, traditional centrifugalpumps are most commonly used, because systemefficiency is increased. The pump is either dry-placed or submerged. When using submergedpumps some heat will be transferred to the waterfrom the pump, because the pump creates heatwhen running (see Chapter 2); when having a highdegree of re-use this can contribute an importantpart of the total heating needs of warm-waterspecies. In addition, the fish in the system will createheat from their metabolism, so the amount of heatthat must be added is reduced even more.

References1. Timmons, M.B., Ebeling, J.M., Wheaton, F.W.,

Summerfelt, S.T., Vinci, B.J. (2002) Recirculatingaquaculture systems. Cayuga Aqua Ventures.

2. Liao, P.B., Mayo, R.D. (1972) Salmon hatchery waterreuse systems. Aquaculture, 1: 317–355.

3. Gebauer, R., Eggen, G., Hansen, E., Eikebrokk, B.(1992) Fish farming technology – water quality amdwater treatment in closed production farms. TapirForlag (in Norwegian).

4. Langvik-Hansen (1995) Recircualtion of water in fishfarming. Master thesis. Norwegian University of LifeScience (in Norwegian).


11Production Units: a Classification

sumed depends on the fish density, amount of feedsupplied and the growth rate.

The following chapters give more informationabout the most commonly used production units,commencing with those for storing and hatchingeggs, continuing with tanks and ponds, and con-cluding with cages.

11.2 Classification of production unitsThere are a number of different designs of produc-tion units (Fig. 11.1). Several classifications systemscan therefore be used, either directly related to theproduction unit or the production method, whichagain influences the design of the production unit.

11.2.1 Intensive/extensive

One classification of production systems describedin Chapter 1, is intensive, semi-intensive and exten-sive. The same can be used for the production units.In extensive units the biomass is lower than is nor-mally the case for cultivation. An example of a veryextensive rearing unit is the utilization of smalllakes for fish farming purposes. Before use the lakes should be cleared of natural predators. In use,fry, for instance, can be released and harvesting can be done with a seine net when the fish have reach the required size. In dry periods suchsystems may be run without any water or oxygensupply, the only available oxygen being that produced by photosynthesis occurring in the lake.No food is supplied, the only food is that naturallyproduced. More advanced and intensive is the use of artificially created ponds (excavated or

11.1 IntroductionThe aim of the production unit is to create arestricted area where aquatic organisms can bereared under the best possible growth conditions.Ahabitat must be created for the aquatic organismswithin a body of water. In the unit the organisms(apart from eggs) need to have access to food andoxygen; in addition, the waste products must beremoved. Optimal performance of the productionunits is of major importance because they consti-tute the production system on the farm.

When starting to develop production units, theaim is to create an environment that resemblesnatural conditions as much as possible. Over aperiod of time after the species have gone throughbreeding programmes and become more adaptedto farming conditions, the production units mayalso be developed to increase cost effectiveness byachieving greater production coupled with reducedinvestment and operating costs.

The design of the units depends on the organ-isms, for instance whether fish or shellfish are beingfarmed, and will also vary with the species. Forinstance, requirements for flatfish are differentfrom those of pelagic fish, the former requiring alarger bottom area. Requirements may also changewith the development stage and will be different foreggs and on-growing fish.

During the past few years, animal welfare con-siderations have been introduced into fish farming.Therefore the production units should be designedto allow the fish to behave as naturally as possible.

In the production unit oxygen and eventuallyfeed are consumed by the organism, and wasteproducts are released. The amount of oxygen con-

144

Production Units: a Classification 145

A

B

C

Figure 11.1 A number of differentdesigns of closed production units arein use: tanks (A), closed sea cages(B), sea cages (C), ponds (E),raceway (F), tidal basin (G).


D

E

FFigure 11.1 Continued.

embankment ponds). At low density such systemsmay also be run without any supply of water, theecosystem in the pond ensuring proper waterquality in the pond. However, when artificial feedingof the fish in the pond is commenced, it is necessaryto increase the oxygen supply and remove the waste products, thus gradually progressing to moreintensive systems. Tanks allowing high fish densityrepresent intensive production units.

11.2.2 Fully controlled/semi-controlled

Another way to create simple and quite extensiverearing units is to fence in a water volume, either inlakes, rivers or the sea, and so create a restrictedvolume (a pen) where fish or other aquatic organ-isms can be reared. Normally the fence is made of net or wire, but the use of electric barriers has also been proposed. If using a net, the fencemay be established by means of a post planted inthe ground. The area that is fenced in will vary withthe geographical conditions, amongst other things.For example, by restricting an area in the seabetween two narrow necks of water may create avery large farming volume. Dams of natural earthmasses or concrete may also be used to restrictsmall natural bays and give a low cost productionvolume.

Classification can also be based on the possibili-ties for controlling the environment inside the pro-duction unit. If the bottom or walls are made of

fixed materials such as concrete, steel or plastic,control over the water environment may be possi-ble, as in a closed production unit. In addition,a light-tight insulated superstructure will give fullcontrol of the environment. Cages floating in sea-water represent an open production unit where fullcontrol of environmental factors is impossible; theonly factor that is controlled is that the fish are col-leted in a restricted area. However, if the cage isclosed with a tarpaulin, for example, more controlwill be possible (Fig 11.2).

11.2.3 Land based/tidal based/sea based

Units can also be classified depending of wherethey are placed: 1, on land; 2, in the tidal zone; 3,in the water (sea or freshwater) (Fig. 11.2). On land the units will be closed. The water supply andexchange is either gravitational or pumped.

Units in the tidal zone are normally closed, butmay also be open. In the second case the water levelin the unit wills vary according to the tide. If theunit is closed the water may either be pumped orthe tide can be used to ensure water supply andexchange. If using the tide there can be a batchexchange, meaning exchange only occurs when thetide is high. Specially designed valves may also beused either taking in only surface water or onlytaking in bottom water (Fig. 11.3).

Sea-based farms float in the water, normally onthe surface, but submerged units may also be used.


Figure 11.2 Production units can beopen or closed, and placed onshore, inthe tidal zone or in the sea.


Normally they have an open construction, like seacages, where water supply and exchange is ensuredby the natural currents. However, the units can alsobe closed with sealed walls and bottom (Fig. 11.2),in which case a pump must be used to ensure watersupply and exchange. The advantage with suchsystems compared to land-based units is that thepumping head is reduced.

11.2.4 Other

A number of other categories are also possible. Oneis based on the means of water supply andexchange, whether continuous or batch. Another isbased on the investment cost per unit farmingvolume or per unit farming area. Sea cages andponds represent relatively low-cost systems,while circular concrete tanks represent high-costsystems.

If freshwater is fed into a seawater pond, becauseof the density difference, it will form a layer on the

top of the seawater for a period of time. This can beutilized in specially designed production units forArtic charr farming during the winter because theydo not then tolerate full salinity. A freshwater/brackish layer can be kept in a sea cage by havinga tarpaulin skirt in the upper part of the net bagwalls with an open lower netted part and sendingfreshwater in through a pipe (Fig. 11.4).

Another way that this density gradient can be uti-lized is by sending freshwater into a closed lagoonor basin of seawater. Because of the density differ-ence the freshwater will float on top in a separatelayer and a ‘greenhouse effect’ can be achieved.Heat radiation will go down to the seawater, but the reflection is reduced because of the border layer between the saltwater and freshwater so thetemperature of the seawater in the basin willincrease more than that of the freshwater on thesurface. This principle can be used for oyster spatfarming in the northern hemisphere, for example(Fig 11.4).

Figure 11.3 Use of a tidal basin wherespecial valves control the water supply andexchange so only bottom water is takeninto the basin.

11.3 Possibilities for controllingenvironmental impact

It will become increasingly important to control the environmental impacts of aquaculture, mainlythe discharge of nutrients, organic substances andmicro-organisms, as well as escape of fish. The pos-sibilities for control will depend very much on thedesign and construction of the production unit.For closed units, control is normally possible by col-lecting and treating the outlet water from the unit.Closed production units may also float in the sea;here treatment of the outlet water is rather moredifficult; it is best to have a small height differencebetween the production unit and the treatmentplant. Floating production units may also bedamaged by waves, so shallow sites must be chosenfor such installations.

For open production units like sea cages the pos-sibilities for controlling the environmental impactare limited. It is therefore important to use sites

that tolerate discharge of nutrients and organic sub-stances without local accumulation.

In open production units in the sea there willalways be possibilities for fish to escape as a resultof construction failure. The weather is unpre-dictable and large waves can result in breakage ofthe production unit. Necessary precautions musttherefore be taken, including optimal design of theunit and units adapted to the site. A typical openproduction unit is the sea cage for which correctmooring is important. In addition, it is importantthat the net bags are strong enough for the site andare regularly inspected. Double net bags have alsobeen used; as have nets around the farm area, all toprevent fish escaping.

Fish escape from closed production units mayalso occur, especially at the fry stage, even if this theoretically can be avoided. The fry follow theoutlet pipes out of the unit. An absolutely securescreen ought therefore to be set in the outlet system,its sole purpose being to prevent escape of fish.


A

B

Figure 11.4 Production units utilize thefact that freshwater has a lower densitythan salt water, either to create a volumeof freshwater in a sea cage (A) or to heatthe water in a basin (B).

12Egg Storage and Hatching Equipment

tives of the first group where eggs prefer to stay onthe bottom, while several marine species such ascod and halibut and several freshwater speciesbelong to the second group. Under farming condi-tions, however, one species may be adapted to usea rearing system different to that used in the wild,such as holding pelagic eggs on a bottom substrate.Normally, systems in which the eggs lie on thebottom or on a bottom substrate are easiest to buildand control. There are also differences in how theegg lies, because some species have single eggs,while in others the eggs lie together in a matrix orwith a ‘cover’ around the egg batch. Egg size variessignificantly between species and this is also ofmajor importance when designing storage andhatching units with the required water supply; thetask is more difficult with smaller eggs.

Egg production can be separated into intensiveand extensive farming, and this also influences the design of the equipment. For more extensivefarming, ponds, net pens or cages may be used, butof course the production per volume unit will bereduced. If using extensive systems, the eggs can becollected and put into intensive hatching systems or the hatching can be performed in the extensivesystem, depending on production strategy. Egg pro-duction is normally based on artificial spawning, butegg production can also be based on collecting wildeggs that are introduced into extensive or intensivefarming systems.

In this chapter the focus is on intensive farmingsystems. Because of the great differences betweenspecies, it is difficult to give a general overview ofthe units for storage and hatching of eggs. There-fore information is given concerning the two basicmethods: 1, systems where the eggs stay pelagic; 2,

12.1 Introduction

The main purpose of units for storage and/or hatch-ing of eggs is to create a restricted area where theeggs can grow under optimal conditions. Separateunits can be used for storage and hatching, or thesame unit can be used for both purposes. It is alsopossible to combine the hatching equipment withlater holding of fry and eventually also for firstfeeding. Units for the incubation of eggs are oftencalled incubators.

In the storage unit, the eggs must be suppliedwith sufficient new water to meet their oxygenrequirements and remove metabolic waste prod-ucts. Continuous addition of new water will alsoachieve the necessary water exchange that is essen-tial to inhibit the growth of fungus which mayincrease mortality. In addition, the quality of thewater is of great importance at the egg stage.As theoxygen requirement is low at the egg stage, it is pos-sible to not have a continuous supply of water, butto exchange water in batches, provided that there isno fungus problem or an antifungal agent is addedto the water. Addition of air through diffusers mayalso be used to supply oxygen.

The design and function of the units depends onhow the eggs need to be stored and the intensity ofproduction. Eggs from different species have dif-ferent storage requirements. Some prefer to lie onthe bottom or on/in a bottom substrate; othersprefers to stay pelagic in the free water mass, whileothers again are stored inside the females: underwild conditions, for instance, Tilapia stores the eggsin the mouth and initially the wolf fish stores theminside the ‘belly’ for fertilisation before laying dem-ersal eggs. Salmonids and catfish are representa-

150

systems where the eggs lie on the bottom or on abottom substrate. The basic principles of the twocategories are however, the same and also used forother species. Many textbooks are available withmore detailed information regarding the require-ments of the various species, as this is of greatimportance for egg storage equipment (see, forexample, refs 1–5).

12.2 Systems where the eggs stay pelagicTwo production methods are commonly used forpelagic eggs, which may influence the design of the incubator. Either the eggs stay in the same unituntil hatching is finished, or there is a two-stepprocess using two separate units where the eggs areremoved before hatching. In both units the eggsstay pelagic (Fig. 12.1). The normal difference is thesize of the incubators; for the second step largeunits with volume of several cubic metres may beused. The system chosen is also species dependent,and there will always be species-dependent adjust-ments to be made.

If the eggs are to stay pelagic in the watercolumn, it is important to create an environmentthat makes this possible. A suitable incubator hasboth a water inlet and a water outlet. Inside theincubator an environment must be created that isas close to natural conditions as possible for pelagiceggs; at the same time oxygen must be supplied andwaste products removed. Most usually a small up-flowing (up-welling) current is created so that theeggs will stay in suspension in the incubator, as in afluidized bed.

12.2.1 The incubator

The traditional shape of an incubator for storageand hatching of pelagic eggs is a cylinder with aconical bottom (Fig. 12.2). With this shape it iseasier to achieve a good flow pattern in the unit.Funnel-shaped units, however, can also be used(Fig. 12.2). Traditional tanks (see Chapter 13) canbe used, but are not ideal.

Having a conical bottom makes it possible toremove the dead eggs easily by tapping them out

Egg Storage and Hatching Equipment 151

A

BFigure 12.1 Incubators for (A) storage of pelagic eggsand (B) hatching of eggs and storage of yolk sac fry.

�


through an outlet in the bottom of the cone. In prac-tice, several methods are used: for instance, stop-ping the water flow or adding a plug with highersalinity and thereafter stopping the water flow.6,7 Inboth systems dead eggs will sink to the bottom,while live eggs will remain in suspension due totheir greater buoyancy.

Incubators are normally constructed of fibreglassor polyethylene (PE). Glass jars may also be used,but this occurs more typically in smaller units. Thenormal size of incubators varies from some litres toseveral thousand litres.

12.2.2 Water inlet and water flow

To create an up-flowing current in the incubator,the inlet is normally placed at the bottom and theoutlet at the top. Various flow patterns can becreated in the up-flowing water.8 It is important thatthe up-flowing water forms a ‘plug’ that equals thecomplete cross-sectional area of the incubator. Theup-flow velocity must be the same at every pointover the cross-sectional area of the incubator;this will enable the eggs to remain in suspension.Although it is common to place the inlet at thebottom of the incubator, vertical spray inlets mayalso be used provided that the outlet is placed atthe top of the water column. An up-flow will alsobe created in this way. Air bubbling through dif-fusers on the bottom of the tank may also be usedto create an up-flowing current, and this can usedto reduce the necessary water flow. Air bubblingmay also be used to prevent aggregation of the eggson the water surface.

The amount of water to be added is dependenton the size of the incubator, the number of eggs andthe species. For example, in a 700 l incubator for codeggs, the water supply must be at least 3 l/min,7

while the recommended water flow for a 250 l incu-bator for halibut eggs is 3–4 l/min.6 It is importantthat the vertical flow velocity is not too high toavoid lifting the eggs to the surface. This can bemonitored if transparent materials are used for theincubator or there are inspection windows.

12.2.3 Water outlet

It is important to avoid having too high a flowvelocity through the outlet screen otherwise theeggs are dragged towards the outlet screen wherethey may get stacked (a large surface area of theoutlet screen is therefore important). Several solu-tions are used to avoid this (see, for instance, Refs.9, 10). Outlet systems can either have a screenaround the total circumference or part of the cir-cumference, or the outlet can be within the incuba-tor; the last may, for example, be an open pipecovered with a screen (Fig. 12.3) or what might be called a banjo screen. A banjo screen is a banjo-shaped screen placed inside the incubator.When using an internal screen it could be placesome distance below the surface to ensure pressureon the outlet screen.

The holes in the screen must be small so the eggs are not lost. Typically, a plankton cloth withsuitable mesh size is used. For halibut, a mesh size of 250μm can be used.6 To avoid blockage

A B

Figure 12.2 Designs of incubator for pelagic eggs: (A) cylinder with a conical bottom; (B) funnel-formed.

of the outlet screen, air bubbles may be blownagainst it.

12.3 Systems where the eggs lie onthe bottomThere is much experience concerning eggs lying onthe bottom and several systems have been devel-oped for intensive fish farming, especially for use insalmonid farming. It is possible to divide these intothree different systems:

• System where the eggs remain in the same unitfor the whole process from spawning up to fryready for first feeding

• System where the eggs lie in thick layers andmust be moved before hatching

• System where storage, hatching and first feedingare carried out in the same unit.

The system chosen depends on the managementstrategy for the hatchery, whether the farm pro-duces eggs for its own use or for sale. If the farm is to produce eggs for sale they may lie in thicklayers, because these will be sold before hatching.This system is suitable for salmonid eggs that are sold as eyed eggs, because they tolerate a lot ofhandling.

12.3.1 Systems where the eggs lie in the sameunit from spawning to fry ready for start feeding

Hatching troughs

A common unit is the hatching trough with traysinside; this is also known as the California system(Fig. 12.4). Trays or boxes are placed beside eachother lengthwise, in the trough. The trays have aperforated bottom and one of the sidewalls is alsoperforated. Water is supplied at one end of thetrough and leaves from the opposite end. A leveloutlet controls the water level in the trough.

Inside the trough the tray is installed so that anundercurrent of water is forced to flow up throughthe perforated bottom, through the layers of eggslying in the tray and then out through the perfo-rated side of the tray.The water is then forced downto the bottom of the trough and up through the perforated bottom of the next tray. In this way anundercurrent is generated in all the trays in thetrough.

After hatching there are large amounts ofeggshells to be removed to prevent obstruction ofthe outlet grating. To increase the grating area anL-shaped outlet grating can be installed in the trayduring hatching.


Figure 12.3 Outlet screen in anincubator for halibut eggs (seen hereas white on the surface).


The typical size of one type of hatching tray is 40cm × 40cm × 15cm. In each trough there may be4–7 trays.The number of eggs in each tray is speciesdependent; normally 1–2 l or two layers of eggs is recommended for salmon. The recommendedwater flow to each trough is 7 l/min and 12 l/min,for troughs with four and seven trays, respectively.If the water supply is too large, the undercurrentmay lift the eggs in the trays; increased mortalitymay occur if the eggs are moved during criticalphases in the incubation period. The troughs and trays are usually made of glass-reinforcedplastic.

There are also simpler hatching troughs wherethe eggs are not distributed in individual trays, butplaced along the whole surface of the trough.The water enters on one side and leaves from theopposite side. Such a system has no up-flowingwater through the egg layers, but a horizontal flow.The capacity per area unit is therefore lower. Inaddition, the hatching results are normally reduced,and the work requirement for production isincreased.

A special type of hatching trough, or smallraceway, can be used in channel catfish produc-tion.1,2 Here the eggs are put into stiff cloth baskets.

Between each basket there is a space where a smallpaddle wheel rotates (Fig. 12.5) to simulate thefanning action of the adult male. As catfish eggs are deposited in an adhesive yellow mass, it mayneed to be broken into smaller pieces before beingplaced in the baskets.

Artificial substrate

Artificial substrate can be placed in the bottom ofthe trays to improve the results (Fig. 12.6).The eggsare laid on top of a perforated plate in the hatch-ing trough, and when hatching occurs the yolk sacfry (alevins) will move down through the perfora-tions. The artificial hatching substrate is locatedbelow the perforated plate. This substrate createssmall spaces where the yolk sac fry can stay in anupright position. In this way only a small percent-age of the yolk sac is used for swimming and main-taining balance. Instead this energy is used forgrowth. The substrate can be designed in manyways, from squared cells to mats made of artificialgrass (for example, AstroTurfTM). The mats areusually made of plastic. Increased growth andreduced mortality have been achieved forsalmonids with the use of artificial substrate. It is

Figure 12.4 Hatching troughs with trays inside aremuch used in salmonid farming where the eggs are lyingon the bottom.

important that the artificial substrate does notrelease toxic substances into the water.

Hatching cabinet

In the hatching cabinet the eggs are placed indrawers or on racks (low boxes) on top of eachother. There are two different designs of hatchingcabinet, either water drops fall from the top, orthere is an individual water inlet and outlet in eachdrawer (Fig. 12.7).

The second design is the most used; the con-struction includes an individual inlet and outlet toeach drawer, with the outlet keeping the water level constant. The drawer has a perforated bottomwhere the water flows up through the layers of eggs.

After passing through one drawer the water is sentinto the drawer below. This system has the advan-tage that it maximizes the space utilization in rela-tion to hatching trough and trays, but it is moredifficult to control and reduced production mayresult.

Systems using the first design comprise a numberof drawers made of perforated plates where theeggs are distributed. Water is supplied from the topand flows down through the layers. However, theeggs must be removed before hatching in thissystem.

12.3.2 Systems where the eggs must be removedbefore hatching

In egg-rearing cylinders the eggs are layered on topof each other (Fig. 12.8), so this system cannot beused through to hatching, and for salmonids only to


Figure 12.5 System for rearing eggs of channelcatfish.

Figure 12.6 Artificial substrate is placed in the hatch-ing tray to improve the development of the yolk sac fry.


the eyed egg stage. In these cylinders the water istaken in at the bottom and a distribution plateensures that it is distributed evenly through thelayers of eggs via an underflow. The water will thenflow up through the layers of eggs and over the topedge of the cylinder to the outlet. To avoid airbubbles or clogging, the distribution plate must beset at an angle and there is an aeration pipe in thecentre of the cylinder. From here air bubbles can goto the surface without passing through the layers of eggs in the cylinder. If the air bubbles must pass through the egg layers they may increase mor-tality by moving the eggs during a critical phase ofdevelopment.

A commonly used egg-rearing cylinder containsapproximately 30 l of eggs; the necessary watersupply is 5–7 l/min. The great advantage of usingthese cylinders is that more eggs can be stored in a restricted area; up to 120 l/m2 floor area. The

cylinders are often made of polyethylene (PE);typically 80cm high and 50cm diameter. Largertanks of up to 200 l, may also be used with an up-flow of water, but here control of conditions isreduced.

Figure 12.7 In hatching cabinets the eggs are placedin drawers or racks above each other.

Figure 12.8 In the egg-rearing cylinder layers of eggsare stacked on top of each other, but must be removedbefore hatching. During incubation the eggs must liecompletely still.

12.3.3 System where storing, hatching and firstfeeding are carried out in the same unit

There is also special hatching equipment that maybe placed directly into the first feeding units (Fig.12.9). The space, water supply and water outlet tothe first feeding units can then also be used forstoring and hatching of the eggs. In one arrange-ment the hatching installation consists of a perfo-rated inner bottom and a fixed exterior bottomwhere the hatching substrate can be attached.It is placed inside the tank on legs and a speciallarvae outlet is directed into the tank outlet. Theadvantage with this system is that the same unit isused for egg storage, hatching and first feeding.Therefore less space is needed and no separatehatchery is necessary. Problems with disease, due tothe lack of a special isolated area, can be a disad-vantage with the system; neither is it as suitable for

specialized egg production. The system is mostcommonly used in smaller farms.

References1. Tucker, C.S., Robinson, E.H. (1990) Channel catfish

farming handbook. Van Nostrand Reinhold.2. Stickney, R.R. (1992) Culture of non salmonid fresh-

water fish. CRC Press.3. Stickney, R.R. (1994) Principles of aquaculture. John

Wiley & Sons.4. Tucker, J.W. (1998) Marine fish culture. Kluwer Aca-

demic Publishers.5. Lucas, J.S., Southgate, P.C. (2003) Aquaculture,

farming aquatic animals and plants. Fishing NewsBooks, Blackwell Publishing.

6. Mangor-Jensen, T. (2001) Klekkeridrift. I kveite-manualen. Havforskningsinstituttet, Bergen (in Norwegian).

7. Van der Meeren, T. (2002) Gyting, innsamling, innku-bering og klekking av egg. I havbruksrapport 2002.Havforskningsinstituttet, Bergen (in Norwegian).

8. Danielberg, A., Berg, A., Lunde, T. (1993) Design ofinlets for yolk sac larvae of Atlantic halibut (Hip-poglossus hippoglossus L.). In: Fish farming technol-ogy. Proceedings of the first international conferenceon fish farming technology (eds H. Reinertsen, L.A.Dahle, L. Jørgensen, K. Tvinnereim). A.A. Balkema.

9. Myre, P., Danielberg, A., Berg, L. (1993) Experimentson upwelling tank systems for halibut larvae (Hip-poglossus hippoglossus L.). In: Fish farming technol-ogy. Proceedings of the first international conferenceon fish farming technology (eds H. Reinertsen, L.A.Dahle, L. Jørgensen, K. Tvinnereim). A.A. Balkema.

10. Reitan, K.I., Evjemo, J.O., Olsen, Y., Salvesen, I.,Skjermo, J., Vadsetin, O., Øye, G., Danielsberg, A.(1993) Comparison of incubator conceps for yolk saclarvae of Atlantic halibut (Hippoglossus hippoglossusL.). In: Fish farming technology. Proceedings of thefirst international conference on fish farming technol-ogy (eds H. Reinertsen, L.A. Dahle, L. Jørgensen, K.Tvinnereim). A.A. Balkema.


Figure 12.9 Tank for combined egg storage, hatchingand start feeding.

13Tanks, Basins and Other Closed

Production Units

requirements. For example, some species prefer tolie on the bottom and do not utilize the entire watercolumn (non-pelagic species). This chapter gives a survey of the design and construction of closedproduction units. The focus is on tanks with a cir-cular flow and species that utilize the entire watercolumn. However, there is a wealth of general information for all types of closed production units.

13.2 Types of closed production unitsThe flow pattern of the water in the unit can beused to classify closed production units (Fig. 13.2):

• Production units with a circulating water flow• Production units with a one-way water flow.

Production units with a circulating water flow mayagain be separated into tanks with a circular flowpattern (as is most common), or oval tanks thathave an oval flow pattern, of which there areseveral types;1,2 for example, Foster Lucas tanks andBurrow tanks, as well as different ovals and pipeconnections. Among the traditional tanks with circular water flow, one type may be defined as afarming silo. This is a circular tank of greater heightthan diameter, i.e. a tower. It is normally difficult to achieve satisfactory water exchange and self-cleaning in silos (see Section 13.7).

Earth ponds belong to the group of productionunits with one-way water flow and represent theoldest type of closed production units used for fish production. They are mainly used for extensivefish farming; i.e. there is a low production per unitfarming volume. Earth ponds are described sepa-rately in Chapter 14 because they have so little in

13.1 IntroductionThe main purpose of a closed production unit is tocreate a restricted volume where the fish or otheraquatic organisms can be fed in a good water envi-ronment. For fish, the units may be used from firstfeeding of fry up to on-growing. Closed productionunits are used for both freshwater and seawaterspecies. A unit includes the water inlet, the produc-tion unit and the water outlet (Fig. 13.1).

To get a closed production unit to function asoptimally as possible, a number of requirementscan be set in addition to the main requirement thatthe unit should produce as much fish as possible atlow cost. These are as follows:

• The supplied water, including the oxygen dis-solved in the water, should be distributed evenlythroughout the entire production unit

• The fish should be evenly distributed in the totalproduction volume

• The fish must be transported effectively in andout of the unit

• The unit should require a minimum of manualcleaning

• The inner surface should be smooth and requirelittle cleaning and maintenance

• The unit should be easy to operate, which meansit should be easy to clean, to remove dead fishand to perform other handling fish tasks

• The unit should have low investment costs perunit effective farming volume.

This list shows that there are a number of require-ments that need to be fulfilled before a productionunit functions optimally. However, compromiseshave to be made because it is difficult to fulfil all

158

Tanks, Basins and Other Closed Production Units 159

Figure 13.1 The main components in aclosed production unit include the waterinlet, the storage unit and the wateroutlet. A circular tank is used as anexample.

Figure 13.2 Production units can beseparated into units with a circulatingwater flow and those with a one-waywater flow.


common with other closed production units such astanks. According to the requirements for closedproduction units, the main interest in using pondsis the low initial cost per unit farming volume. Inaddition, normally a natural ecosystem that can beutilized is created inside the ponds. In some coun-tries such as Norway, no new permits for earthponds are given unless they are dried once a year.This is because it is quite difficult to control diseaseas pathogenic micro-organisms may survive in theearth. If the ponds are dried during the winterseason these micro-organisms will probably bekilled. A layer of lime which increases the pH mayalso be used as a disinfectant. If the ponds arecovered with a plastic tarpaulin (polyvinyl chloride(PVC) or polyethylene (PE)) the problems areavoided because such ponds may be cleaned inside,but this is not a normal pond design, and suchinstallations are more like a traditional tank withno ecosystem inside.

A further development of the earth pond is theraceway, which also uses the one-way flow pattern.This is a fixed construction often of concrete, builtas a long rectangle.The water is supplied at one endand the outlet is located at the opposite end. Race-ways are quite commonly used for various speciesthroughout the world. However, raceways requirequite large amounts of water to have effectivehydraulic self-cleaning of their total volume,3,4 andeven then it is very difficult to get good cleaningresults. Normally, some kind of mechanical equip-ment is necessary for additional cleaning of theraceway. It is therefore important to create a good flow pattern inside the unit, with a correctlydesigned flow inlet and outlet, to ensure uniformwater flow through the entire cross-sectional areaand length of the raceway to reduce the require-ment for manual cleaning. During the past fewyears, a specially designed raceway with a very lowwater level (10–50cm) has been developed.5 Theunit is specially designed for fish species that needa bottom to lie on and do not utilize the entirewater column, for example halibut and wolf fish.The unit is constructed so that it can be installed intiers, one above the other.

Closed floating cages are also a type of closedproduction unit. In this case, both circulating water flow and one-way flow systems have beentried in different variants that have been con-structed using different materials. In one variant the

traditional net bag in a sea cage has been substi-tuted by plastic sheeting. Water is pumped into thebag tangential to the edge and the outlet is placedin the centre of the unit. This creates water circula-tion inside the bag (see, for example, Ref. 6). Theadvantage of this type of unit is that it lies on thewater surface and there is only a small head to over-come to pump the water into the cage, comparedwith, for instance, closed production units placed onshore and where seawater is normally pumpedseveral metres.

As shown, a closed productions unit can be builtin several ways and can have different water flowpatterns. The design of the production units,however, depends on the type of aquatic organismto be grown and its requirements regarding waterdistribution and bottom conditions.

13.3 How much water should be supplied?The reason for adding water to a closed productionunit is to give the fish or shellfish access to oxygenand remove waste products excreted by the fish. Inthis way, a water environment that creates optimalconditions for growth is established. The amount ofwater that must be supplied to a closed productionunit depends on a number of variables, includingspecies, fish density, growth stage and rate, watertemperature, whether adding pure oxygen or not,and whether hydraulic self-cleaning occurs ormanual cleaning is required.

Tables showing the supply of water necessary to satisfy the oxygen requirements of fish and shellfish at different water temperatures have beendeveloped for many species. The amount of water that must be added to the production unit tocover the oxygen consumption of the fish can be calculated based on the amount of fish and watertemperature.

Provided the quality of the water supplied isacceptable, one way to regulate the necessarysupply of new water to a closed production unit canbe to monitor the oxygen concentration in theoutlet. This must be at an acceptable level foroptimal growth of the actual species, and is nor-mally around 7mg/l. By using electronically con-trolled actuator valves for controlling the inlet flow,

it is possible to have fully automatic flow controlbased on the oxygen concentration in the outletwater.

The amount of inlet water can be reduced byadding pure oxygen to the incoming water or insidethe production unit. With high fish density andrapid growth, consideration must, however, begiven to the concentrations of CO2, suspendedsolids (SS) and NH3 that might become excessivewhen adding oxygen. If this is the case, these sub-stances need to be removed; this requires watertreatment as is carried out in a water re-use system(see Chapter 10).

The hydraulic forces in the water supplied to thetank may also be used to clean it; this is known ashydraulic self-cleaning. Extra requirements thenapply to the amount of added water; these aredescribed in Section 13.7. This gives anothermethod for calculating the necessary supply of newwater. If this water supply does not fulfil therequirements for oxygen, pure oxygen gas must beadded to make up the shortfall.

13.4 Water exchange rateThe water exchange rate indicates how quickly thewater in closed units is exchanged. This can bedefined as the period for which a specific water mole-cule stays in the unit before leaving via the outlet(Fig. 13.3). As the new incoming water will bemixed with the ‘old’ water in the tank, the outletwater will always contain both new and old water.It is important to realize this, and means that if one

tank volume is run into a tank full of old water,only part of the old water is exchanged, not all. Todescribe this, the term ideal water exchange is used.When expecting ideal water exchange and adding 1 l of new water to 100 l of old water, the new andold water will be mixed immediately; for example,the addition of 1 l red water to 100 l of clear waterinstantly results in pink water. This is a simplifica-tion, but it helps us to understand better what ishappening with the water exchange, and does notrequire difficult equations.

Mathematically, the water exchange rate can becalculated as follows (based on developing a dif-ferential equation):

F = (1 − e−t/th) × 100

where:

t = time after start of filling water into the unitth = time necessary to fill one tank volume at the

actual water flow rate; also known as the the-oretical retention time

F = water exchange rate (proportion of the watervolume in the unit that is exchanged after timet)

ExampleFifty litres of new water is added to a tank that con-tains 100 l water, described as old water, over a periodof 5 minutes (i.e. 10 l/min). The same amount of oldwater flows out through the outlet because the watervolume and level are constant. How much of thewater volume is exchanged after 5 and 10min,respectively?


Figure 13.3 Amount of water exchangedin a tank in relation to time.


Setting t = 5min

th = 100 l/(10 l/min) = 10min

F = (1 − e−5/10) × 100= (1 − 0.605) × 100= 39.5%

Setting t = 10min

F = 63.2%

This means that by adding a water volume equal tothe tank volume only 63.2% of the water isexchanged, not all as might be expected, the reasonbeing that new and old water are mixed.

13.5 Ideal or non-ideal mixing andwater exchangeThe aim in designing a closed production unitincluding the inlet and outlet, is to achieve the mosteffective mixing as possible of the new incomingwater to the old water together with a goodexchange of water in the entire farming volume. Nonew water must go directly to the outlet, by whatcould be called a short cut; neither must there beareas or zones in the unit were there is small or noexchange of water, so called ‘dead zones’ (Fig. 13.4).Since water exchange does not occur in the com-pletely dead zones, no new water or oxygen is sup-plied, so the fish will prefer not to stay there andthe effective farming volume will be reduced belowthe real tank volume. If there are short cuts, part ofthe inlet water will go directly to the outlet without

having been properly utilized by the fish; this resultsis non-ideal mixing. Short cuts will also give zonesin the unit where the water flows much more slowlyand than elsewhere water exchange will thereforenot be satisfactory.

A picture of the velocity gradients within the pro-duction unit can be found by using a speciallydesigned small propeller – a velocity meter – tomeasure the flow rate at different points in the tankboth horizontally and vertically (different depths).Eventually, dead zones and zones where the waterflow is too fast will also be identified.

A number of factors, including design of the tank,the inlet and the outlet, will affect water exchange.Before starting to use a new tank design or inlet oroutlet system, it is advantageous to test the flowpattern in the tank and find the velocity gradients.

13.6 Tank designSeveral designs of tanks with circular water flow arein use. What is important when choosing is that thenew water is uniformly distributed throughout the entire tank volume. Round or polygonal (6–8edges) tanks with a circulating flow pattern are suit-able because they have no dead zones provided thatthe inlet and outlet are correctly designed. Squaretanks will, however, have dead zones in each cornerand the effective farming volume is therefore notso large; for this reason square tanks are not rec-ommended. Square and rectangular tanks with cutcorners have, however, been shown to be quitegood; experience with tanks having a total side

Figure 13.4 Water exchange rate in tankswith dead zones.

length of a and corners of size a/5 shows they arewell suited7 (Fig. 13.5). For other tank designs suchas raceways or earth ponds, it is far more difficultto avoid dead zones, and the effective fish produc-tion volume is normally less than the actual tankvolume.

When selecting a tank design, it is also importantto take into account the utilization of the area;square tanks with cut corners utilize this well,achieving several m3 of farming volume per m2

surface area. Raceways will also utilize the area satisfactorily.

Utilization of the tank construction material isanother factor that must be considered when decid-ing the shape of the tanks. Circular tanks will havethe best utilization of the construction materials.The pressure of the water is equally distributed allaround the circumference of the tank and thereforea thinner wall may be used than for square tanks.

In square tanks the forces are greatest in the middleof the sides, and there is an accumulation of forcesin the corners. The height of the tank will also beimportant because the pressure on the tank wallsand hence the necessary thickness will increase.

The bottom of the tank could be horizontal orhave a small slope towards the outlet which isusually in the centre of the tank; however, a partoutlet might be in the tank wall, see Section 13.10.There is, however, little benefit from sloping thebottom towards the grating and outlet of the tankwhen having a correct flow pattern inside the tank.This is because the most important mechanisms fortransport of the settled solids (faeces, feed loss) arethe water flow and its hydraulic force, not gravity.Even a small upward slope (2–5%) to a centrallyplaced outlet has been used by the author withgood results in tanks with a circulating flow pattern.This also confirms that the most important factorfor transport of settled solids to the outlet is theforce created by the water flow, rather than thebottom slope and force of gravity. When havingnon-self-cleaning flow conditions, for instancebecause fry production requires water flow of lowvelocity, it is important to have a slope to the outletgrating to be able to utilize graviditational forces.However, this slope must be quite large to really getan effect of gravity. In a filter unit the angle is rec-ommended to be above 55° to utilize the force ofgravity to get the settled solids to slide; this isbecause the density of aquaculture solids is low(1.05–1.2) and almost equal to that of water.8,9

The height of the tank compared to the diameterwill also affect the water exchange. For tanks witha circular flow pattern, a tank diameter: height ratioof between 2 and 5 has been successfully used. Ifthe tank diameter is 10m, the height could there-fore be between 2 and 5m. For tanks that do notfall within these ratios, special attention must begiven to the design and placement of the inlet andoutlet. If the ratio is lower, the inlet should beplaced some distance away from the tank wall,closer to the centre of the tank. Tanks where theheight is greater than the diameter are often calledsilos; by using such tanks high production can beachieved per unit area. It is, however, difficult to geta proper water exchange throughout the entirevolume in such constructions.

Various materials are used to fabricate tanks(Fig. 13.6). It is important that there is a smooth


Figure 13.5 Example of an octagonal tank (a squaretank with cut corners).


surface inside the tank to reduce problems withfouling, and that the material does not release anytoxic substances into the farming water. Glass-reinforced plastic is a commonly used material for tanks, because it can be produced with a verysmooth surface; small tanks are also light and easyto move. The tanks are either delivered completelyfinished or as elements that are screwed togetheron site. Plastic (polyethylene, PE) may also be used;

this is also a light and cheap material. New mater-ial has a very smooth surface; however, is it moreprone to ageing and the surface gradually becomesless smooth. The surface also scratches more easily.The tanks are either made of plates weldedtogether into tanks or the tanks are rotation cast (aspecial casting process). Concrete is much used forlarger tanks, where the price is competitive; con-crete may also be mixed on site or prefabricated

A B

DC

E

Figure 13.6 Tanks of different materials: concrete (A),tarpaulin with a framework of wood plates (B), fiberglass(C), stainless steel (D), coated steel plates (E).

elements can be joined together. The other material that has been used to some extent is metal; for example steel plates (stainless, acid-proofor coated) or aluminium (special quality). Tanks oftarpaulin with a frame of steel or wood represent alow cost easily movable construction.

13.7 Flow pattern and self-cleaningA flow pattern will be created inside a productionunit having a water inlet and outlet. It is importantthat this flow pattern encompasses the entire unitso that all the fish can come into contact withflowing water. The flow pattern depends on thedesign of the production unit.

In a tank with a circulating water flow and cor-rectly designed inlet and outlet, two flow patternswill occur: the primary flow and the secondary flow(Fig. 13.7). The primary flow causes even distribu-

tion of the water in the horizontal plane, while thesecondary flow will clean the tank walls and bottom.

In well-designed tanks with correctly designedand constructed inlet and outlet, the incomingwater may therefore be used to clean the tank wallsand bottom. This process is known as hydraulic self-cleaning. To achieve self-cleaning in a tank, acertain amount of water has to be added; theamount depends on the tank construction. Thewater velocity at the bottom of the tank must be sohigh that the settled solids are removed. To ensuretransport of settled solids in circular tanks, the rec-ommended bottom velocity to ensure self-cleaningis above 6–8cm/s.7 This will also remove algalgrowth from the tank sides. Inside the tank therewill be a velocity profile equal to that in a channel,where the lowest velocity occurs near the bottomdue to friction (Fig. 13.8). Bottom water velocitiesof between 6 and 8cm/s normally represent a water velocity in the free water mass of between 12 and 15cm/s.7 Practical experience has also shown that high fish density promotes self-cleaningof the bottom. A lower velocity could therefore be accepted when the fish density is increasedbecause the movement of the fish results in resus-pension of settled solids, so the secondary flowpattern could more easily transport the particles tothe drain.

In a correctly designed flow through tank withinlet and outlet, and a circular flow pattern, thewater retention time should be between 30 and 100min for satisfactory self-cleaning.10 This means aflow through of between 10 and 33 l/m3 farmingvolume. A retention time of less than 30min maycreate a vortex around the centre drain.The periph-eral velocity in the tank may also be so high thatthe fish will have problems staying there. Whenusing low retention times, a specially designed inlet and outlet are necessary. With retention timesabove 100min the self-cleaning effect is decreasedand additional cleaning is necessary.

To attain hydraulic self-cleaning, a high volumeof water is needed to create a high water velocityinside the tank. Even if the water velocity hasyielded improved growth results,11–13 there is amaximum velocity that not must be exceeded.14

This will vary according to species and growthstage. Examples here are fry of marine or freshwa-ter species, where only very low velocity is toler-ated; to maintain satisfactory water quality for


Figure 13.7 In a tank with a correctly designed inletand outlet, both a primary and secondary flow will begenerated.


these species is therefore a challenge. Settled parti-cles and fouling on the tank bottom and sides willcreate a sub-optimal environment, and be a goodsubstrate for unwanted bacterial growth. Regularremoving of fouling is therefore absolutely essen-tial. If this is done manually, it is labour intensive,and therefore commercially available automaticsystems are preferred. Rotating brushes on the

bottom of the tank, powered either by electricmotors or by the pressure of the incoming water areone solution. Other solutions include a small turtle-like unit moving around on the tank bottom orinstalling a washing arm half the pipe diameter inlength. Addition of chemicals that remove thefouling, such as oxidizing agents, has also been triedby the author and colleagues.15

Figure 13.8 An idealised velocityprofile matrix across A–A in a 1800 ×600 mm tank.

13.8 Water inlet designCorrect design of the inlet flow arrangement to thetank is necessary to ensure even distribution andmixing of the new incoming water and if self-cleaning is to be attained. This requires the inletwater pipe to enter below the water surface in thetank and the water must pass through a narrownozzle. The force of the inlet water can then be uti-lized to create a flow pattern inside the tank (Fig.13.9). It is also important to spread the incoming

water throughout the water column. This can beachieved by the use of several holes, splits ornozzles in the inlet pipe below the water surface.The force of the incoming water will now be distributed throughout the water column, not justin one place. Improved distribution of the newoxygen-rich incoming water is also achieved.

The impulse, as the force caused by the inletwater is called, depends on the water flow andwater velocity; it can be expressed as follows fortanks with a circular flow pattern:

F = rQ (v2 − v1)

where:

F = impulser = water densityQ = water flowv1 = velocity of the water in the tankv2 = velocity out of the holes in the inlet pipe.

This equation shows that by increasing Q, theimpulse will increase. The same result is achievedby increasing the velocity of the water emergingfrom the holes in the inlet pipe. Decreasing thecross-sectional area of the holes will increase thevelocity of the water. However, the increased veloc-ity will increase the turbulence and hence the headloss. Recommended values are below 1.5m/s in theinlet pipe, while the velocity in the hole (or split ornozzle) should be below 1.2m/s.7

The inlet pipe can be arranged in several waysdepending on the tank design (Fig. 13.10). In tankswith a circular flow, a horizontal spray inlet has theadvantage of creating good water distribution(primary flow), but the secondary flow is notoptimal. The vertical spray inlet creates both goodprimary and secondary flow, and is therefore pre-ferred. It is also possible to use a combined verticaland horizontal inlet with good results.

Normally a vertical spray inlet will be placedabout one fish width away from the tank wall, so


Figure 13.9 Having the tank water inlet below thewater surface and using nozzles in the inlet pipe willcreate velocity gradients, and mixing of the water in thetank is achieved.

Figure 13.10 Various designs ofinlet pipe in tanks with a circular flowpattern.


that the fish can pass behind, and to avoid too muchfriction from the tank walls. If it is too close to thewall, friction against the wall will reduce theimpulse. However, in a low tank designed with adiameter :depth ratio of less than 0.2, the inlet hasto be placed further into the tank closer to thedrain, to create a good flow pattern. In silos is itespecially difficult to get good inlets and effectivetransfer of the impulse, and hence effective waterexchange throughout the water volume. However itmay be possible to use several water inlets in thetanks to improve the flow pattern; testing of thevelocity profile is recommended in such cases.Depending on the current velocity from the holesin the inlet pipe, the current velocity in tanks withcircular flow is normally in the range 0.15–0.25m/s.7

In raceways it has proved to be difficult to createan inlet that distributes the water in a uniform waythroughout the entire cross-sectional area and totallength. It is important that the impulse is distrib-uted over as large a part of the cross-sectional areaas possible. Because of the continuous reduction inwater flow velocity close to the bottom due to fric-tion, there have been experiments in which waterwas added at several places over the length of theraceway to improve the velocity over the total area;this, however, increases the costs.

The velocity of the inlet water out of the holes or nozzles in the inlet pipe (V2) depends on the design of the nozzle (hole), the area and theamount of water that has to pass. It can beexpressed as follows (based on the continuity equa-tion; see Chapter 2):

where:

V2 = velocity out of the nozzlesQ = water flow out of the nozzlesΣA = total cross-sectional area of all the nozzles.

The relation between the water velocity in the inletpipe and the velocity out of the nozzles will be asfollows:

where:

V0 = velocity in the inlet pipeA0 = area of the inlet pipe.

VV A

A2

0 0=∑

VQ

A2 =

∑

ExampleAn inlet pipe to a circular tank is designed for awater flow (Q) of 50 l/min. Suggest an appropriatepipe diameter and area of the nozzles (holes).

First, calculate the area of the inlet pipe:

A = Q/V

Transform the units so that they correspond and takea maximum water velocity of 1.5m/s.

50 l/min = 0.00083m3/s

Calculate the diameter:

A = pd2/4

In practice the nearest standard dimension will beused.

Calculate the total nozzle/area (cross-sectional area)using a velocity of 1.2m/s.

This must then be divided by the number of holesused in the total water column.

To force the water through the inlet pipe resultsin a head loss. Because of this, a minimum head(available water pressure) is necessary to get thewater to flow through the nozzles/holes. Differentshapes of the nozzles/holes will result in differenthead loss because of different degrees of turbu-lence created in the nozzles. This must be taken into consideration when constructing the inlet pipe.

A Q V=

=

==

∑ /

//

0.00083m s1.2m s

0.00069m

6.9cm

3

2

2

dA=

= ×

=

4p

4 5.533.1415

2.65cm

A =

==

0.000831.5

0.000553m

5.53cm

2

2

ExampleHead loss in the inlet pipe

The inlet pipe has a diameter of 63mm and a waterflow of 300 l/min is used. This gives a water velocityof about 1.6m/s. Find the head loss when increasingthe water velocity from 1 to 2.5m/s (f = 0.024).

ment of the particles which may break up. Thismakes later filtration of the outlet water more difficult.

It is important to treat the outlet water as gentlyas possible to avoid increased particle breakage.Elbows and other pipe parts that create additionalturbulence should therefore be avoided. It is impor-tant to use long bends and small bend angles in thepiping. It is also important to understand that theinlet and outlet of a tank are designed for a givenflow rate, but with some latitude outside which theinlet and outlet will not function optimally becausethe velocity is either too low or too high.

The head loss in the outlet system can be seen bycomparing the water level inside the tank with thetop level of the water in the outlet system. A largedifference means a high head loss.

ExampleThe supply of water to a fish tank with a farmingvolume of 6m3 should be between 60 and 200 l/minto ensure sufficient self-cleaning. Find a suitablediameter for the outlet pipe.

Choose a velocity of 0.5m/s and calculate for thesmallest amount of water.

A = Q/V

Q = 60 l/min = 0.001m3/s

A = pr2

r = 0.025m= 2.5cm

The next task is to calculate the water velocity in thepipe with the largest amount of water when theradius of the pipe is 2.5cm.

V = Q/A

Q = 200 l/min = 0.0033m3/s

A larger dimension for the outlet pipe is recom-mended because this velocity is rather high. This

V =

=

0.0033m /s0.002m

1.67m/s

3

2

A =

=

0.001m /s0.5m/s

0.002m

3

2


Water velocity Head loss per metre of pipe(m/s) (mH2O)

1.0 0.019

1.6 0.05

2.0 0.08

2.5 0.12

With a water velocity of 2.5m/s in the inlet pipe, it isnecessary to have a pressure in the pipe of 0.1mH2Oper metre of pipeline to achieve the necessary waterflow.

13.9 Water outlet or drainThe tank outlet or drain normally has two func-tions: 1, to remove the waste from the tank asquickly as possible, before the leakage of nutrientstarts; 2, to maintain the correct water level insidethe tank. Correct design of the outlet is also impor-tant for the water exchange rate and to ensureeffective self-cleaning. Incorrect specification of the outlet system or the outlet pipe may result insettling of uneaten feed particles and faeces, andthe outlet system will function as a settling basin.This situation can be rectified by shock draining thetank, by removing the device that controls the tankwater level. Shock draining will increase the watervelocity through the outlet above normal and even-tually settled solids will be dragged out. If there arelarge number of particles coming out, there may beplaces in the outlet pipe system where settling ofsolids occurs if the design of the outlet system issub-optimal. It should not be necessary to shockdrain a well-designed outlet system to remove particles that have settled in the outlet. The outletpipe should be designed for water velocities above0.3m/s to ensure no settling of solids. Velocitiesabove 1.5m/s in the outlet will result in rough treat-


also shows that the outlet system is not designed optimally if there is too large a variation in the waterflow.

When designing the outlet system, it is importantthat the outlet creates a drag on the water in thetank. Particles lying close to the outlet will beforced against the outlet by the reduced pressurethere.

The design of the outlet can be either of the flator tower type (Fig. 13.11). The flat outlet contains ahorizontal screen inside the fish tank, adapted tothe fish size. This normally covers an outlet pot.From the outlet pot the outlet pipe goes to a verti-cal standing pipe where the water level inside thetank is controlled. It is important to increase thespeed through the holes in the outlet screen toavoid fouling and blockage. Velocities above 0.3–

0.4m/s are recommended. The outlet screen is aperforated sheet of stainless steel, aluminium orplastic. Good results have been achieved by usingoblong slots instead of holes in the outlet screen.14,16 Slots do not clog as quickly and areeasier to clean.The holes or slots are recommendedto be as large as possible, but of course not so largethat fish escape. It may also be possible to run thetank without any outlet screen, because the fishprefer not to go down in the outlet pipe. However,this is species dependent, and some species thathave a crowding behaviour, such as eel, will godown. For salmonids, the author’s experience is thatit is possible to run without any outlet screen. If thisis done it is necessary to install a trap in the outletsystem to collect the dead fish (see Section 13.11).

The outlet pot, if used in the tank, has beenshown to be a critical part of the construction (Fig. 13.12). Settling of solids is very common here.It is therefore important not to reduce the watervelocity in this pot too much, both to prevent set-tling of particles and also to reduce the amount offouling. Good results are obtained by using aneccentric outlet from the pot; here the outlet pipefrom the pot is not in the centre but to one side, anda swirl is therefore created inside the pot. Anotherpossibility is to avoid the use of the outlet pot

Figure 13.11 Flat or tower outlets are typical designs.Figure 13.12 Correct design of the outlet pot is important.

totally by having the outlet pipe directly into thetank bottom. This requires the use of another typeof outlet screen inside the tank (not a horizontalone) to ensure enough hole or slot area. Outletscreens may, for instance, be shaped like a cone,pyramid or a small tower, not reaching the watersurface.

There are two types of tower outlet, with eitherinternal or external regulation of the water level. Inboth cases there will be a tower inside the tankgoing from the bottom to the water surface, con-sisting of a screen with holes or slots that preventthe fish leaving the tank. The disadvantage of thisoutlet is that the presence of the tower in themiddle of the tank may make it difficult to handlethe fish in the tank. With regard to self-cleaning, nodifferences have been observed between flat andtower outlet systems.

Instead of taking the outlet from the bottom ofthe tank, a siphon can be used. Here the outlet fromthe tank goes over the top edge of the tank;however, this will require a fixed water level. Theadvantage with the system is that it is not necessaryto have any pipes going out of the bottom of thetank, which is therefore only a basin. This reducesthe cost and increases the second hand value. Thedisadvantage is that it is not possible to reduce thewater level because it is fixed; a pump must be usedto drain the tank completely.

Both flat and tower outlets, include a levelcontrol to keep the water inside the tank at a givenheight. Level controls can be adjustable or fixed.Adjustable level controls can be incremental orcontinuous. Fixed level controls have no possibili-ties for adjustment and therefore there must be abypass for complete drainage of the tanks. For largevolume tanks (>100m3) adjustable level controlsare quite expensive and seldom used. Usually fixedor perhaps incremental controls are installed. Acontinuous level control may be built by having asliding pipe inside another pipe. The same principlemay be used for a stepwise control, but here thelevel is controlled by addition and removal of pipeparts.

13.10 Dual drainIt is also possible to use a double outlet, i.e. a dualdrain (Fig. 13.13).14,17–19 Such an outlet allows partof the water flow (up to 5% of the total flow) to be

taken through a separate centre drain located in thebottom of the tank, while the main water flow istaken out from the water column above the tankbottom through some type of tower outlet. This canbe a traditional tower outlet or a small tower reach-ing 5–10cm up from the tank bottom. In this waythe tank is used as the first effluent treatmentsystem. A particle purification step is achievedbecause the tank functions as a swirl separatorusing the ‘tea-cup’ principal. The circular flowpattern in the tank will cause the solids to be colleted in the centre of the bottom where the particle outlet is located. Normally over 90% of thesolids be taken out via the bottom outlet. Hence theparticle concentration will be significantly reducedin the main outlet that is taken out slightly higherup the water column.

The main water flow may also be taken out via asidewall.4 By taking out up to 20% through thebottom and the rest through the outlet in the upperpart of a tank wall, good separations have beenachieved. A specially designed square tank with cutcorners, with the main inlet and main outlet in thecorners and the particle outlet in the centre, hasalso been used with good results.20 Here less than1% of the total water flow was taken out via theparticle outlet. In both the above two systems theflow pattern inside the tank gave satisfactory self-cleaning. In the last system the place on the tankbottom where the main effluent collected was easilyregulated by changing the direction of the nozzlesin the inlet pipe; these were adjusted to force the


Figure 13.13 The double or dual drain outlet will sep-arate particles from the water.


effluent towards the middle of the tank where theparticle outlet was located.

By using dual drain systems the cost of particlepurification will be reduced because only a minorpart of the total water flow has to be purified; thesize of the particle filter will be reduced dramati-cally. If using a water re-use system (Chapter 10),the water in the main outlet may be re-used withoutgoing through a particle separation step as the con-centration of particles is already low enough.

13.11 Other installationsThere may also be other installations that can beintegrated into the tank. By using a speciallydesigned outlet screen that can be opened from thesurface, it is possible to knock dead fish lying on theoutlet screen out of the tank and through the outletsystem; it is then necessary to have a collector forthe dead fish in the outlet piping system.This avoidsdragging dead fish up through the productionwater.

Handling facilities can also be an integral part ofthe tank construction. Additional outlets used onlyfor guiding the fish to a common centre (seeChapter 17), or hatches in the tank walls for trans-port of fish through channel systems, can be inte-grated.21 The tank may also be constructed toinclude guides for adapting tank internal gradinggrids (see Chapter 17).

A feed detection unit may also be integrated intothe tank outlet system.22

References1. Wheaton, F.W. (1977) Aquacultural enginering. R.

Krieger.2. Cripps, S.J., Poxton, M.G. (1992) A review of the

design and performance of tanks relevant to flatfishculture. Aquacultural Engineering, 11: 71–91.


4. Timmons, M.B., Riley, J., Brune, D., Lekang, O.I.(1999) Facilities design. In: CIGR Handbook of Agri-cultural Engineering, Part II Aquaculture Engineering(ed. F. Wheaton), pp. 245–280. American Society ofAgricultural Engineers.

5. Øiestad. V. (1995) Shallow raceways as the basis for industrial production centers of seafood. In:Quality in aquaculture. European AquacultureSociety special publication no. 23.

6. Solaas, F., Rudi, H., Berg, A., Tvinnereim, K. (1993)Floating fish farms with bag pens. In: Fish farmingtechnology, Proceedings of the first international conference on fish farming technology. (eds H.Reinertsen, L.A. Dahle, L. Jørgensen, K.Tvinnereim). A.A. Balkema.

7. Tvinnereim, K. (1994) Hydraulisk utforming av oppdrettskar. Brukerrapport. SINTEF report STF60A94046 (in Norwegian).

8. Chen, S., Coffin, D.E., Malone, R.F. (1993) Produc-tion, characteristics, and modeling of aquaculturalsludge from a recirculating aquacultural system using a granular media filter. In: Techniques forModern Aquaculture, Proceedings of aquaculturalengineering conference, Spokane, Washington, (ed.J-K. Wang), pp. 16–25. American Society of Agricul-tural Engineers.

9. Wong, K.B., Piedrahita, R.H. (2000) Settling velocitycharacterization of aquacultural solids. AquaculturalEngineering, 21: 233–246.

10. Skybakmoen, S. (1991) Kar og karmiljø, temahefte.Aga as (in Norwegian).

11. Jobling, M, Jørgensen, E.H., Christiansen, J.S. (1993)Growth performance of salmonids exposed to differ-ent flow regimes. In: Fish farming technology. Pro-ceedings of the first international conference on fishfarming technology (eds H. Reinertsen, L.A. Dahle,L. Jørgensen, K. Tvinnereim). Balkema.

12. Jobling, M., Jørgensen, E.H., Arnesen, A.M.,Ringø, E. (1993) Feeding, growth and environmentalrequirements of Arctic charr, a review of aquaculturepotential. Aquaculture International, 1: 20–46.

13. Davidson, W. (1997) The effects of exercise training on teleost fish, a review of recent literature.Comparative Biochemistry and Physiology, 117a:67–75.

14. Timmons, M.B., Summerfelt, S.T., Vinci, B.J. (1998)Review of circular tank technology and management.Aquacultural Engineering, 18: 51–69.

15. Lekang, O.I., Andreassen, I., Nergård, R. (2003)Design of start feeding tanks for wolf fish. ITF con-ference report 194. Norwegian University of LifeScience.

16. Piper, R.G., McElwain, I.B., Orme, L.E., McCraren,J.P., Fowler, L.G. Leonard, J.R. (1982) Fish hatcherymanagement. US Fish and Wildlife Service.

17. Lunde, T., Skubakmoen, S. (1993) Particle separationintegrated in enclosed rearing units. In: Fish farmingtechnology. Proceedings of the first international con-ference on fish farming technology (eds H. Reinert-sen, L.A. Dahle, L. Jørgensen, K. Tvinnereim) A.A.Balkema.

18. Twarowska, J.G., Westerman, P.M., Losordo, T.M.(1997) Water treatment and waste characteriza-tion evaluation of an intensive recircualting fish production system. Aquacultural Engineering, 16:133–147.

19. Davidson, J., Summerfelt, S. (2004) Solids flushing,mixing, and water velocity profiles within large (10

and 150 m3) circulation ‘Cornell-type’ dual draintanks. Aquacultural Engineering, 32: 245–271.

20. Lekang, O.I., Bergheim, A. & Dalen, H. (2000) Anintegrated waste treatment system for land-basedfish-farming. Aquacultural Engineering, 22: 199–211.

21. Lekang, O.I., Fjæra, S.O., Thommassen, J.M. (1996)

Voluntary fish transport in land based fish farms.Aquacultural Engineering, 15: 13–25.

22. Summerfelt, S.T., Holland, K.H., Hankin, J.A.,Durant, M.D. (1995) Hydroacoustic waste feed con-troller for tank systems. Water Science and Technol-ogy, 31: 123–129.


14Ponds

night they consume oxygen. Thus there will be adaily fluctuation in the oxygen level in the pond,and special care must be taken during the nightwhen it may be necessary to supply additionaloxygen. Similarly, the pH may fluctuate becausephotosynthesis fixes carbon dioxide and thereforethe pH will increase during the daytime, while atnight the algae release carbon dioxide and the pH will drop.

The ecosystem created will affect all nutrientssince a nitrogen cycle will occur in the pond:nitrification will transform TAN to NO2

− andfurther to NO3

−. If there are areas in the pondlacking oxygen, dentrification of NO3

− to N2 will alsooccur.

A major benefit of a pond is therefore that it is possible to utilize this biological production,which includes prey that is food for the fish. In addi-tion, there will also be decomposition of waste.However, to achieve this state the water exchangerate must not be too high (see Section 14.3). Themajor disadvantages with production in ponds isthe low production per unit volume, and the diffi-culties of maintaining control over the waterquality and the actual fish biomass.

14.3 Different production pondsPonds can be separated into those for fry produc-tion and those for on-growing production (Fig.14.1); the difference is normally the size of theponds. However, full production ponds are alsopossible. In such ponds, spawning, fry productionand on-growing all occur, although harvesting canbe quite difficult. Full production ponds may, forinstance, be used in crayfish production (Fig. 14.2).

14.1 IntroductionEarth ponds are the most used unit for fish pro-duction worldwide, and more than 40% of worldaquaculture production is performed in ponds.1

Ponds are used both for fish and shellfish and at different life stages. Important species grown inponds, include different types of carp, catfish,shrimps and prawns.2,3 Ponds are normally used inextensive production4,5 and to some extent in moreintensive production; here, however, the construc-tion is less optimal (see Chapter 13).

An earth pond for aquaculture farming is usuallydefined as a pond where a natural ecosystem iscreated inside. This is the major difference betweenearth ponds and other closed production units,and the reason why they are described separately.The water exchange in the pond is normally verysmall and it will also function as a settling pond,so faeces and particles will settle on the bottom.There will be none or very little self-cleaning in thepond. When establishing the pond this must betaken into consideration, so that there is some sparecapacity.

Inside the pond there can be a monoculture orpolyculture. If using a polyculture, the natural foodcreated in the pond (phytoplankton, zooplankton,aquatic insects, benthic organisms and the vegeta-tion) can be utilized optimally by different species.

14.2 The ecosystemAn established ecosystem in the pond includes fullalgal photosynthesis. During the day the algaeproduce oxygen by photosynthesis, while during the

174

In a pond for fry production, it is especiallyimportant to have a well functioning ecosystem,including photosynthesis. Eggs or newly hatched fryare released into the pond where the on-goingecosystem will produce natural prey for the fry. Asthe fry grow they will gradually feed on other preythat are also available in the pond. Depending onthe desired production, additional feeding of the frymay not be necessary.

To stimulate and increase the development of thenatural ecosystem, it is possible to fertilize thepond.6–8 This increases production of algae andhence production of higher organisms that functionas natural prey for the growing fry. It is, however,easy to lose control of the ecosystem, and totalbreakdown may occur. If fertilizing, it is thereforeof major importance regularly to monitor andcontrol changes in the water quality, for instance bymonitoring the oxygen content in the pond water.

In on-growing ponds, there is often some type ofadditional feeding, but this depends on the species.

Some species will utilize the plants growing in thepond and the organism created by the ecosystem,but this is normally not enough if high productionis wanted; an example here is grass carp. Otherspecies may only use supplied artificial feed, suchas catfish and rainbow trout. In on-growing pondsit is easy to overload the system when adding for-mulated feed and cause problems in the ecosystemwhich will be put out of balance so that the pondfunctions in an uneconomic way.

The water flow through a pond having a naturalecosystem must not be too high, otherwise algaeand natural prey may flow out with the outlet water,and an imbalance in the ecosystem will occur.Many earth ponds are, however, used in this way.If the fish densities are high the water requirementsincrease with the fish density and the ponds func-tion as raceways, for which the pond construction issub-optimal (see Chapter 13). The results are largevariations in pond water quality and accumulationof faeces and feed loss.

Ponds 175

A

C

B

Figure 14.1 Ponds for on-growing production during (A) the summer (catfish) and (B) the winter (rainbowtrout). Warm groundwater is utilized in the winter seasonto keep the pond free from ice. (C) Empty ponds revealthe construction more clearly.


14.4 Pond typesThere are several ways to classify ponds: one isbased on construction and another on whether it ispossible to drain the pond or not.

14.4.1 Construction principles

Based on their construction three types of pondscan be identified: watershed, excavated andembankment or levee ponds7 (Fig. 14.3).Watershedponds utilize the terrain features; for instance, aravine can be dammed so the construction is quitesimple. However, there are few sites that satisfy the

requirements for a watershed pond, so this is not avery common pond type. An excavated pond issimply a hole in the ground which is filled withwater. Part can be below the water table and in thisway water infiltrates into the pond, but this con-struction is little used.

The main type of construction is the embank-ment or levee pond. There are several ways toestablish such ponds: they can be at ground level,or the levee can be above and the bottom belowground level. For the first type it is necessary tosupply material; for the second type the excavatedmaterial can be used to construct the leveess whichwill reduce the cost of establishing the pond. Levee

Figure 14.2 A crayfish pond in cross-section and from above.

ponds can be constructed in a flat landscape, andlarge areas can be used for pond production. It isimportant that the levee is sufficiently wide to carrytraffic, for instance for feeding, maintenance or harvesting.

When constructing the pond it is an advantage toensure that it is possible to tap the water to a lowerlevel to make drainage possible. Eventually a canaldrainage system can be excavated in conjunctionwith the pond area.

14.4.2 Drainable or non-drainable

Depending on the construction of the outletsystem, traditional earth ponds can be divided intodrainable and non-drainable ponds (Fig. 14.4); non-drainable ponds are normally larger, up to severalhectares. Small natural lakes used for aquacultureproduction function like a non-drainable pond. Thesame is the case if a ravine is dammed, watershedponds are used or if a hole is excavated in the

ground. Low establishment cost is the main advan-tage with non-drainable ponds. Natural lakes canalso be used and create a low cost farming volume.

The advantage with drainable ponds is the possi-bility for a more effective harvesting process and tohave control over the water level, because the watercan be drained out of the pond. It is simpler to fer-tilize/feed, and also to supply additional air. Whilenon-drainable ponds are normally run extensively,drainable ponds can be run more intensively,depending on the amount and growth rate of thefish. In intensive drifted ponds the fish can be fed(Fig. 14.5), and additional air supplied periodically.

Ponds 177

Figure 14.3 Ponds can be divided into: 1, excavated;2, embankment or levee; 3, watershed ponds.

Figure 14.4 Drainable and non-drainable ponds areconstructed differently.

Figure 14.5 Feeding from a tractor-trailer with a feedblower to increase production in the pond.


This may also be done without damaging theecosystem in the pond, if extra care is taken.

In ponds the production is given in kg per hectareof pond surface area. This varies with water tem-perature, environmental conditions, pond type andthe fish species, so it is difficult to give a generalvalue.9,10 A rough estimate is 1000kg/ha, but this canvary by a factor of 10 in both directions: over 15000kg/ha can be achieved for channel catfish by use ofadditional feeding and continuous aeration.2,7

14.5 Size and constructionThe size of the pond varies with species, fish sizeand site conditions, from fractions of a hectare toseveral hectares. As the ponds become larger,control becomes more difficult.The same is the casewith harvesting. To carry out the harvesting thepond must normally be emptied and/or drainedseveral times, otherwise the fish density will be toohigh when lowering the water level and harvesting.A seine net may also be used for harvesting. Com-monly, relatively small ponds are used for broodstock, fry and juvenile production, while largerponds are used for on-growing.

Pond depth is usually between 0.5 and 2.4m,depending on what the pond is used for. For on-growing fish it is normal to choose a depth sufficientto prevent any light reaching the bottom of thepond. In this way growth of vegetation at thebottom is prevented and harvesting is easier. Pondsfor fry are normally shallower because the bottomvegetation may function as shelter. However, thedepth must not be so great that temperature layersoccur; deep water will also increase the pressure on the sides and bottom of the pond together with possibilities for increased seepage. This increasesdemands for compaction of the material used for the bottom and sides when constructing thepond.

It is important to have a slope towards the outleton the bottom to make drainage possible and har-vesting easier: this can be in the range 1/1000 to1/100, with the largest slope in the smallest ponds.7,10

The slope of the walls inside the pond is between2.5 :1 and 4 :1; 3 :1 is quite common, but this varieswith the composition and angle of repose of thematerial used for construction. Exterior walls arenormally slightly steeper than 2 :1, but of coursethis also depends on the angle of repose for the

material.The recommended width of the pond crestvaries with depth; for 3m deep ponds a crest widthof 1.8m is recommended;11 for shallow ponds thecrest width can be reduced depending on the mate-rial used for construction.

The length–width ratio of ponds is normallyabout 2 :1, but of course is adapted to the site con-ditions. The shape of watershed ponds depends onthe terrain. Harvesting with a seine net is easier ifthe pond is rectangular. If ponds are too wide, har-vesting will be more difficult.

14.6 Site selectionPonds should be as close to the water source as pos-sible, to avoid long inlet pipes or channels. In addi-tion, there must be enough clay in the earth toprevent leakage. A rule of thumb is that the mate-rial must consist of at least 20% clay particles withdiameter below 0.002mm in a 1.5m deep coretaken where the pond is to be established.12 If thematerial contains too much sand, it will be porousand water will drain out much faster. The seepageloss in sand is reported to be between 25 and 250mm/day, in loam 8–20mm/day and in clay1.25–10mm/day.13 Furthermore, the earth must befree of toxic substances, for instance copper.

There are several methods to prevent leakagefrom ponds. If the leakage is only slight, a solutionis to break down the earth structure, reduce theaggregate size and puddle the bottom. Breaking upthe lumps in the surface layer achieves this and isquite commonly done on rice fields. Addition ofchemicals may also reduce the aggregate size.14

Compression of the surface may also be used toreduce the water loss, for example by using a roadroller. Several thin layers of compressed earth arebetter than one thick layer. If the natural soil isunsuitable, a membrane of clay or plastic may beused. A clay layer transported to the site must beabout 30cm thick for a 3m deep pond.11 This,however, represents increased costs for establishingthe pond. To avoid the layer of clay crumbling as aresult of drying or freezing, a covering layer of sandor gravel may be used; this can be from 30 to 45cmthick on clay and 15–20cm thick on plastic.14 Theplastic membrane should also be covered to avoidbreakage from plants growing through it. Materialused to construct the pond may also be sprayedwith plant poison before laying the membrane.

After ponds have been in use for a time, waterleakage will normally be reduced because settledmaterials block the cracks in the earth.

14.7 Water supplyThe volume of water supply depends on theamount of fish in the pond and the intensity offarming in relation to evaporation.15,16 In addition,the acceptable time for filling the pond with wateris important. Normally this is the value specified forthe water supply, because the water flow into apond under normal farming conditions is very low.

The value used to calculate the necessary watersupply should be sufficient to fill the total pondvolume within 2–3 days. The whole farm, includingall the ponds, should be filled during a 20 dayperiod.12 The amount of water needed is, of course,affected by evaporation. The total evaporationdepends on temperature, cloud conditions, windconditions and pond construction; Normally it is inthe region of 0.25 to 1cm per day in temperateareas.9

ExampleCalculate the necessary water supply to cover theevaporation loss to a 100m2 pond where the evapo-ration is 0.5cm per day. The water supply is equal tothe evaporation volume.

Evaporation volume = 100m2 × 0.005m = 0.5m3/day = 0.35 l/min

If the water supply and exchange are too high,the production of algae will not be adequate. Thealgae may flow out through the outlet and theecosystem will not flourish. The ponds are nowfunctioning as raceways, for which the constructionis sub-optimal. The ecosystem inside the pondbecomes unstable when the daily water exchange istoo high, i.e. between 25 and 33% of the totalvolume, which equates to retention times of 3–4days.9

In addition to supplying water, it may be neces-sary to supply extra air or oxygen17 (Fig. 14.6).Control of concentrations CO2 and ammonia incritical situations may also be necessary; additionalinlet water may be required in such circumstances.

14.8 The inletThe water can either be supplied by a pump orunder gravity; the latter is the best solution. Sea-water ponds may be filled at high tide.

Channels or pipes can be used to distribute thewater from the source to the ponds. If channels areused, gravity flow is necessary. If pumps are usedthey must eventually lift the water from the sourceinto the distribution channel.

Ponds 179

Figure 14.6 A surface aerator can be used to increase the supply of oxygen to the pond. The photograph showsa paddle aerator on shore.


Since the amount of inlet water is normally quitelow, the actual design of the inlet pipe is of lessinterest than for other closed production units.Therefore where and how the inlet water is takeninto the pond is less important than for other closedunits. If using channels, the inlet from the distribu-tion channel into the pond is just a hatch in a smallchannel branching from the distribution channel.Water is usually supplied to the ponds continuously,but batch filling is also possible.

14.9 The outlet – drainageHow the outlet is constructed depends on whetheror not there is a collection basin for the fish, and ifthis is inside or outside the pond. The design of thelevel control will also influence the construction ofthe outlet system. If open channels are used, thereare separate ones for the outlet water and levelcontrol.

Normally a standpipe inside the pond functionsas a level control.The water has to pass through thisstandpipe, which may be variable or fixed, to flowout of the pond.A swivel can be used to control thelevel of the standpipe; this will again control thewater level in the pond. If using a double standpipe,it is possible to take the outlet water from somedepth in the pond. The standpipe and level controlmay also be placed outside the pond.

Special material must be used at the end of theoutlet pipe to prevent erosion when tapping down

the pond (shock tapping); concrete is usually usedhere. Concrete can also be used to construct the col-lecting basin for fish so that fry can be harvested orcollected. When the outlet pipe is laid through thepond levee it is important to use mooring blocks,for instance of concrete, which are clamped on thepipe to prevent it being dragged out by the fric-tional forces of the water on the pipe. When boththe flow rate and velocity are high, this is especiallycritical.

If channels are used for drainage, a speciallydesign outlet is commonly placed in a small channelwhere the water leaves the pond; this is known asa weir gate or monk15,18 (Fig. 14.7). It is normallyconstructed with two plates vertically installed inthe channel, like hatches. The water has to passbelow the first and above the second. The level ofthe second controls the level in the pond. From thepond the outlet water continues through a commondrainage channel out of the farm. For large waterflows open channels represent a simple cost effec-tive solution.

The methods chosen for handling the fish mayinfluence the construction of the pond and hencethe outlet system. To collect the fish from the pondit is normal to either reduce the water level with thehelp of the outlet or to use a seine net, or eventu-ally a combination of both methods. When usingwater reduction as a collection system, a collectionbasin, either inside or outside the pond, can be usedto collect the fish. If the basin is inside the pond it

Figure 14.7 Outlet and inlet con-struction of a pond.

is normally small and at the bottom of the pond; itis commonly made of concrete or wooden planks.The outlet pipe from the pond is taken out from thisbasin. When the fish have been collected they canbe removed from the basin using a net, pump orscrew. The other possibility is to have the collectionbasin outside the pond. In this case the fish aretapped directly out of the pond together with thewater and collected in the external collection basin.

Both these solutions improve capacity and reducethe cost of fish handling.

A seine net may also be used for collection. Typ-ically it is dragged through the pond and the fishare collected. The net can be hauled from the pondlevees. If the pond is too wide, this will, however, bequite difficult, so widths of more than 20m are notrecommended.19 The seine net may also be draggedmechanically, for instance by a tractor.

Ponds 181

Figure 14.8 Several pond layouts areused: series, parallel and radial. The parallel layout is the most common.


14.10 Pond layoutA farm normally comprises several ponds. Thearrangement of the ponds is important for optimalutilization of the area, and to ensure efficient watertransport, fish handling and (eventually) feed han-dling. If watershed ponds are used, they must beadapted to the ground conditions and the layout isnormally predetermined. If using levee or embank-ment ponds the layout is more important. Rectan-gular ponds are usually best regarding utilization ofthe area. Four main layouts may be used (partlyfrom ref. 19) (Fig. 14.8):

• Series ponds are constructed so that the waterflows from one pond into the next.The advantageis that gravity can be used to ensure the waterflows though the entire farm. The main disad-vantage is that the effluent water from one pondis the inlet water to the next pond and waterquality decreases from pond to pond. Eventuallydisease pathogens will also follow the water andspread disease from pond to pond as isolation ofa single pond is impossible. The water may beaerated when flowing from one pond to the next.

• Parallel ponds are set out beside each other, witha common water supply canal and a commoneffluent water canal. This is the most usual layoutfor a pond farm. The advantages are that thewater quality is the same in each pond and it isalso possible to increase and reduce the waterflow to the separate ponds.

• Radial ponds are in a circle with the smallestclose to the centre and the larger ponds outsidethis. In this way the size of the ponds increaseswith the radius of the circle. The great advantagewith this system is that fish handling is very easy. If the fry are in the inner ponds they can gradually be moved to larger ponds whenthey grow and need a greater volume of water.The empty ponds can then be restocked with new fish.

• Inset ponds are small ponds placed inside a largerpond. This method can provide a nursery pondinside a grow-out pond.

References1. Nash, C.E. (1988) A global overview of aquaculture

production. Journal of World Aquaculture Society, 19:51–58.

2. Stickney, R.R. (2000) Pond culture. In: Encyclopediaof Aquaculture (ed. R.R. Stickney). John Wiley &Sons.

3. Stickney, R.R. (1994) Principles of aquaculture. JohnWiley & Sons.

4. Coche, A.G., Muir, J.F. (1992) Pond construction forfreshwater fish culture, pond farming structures andlayouts. FAO Training Series, FAO.

5. Coche, A.G., Muir, J.F., Laughlin, T. (1995) Pond con-struction for freshwater fish culture, building earthenponds. FAO training series, FAO.

6. Kwei Lin, C., Teichert-Coddington, D.R., Green,B.W., Veverica, K.L. (1997) In: Dynamics of pond aquaculture (eds H.S. Egna, C.E. Boyd). CRCPress.


8. Brunson, M.V., Haregreaves, J., Stone, N. (2000) Fer-tilization of fish ponds. In: Encyclopaedia of aquacul-ture (ed. R.R. Stickney). John Wiley & Sons.

9. Timmons, M.B., Riley, J., Brune, D., Lekang, O.I.(1999) Facilities design. In: CIGR handbook of agri-cultural engineering, part II aquaculture engineering(ed. F. Wheaton). American Society of AgriculturalEngineers.

10. Hargreaves, J.A.,Tucker, C.S. (2003) Defining loadinglimits of static ponds for catfish aquaculture. Aqua-cultural Engineering, 28: 47–63.


12. Summerfelt, R.C. (ed) (1996) Walley culture manual.North Central Regional Aquaculture Centre, IowaState University.

13. Coche, A.G., Van der Wal, H. (1981) Water, for fresh-water fish culture. FAO training series. FAO.

14. Wheaton, F.W. (1977) Aquacultural engineering.R. Krieger.

15. Yoo, K.H., Boyd, C.E. (1994) Hydrology and watersupply for pond aquaculture. Chapman & Hall.

16. Kelly, A.M., Kohler, C.C. (1997) In: Dynamics of pond aquaculture (eds H.S. Egna, C.E. Boyd). CRCPress.

17. Boyd, C.E. (1998) Pond water aeration systems.Aqua-cultural Engineering, 18: 9–40.

18. Landau, M. (1992) Introduction to aquaculture. JohnWiley & Sons.

19. Lucas, J.S., Southgate, P.C. (2003) Aquaculture,farming aquatic animals and plants. Fishing NewsBooks, Blackwell.

15Sea Cages

pilings driven into the seabed to which the net isthen fixed to fence in an area.

Another classification of sea cages divides theminto two categories depending on the nature of thebag that makes up the cage; it may be an open bag ofnet, or a closed bag of plastic, for instance. A closedbag will normally require water to be pumped intoit, and there is an outlet pipe from the bag.Actually,a closed production unit has been created.

Open offshore cages can be classified as follows:4

• Class 1 Gravity cages that rely on buoyancy andweight to hold their shape and volume againstenvironmental forces (the focus of this chapter)

• Class 2 Anchor tension cages that rely on theanchor tension to keep their shape and volume

• Class 3 Self-supporting cages that rely on a com-bination of compression in rigid elements andtension in flexible elements to keep the net inposition so the shape and volume are maintained

• Class 4 Rigid self-supporting cages that rely onrigid constructions such as beams and joints tokeep their shape and volume.

In this chapter the focus is on open floating seacages which are those most used for intensive aqua-culture. A traditional open cage comprises the fol-lowing main parts (Fig. 15.1):

• Net bag with weights in the bottom to spread thebag

• A jumping net above the surface fixed to the netbag to prevent fish escaping

• Cage collar for spreading out the net bag andgive buoyancy to keep the bag in the correct posi-tion in the water column

• Mooring system.

15.1 IntroductionA cage represents a delineated volume in the bodyof water where the aquatic organisms can be farmed.Cage aquaculture may date back to as early as the 1200s in some areas of Asia,1 and is currently amajor form of aquaculture in countries includingCanada, Chile, Japan, Norway and Scotland, whereit has been successfully used, mainly for salmonidfarming. However, a large variety of species aregrown in cages today and include seawater, fresh-water and diadromous species. Therefore todaycages are used worldwide in the sea, in lakes andlarge rivers.2 The main differences are in the sizeand construction for withstanding waves and cur-rents. Trends today are that new more weather-exposed sites are taken into use to ensure continuousgrowth in the cage farming industry. The number ofgood sites in less exposed locations is limited.

There are a number of approaches to designing a cage and also classifying the various cagesystems.1,3,4 One classification is based on where inthe water column the cage floats. Three categoriescan be used: floating, submerged, or submersible.The last two types consist of a frame that can floaton the surface and that maintains its shape whenlowered below the water surface.

Another classification is according to the type ofnet used in cages: rigid or flexible. Rigid nets maybe created by using a flexible net attached to a stiffframework to distend it. Instead of using a flexiblenet a rigid metal net may be used. A rigid net cagewill maintain its original shape regardless of thewaves.

Instead of using a floating construction, a fixedconstruction may be used. This can, for example, be

183


When choosing technology and systems for farm-ing in traditionally open sea cages, there are manyconditions to be evaluated. This is also the casewhen the actual type of cage and mooring systemis chosen and designed. The following list can be used to help when establishing a new sea cagefarm:

(1) Choose a site that is suitable for farming(2) Describe and calculate the environmental con-

ditions on the site(3) Choose farming systems, i.e. the cage and moor-

ing system, adapted to site conditions(4) Design the cages (normally done by the cage

manufacture) and mooring system(5) Set out the cages and mooring system(6) Establish requirements for operational control

of the system.

15.2 Site selectionSelecting a good site is of major importance for thefuture economic viability of the cage farm. A suit-able site for cage farming must fulfill a number ofrequirements.1,3 It is normally difficult to fulfill allof these, and they will depend also on the cage tech-nology used; for example, the extent of wave toler-ance. There are a number of ways to classify thefactors that must be evaluated when selecting asite.1,3 Several of the factors also affect each otherdirectly.

The main factor is, of course, the water quality.This must be satisfactory for the cultured species,including temperature, salinity and oxygen content.A continuous supply of oxygen requires a currentto exchange the water. This is also required toremove metabolic products from the cage area.A good water exchange will occur with a water

velocity above 0.1m/s. This is normally sufficient to supply enough oxygen and to remove fish excrement. However, the water currents ought to be below 1m/s because velocities above thisresult in very large forces on the cage structures andmooring system; in these situations specially de-signed systems must be used. Fjords with a sill arenot recommended because the water current and water exchange are normally too low to trans-port the faeces and eventual feed loss away so thiswill collect on the bottom below the cages anddecomposition under anaerobic conditions mayoccur. Hydrogen sulphide (H2S) gas which is toxicfor the fish may then be released from the bottomsediments (Fig. 15.2). Areas where the water can be polluted with toxic substances must also beavoided; this can, for instance, be near variousindustries. Some areas are also more exposed toalgal blooms and some sites are particularlyexposed to fouling; this must be taken into consid-eration when selecting a site.

Shelter from the weather is also important. Waveheight is normally the most critical parameter. It is usual to avoid areas with high waves, even if it is theoretically possible to build farms and

Figure 15.1 Major components in atraditional open sea cage farm.

Figure 15.2 A sill fjord is not recommended for cagefarming.

mooring systems that can tolerate very large waves.However, these farms are difficult and expensive to operate when the waves are large and opera-tional access is reduced. In addition, large expen-sive boats have to be used to operate such farms.If the wave height is below 2m the cage is easy to operate, and many available cages are con-structed to tolerate such wave heights. Several sup-pliers deliver cages that may tolerate 4–5m waveheight. Ocean cages can tolerate up to 7–8m, butspecial routines for operation of such farms mustbe taken into consideration before selecting suchsites.

Another factor included in the geographical con-ditions on site involves water depth; a distance ofmore than 5m from the bottom of the net to the seabottom is recommended, but this depends on thecurrent conditions. Depths above 100m will greatlyincrease the costs of the mooring system becauselong mooring lines will be needed. Areas with fre-quently shipping traffic should be avoided becauseof disturbance to the fish and creation of waves.When selecting a site, good infrastructure, forexample, proximity to roads, available electricity,will also be of benefit.

The legal requirements for fish farming in an area must be satisfied. There may be areas publiclydesignated for other purposes, or where cagefarming is unwanted from an environmental pointof view. For example, it will be difficult to establisha cage farm for salmonid production just outside an important salmon river because of the risk ofescape. The legal requirements for access to landfor an on-shore base and on-shore mooring are alsoincluded here.

Before choosing a site, the environmental condi-tions must be clearly known. This information may

be obtained from government maritime depart-ments or by the use of special oceanographic buoysthat automatically monitor environmental condi-tions on the sites. Talking with people living in thearea and local fishermen could also give valuablesupplementary information.

15.3 Environmental factors affecting afloating construction

15.3.1 Waves

Waves are normally the limiting factor for siteselection for cage aquaculture. If the wave height istoo great it is very probable that this will affect thecage structure. Knowledge of the wave climate onthe site will also be an important tool in choosingthe correct cage technology and mooring system toavoid later breakages in cages and moorings.5,6 Thetrend towards using an increasing number of wave-exposed sites for marine cage farming is proving the importance of this.

Several terms are used to describe a wave (Fig. 15.3):

• Crest: the high point of the wave• Trough: the low point of the wave• Wave height: vertical distance between trough

and crest – H (m)• Maximum wave height: highest measured wave

height – Hmax

• Significant wave height: average wave height ofthe highest one-third of the waves recorded in a period. During a recording period there willalways be a variation in the wave height. The sig-nificant wave height corresponds quite well withwhat an observer will record as the approximate

Sea Cages 185

Figure 15.3 Important parametersdescribing a wave.


wave height when looking at the waves over aperiod – Hs

• Wave amplitude: distance from calm water levelto crest or trough; wave height divided by two(H/2) – a (m)

• Wavelength: the horizontal distance between twofollowing crests – L (m)

• Wave period: the time taken for a wave crest totravel a distance equal to one wavelength – T (s)

• Wave frequency: the inverse of the wave period;the same as the number of waves passing a givenpoint per unit time – f (s−1).

Wave calculations

The description and calculation of waves and waveforces are quite difficult, and in this chapter only abrief survey is given. Several textbooks are avail-able and may be consulted for further information(see, for example, refs 7–12). To illustrate what awave is and how it moves, the water volume may berepresented by many single water ‘particles’ whichwill be transported both vertically and horizontallywith the wave (Fig. 15.4); they will move up with thecrest and down with the trough. There is no nettransport of water particles as long as the wave isnot breaking (see below); they stay in the sameplace, but rotate in an orbit depending on the waveheight and wave description. It is normally thecurrent that causes the net transport of the waterparticles in the sea.

To understand how a wave is created the follow-ing simplified explanation can be used. Imagine that a stone is dropped into the water. As it dis-places the water, the water particles are forceddown and away. The energy that the stone adds to

the water will be used to force neighbouring parti-cles up. In this wave a wave is created. The wavecontinues to disperse until the wave motion isdamped by friction between the water particles andno energy is left.

If an object such as a sea cage is lying in thewater, the energy from the wave will also be trans-ferred into this, and it will follow the wave motion.If an object is lying in the sea, however, the wavemotions will be reduced because energy is used tomove the object (see section 15.4.4).

To describe the wave motion, actually how asingle water particle moves, several theories areused. The linear wave theory is the easiest and isalso quite easy to understand, but several simplifi-cations are made compared to real waves.11 In mostcases, however, this theory will suffice. The wavesare described as sine waves, and all the standardgeometrical knowledge of sine waves can be usedto describe them. The single water particle rotatesin a circular orbit where both the acceleration andvelocity vary depending on where the particle is inthe orbit: on top of the crest or down in the trough.However, under practical conditions the waves willnot behave as sine waves. The wave is the sum ofseveral wave systems coming from different directions, with different wave periods, height andphases (Fig. 15.5).This can be described by an irregu-lar wave spectrum. Such spectra will vary from seaarea to sea area. Spectra that are fitted to describethe wave for the different sea areas can be madebased on actual measurements and calculations. Forcalculating wave motions according to developedwave spectra computer programs are used. In thelinear wave theory, which is a simplification usingsine waves, a set of formulae have been developed

Figure 15.4 During a wave cycle thesingle water particles move in differentdirections.

to calculate the wave motion and the total energyin the waves.

Depending on the depth, the wave will ‘travel’ indifferent ways. In shallow water the waves will havean effect all the way to the bottom; at intermediatedepths the wave motion will be reduced closer tothe bottom. In deep water the wave motion willquickly decrease; half a wavelength down there isalmost no wave motion left because so much energyis used to move the surrounding sea. If the cages

are located in the sea, deep-water conditions arecommonly chosen. Here also is the explanation forthe advantages of using submerged cages in waterwith high waves; at a depth of one wavelength thereare no effects of the wave in deep water conditions.

Breaking, reflecting and diffraction of waves

A wave ‘breaks’ when the height increases in rela-tion to the wavelength (Fig. 15.6) and the wave getssteeper. White foam crests characterize a breakingwave visually. Much more energy is consumedwhen a wave starts to break. Under deep-waterconditions wave breaking will occur under the fol-lowing conditions:

H/L > 1/7

where:

H = wave heightL = wavelength.

When a wave breaks there is net transport of waterin the direction that the wave breaks.

ExampleThe wavelength is 30m and the wave height is 5m.Does the wave break?

H/L = 5/30= 1/6> 1/7

Therefore the wave will break.

If the wave height decreases to 3m will the wavebreak now?

3/30 = 1/10< 1/7

Therefore the wave does not break.

Sea Cages 187

Figure 15.5 A wave comprises a number of singlewaves which create an irregular wave spectrum.

Figure 15.6 When the wave heightincreases relative to the wavelengththe wave will start to break.


Reflection of a wave occurs when it hits anobstruction, such as the beach, a rock wall, a rockawash or a floating construction. If a vertical rockwall is hit by the wave an opposite wave motion iscreated and the energy in this wave will send theenergy backwards. Under special conditions it cantherefore be quite calm just outside such rocks,because the wave motions from the two differentwaves neutralize each other. If the wave hits thebeach the depth will decrease and the wave willgradually start to break because it is forced up fromthe bottom and gets steeper. Breaking of waves caneasily be seen when they hit the beach. Waves maynot break before the beach, but when they hit thebeach they start to break. The wave must get rid ofall its stored energy when it reaches the land. Howmuch is reflected is determined by the angle to theshore. Much energy is dispersed when the wavestarts to break.

Diffraction occurs when a wave hits an obstruc-tion; the angle at which the wave hits determinesthe direction of the diffracted wave. Diffractionmay cause waves to be sent into areas that shouldbe sheltered.

What creates waves?

Several factors may create waves but the mostimportant are:

• Wind• Human activity, such as shipping• Special natural phenomena such as earthquakes,

land slips and underwater volcanic eruptionscreate waves known as tsunamis

• Tide; waves with extremely long wavelengths arecreated.

Waves created by the wind are the most relevant to aquaculture facilities and are further describedbelow. Shipping may also create waves that areunwanted on fish farms. To avoid such waves, sitesclose to heavily trafficked sea routes should beavoided.Waves created by exceptional natural phe-nomena (tsunamis) are difficult to avoid even ifsuch phenomena occur more frequently in someareas than others. Tsunamis have a very long wave-length (>100m) and a long period (around 1000s).In such waves an enormous amount of energy isstored. They do not represent a great danger if aboat is on the sea, because the wavelength is so

long. However, when they reach shallower water,and especially when they reach the shore and startto break, all the energy that is stored in them isreleased, and the consequences can be fatal. Suchwaves can be up to 30m high when the shorelinehas forced them to increase in height to dissipatetheir energy; they cause enormous destructionwhen they hit the shore. Waves created by the tide normally present no problems for cage farms.The wavelength here is so long that it is not interpreted as a wave, and the wave period is 12.5h. Such waves can, however, create very strongtidal currents.

Wind created waves: The main ingredient in theformation of waves on the open ocean is wind.Wind created waves are normally also what inhibitssite selection for cage farming. When winds blowacross water, a drag is applied on the surface andpushes the water up, creating a wave. The heightwill increase as long as the wind is strong enoughto add energy to the wave. After a period of timethere will be equilibrium between the energy in thewind and the energy in the waves; the wave heightwill now be stable. Once a wave is generated, it willtravel in the same direction until it meets land or is dampened by an opposing force such as windsblowing against it in the opposite direction, or byfriction.

The height of wind created waves depends on thewind velocity (Uv), the duration of the wind (tv), thefetch length (F) and the presence of other waveswhen the wind begins to blow. The fetch length isthe distance where wave development can takeplace (Fig. 15.7).

The Beaufort wind scale gives the expected wind velocity for the different wind strengths (seebelow). Some scales also present the normal waveheight with different wind strengths; however, thisis on open sea with no protection from land orislands.

In protected water the fetch length where thewind can blow will limit the ability of the wind tocreate waves. The fetch length can be read from achart and is the length of the free water surface. Ifthe fetch where the wind is blowing is narrow, as ina fjord, the wind effect will be less because of thefriction against land on both sides reducing thevelocity. To calculate the effective fetch length acompensation factor called the fetch length factor

is used13 (Fig. 15.8). Calculations of effective fetchlength in narrow fjords are given in the followingtwo examples.

ExampleThe length where the wind blows is 10km and thewidth of the fjord is 2km.

Effective fetch length = fetch length ×fetch length factor

Fetch length/Fetch width = 2/10= 0.2

From Fig. 15.8 the fetch length factor is found to be0.4.

Therefore the effective fetch length is

0.4 × 10 = 4km

ExampleLength where the wind blows is 10km and the widthof the fjord is 10km.

Fetch length/Fetch width = 10/10= 1

From Fig. 15.8 the fetch length factor is found to be0.9.

Therefore effective fetch length is

0.9 × 10 = 9km

In shallow water the Sverdrup–Munk–Bretsnei-der (SMB) method may be used to estimate waveheight. Formulae and diagrams have been devel-oped to find wave heights based on wind velocity(Uv), wind duration (tv) and effective fetch length(Fe) (Fig. 15.9). It must be remembered that somediagrams use the traditional sea units of foot (ft),knot (kn) and nautical mile (nm) (1 ft = 0.3048m;1kn = 0.5144m/s; 1nm = 1852m).

To use the diagram in Fig. 15.9, knowledge of thethree factors, wind velocity (Uv), wind duration (tv)and effective fetch length (Fe) is required. First thewave height is found based on Uv and tv, and after-wards based on Uv and Fe. Of the two differentvalues found, the lower will be the wave height onthe site under the specified conditions becauseeither wind duration or fetch length will limit the maximum wave height. For instance, if the wind duration is short a maximum wave height will not be attained; if the fetch length is also short it will also inhibit maximum development ofwaves, even if the wind duration indicates higherwaves.

ExampleUse SMB to estimate the wave height and waveperiod for a site if a fresh breeze of 20kn blows for2h and the fetch length is 10nm.

First calculate wind velocity and fetch length in SIunits:

Sea Cages 189

Figure 15.7 The fetch length is the length of the freewater surface where the wind blows and can createwaves.

Figure 15.8 If the fetch is narrow the wind velocity will be reduced because of friction against land. Theeffective fetch length is calculated by the use of a compensation factor, the fetch length factor. (Adaptedfrom ref. 13: Saville, 1954.)


20kn = 20 × 0.5144m/s= 10.29m/s

10nm = 10 × 1.852km= 18.5km

Using wind velocity and duration criteria and Fig. 15.9

Significant wave height = ca. 0.6m

Significant wave period = ca. 3.3s

Using wind velocity and fetch length criteria and Fig. 15.9

Significant wave height = ca. 0.8m

Significant wave period = ca. 3.9s

This means that wind duration is the limiting factorfor development of waves; the wind does not blowlong enough to create maximum wave height in pro-portion to the fetch length. Critical values for the sitewill therefore be:

Significant wave height = 0.6m

Significant wave period = 3.3s

The SMB method with the values and diagramgiven above may be used for depths greater than 15m, which is normal in sea cage aquaculture. Forintermediate depths and shallow water other for-mulae and diagrams apply.11

In open sea conditions with no limitation of thewind duration, the wave height will only depend onthe effective fetch length and the wind velocity.Thewave height created is the maximum possible withthe given fetch length. A simplified method canthen be used to calculate wave height in shallowwater11 (Table 15.1):

Hs = 5.112 × 10−4 × UA F1/2 (m)

Ts = 6.238 × 10−2 (UA F)1/3 (s)

UA = 0.71U1.23 (m/s)

where:

U = wind velocity (10min average value 10mabove sea level) (m/s)

UA = adjusted wind velocity (m/s)F = fetch length (m)Hs = significant wave height (m)ts = significant wave period (s).

Figure 15.9 Diagram for calculation of wave height in relation to wind velocity, wind duration and effective fetchlength (adapted from ref. 11 – see here for more details).

The following may be used to estimate themaximum wave height from the significant waveheight:14

Hmax = 1.9Hs

ExampleUse the simplified method to calculate the significantwave height and wave period for a near gale withwind velocity of 15m/s and fetch length of 3km.

UA = 0.71 × 27.96= 19.9m/s

Hs = 5.112 × 10−4 × 19.9 × 30001/2

= 0.56m

Ts = 6.328 × 10−2 × (19.9 × 3000)1/3

= 2.44s

Swell: Swell comprises wind generated wavescreated far away, which can also be called oceanwaves; these may also affect cage farms when theycome in from the sea. This is another reason forsitting cage farms in sheltered positions behindholms and breakwaters. Swells are characterized byquite large wave heights and long wavelengths. Aswell can be recognized by its higher wave periodcompared to a local wind generated wave: typicalswell periods are in the range 9–20s, compared with2–11s for wind generated waves.

15.3.2 Wind

Wind is normally not directly harmful to sea cagefarms.The area of the farm above the water surface,where the wind blows, is small. On large operationalplatforms with buildings, however, the wind willhave some effect.

Wind can be separated into two components: onenormal and the other fluctuating (gusts). Theamount of gusting depends on the local topography.It is normally gusts of wind that cause damage, forinstance to houses.The wind velocity increases withheight above the ground. For meteorological pur-poses, wind velocity is measured 10m above theground (V10). Because it varies continuously it isgiven as an average over a period, normally 10min.Many meteorological stations measure the windvelocity; however, a large number are located at air-ports or lighthouses, where the landscape is quiteflat with few mountains to create gusts.When trans-ferring these wind data to other sites, this must betaken into consideration. An easy method topresent the wind conditions on a site is by usingwind roses; these show where the major wind iscoming from and may also show the averagestrength of the wind from the various directionsover a given period (Fig. 15.10).

15.3.3 Current

Water current is normally the dominant environ-mental force on a sea cage farm. Several factorsmay create a current in the water, including:

• Wind• Tide• Local water flows, such as rivers• Large global oceanic currents or coastal streams

such as the Gulf Stream.

Currents create both horizontal and vertical move-ments in the water. In fish farming the focus is nor-mally on the horizontal currents.

There are large variations in the current from site to site, and the overall current picture com-prises all the different single currents. A properdescription of the current conditions on site musttherefore be included in the site measurements (seebelow).

Wind generated current

Current is created in the water when the windblows over the surface, because there will be a dragfrom the wind on the water surface. The velocity ofthe created water current depends on the strengthof the wind. Because the drag is on the watersurface, wind created current will be highest nearthe surface and decrease with depth. The following

Sea Cages 191

Table 15.1 Examples of wave heights with fully devel-oped sea with different wind velocities and effective fetchlengths (F ). Wind duration is without limitation.

Wind velocityWave height (m)

(m/s) F = 1 km F = 3 km F = 5 km F = 10 km

5 0.08 0.14 0.19 0.2610 0.20 0.33 0.44 0.6215 0.32 0.56 0.72 1.0120 0.46 0.79 1.02 1.4530 0.75 1.30 1.68 2.38


equation may be used to calculate the currentcreated by wind in open water:12

where:

U10 = wind velocity measured 10m above the watersurface

U z Uz

sv( ) = −⎛⎝

⎞⎠0 02

5050

10.

z = distance downwards from the surface to thedepth at which the velocity of the wind gen-erated current is to be found.

ExampleFind the current generated by wind at depths of 5mand 10m. There is open water and the wind is neargale force.

First the near gale force wind velocity is found fromthe Beaufort wind scale to be between 13.9 and 17.1m/s. Therefore use a velocity of 15m/s.The current velocity at 5m depth is then

At 10m depth the wind creates a current of:

Usv(10) = 0.02 × 15 × (40/50)= 0.24m/s

These calculations also show that the velocity isreduced by the depth.

As can be seen from the first equation, in thissubsection, the surface current velocity is 2% of thewind velocity (U10) calculated for open sea. Inshallow water the wind generated current velocitywill normally be somewhat higher, partly due tostratification and reduced thickness of the waterlayers being dragged by the wind, and under prac-tical conditions may be up to 5% of the wind velocity.

Tidal current

Tidal current is created by tidal range. The tidalrange varies from site to site around the world. InCanada, a tidal range of up to 14m occurs. Innarrow fjords and narrow necks the tidal current isespecially high, because the tide forces the water inand out. In North Norway a tidal current of up to16kn (8.23m/s) has been measured in Saltstraumenin the inlet to Skjerstadfjorden, which is the world’sstrongest tidal race.

In open waters the current caused by the tide isnot as high, and depends on the size of the tidalrange in the area. If Norway is taken as an example,the tidal current varies from 0.2 to 0.8m/s along the coast.14 Tidal current is approximately equal

Usv 5( ) = × × −⎛⎝

⎞⎠

=

0.02 1550 5

500.27m/s

Figure 15.10 A wind rose may be used to show wherethe wind is coming from and how strong it is.

through the total water column and does not varywith depth.The strongest tidal current occurs in themiddle between high and low tide.

Oceani currents

Solar heating, variation in water density, wind,gravity and the Coriolis force have influences overthe large ocean currents. The Gulf Stream is one ofthe strongest currents known. It starts in the Gulfof Mexico, passes along the east coast of the USAand crosses the Atlantic Ocean past Ireland andGreat Britain and continues up past the west coastof Norway. Where its velocity is 0.4–0.5m/s. Coastalcurrents present the largest velocity close to landand decrease with depth.

Measuring current

When evaluating a site, the water current must bemeasured. Specially designed instruments are usedto measure the direction and the velocity of thecurrent.They are an integral part of a floating buoy,and also include a recorder to store the results.Some also have a transmitter for downloading themonitored results. It is recommended that the measuring buoy stays out for quite a long period,preferably for a whole year or at least in the periodswhen the strongest and weakest currents occur.Theresults of the measurements can be shown in acurrent rose in the same way that the wind data canbe presented.

15.3.4 Ice

In northern and southern regions near polar areas,ice in the water may be a problem for the develop-ment of cage aquaculture, especially in fresh andbrackish water, but also in the sea. The problemsare of three types:

• Surface ice• Drift ice• Icing up.

Surface ice is mainly a problem where there is asupply of fresh water which reduces the salinity ofthe top layer of the water. This causes the freezingpoint to change to close to 0°C, and the watersurface might freeze.This will, of course, also be thecondition on a freshwater site. From time to time

there may be a thin layer of ice on the surface thatis so sharp that it is able to cut the nets on the cages.

Drift ice does not normally present any problemfor traditional aquaculture sites, but may occur nearthe polar areas from where it is released.

Icing up of parts of the farm above the surfacemight be a problem. Icing up occurs when sea sprayor supercooled rain hits a construction cooled tobelow freezing point. When the water hits the construction, it will immediately freeze and coat theconstruction with ice.This can also happen with air-craft under certain weather conditions and is thereason they are de-iced before take off. The samephenomenon can be observed on fishing vesselsworking in polar areas; under unfavourable condi-tions the vessel may become totally covered withice, the amount and weight which can be so largethat the vessel will sink. When sea cages ice up thesame thing can happen; the loads may be so largethat construction breakage occurs. The weight of the ice can exceed the buoyancy of the collar andthe cage will sink below the surface. Here, however,the ice will melt after a time and the cage willsurface again. Windy conditions and relatively lowtemperatures may cause icing by sea spray. If thetemperature is very low the water may freeze in theair before it hits the construction. Ice may causesupercooling of the water and possibly also fishdeath.

15.4 Construction of sea cagesA typical sea cage comprises several parts: the cagecollar or support system (framework), the flotationsystem, the net bag, a jumping net, and weights tostretch out the net bag at the bottom and to stabi-lize it in the water column.

Three different methods may be used to construct the framework/collar for a sea cage (Fig. 15.11).

(1) Stiff framework The framework does not followthe wave movements.An example of a stiff con-struction is a boat. Some specially designedsteel cages use a stiff framework. The construc-tion is characterized by large forces transferredto the framwork.

(2) Framework with movable joints The frameworkwill to some extent follow the wave move-ments. An example is a traditional steel cage

Sea Cages 193


system, where joints are used to connect thesingle elements in the framework.

(3) Flexible framework The framework is quiteflexible and will follow the wave movementswell. These include frames made of plastic (forexample, polyethylene, PE) which are flexibleto some degree and those made of rubber (forexample, ocean cages).

15.4.1 Cage collar or framework

The collar or framework may have several func-tions. It helps to support the cage safely in the watercolumn, it helps to maintain the shape of the netbag, it may help with buoyancy and it may serve asa work platform.

The framework construction for stretching outthe net bag can be combined with the buoyancy, asseen in plastic floating ring cages (PE or polypropy-lene (PP) pipes).Alternatively, the buoyancy can beindependent of the cage collar as can be seen whenusing wood or steel for support systems with blocksof expanded polystyrene (PS), such as StyrofoamTM,as buoyancy.

The buoyancy is necessary to keep the cage bagin the correct position in the water column. It musthave a smooth surface to inhibit the accumulationof fouling. Fouling increases the weight of thecollar, which results in increased requirements forbuoyancy; furthermore, fouling will increase thefriction between the flowing water and the seacages which again increases the forces on themooring system. Expanded polystyrene is com-

monly used as buoyancy; if not covered with PE,exposure to sunlight causes it to age. It turns yellowand becomes brittle. Uncovered polystyrene willalso be very prone to fouling, because the surfacebecomes so rough. The use of uncovered poly-styrene in sunlight is not recommended; to increaseits durability it is quite common to put it into PEcylinders or rhombs.

If too much buoyancy is added the cage collarwill float high up in the water column and fullyfollow the wave motions, floating on top of thewater column throughout. Unnecessarily largeforces on the cage bag and mooring system fromthe induced vertical motion result if there is muchwave action in the area. The cost of the buoyancywill also be unnecessarily high. Buoyancy elementsought to have an aerodynamic shape to reduce theforces transferred from the water current. Thecurrent forces on the collar are, however, muchsmaller than the forces on the net bag.

The framework or collar can be of circular, polyg-onal or square construction. It is best to use a roundframework because the forces are equal all aroundthe circumference; polygonal or square frameworkswill have large forces in the corners and eventuallybreakages in the construction will occur here (see,for example, ref. 15). For this reason, good connec-tions at these points are important. Wooden frame-works are sometimes used to construct sea cages;only bolts, nails or ropes are used to connect theplanks at the corners. If these cages are used inexposed sites with fast currents and high waves, theframework will break at the weak points in thecorners.

Polygonal collars are better than square collarsbecause there are more corners to share the totalforces, and the force in each corner is thereforereduced.

Different materials may be used in the frame-work (Fig. 15.12), ranging from steel, aluminum,wood and concrete which are rigid, to more flexiblematerials such as PE and rubber. The modulus ofelasticity for the material is a measure of its rigid-ity and is given by its E value, the load in relationto non-permanent deformation. Steel has a high Evalue, while wood has a rather lower value;1 that forPE is even lower. Bamboo is also used in cagecollars, but only on low exposed sites.

The risk of corrosion of the framework in sea-water when using steel or aluminum must be taken

Figure 15.11 Different methods of construction forframeworks of traditional surface sea cages.

into consideration. If steel is used it must always becovered, for example with paint or zinc, to avoidcorrosion.

15.4.2 Weighting and stretching

Weights on the bottom of the net bag are used tokeep the net bag down, and to maintain as mucheffective volume as possible for the fish. Lead ropemay be used for the bottom line in the net bagtogether with lump weights. The lump weights arenormally added at the corners and in the centre. Forexample, on a 15m × 15m square cage, the totalamount of weights can be 150–200kg, divided intolump weights of 25kg in each corner and in thecentre, and the rest evenly spread along the bottomline as lead rope.

Rapid currents will decrease the effective volumeand adding more lump weights can inhibit this;however, care must be taken because this willincrease the current forces on the net bag.The needfor buoyancy will also increase, and the same will

be the case for the size of the mooring system. Useof weights will also increase the dynamic forces on the net bag caused by the waves (stretch andslack).

Instead of using weights, stays may be used tostretch out the net bag; this gives a rigid construc-tion. The environmental forces will be greatlyincreased by doing this and the bag will require amuch larger mooring system.

15.4.3 Net bags

Net bags can be constructed in different ways andwith different materials.16–18 In the past materialssuch as cotton and flax were used. These materialsget heavy in water and their strength is rapidlyreduced; in addition they are not very durable.Today synthetic plastic materials, such as polyamide(PA; nylon) predominate. This material is cheap,strong and not too stiff to work with. PE is also usedto some extent because it is more resistant tofouling as the surface is smoother; it is however,

Sea Cages 195

Figure 15.12 Different designs of framework.


stiffer to work with. Polyester (PES) has also beentried.

Nylon used for nets is made as a multifilamentconsisting of several thin threads spun together tomake a thicker one. The advantage with multifila-ment is that the thread is easy to bend, easy to workwith, tolerates more loads and is more resistant tochafing. In contrast, monofilament is a single threadas used in a fishing line. It can be made of PE; it is stiffer and more vulnerable to chafing than a multifilament.

Nets are either knotted or knotless, in which case they are sewn together. In the past there wereproblems with knotless nets because they cameunstitched, but today this problem has been over-come and both knotted and knotless nets are incommon use.

The normal mesh shape is square; hexagonalmeshes are also used, but to a lesser extent. Hexag-onal meshes are more commonly used for trawlingbags on fishing vessels.

A number of dimensions are used to describe themesh. Bar length is the distance between two knotswhile mesh size is the distance between the knotson a stretched mesh. Mesh size may, however, alsorefer to bar length, which makes this expressionrather confusing. In this chapter mesh size is givenas stretched mesh. In a hexagonal mesh, the meshsize is given as the distance between the two longestparallel bars (Fig. 15.13).

Another expression that indicates how the netpanel is standing in the sea is how the net isstretched in the x and y directions. This can becalled the hanging ratio of the net (E). This is theratio between the length of the stretched net panel(Ly) and the length of the line where the net is fixed(top line) (Lx):

E = LX/Ly

Normally E for net bags for fish farming is in therange 0.6–0.9, while for a fishing net, for instance,E is between 0.4 and 0.6, meaning that fishing nets have meshes that are more stretched out (Fig. 15.14).

Solidity is used to describe how ‘tight’ a net is andis the ratio between the total area that the netcovers, compared to the area covered with threadsincluding knots. This relation is important when theresistance against water flow through the net is to

be calculated. Fouling on the net will increase thesolidity, because the covered area is increased.

The material strength of net panels exposed tosunlight (UV), wind, rain, acid rain, etc. is reduced.This process is called weathering. Polyvinyl chloride(PVC) is the material that is most resistant to

Figure 15.13 Important dimensions used to describe amesh.

Figure 15.14 How the net is stretched in the x and ydirections will result in different shapes of the mesh, andthe net panel will appear different in the sea.

ageing, followed by PE and PA; PP has the shortestlifetime.16 The lifetime can be increased by addinga coloured (black) antioxidant, so the developmentof weathering is reduced. Today, however, whiteuntreated material is commonly used for net bags. As it is usual to add some type of antifoul-ing agent to the nets, this will also cover the multifilament.

The normal lifetime of a net bag will vary withthe site conditions; in Norway, for example, the life-time of a net bag is usually set as 5 years.19 Anotherway of controlling the duration of a net bag is tocarry out a strength test. In Norway, the breakingstrength of the net bag below the surface must notfall below 65% of the initial strength.

When the bag is exposed to water currents thevolume is reduced by deflection. Because of this, thenet bag must be correctly constructed. Narrow deepnets are not recommended on sites exposed to cur-rents. Recommended net depths are 0.8–1.25 timesthe diameter of the bag.20

Because wave motions decrease significantlywith increasing depth, it is an advantage to placethe cage bags at some depth on exposed sites; 15–25m deep is normal for large cages with a circumfer-ence above 60m. The recommended depth of cagebags in some exposed sites is more than eight timesthe significant wave height.

The merits of vertical or sloping sides in the netbag are as follows. Volume reduction is limited byusing sloping sides, but if the amount of lumpweights has to be increased to maintain bag volume,the forces are reduced if the bag has vertical sides.

15.4.4 Breakwaters

On sites exposed to waves, breakwaters may beused to reduce wave height and impact. Hence theenvironmental loads on the cages lying behind thebreakwater will also be reduced. Breakwaters maybe constructed in different ways.1 One method is touse concrete blocks or a steel construction fixed tothe bottom; however, these are expensive to installand little used for protection of sea cages in deepwater, although they may be used in shallow water.Pneumatic barriers with air bubbles may also occa-sionally be used. Most usually a breakwater madeof rubber tyres is used. Old tyres from trucks or carsare tied up with wire to form a fleet. The width of

the breakwater is important for its effect on thebreaking waves. It is normal to use several break-waters and their total width must be at least as greatas the width of the farm to be protected. The widthof a breakwater is recommended to be at least 1.5times its length.21 This is because the most damag-ing wavelengths are 0.5–1.25 times the length of thestructure. A simple pipe will to some extent func-tion as a breakwater, but it is much less effectivethan specially designed breakwaters; eventuallyseveral pipes can be placed adjacent to each other.

Breakwaters can be moored like sea cages. Somedistance is necessary between the breakwater andthe farm: up to four wavelengths is recommended.1

The breakwaters will then create a shadow wherethe cages are placed.

The breakwater decreases the wave height byreducing the energy in the wave. This is becausethere is:

• Reflection and waves travelling in the oppositedirection are created

• Distribution in the breakwater• Transfer of energy to the breakwater• Deflection of waves hitting the corners of the

breakwater.

15.4.5 Examples of cage constructions

Below, examples of designs used for sea cages areshown to give some idea of the dimensions.

Plastic cages

Plastic collar cages made either of PE (actually highdensity polyethylene (HDPE)) or PP are often circular, but may also be made quadrangular and be used as a system farm in less exposed sea areas.

In circular cages it is normal to have two pipes ofdiameter 200–315mm. Both may be filled with PS,or one filled with PS while the other is air filled. Awide range of circumferences are available, com-monly between 30 and 120m. Between the twopipes some type of fitting is used, either of plasticor steel, to ensure strength and make a base for thewalkway. Some also use a chain all round the cir-cumference to improve the strength. Examples ofrequired buoyancy are from 40 to 120kg/m depend-ing on the cage dimensions.

Sea Cages 197


Steel cages

Steel cages are constructed with pontoons to ensure buoyancy, while the steel framework givesstrength and stretches out the net bag. The steelconstruction is normally galvanized but can also bepainted. Typically there is a 2–3m wide walkwaythat eventually can be used by a small forklift truck.Normally the buoyancy is in the range 800–4000kg/m2, the highest value being for walkways fordriving. The walkways around the cages are smaller(up to 1m wide) and have lower buoyancy of around500kg/m2. Between the centre gangway and side-ways there are special movable hinges. The pontoons are normally made of PE infilled withexpanded PS.

Ocean cages

One type of cage suitable for large waves is madeof rubber pipes with a typical exterior size largerthan 400mm. The cages are made as a quadrangle,hexagon or octagon. Steel pipes are used in thecorners and to connect the parts made of rubberpipes.The rubber pipes will follow the wave motionvery well. The cages are reported to tolerate veryrough weather conditions, such as wave heights ofup to 8m.22

Ocean spar technology is another technologyavailable for ocean cages, and these have no typicalcage collar.23 In one system vertical cylinders(spars) are placed in each corner of a quadrangularcage bag to stretch it out; the horizontal areasaffected by the waves are thus reduced. Anothertype includes a central spar and rim held togetherwith tension stays. It forms a cube-like constructionthat is only partly above the water surface whichwill be dragged below the surface when there ismuch wave activity or strong currents.

15.5 Mooring systemsThe function of the mooring system is to keep thefarm in a fixed position and to avoid transfer ofexcessive forces to the cages, especially verticalforces. Different methods are used for mooringdepending on the type of cage, how exposed thesites are to the weather, and the requirement forposition exactness. Two major systems are used for

mooring: pre-stressed and slack (Fig. 15.15). Slackmooring is used to moor ships which drift aroundone anchorage point. Such mooring systems arewell adapted for stiff constructions such as ships.Few cage farming systems are stiff and thereforepre-stressed mooring systems are most often used,but slack mooring has also been tried.24 Pre-stressed systems are well adapted for use in flexibleconstructions, and in correctly designed systems theforces will be equally spread over the entire farm.Pre-stressing of the mooring system is performed athigh tide and forces can be up to several tens ofkilonewtons.

A special type of mooring is needed for tensionleg cages25 in which the forces are taken up in thetension legs which, regardless of the weather con-ditions, are always under tension. In this way thedynamic loads resulting from the weather, thataffect traditionally moored cages and create slackand tension in the lines, are avoided. The challengewhen constructing tension leg moorings is to findtechnical solutions where the legs will always betensioned.

A pre-stressed mooring system contains threemajor parts (Fig. 15.15):

(1) Mooring lines which include the point ofattachment to the cages

(2) Buoys(3) Anchors.

Later in this chapter the design and constructionof different mooring systems are described withprimary emphasis on mooring of seawater cagesexposed to some waves (>1m) and current (up to 1m/s). In well protected seawater sites and fresh-water sites the environmental loads transferred tothe cages are reduced and a smaller mooring systemcan be used. However, the same basic principles canbe used for design and construction.

15.5.1 Design of the mooring system

The design used for the mooring system dependson the type of cages to be moored: these may be asfollows (Fig. 15.16):

• Single cages• System for mooring several single cages• Single cages with walkway• Single cages with walkway and landing

• Group of cages – platform cages• Ocean cages• Cages lying on sway.

A single cage may be moored by between fourand six single buoys attached to anchors and to thecage by mooring lines. It is recommended that the single mooring line be divided into two beforethe fixing point to the cage; this is known as the henfoot mooring; it reduces the forces at the points ofattachment to the cage framework, because thenumber of fixing points is doubled. The points ofattachment are critical in the mooring system,because all forces are transferred through thesepoints. The cost of mooring single cages is high; the

system is most viable for circular and polygonalcages.

Today, systems for mooring several cages aremore usually used. Two or three longitudinalmooring lines are attached to each other with trans-verse mooring lines; in this way a frame where thesingle cages can be fixed is built. By lowering theframe 1–2m below the water surface, access to the cages by boat is improved. Here again, the singlecages are attached to the frame with a hen footmooring. One single cage may be removed, takenout from the system and transferred to another site.

Single cages may also be moored to walkways.If the walkway has a landing, the requirement tokeep the cages in the same position is greater (high

Sea Cages 199

Figure 15.15 The mooring system of sea cages con-sists of three major parts; the system can be pre-stressedor slack. Pre-stressed systems are commonly used whenmooring sea cages.


Fig

ure

15.

16V

ario

us s

yste

ms

for

moo

ring

sea

cage

s.

position exactness); the pre-stress in the mooringlines must also be higher. With such moorings theproblem is that the wave is transmitted through thecage and the walkway with different velocities,resulting in large forces at the connection points.The connection point may therefore be subject tomaterial failure. To reduce the forces it is advanta-geous to use several connection points. Alterna-tively, the connection point must be flexible. Fixedconnections to walkways are not recommended inexposed water.

In ocean cages the requirements for the moor-ings system are large because the forces are sogreat. Ocean cages are moored individually and theproducer will normally have their own mooringsystems, which could be an integral part of the cageconstruction.

In a platform cage farm the mooring line is typ-ically withdrawn from each corner and from themiddle of the farm. Such farms are normally sitedin protected water because they tolerate lowerwaves, due to their construction with linkages.

If the sway principle is used, only one mooringpoint is necessary; it is actually a slack mooring.Thismooring point is exposed to large forces and a biganchor and buoy are necessary.24 Such farms needa large area because the unit will drift around themooring point. Ageing of the sites is, however,delayed when using this arrangement. However, themethod is seldom used, mainly because such a largearea is necessary.

15.5.2 Description of the single components in apre-stressed mooring system

Fixing point

The mooring line is fixed to the collar by a shackle.If a rope is used this should be spliced and a thimbleused to reduce the bending of the rope.All bendingwill weaken ropes to some extent. A bending diam-eter of three times the rope thickness is necessaryto avoid significant weakening. A knot can reducethe strength of the rope by 50%.19 Rings may alsobe used in the connections because they toleratechafing better. It is an advantage to over-dimensionthe connection, for instance by doubling the size ofthe maximum transferred forces.26

The force through the connection point can bedivided into horizontal and vertical components.To

avoid breakage of the cage collar by the tide, thevertical component should be transferred so that it is as low as possible, almost negligible. By usingseveral connection points to the collar, the forcestransferred at each point will be reduced, althoughthis will increase the mooring costs. Normally atleast four points are used. As said previously, it isadvantageous to split the single mooring line beforeattaching it to the collar, so more connection pointsare achieved; such an arrangement is known as ahen foot mooring from its design.

To secure the fixing point against breakage, a sec-ondary fixing may be used (Fig. 15.17). An extrarope or wire transfers forces directly from the collarto the mooring lines if the main fixing point breaks.

Mooring lines

Different materials and designs can be used for themooring lines which are often made of syntheticrope. When choosing rope, the breaking strength isthe most important factor, but price and durationare also major determinants.The elasticity, given bythe E modulus, must also be taken into account.How much a line stretches lengthwise when loadedcan be given in the mooring line characteristics.

Lines of synthetic rope are often stretched per-manently after the first load so are beyond theirelastic range and do not return to their originallength. It may be advantageous to pre-stretch therope before it is used; alternatively, the rope mustbe tightened after exposure to the environmentalloads.

Sea Cages 201

Figure 15.17 Double fixing is recommended where themooring lines are connected to the collar.


Different materials are used for synthetic ropes,such as PA (nylon), PE, PES and PP, all of whichhave advantages and disadvantages. PA toleratesthe highest forces with a given diameter, while PPropes have the lowest weight. Metal or Kevlarthreads may also be integrated into the rope to increase its strength, but this also increases the cost.

Ropes may be exposed to chafing and this mustbe avoided. The rope may be covered, for instancewith a PE pipe, to reduce chafing. Variable loads,such as occur in mooring lines, will decrease thestrength of the rope more than constant loads.19 Thelifetime of synthetic rope in a mooring system canbe set as 4 years as a starting point, but will ofcourse depend on the rope and the loads on it.

If a chain is used in the mooring system, it will be exposed to corrosion. Good quality thereforeconfers an advantage, but this will be a question ofprice compared to duration. Chain is heavier tohandle than rope and therefore more handling timeis necessary. Chain tolerates more mechanical influ-ence (chafing) than rope. For this reason chain isoften used near the bottom, close to the anchor,normally for the first 15–20m of the mooring lines.Afterwards rope is used because it is so much easierto handle. To avoid the connection point beingexposed to chafing against the bottom, a small buoymay be added here.

Wire or metal rope, may also be used formooring, but even though it is strong, it is expen-sive and difficult to work with, so it is seldom used.It may, however, be a good alternative secondaryfixing for cages.

Buoys

Buoys are used to hold the mooring lines up so thatvertical forces on the collar are avoided (Fig. 15.18).

They will also take up the weight of the mooringsystem so this is not transferred to the cage if placedsome distance (normally 15–20m) away from thecollar. By limiting this distance, spreading of themooring system will also be avoided. Furthermore,to ensure horizontal transference of forces, thebuoys have a major roll in pre-stressing the farm bypre-stretching the mooring lines. Buoys will alsofunction like extension springs and damp wavemovements, for instance. When a load is added tothe cage the buoy will be dragged down and not somuch will be seen on the surface; if the load isremoved it will float up again.

Typical buoy sizes are from 200 to 700 l. Thebuoys can be filled with air or foam. To avoid puncturing reducing the buoyancy, the use of foamfilled buoys is highly recommended, and in someplaces is mandatory. Expanded PS or polyurethane(PU) is commonly used. If an air-filled buoy isdragged below the water surface the buoyancy willdecrease rapidly because the buoy is compressed;reduced volume results in reduced buoyancy.Foam-filled buoys will also be compressed whenbeneath the water surface. PVC foam filling will tol-erate more pressure, but is more expensive thanPU. The buoyancy of the buoys is recommended tobe somewhat higher than the calculated require-ment, twice what is necessary, for instance. Round,cylindrical and polygonal buoys are used inmooring systems. Experiments in which singlebuoys were replaced with multiple floats coveringa distance of the rope showed that these gave amuch smoother dynamic response on the anchorlines.24

A lump weight can be added to the mooring linebetween the cage collar and the buoys. This makesit possible to go over the lines by boat, and will alsofunction as an additional pre-stressor of the farmand reduce jerks in the lines. However, the wear

Figure 15.18 Buoys are used toreduce the vertical forces on thecollar, and birdnets are used to avoidbirds taking small fish.

on the mooring lines is increased by this method,which is therefore not recommend.

From the buoys the mooring lines go to themooring points. Depth–length ratios are commonlyset to 2.5–4. Long anchor lines reduce the verticalforces and reduce the required buoyancy in thebuoys, but have the disadvantages that a greaterarea is needed and the cost of the mooring lines isincreased.

The magnitude of the pre-stress in the mooringlines for a cage farm depends on the site and otherfactors. One method is to pre-stress at high tide sothat 75% of the volume of the buoy is below thesurface and 25% is above to take additional loads.Inspection of the buoys could then constitute asimple control of the mooring system. Since allmooring lines are equal and equally pre-stressed, allbuoys will have the same volume above the surface.It can be an advantage to have distinct marks onthe buoys to control more easily how much of thebuoy volume is below the surface. This check must,however, be done when no environmental forcesaffect the cage. If, for instance, a current is comingin from one side, the buoys on this side will bedragged down and those on the opposite side willfloat higher in the water as a result of the unequalloads on the mooring lines.

Anchors

The simplest type of anchor is the dead weight orblock anchor. Theoretically all heavy objects, suchas old engines may be used, but they may be diffi-cult to handle. In addition, this may result in chafingof the mooring lines. Concrete blocks are mostcommonly used as weight anchors and vary fromsome hundreds of kilograms to several tonnes. Asthe density of concrete is quite high (around 2.4t/m3) it will easily sink to the bottom, stay there andtake up forces. The great disadvantage with blockanchors is that they are heavy to handle. Largecranes are needed on the boats setting them outand precautions must be taken to avoid tilting theboat during this operation.

To prevent block anchors being displaced on the bottom, good friction between anchor and thebottom is necessary; this depends on the bottomconditions and is given by the friction coefficient.Sand and clay have high friction coefficients whilethat of rock is low (between 0.1 and 0.5), meaning

that an anchor will slide easily on rock. Blockanchors are not recommended for use on rockyground; here either drag anchors or bolts should beused. Friction coefficients of 0.5 for sand and 0.3 forclay may be used as a basis, for anchor choice if nomeasurements are done.26

To increase the friction coefficient, the bottom ofthe anchors can be rough, even having iron barsemerging from the base. Another method ofincreasing the friction is to add an iron jacketaround the bottom of the block. This will increasethe forces holding the anchor on the bottom,because it is sucked to the bottom like a suckingdisc, assuming suitable ground conditions such asclay. The angle at which the mooring line is joinedto the block is, of course, also important in inhibit-ing horizontal movement (see section 15.7.2). Toavoid tilting of block anchors, the width of theblocks ought to be large in relation to the height(>2 :1), or the block anchor may tilt over themooring lines and cause chafing.

A drag anchor or ebbing anchor is another typemuch used (Fig. 15.19). The aim of the design is forthe anchor to be dragged down into the ground likea plough and become fixed. Various designs ofebbing anchors are available, and the different suppliers normally have their own designs. The oldtraditional one is the stock anchor as used on boats;today more effective designs for mooring of fishfarms are available. How well fixed an ebbinganchor is to the bottom depends on the bottom con-ditions and the design of the anchor; the angle ofthe mooring line is also important. The optimumangle depends both on the bottom conditions andanchor design. The angle for ebbing anchors usedin sand can be 30–35° and in clay 30–50°.1

An ebbing anchor will tolerate a large horizon-tal force, but tolerance of vertical forces is low. Toimprove this, a heavy chain may be used before theanchor. Another method is to use a small blockweight on the mooring line before the anchor. Anebbing anchor is of much lower weight than a blockanchor, because it is based on a totally differentprinciple, but is more expensive. From the backedge of the ebbing anchor there should be a ropeto the surface; to keep the rope on the surfacewhere it can be reached, it is attached to a buoy.This must be done to ensure the anchor can bereleased when removal is necessary, otherwise thismight be impossible as the anchor is normally

Sea Cages 203


completely buried in the bottom after the mooringsystem has been pre-stressed.

Bolts can be used to advantage where it is possi-ble to fasten the mooring lines in rock, usuallywhere the mooring lines lead to land, but bolts mayalso be use under the water surface. Galvanizedbolts are set into drilled holes; either they can havean expanding construction or expanding slurry canbe injected into the holes.The bolts are either of theeye or T-type.

Piles are an alternative in sand and clay bottoms,but the forces they can take up are normally quite

low; otherwise they must be very large whichincreases the cost considerably.

There are several other special anchors thatcould be used, but they are normally more expen-sive. One type used in the offshore oil industry isthe suck anchor. In principle the anchor is a largespecially designed sucking cup that sucks down tothe bottom. This can be made from a wire-spokedwheel with a plastic cover.

15.5.3 Examples of mooring systems in use

The mooring system will naturally be individual toeach site, and depend on the environmental forces.Some examples are given below for some exposedsea sites.The system will also depend on the chosenquality of the rope, the chain and the type of draganchor.

A system mooring is used to moor eight circularplastic cages with a diameter of 15m. The systemmooring is prefabricated and consists of a frame ofchain (13mm × 88mm alloy). Within the frame a 36mm synthetic rope is used. The three main longitudinal mooring lines are of 58mm rope; thisis also used in the ten side moorings. Anchors are32mm eye-bolts and drag anchors with a weight of1.2–1.8t In the last part of the mooring lines to theanchors/bolts chain is used to avoid chafing.

To moor a platform cage farm with a total of 16cages with a diameter of 15m × 15m and a bagdepth of 10–15m, 32mm synthetic ropes are used.27

16mm chain with a weight of 3.8kg/m is used on thelast part to the anchors. A lump weight of 100kg isused at the connection point between the chain andthe rope. Eye-bolts or drag anchors of 800kg areused as anchors and the buoys are 500 l.

Typical anchor sizes for mooring offshore rubbercages are 10–30t for block anchors or 3–5t for draganchors. The mooring ropes are typically 50–70mmdiameter.22

15.6 Calculation of forces on a seacage farmTo be able to design a cage with its mooring system,it is necessary to calculate the environmental forces affecting the farm. On ordinary sites it willbe the current that causes the highest forces, whileon more exposed sites the wave forces will also be

Figure 15.19 Different anchors are used for mooringsea cages: block (A), drag (B), pile (C), bolt (D). Thephotograph shows drag anchors.

considerable.The direct wind forces will be low dueto the small area above the water. Both static anddynamic (from the waves) forces will affect thecages. The size of the forces must be known to findwhether the cage construction will tolerate theforces on the site, or if the construction will breakdown. The supplier of the cage ought to have donethese calculations and know whether the equip-ment will tolerate the forces at the site. However,the environmental forces will also determine thesize and design of the mooring system, and thesemust be determined on site, because the environ-mental loads will vary with the site.

Calculation of the forces that affect the sea cagesis quite difficult, especially for open ocean cages;normally computer programs utilizing numericalmethods are employed (for the latest informationsee, for example, refs 28–30). In this section only abrief overview is presented, focusing on simplifiedsolutions to increase the reader’s knowledge of howenvironmental forces affect floating constructions.

15.6.1 Types of force

A construction floating in the water, for instance aboat or a sea cage, will be exposed to forces as aresult of current, waves and wind.The forces can beresolved into three linear components in the x, yand z directions; the torque occurs around the samethree axes (Fig 15.20):1

• Linear movements– heave: vertical motion– surge: horizontal motion along longitudinal

axis– sway: horizontal motion along the transverse

axis• Rotating movements

– yaw: rotation about the vertical axis– roll: rotation about the longitudinal axis– pitch: rotation about the transverse axis.

When doing calculations, the six forces may beresolved into two forces: a resultant horizontalforce and a resultant vertical force. Even if theaccuracy is less, the calculation is greatly simplified.Thus there are two forces affecting the farm (Fig. 15.21):

Sea Cages 205

Figure 15.20 Like a boat, the top of asea cage top will be subjected to forcesas a result of current, waves and wind.(Adapted in part from ref. 1.)

Figure 15.21 Resolution of the forces affecting thecage farm into horizontal and vertical components.


(1) Drag force parallel with the current direction,FD

(2) Lift force normal to the current direction, FL.

To give some idea of the environmental forcesinvolved, the following minimum values have beenrecommended in Norway26 when designing amooring system for sheltered offshore farms:

• Significant wave height: 1.5m• Current velocity: 1m/s• Wind: 30m/s over 10min period• Storm flood tide: 1m• Fouling: 30% of the area underwater is fouled

completely.

These values will, however, vary from site to sitearound the world, depending of the location.During the past few years an increasing number ofexposed sites have been used for farming purposeswhere environmental forces are greater than thosegiven above.

15.6.2 Calculation of current forces

General methods

All parts of a cage farm below the water surface will be affected by the current.This includes the netbag, the cage collar (pontoons), the buoys and themooring lines. The size of the forces on the ele-ments depends on the area affected by the current.Thus the net bag is affected the most because it hasthe largest area.

To calculate the current forces affecting a sub-merged construction, the first part of Morrison’sequation can be used. The same equation may alsobe used to calculate the wave forces, but here anacceleration term is also included. As the currentvelocity is constant, this term can be neglectedwhen calculating the forces.

Morrison’s equation without the accelerationterm and with forces only in the x direction can bewritten as follows:

where:

FD = dragr = water density

F C U AD D c= 12

2r

A = area affected by the currentUc = current velocityCD = drag coefficient.

The drag coefficient gives a picture of the currentresistance between the object and the water currentand is determined by experiment. If the object issquare the coefficient will be higher than for anaerodynamically shaped construction. The dragcoefficient can be found from the bar diameter andbar length of the net used in the cage bag.31 It canalso be found from the solidity, Reynolds numberand the angle of attack of the current.32

Fouling on the net bag will increase the currentforces because the area of the net is increased, andby this the area that effects the current. Solidityincreases with fouling because the bar diameterincreases. To calculate the current forces on the netbag is, however, not easy. Not only does the waterpass through the net bag, but the current will alsocause the net bag to deflect (Fig. 15.22). By havingseveral net bags one behind another, the currentvelocity will be reduced from net panel to net panelas a result of the resistance to flow through the netbags; therefore more and more water will graduallyflow below and around the bags behind the firstone.

Figure 15.22 A normal net bag will be deflected by thecurrent.

The net bag will be exposed to a lifting force dueto the deflection because it is fixed on the surfaceby the mooring system. More accurate equationsfor the current forces affecting the net bags whichare not rigid, and deflected by an angle a, will there-fore be as follows:

where:

FD = dragFL = lifta = angle of deflection of the netr = water densityA = area affected by the currentUc = current velocityCD = drag coefficientCL = lift coefficient.

These equations will also only be partly correctbecause the net bag is not deflected by a fixed angletowards the surface, but in a curve (Fig. 15.23). Bydividing the bag into several sections with differentangles it is possible to compute the drag and the lift,but this requires a lot of calculations. The coeffi-cients will depend mainly on the solidity of the netand the attack angle.

Because it is fairly complicated to calculate theforces on the entire net bag in one go, a methodwhich divides the net bag into single net panels hasbeen utilized.33,34 For example, a square cage con-tains four side panels and one bottom panel. Theforce is then calculated for each panel and the totalforce represents the sum of the forces on the sepa-rate panels: a trial showed that the calculated forceswere in the range 0.9–1.3 of the measured forces.

The amount by which the net is deflecteddepends on the current velocity, the weights in thebottom of the bag, and the bag design. Experimentshave shown that the reduction in volume of the netbag can be over 90%, when the velocity is increasedfrom 0 to 1m/s.35 To reduce the deformation,increased lump weights in the bottom could beused. However, this will increase the horizontalforce on the net several times over and may not beadvisable.33 In addition, if there are waves extraweights in the bottom of the bag will increase thewave loads on the net bag.

F C U A

F C U A

D D c2

L D c2

= ( )

= ( )

1212

r a

r a

Reduction in velocity

When the water current passes one net panel thecurrent velocity will be reduced by friction from thenet; much of the water is then forced to go underand around the following nets. When the watercurrent hits the next panel, the force on this panelwill be reduced because the water velocity isreduced. To be able to calculate the forces on thefollowing net panels it is necessary to know the sizeof this reduction, particularly since the velocity issquared in Morrison’s force equation so gives alarge contribution.

The following equation may be used to estimatethe current velocity Ui after i net panels:33

Ui = Ucri

Sea Cages 207

Figure 15.23 The cage is not rigid but will be deflectedat an increasing angle by the current.


where:

i = number of net panels that the water has topass

Uc = current velocity of the water when it hits thefirst panel

r = reduction factor (depending on the solidity,Sn, for the net).

The velocity reduction factor may also be expressedby the drag coefficient CD, which again is affectedby the solidity, Sn:

r = 1 − 0.46CD

If Sn is between 0.1 and 0.35 CD may be calculatedas follows:

CD = 0.04 + (−0.04 + Sn − 1.24Sn2 + 13.7Sn

3)cosa

where:

a = angle of deflection of the net.

If using rigid cage bags the equation will be asfollows:

CD = Sn − 1.24Sn2 + 13.7Sn

3

As seen there is a dramatic reduction in velocity forevery net panel the water has to pass.

In addition, to reduce the forces on the net bagsthat lie behind the first one will result in a reduc-tion in supply of new oxygen-rich water. Whenhaving several cages, one after another in the direc-tion of the current, lack of oxygen might occur inthe last cages that the water passes. As the waterexchange is reduced there will also be a reductionin removal of metabolic waste substances, whichalso shows the importance of correct individualplacing of the cages in relation to the current direction.

By increasing amount of fouling the currentvelocity will be reduced even more, as a result ofincreased solidity, so it is important to minimizefouling on the nets.

ExampleThree cages are lying behind each other. The veloc-ity of the current that hits the first net panel is 0.7m/s;CD is 0.32. Calculate the current velocity in each ofthe three cages assuming rigid nets.

First the velocity reduction factor (r) is calculatedfrom CD

r = 1 − 0.46(0.32)= 1 − 0.15= 0.85

Then the velocity inside cage 1 (after 1 net panel),cage 2 (after 3 net panels) and cage 3 (after 5 netpanels)is calculated:

U1 = 0.7 × 0.851

= 0.6m/s

U2 = 0.7 × 0.853

= 0.43m/s

U3 = 0.7 × 0.855

= 0.31m/s

Simple method for calculating the current forceswith rigid nets

As shown, there is a lot to consider when calculat-ing the current forces on sea cages, and usually spe-cially designed computer programs are used for thispurpose. For rigid nets, however, it is possible to cal-culate the forces employing quite simple methods.Use of rigid nets will, however, result in overesti-mation of the forces compared to the real situation,but will show the principles. Using the valuesobtained to calculate the size of mooring lines and anchors will also ensure that the mooring islarge enough, although the mooring system will bemore expensive than necessary because it is over-specified.

A simple set of equations and diagrams havebeen developed for calculation of the current forceson rigid sea cages.33 The method assumes rigidmesh/bags with no deflection, so therefore no lifting forces. In addition, only forces normal andparallel to the current on the farm are taken intoconsideration.

First the drag on the net panel normal to thecurrent direction, FDN, must be calculated; thenforces parallel to the current direction FDP are cal-culated; these are then added to give the total forces FDtot:

F C U BD mrr

F C U B D lmrr

r

n

n

DN DN c2

DP DP c2

= ( ) −−

⎛⎝⎜

⎞⎠⎟

= +( ) −−

⎛⎝⎜

⎞⎠⎟

12

11

12

211

4

2

4

42

r

r

where:

FDtot = FDN + FDP

r = density of the liquidCDN = drag normal to the current direction

= Sn − 1.24Sn2 + 13.7Sn

3

CDP = drag parallel to the current direction (set to0.04 based on the shape of the bag)

Uc = current velocityB = width of the cage bagD = depth of the cage bagl = length of the cage bagn = number of cages parallel the current directionm = number of cages normal to the current

directionr = reduction factor

= 1 − 0.46CDN.

The above equation is used to find the force whenthe farm is parallel to the current direction. To useit when the farm is normal to the current directionrequires only that the values of n and m in the equa-tions be exchanged.

ExampleA rectangular system farm includes two cage bags,one after another and eight side by side in relationto main current direction, a total of 16 cages.The sizeof the single cage is: width 10m, length 15m, depth12m. The design water flow is 0.8m/s. Calculate thecurrent forces that affect the farm, presuming rigidnets.

The drag coefficient CDN must be calculated first:

CDN = Sn − 1.24Sn2 + 13.7Sn

3

= 0.3 − 1.24 × 0.32 + 13.7 × 0.33

= 0.56

Calculate r

r = 1 − 0.46CDN

= 0.74

Then the forces on the net panels normal (FDN)on parallel (FDP) with the current direction can becalculated.

S solidity

bar diameterbar length

n =

= ×2

Drag normal to flow direction:

Drag parallel to flow direction:

Total drag:

FDtot = FDN + FDP = 354724 + 38103 = 392827N= 392.8kN

Another very simple method is to use Morrison’sequation on the total length of the rope used tocreate the net panel. By calculating the currentresistance from the total length of the rope the totalforces affecting the bag can be found.18

where:

N = number of meshes in the length of the panelIm = bar lengthf = coefficient of decrease (normally around 0.7)L = length of the panelT = number of meshes in the depth of the panelD = depth of the panel.

Next the length of the filament (rope) per side (LT)is calculated:

LT = 2NTImkn

where:

kn = knot factor.

TD

I f=

m

NL

I f=

m

F C U B 2D lmrr

rDP DP c2

4n

42

2

4 2

42

12

11

12

1025 0.04 0.8 10 2 12 15 8

1 0.741 0.74

0.74 38103N

= +( ) −−

⎛⎝⎜

⎞⎠⎟

= × × × × + ×( ) × ×

× −−

⎛⎝⎜

⎞⎠⎟ × =

×

r

F C U BD mrr

DN DT c2

n

2

4 2

2

12

1025 0.56 0.8 10 12 8

1 0.741 0.74

354724 N

= ( ) −−

⎛⎝⎜

⎞⎠⎟

= × × × × × ×

× −−

⎛⎝⎜

⎞⎠⎟ =

×

12

11

4

2r

Sea Cages 209


Therefore the total drag force on the net panel canbe calculated by Morrisons equation:

where:

d = rope diameter (bar diameter)CD = drag coefficient for a cylinder is used since this

is the same as that of the rope, and this is 1.2.

15.6.3 Calculation of wave forces

Waves are important when designing sea cages.Thewave forces will influence design of both the cagecollars and the net bag. If the wave forces are toohigh the collar may break. The wave forces will,however, also affect the mooring systems and mustbe taken into consideration when calculating thesize of the mooring system, even if they are smallerthan the current forces.

Calculation of wave forces normally involves theuse of computer programs and numeric solutions.Methods for calculating the wave forces include, forinstance, Morrison’s equation, diffraction theoryand Froude–Krylov forces.7,12 Calculations for tra-ditional fixed offshore constructions will overesti-mate the forces for a pre-stressed cage farm floatingon the surface and following the sea.

Compared to currents waves apply dynamicforces to the construction. Use of Morrison’s equa-tion can illustrate this. Morrison’s equation has twoadditional terms, one for the velocity of the waterparticle in the wave and one for the acceleration ofthe water particle:27

where:

Fi = forces on the object in the x, y or z directionr = density of the liquidA = area of the objectCM = mass coefficientCD = drag coefficientai = water particle acceleration in direction iui = water particle velocity in direction iV = volume of the object.

Without going deeper into this equation, it showsthat in addition to the drag term, as there was for

F C Va C u u Ai M i D i i= +( ) +112

r r

F C v L dD D T= 12

2r

current forces, there is a mass force term that is pro-portional to the acceleration. The total force is thesum of these two forces.

Wave forces are dynamic, which means that theycome again and again. This will result in reducedloads being tolerated by structures before breakageoccurs. Experiments have shown that a cage mayonly tolerate a dynamic load which is 10% of onesingle static load.5 Waves impose additional forceson the mooring system. On normal, partly pro-tected off-shore sites, the wave forces will be muchlower than the current forces. Usually waves add upto 20–30% to the current forces, but this is of coursesite-dependent.

Due to the complexity of calculating wave forcesand measuring water particle velocity and accelera-tion, this will not be further described here. Spe-cialized literature is recommended, as mentionedearlier.

15.6.4 Calculation of wind forces

Since such a small part of a cage farm is above thewater surface where the wind is blowing, the forcestransferred to the cage and further to the mooringsystem will be very low compared to the forcesfrom the current. To calculate the forces on thejumping net the following equation may be used:32

where:

r = density of airU = wind velocitym = drag coefficientA = effective area of the jumping netn = protection factor (for instance, distance from

other jumping nets).

The wind forces will be considerably higher if, forinstance, a feed barge is moored together with thecages.

15.7 Calculation of the size of themooring system

15.7.1 Mooring analysis

A calculation of the size of a mooring system mustinclude a performance analysis under extreme con-

F U Anw = 12

2r m

ditions,36 for instance for an intact mooring system,with a break in one of the mooring lines and withan increase in the water level compared to normaldue to a storm tide of 1m for example (require-ments in the Norwegian standard for cages).14 Theanalysis must show that the mooring system willwithstand such situations without breakdown.

What is then happening when the environmentalforces affect a pre-stressed moored sea cage farm?A total environmental force (F) will try to move thecages out of their original position. In all mooringlines on the side where the forces F is acting therewill be additional tension, and a correspondingreduction in the lines on the opposite side. Depend-ing of the degree of pre-stress and the elasticity ofthe lines, there will be a drift away from the equi-librium position caused by the acting forces, but thefarm will not drift freely. The mooring lines willgradually create an opposing force towards F thatprevents the cage farm drifting freely. A new equi-librium position will be established where forcesfrom the mooring lines are opposite and equal tothe forces created by the environmental factors.The tension in the mooring lines on the side wherethe environmental forces are acting is now muchhigher than when the cages were in their originalpositions; there might also be slack in lines on theopposite side.

If, in addition, there are waves additionaldynamic forces will be imposed on the farm and itwill oscillate around the equilibrium position aslong as the mooring system does not break. Themaximum load in the mooring lines will thereforebe higher than from the static current force.

If there is a break in one of the mooring lines dueto unforeseen circumstances the farm will drift intoa new equilibrium position. This movement will bedetermined by the tension in the broken line, theweight of the farm and the resistance against move-ment from the net bag and cage collars. When thefarm is drifting and needs to be stopped, it is impor-tant that the tension in the remaining lines does notexceed their breaking strength. If this happensother lines will break and most probably this willresult in progressive breaking of all the remainingmooring lines so the cages will drift freely. If onemooring line to a cage breaks, there is also the pos-sibility that the cage will crash into another cage or fixed construction such as the walkway whilstdrifting towards its new equilibrium position. This

may cause a material break in the cage collar, forinstance. When doing mooring calculations a breakin one of the mooring lines will normally be toler-ated, but progressive breaking must not occur insuch situations.

15.7.2 Calculation of sizes for mooring lines

There will always be some inaccuracy whendescribing and calculating the environmental loadsaffecting cage farms. For example, this can be thatthe current velocity or wave height varies and mightbe slightly higher than expected. A load factor isrecommended to compensate for the possible inac-curacy. Normally the load factors are between 1 and1.5 depending on the uncertainty in the calculationsof environmental loads. A load factor of 1.5 meansthat the environmental loads can be up to 50%more than those calculated and the mooring willstill hold. The total force that is used in further cal-culations is then:

Ft = giFE

where:

Ft = total forceFE = calculated environmental forcesgi = load factor.

In the new Norwegian standard for mooring analy-sis, the standard load factor is set at 1.15 forunmanned farms and 1.3 for continuously mannedfarms when doing static analysis i.e. the safetyfactor is larger.14

The breaking strength of the different types ofmooring line is also given with some accuracy bythe suppliers and is based on a number of mea-surements.To take care of possible inaccuracy whentesting the material, and minor variations in thematerials, it is recommended that a material factor(gm) be used. However, will this vary with the mate-rial and the degree of testing of the material thathas been performed. Normally it lies between 1.1and 5. To find size of the mooring lines necessarythis must be taken into consideration, which givesthe following equation:

FR = FTgm

where:

FR = mooring forces (the force that the mooringlines shall tolerate)

Sea Cages 211


FT = calculated total forces including load factorgm = material factor.

The following material factors are used in the Norwegian standard for cage farms:14 chain, 1.5;synthetic rope with knot, 5.0; synthetic rope, 3.0;synthetic rope specially resistant to ageing, waveand water absorption 1.5.

ExampleTo show how a mooring analysis can be performedan example is shown where the mooring line, buoyand anchor are to be dimensioned (Fig. 15.24). Ona cage farm the calculated environmental force thatan anchoring line must take up is 1 t or approxi-mately 10kN. The length and type of mooring line,buoy type and size, and anchor type and size shouldbe found to keep the farm in position. The depth ofthe site is 30m and the bottom is sand. Current veloc-ity is set to 0.2m/s.

First the design of the mooring system must befound. Choose to have the buoys 15m away from thecage collar. The mooring line is set to 3× the depth,and becomes 90m. The total length of the mooringline is therefore 90 + 15 = 105m. The forces on thebuoys can then be found. First the angle a that themooring line has from the bottom and to the surfacemust be found:

a = 19.47°

sina = 3090

Then calculate the force in the direction x on themooring line:

The mooring line must therefore tolerate a force of10.61kN.

Calculate the force in the y direction:

y = sinax= 3.54kN

The buoy will be dragged down with a force of 3.54kN.

Now the buoy can be described. The requirementfor buoyancy is set to twice the force F in themooring line which is 7.08kN. Archimedes law isused to calculate the buoyancy; the density of sea-water, rw = 1025kg/m3.

F = rwgV

VF

g=

=×

=

rw

70801025 9.810.70

sina = yx

x

x

=

=

10

10 61

cos

.

aKN

cosa = 10x

Figure 15.24 Environmental forceswill affect the forces in the mooringsystem.

where:

F = buoyancyrw = density of the displaced liquidg = acceleration due to gravityV = displaced volume.

This means that the buoy needs a volume of 700 l ormore to stay in the correct position on the surface.In addition the buoy must have buoyancy that coversits own weight; this depends on buoy type and isgiven by the supplier.

The next step is to calculate the size of the anchor.A block anchor is chosen. First the size to withstandthe vertical lifting force (y) is calculated:

y = 3.54kN

Choose a concrete block anchor with density fc of2500kg/m3.The weight of the anchor (G) is given by:

G = mg= rcVg

where:

m = mass of anchorV = volume of anchorg = acceleration due to gravityrc = density of cancrete.

The buoyancy (F0) that will lift the anchor is, fromArchimedes law:

F0 = rwVg

where:

rw = density of seawater.

The following equation may be used to find the nec-essary volume of the block anchor (y):

y = G − Fo

= rcg − rWVg= Vg(rc − rW)

The horizontal force is calculated using a frictioncoefficient of 0.5 for the sand bottom. The horizon-

Vg

=−( )

=−( )

==

Y

35409.81 2500 1025

0.245m

245l

C W

3

r r

tal force (F) that will try to move the anchor is 10kN, while the weight G will keep the anchor inplace. In addition the buoyancy F0 of the anchor willhave effect, because it will reduce the weight com-pared to when it is on shore. The following equationcan be set up:

F = f(G − F0)= fVg(rc − rw)

where:

f = friction coefficient for the block anchor.

Therefore the volume of the block anchor mustexceed 1.382m3 or the mass be above 3.46 t. In prac-tice two or three anchors will be used.

15.8 Control of mooring systemsAfter setting out the mooring system, it is impor-tant to do necessary checks to avoid breakages.Insurance companies will normally require sometype of checking of the mooring system. There mayalso be national standards to prevent breakage andpossible escape of fish. For instance the followingmay be used:

• The supplier’s specification must be used to setout the cage

• Parts in the mooring system above the surfacemust be checked daily

• The whole mooring system including underwater installations must be checked every year.

Some time after the mooring system was first setout, a more comprehensive and systematic checkshould be carried out, for instance every 4 years.This includes load tests on important and heavilyloaded parts. Components exposed to hard wearshould be replaced, such as mooring ropes, chain,wire, fixing points and eventually links.

References1. Beveridge, M. (1996) Cage aquaculture. Fishing News

Books, Blackwell Scientific.

VF

fg=

−( )

=× −( )

=

r rC W

3

100000.5 9.81 2500 1025

1.382m

Sea Cages 213


2. Liao, I.C., Lin, C.K. (eds) (2000) Cage aquaculture inAsia. In: Proceedings of the first international sympo-sium of cage aquaculture in Asia. Asian FisheriesSociety and World Aquaculture Society – SoutheastAsian Chapter.

3. Huguenin, J.E. (1997) The design, operations andeconomics of cage culture systems. AquaculturalEngineering, 16: 167–203.

4. Loverich. G.F., Gace, L. (1997) The effect of currentsand waves on several classes of offshore sea cages.Paper given at International conference on openocean aquaculture 97. Maui, Hawaii, USA.

5. Cairns, J., Linfoot, B.T. (1990) Some considerations inthe structural engineering of sea-cages for aquacul-ture. In: Engineering for offshore fish farming,pp. 63–77. Thomas Telford.

6. Pérez, O.M.,Telfer,T.C., Ross, L.G. (2003) On the cal-culation of wave climate for offshore cage culture siteselection: a case study in Tenerife (Canary Islands).Aquacultural Engineering, 29: 1–21.

7. Sarpkaya, T., Isaacson, M. (1981) Mechanics of wave forces on offshore structures. Van NostrandReinhold.

8. Sawaragi, T. (1995) Coastal engineering – waves,beaches, wave–structure interactions. Elsevier.

9. Sorensen, R.M. (1993) Basic wave mechanics.Wiley-Interscience.

10. Boccotti, P. (2000) Wave mechanics for ocean engi-neering. Elsevier.

11. US Army Corps of Engineers (1984) Shore protectionmanual, vols I and II. US Government PrintingOffice.

12. Faltinsen, O.M. (1990) Sea load on ships and offshorestructures. Cambridge University Press.

13. Saville, T. (1954) The effect of fetch width on wavegeneration. Technical memo. 17. Beach ErosionBoard, US Army Corps of Engineers.

14. NS 9415. Marine fish farms. Requirements for design,dimensioning, production, installation and operation.Norwegian Standardization Association.

15. Linfoot, B.T., Cairns, J., Poxton, M.G. (1990) Hydro-dynamic and biological factors in the design of sea-cages for fish culture. In: Engineering for offshore fishfarming, pp. 197–210. Thomas Telford.

16. Klust, G. (1982) Netting materials for fishing gear.Fishing News Books, Blackwell Science.

17. Klust, G. (1983) Fibre ropes for fishing gear. FishingNews Books, Blackwell Science.

18. Karlsen, L. (1989) Redskapsteknologi i fiske. Univer-sitetsforaget (in Norwegian).

19. Tygut (1997) Regelverk og veileder for dimensjoner-ing og konstruksjon av flytende oppdrettsanlegg.Fiskeridepartmenetet (in Norwegian).

20. Rudi, H., Oltedal, G. (1993) Metode for vurdering avlokalitet for matfiskoppdrett. Rapport marintek (inNorwegian).

21. Kerr, N.M., Gillespie, M.J., Hull, S.T., Kingwell, S.(1980) The design construction and location ofmarine floating cages. In: Proceedings of the instituteof fisheries management cage fish rearing symposium,

University of Reading, 26–27 March 1980, pp. 23–49.Janssen Services.

22. Gunnarson, J. (1993) Bridgestone Hi-Seas fish cage: design and documentation. In: Fish farmingtechnology. Proceedings of the first international conference of fish farming technology (edsH. Reinertsen, L.A. Dahle, L. Jørgensen, K.Tvinnereim). A.A. Balkema.

23. Loverich, G., Forster, J. (2000) Advances in offshorecage design using spar buoys. Marine TechnologySociety Journal, 34: 18–28.

24. Lien, E. (2000) Offshore cage systems. In: Cage Aqua-culture in Asia. Proceedings of the first internationalsymposium of cage aquaculture in Asia (eds I.C. Liao,C.K. Lin). Asian Fisheries Society and World Aqua-culture Society – Southeast Asian Chapter.

25. Lien, E. (1993) Tension leg cage, a new net pen cagefor fish farming. In: Fish farming technology (eds H.Reinertsen, L.A. Dahle, L. Jørgensen, K. Tvinnereim),pp. 251–258. A.A. Balkema.

26. Det Norsk Veritas (1988) Det Norske Veritas tentativeregler for sertifisering av flytende fiskeoppdrettsanlegg.Rapport Det Norske Veritas (in Norwegian).

27. Rudi, H., Lien, E., Slaatelid, O.H. (1994) Håndbok fordesign og dokumentasjon av åpne merdanlegg.Rapport Marintek (in Norwegian).

28. Tsukrov, I., Eroshkin, O., Fredriksson, D., RobinsonSwift, M., Celikkol, B. (2003) Finite element model-ing of net panels using a consistent net element.Ocean Engineering, 30: 251–270.

29. Fredriksson, D.F., De Cew, J., Robinson Swift, M.,Tsukrov, I., Chambers, M.D., Celikkol, B. (2004) Thedesign and analysis of a four-cage grid mooring foropen ocean aquaculture. Aquacultural Engineering,33: 77–94.

30. Suhey, J.D., Kim, N.H., Niezrecki, C. (2005) Numeri-cal modeling and design of inflatable structures –application to open-ocean-aquaculture cages. Aqua-cultural Engineering, 33: 285–303.

31. Milne, P.H. (1972) Fish and shellfish farming in coastal waters. Fishing News Books, Blackwell Scientific.

32. Rudi, H., Aarsnes, J.V., Dahle, L.A. (1988) Environ-mental forces on floating cage systems: mooring con-siderations. In: Aquaculture engineering, technologiesfor the future. ICemE symposium series no. 111.Hemisphere.

33. Løland, G. (1993) Current forces on, and water flowthrough and around, floating fish farms. AquacultureInternational, 6: 33–37.

34. Løland, G. (1993) Water flow through and around netpens. In: Fish farming technology (eds H. Reinertsen,L.A. Dahle, L. Jørgensen, K.Tvinnereim), pp. 177–183.A.A. Balkema.

35. Aarnes, J.V., Rudi, H., Loland, G. (1990) Currentforces on cages, net deflection. In: Engineering for off-shore fish farming. Thomas Telford.

36. Lien, E., Rudi, H., Slaatelid, O.H. (1996) Håndbok fordesign og dokumentasjon av åpne merdanlegg.Rapport Marintek (in Norwegian).

16Feeding Systems

small feed particles, for example for marine orfreshwater fry, might be a problem. If the feed islike meal, it might be difficult to get it through thefeed dispenser; it might clog inside the hopper, andthe sliding angle is very high.

16.1.3 Selection of feeding system

Today is it normal to use automatic feeding systemsin all types of intensive fish farming. Which type ofautomatic feeding system to choose, however,depends on a number of factors of which the mostimportant are: feed type, production species,production type, production size and access to electricity.

A feeding system could range all the way fromsimple dispenser with no need for electricity to anadvanced computerised feeding system which con-trols the feeding on the basis of the appetite of thefish.

16.1.4 Feeding system requirements

The requirements for the feeding system depend onthe chosen type. In the following, some generalclaims especially adapted to dry feed are presented:

• Simple operation• Low maintenance• Tolerate wind and sea (offshore farms)• Tolerate high humidity• Simple to fill with feed• Simple calibration (to control the amount dis-

pensed)• High dispersion accuracy• Cause few breakages

16.1 Introduction

16.1.1 Why use automatic feeding systems?

Feeding can be done by hand, or by automaticfeeders or feeding systems. The time used forfeeding can be considerable for large farms withintensive production, and can justify the investmentin a system for automatic feeding. For instance, thedaily requirement of feed for a rainbow trout farmwith a standing biomass of 100t of 100g fish is atleast 3500kg per day with a water temperature of16°C.

For intensive fry production, several speciesrequire an almost continuous supply of food, espe-cially in the first feeding stage. This requires atremendous amount of work, and is therefore nor-mally done by automatic feeders. Feeding systemsare of most interest for intensive aquaculturesystems because of the importance of getting asmuch feed as possible into the fish.

16.1.2 What can be automated?

How easy it is to automate the feeding depends onthe feed type used.1–3 Dry, extruded or pelleted feedis quite easy to deal with: the particles are fixed andhard. Wet feed or moist feed is rather more difficultto feed automatically. To find good systems for dis-tribution of dense particles is also difficult.Wet feedmay be fed through pump systems, but here it is dif-ficult to obviate the possibilities for over feeding;possible environmental impacts are also muchhigher with this type of feed.

The size and shape of the dry particles will alsoinfluence the feasibility of feeding automatically:

215


Of the more general engineering subjects that areof interest for increasing the basic knowledge ofautomatic feeding there are, for instance, solidshandling, solid conveying and bulk solids handling(for example, refs 4–6).

16.2 Types of feeding equipmentFeeding equipment can be divided into groupsbased on its construction and function. One classi-fication is as follows:

• Feed blowers• Feed dispenser• Demand feeders• Automatic feeders, feed machines• Feeding systems.

This is specifically for the dry feed that is most com-monly used in intensive fish farming. For wet andmoist feed other separations can be made, but onlya few methods are used for this type of feed.Feeding equipment for wet and moist feed will notbe dealt with here.

16.2.1 Feed blowers

A feed blower is only a tool to simplify hand feeding(Fig. 16.1). There are different blower types basedon the ‘carrier’ used for the feed particles which isnormally either air or water. The feed can either besucked up from a tank or a bag by vacuum, or the

feed can be filled into a hopper standing over a pipewith flow of air or water.The hopper can be fixed ona boat or be movable.

16.2.2 Feed dispensers

A feed dispenser is often confused with a feedingmachine, but does not have the distribution unit.Actually it is therefore something between a feedmachine and hand feeding. A weighed portion offeed is placed on the dispenser and the dispenserwill empty it during a fixed period, normally fromone to three days. It either goes continuously orstepwise controlled by a control unit. To get thewanted feed ration, the actual amount of feed mustbe put in the dispenser. This is normally weighedout.

A great advantage with the feed dispenser is itssimple and robust construction. It is also easy tomonitor visually whenever it is functioning and theamount of feed that has been dispensed. The con-struction is favourable to use in research operationsbecause if you weigh out the feed exactly you canbe sure that the dispenser will supply the exactamount to the fish. The great disadvantage, com-pared to a feeding machine, is that it takes quite along time to measure and/or weigh the feed thatshould be placed in the dispenser.

Several designs of feed dispenser are used (Fig. 16.2). In a disc feeder a scraper rotates on ahorizontally fixed circular plate, and the feed falls

Figure 16.1 A feed blower beingused to simplify hand feeding.

off the edge of the plate and into the fish tank. Adisc dispenser needs electricity to run the motor,normally 24V a.c. The feeder normally goes step-wise, controled by a unit that regulates the start andstop intervals. Another much used construction is arubber conveyor belt that is dragged along onrollers. When starting, the belt is dragged back-wards so it creates a surface where feed is supplied.The end of the belt is fixed to rollers; when theserotate the belt will be dragged up and the surfacewhere the feed is lying will gradually be decreasedso that the feed falls off and into the fish tank.This type is either powered by electricity or byclockwork. An advantage with this type of feed dispenser is the possibility of running it withoutelectricity. Feed is either dispensed continuously orstepwise.

16.2.3 Demand feeders

A demand feeder is normally a mechanical con-struction. A stick is attached to a slightly bowedplate sitting under a feed hopper (Fig. 16.3). Thestick goes from the feeder down into the water.When the fish touch the stick, feed will be dispensedfrom the hopper. At the end of the stick is a knob,or something similar, which the fish touch. A great

advantage with using demand feeders is that thereis no need for an electricity supply. Furthermore,the design is simple with few moveable parts.

The fish operate the demand feeder themselves,and can therefore theoretically be fed according toappetite (ad libitum). However, some feed loss hasbeen registered. The fish may use the demand stick

Feeding Systems 217

A B

Figure 16.2 Typical feed dispensers: (A) disc feeder and (B) conveyor belt feeder.

Figure 16.3 The fish operate a demand feeder whenthey move the stick hanging down in the water.


as a toy and feed may be lost; demand feeders arealso sensitive to movements in the water, such aswaves; wind may also affect the demand feeder soshielding may be included.

Demand feeders are used for almost all species,even if some species, such as Atlantic salmon, areslow to learn the system. The fish need a trainingperiod to learn how to operate the system.1 Com-pared to hand feeding, both improved and less goodgrowth results have been shown.7

In electronic demand feeders feeding is triggeredby electric signals. The mechanical stick is replacedby an electric cable with a pressure sensor at theend.When the fish touch this sensor a signal is givento the feeder which starts. This system allows extracontrol over the demand feeding, for instance bysetting fixed interval for the operation of the feederor by setting a maximum limit of distributed feedper portion or per day. A more advanced controlsystem is, however, required in this type of feeder.

Much literature is available on the use of demandfeeders, including the possibility for controlling theappetite of the fish experimentally.8–12

16.2.4 Automatic feeders

A feeding machine or an automatic feeder consistsof four major components (Fig. 16.4): a feed con-

tainer (hopper), a mechanism for feed distribution,an electrical power supply for the distributionmechanism and a control unit for starting and stopp-ing the distribution mechanism. The feed distribu-tion mechanism is the main component in anautomatic feeder and distinguishes it from a feeddispenser. The feeders are fixed in a rack on thetank or on the cage, but may also be included in abuoy (see, for example, ref. 13). When using anautomatic feeder, the amount of feed that has to bedistributed over a period of time is known and thedistribution unit runs for the period that satisfiesthis requirement.

Feed is distributed by volume. In speciallydesigned and more expensive feeders, the systemsmay also use feed mass.When using volume for dis-tribution, the volume: mass ratio (litre/kg), i.e. thedensity, of the feed must be known. The density ofthe feed varies with formulation, from producer toproducer, and also depends on the size of the feedparticles. Because volume distribution feeders onlydistribute a certain volume of feed, and the in massfeed is of interest, calibration of the feeders is nec-essary. To calibrate the feeder, it is run for a knowntime period; then the exact amount of feed that hasbeen dispensed is weighed so that the feed distrib-uted per unit time can be calculated. This informa-tion is then used to find the necessary time that thefeeder has to run to distribute a certain mass offeed.

ExampleA fish tank requires 3kg feed per day. For how longmust the feeder be run to deliver this amount?First the amount of feed delivered from the feederper unit time must be found. The feeder is run con-tinuously for 1min and the amount of feed deliveredis weighed and found to be 1kg. The feeder is there-fore delivering 1kg/min; to deliver 3kg to the fishtank, the feeder must be run for a total of 3min perday.

If the feeder starts every 30min throughout the dayand night, it starts 48 times in total. Each time it musttherefore run for: 3min (= 180s)/48 = 3.75s.

Distribution mechanisms

Many mechanisms for feed distribution are avail-able.3,14,15 Some important types are describedbelow.

Figure 16.4 An automatic feeder consisting of a feedcontainer (hopper), a mechanism for feed distribution,an electrical power supply and a control unit.

Screw: A screw allows a specific batch of feed to bedispensed for every rotation (Fig. 16.5). The screwis installed under a hopper from which it is filled.The amount dispensed per unit time is related toscrew diameter, design of the screw (rise of thescrew thread), the speed of rotation, the degree offilling and the angle of the screw.

Vibrator: A plate that vibrates may be used to dis-tribute a volume of feed. When the almost hori-zontally fixed plate starts to vibrate the feed on theplate will fall over the edge. One method of gettingvibration is to attach a weight to one side of a ver-tically fixed shaft that rotates. When the shaftrotates there will be imbalance in the shaft and inthe plate fixed to it. Another method is to use anelectromagnet with an anchor fixed by a leaf or coilspring.When the electricity is turned on, the anchoris dragged towards the magnet by a varying mag-netic field, making the whole vibrator shake; a slopeon the feeding plate causes the feed to be shaken

over the edge. Advantages of the electromagneticvibrator are its simple construction and that it stopsimmediately the electricity is turned off. Theamount of feed distributed is controlled by adjust-ing the voltage and hence the amplitude of thevibrator.

Cell wheel: A vertically installed rotating wheelwith wings, cells or chambers sitting under a feedhopper may also be used to distribute feed (Fig.16.5). When the wheel rotates it transports the feedin the cells; when the cells approach the lowest posi-tion, the feed is released. The amount of feedreleased depends on the number of chambers in thewheel and the speed of rotation. The leaf dispenserand drawer dispenser are similar in principle to thecell wheel.

Others: A number of other mechanisms might beused for feed distribution, of which a rotating discwith a scraper at the bottom of a cylindrical hopper

Feeding Systems 219

A B

C D

Figure 16.5 Different types of feed distribution mechanisms: (A) screw feeder;(B) vibrating feeder; (C) cell wheel feeder;(D) rotating belt feeder.


is one. When the disc rotates feed will be distrib-uted with the help of the scraper fixed to thehopper. This system must not be confused with thedisc dispenser which has no hopper. The distancebetween the rotating disc and the feed hopper regu-lates the amount of feed that is distributed. Use ofa conveyor belt system is another method for feeddistribution; the rotating belt is placed under thefeed container and a distribution bar regulates thethickness of the feed layer on the belt and by thisthe amount of feed distributed.

For all methods the amount of distributed feedis, to various extents, dependent on the height of thefeed in the hopper. When the hopper is full thepressure of feed is increased and more feed is dis-tributed because the distribution mechanism issited at the bottom of the hopper.

Another fairly new method used for distributingfeed is a bowed screw with an open centre – actu-ally a spring. The advantage with this arrangementis that the amount of feed dispensed does notdepend so much on the level in the hopper becausethe screw is filled with feed from the side.

If the feed particles are very small (as in meal), thefeed particles may clog around the inlet to the distri-bution unit.This is a particular problem for particlesthat have a high sliding angle. In fish farming today,the particles are larger and the sliding angle is quitelow, so this is normally not a problem.

To achieve more exact dispensing and to avoidthe chore of calibration, automatic weight control

can be used as a supplement. However, this is quiteexpensive and has only been used in fish farming toa limited extent. This can, however, be a solution ifrequirements for dispensing are very precise. Elec-tronic weight cells (tension and pressure) havebeen used, especially in feeding systems. The prin-ciple of weight cells is that they measure the tem-porary deformation of the material, which is relatedto the weight of feed/in them. When this system isused, a volume dispenser adds the feed, but is con-trolled by the weight cell, signals from which regu-late the running time for the volume dispenser.

Feed hopper

Above the feeding mechanism is sited the feed con-tainer or hopper. Hopper size varies from somelitres to several hundred litres, depending on thesize of the fish to be fed. Hoppers are usually con-structed of plastic or metal (aluminium) and mustbe designed in a way that gives easy access for refill-ing and ensures that all feed slides out easily.

Spreading of feed

On some feeders a unit for spreading of the feed isattached underneath the distribution mechanism.The purpose is to distribute the feed over a largerpart of the pond, tank or cage. Three spreading pat-terns are used: point feeding (no spreading mecha-nism), sector feeding and circle feeding16 (Fig. 16.6).

Figure 16.6 The feed can be spreadin sectors of the cage, or in a circularpattern.

The sector feeding system normally consists of avertical rotating plate or brush. When this rotatesthe feed is spread out in a sector in the cage or thepond. A circle feeding technique employs a cen-trifugal scattering system. The feed particles dropdown from the distribution unit and hit a horizon-tally placed rotating disc. Because of the centri-fugal forces the feed is spread out in a circle. Thespreading unit may be integral part of the distribu-tion system or it may be installed separately under-neath the distribution unit. The velocity of the discor plate determines the area of the sector or thecircle where the feed is spread. However, a highvelocity will increase the amount of breakagebecause the forces transferred from the spreadingunit to the feed particles will increase.

Control units

The control unit manages the current to the motoron the dispensing system; this also controls thefeeding. The simplest control unit sets the timeinterval between each meal and the running time,i.e. the length of each meal. Some control units areequipped with a photocell which only permitsfeeding during daylight. In more advanced units adaily increase in the running time may be addedrelated to the expected growth rate of the fish inthe production unit.

There may be an individual control unit for eachfeeder or the unit may control several feeders with

the same feeding regime. There can also be onecontrol unit with several channels, which meansthat it can control several feeders individually. Thecontrol unit can be a simple interval relay, whererunning time and time interval between each startare set (Fig. 16.7). Several relays may be settogether in a multichannel control unit. The pro-grammable logic controller (PLC) is a moreadvanced system carrying out the same tasks as amultichannel control unit.The input and output canbe switched on or off, and it is quite easy to extendthe system with more inputs and outputs. Eachoutput channel can be programmed individually. Inaddition input signals can also be used to controlthe output signal. For instance, the output can onlybe started when the input signal from a light sensorregisters that it is daylight; this means that feedingis only permitted during daylight.

A personal computer (PC) equipped with special‘cards’ may also be used to control larger feedingsystems. The PC can also collect data that can beused to control the feeding, such as water tempera-ture and light intensity. PCs may also be used as adata logger to store, for instance, how much is fedevery day.

Electric current

Both distribution and spreading mechanismscontain motors that normally need an externalsupply of power. This is normally electricity, either

Feeding Systems 221

Figure 16.7 A feeder control unitwith a transformer and relay.


from a central electric power station (alternatingcurrent) or from batteries (direct current). A clock-work mechanism may also be used on small feeders.Direct current motors normally require a supply of12V or 24V.Alternating current motors for feedersare either low voltage (12V or 24V) or normalvoltage (110V and 220V). The advantage with highvoltage is that thinner cabling is required andcurrent loss in the cables is reduced, particularlywhen using long cables. The disadvantage is thatnormal voltage, high humidity and free water sur-faces can be a dangerous combination. When usingnormal voltage it is therefore important to becareful when laying electric cables and ensure thatthe feeding system is correctly insulated andearthed to avoid jump sparks. Qualified profes-sionals must perform such work.

Normally the mains voltage is used as the elec-tricity source. If low voltage feeding systems are tobe connected, the use of a transformer is necessary.If direct current is to be used, a rectifier is alsoneeded.

If there is no electricity supply, for instance on asea cage, either batteries or a diesel-powered elec-trical generator must be used. Use of batteriesrequires direct current motors on the feeder. Inthese cases the battery must be taken out regularlyfor recharging; solar panels or windmills may alsobe used to charge batteries.

Direct current motors used on feeders have theadvantage that the speed of rotation of the motorcan easily be regulated. By adding a variable resis-tance, the size of the incoming current that isrunning the motor will be regulated. Regulation ofthe speed of rotation will control the amount offeed coming from the feeders.

The motor output should be matched to the needfor forces to run the distribution unit. Motors thatare too powerful can result in more breakage offeed and are not recommended. The main reasonfor adding a larger motor is to avoid wedging sothat the system is more reliable. Care should,however, be taken regarding the possibility ofbreakage when feeders are equipped with largemotors.

16.2.5 Feeding systems

The term feeding system refers to a completesystem that takes the feed directly from the feed

silo or hopper, transports it to the fish productionunit, and at the end distributes it to the fish. A com-plete feeding system may comprise three parts: astorage unit, a transport unit and a feed distributionunit. Today feeding systems can be divided into twotypes:

(1) Systems where the feeder is centrally placedand feed is transported to the single fish pro-duction units (tanks, ponds or cages) throughpipes, normally known as feeding systems.

(2) Systems where the feeder is installed on a railsystem that covers several units, normallycalled feeding robots.

Central feeding system

A central feeding system consists of storage silos, asluice valve, tubes with a flow of water or air fortransporting of feed, a selector valve, and eventu-ally a distribution unit (Fig. 16.8).

In this system the feed is delivered from the siloand into an auger that brings the feed particles intoa hopper placed above a sluice valve. The sluicevalve brings the feed particles from the hopper andinto the pipes for further transport to the tanks orcages. To transport the feed particles, water or air isused as a medium and the sluice valve thereforealso provides an air or water lock between thehopper and the transport medium; the sluice valvealso represents the feed distribution unit.

Whether air or water is used as a transportmedium, the velocity in the tubes is such that thefeed particles will always stay in suspension. Ablower or a pump ensures adequate velocity insidethe tubes. During the past few years, air has becomethe major transport medium. After being trans-ported for a short distance in the pipes (somemetres) the feed enters the selector valve whitchdetermines the production unit to which the feedportion is sent. There are several designs of selector valve; normally a rotating or a horizontalmoving selector is used.

After the selector valve the feed is transported tothe production unit through the tubes. In sea cagesthe tubes may be up to several hundred metres inlength. The silos and selector valve may be placedon-shore or on a barge. If the system is for largecages, a unit for spreading the feed may also beincluded.

Correct design and use of the feeding system isimportant to avoid feed breakage and dust produc-tion. Important factors are air temperature, pick-upvelocity, material in the pipes, design and use of theselector valve and pipeline routings.17,18

A centrally placed computer controls this type offeeding system. The amount of feed to the differentunits can be set as fixed or be created automatically.Addition of the initial weight, water temperature,expected growth and mortality in to the computa-tion ensures correct feeding. The computer alsostores the inputs and is an important tool for pro-duction planning and production control.

Central feeding systems are also available forautomatic feeding of moist feed.

Feeding robots

Put simply, a feeding robot is a feeder suspendedfrom a rail system hanging above the fish tanks (Fig.16.9). A motor to push the feeder along the rail

Feeding Systems 223

A

B

C

Figure 16.8 A feeding system used on a fish farm: (A)feed silos with a rotating selector valve; (B) selectorvalve; (C) the end of the tube from which feed is spreadinto the tank.

Figure 16.9 A feeding robot is simply a feeder suspended from a rail system hanging above the fishtanks.


system is included. The rail system is laid over theproduction units and under the feed silos.The robotmay have its docking station under the silos whereit enters for automatic refilling with feed when thehopper on the feeder is empty. When the robot isfeeding it moves along the rail until it hits a chipattached to the rail over each tank. Based on infor-mation in this chip, the feeder (robot) recognizesthe tank. On the robot there is a computer wherethe amount to be fed to the actual tank is pro-gramed. When the robot hits the chip it will there-fore feed the programed amount of feed to thetank. After this it continues to the next tank, and soon. When the hopper on the robot is empty it auto-matically goes back to the silos for refilling. Severalindividual feeders can be attached to the samerobot, so that it can deliver several feed sizes in thesame operation. The electricity supply to the feederand the motor may be an integrated part of the railsystem, or it can be a battery which is recharged inthe docking station. The great advantage with thissystem is that the same feeding mechanism can beused for feeding of several tanks. In this way moreinvestment in this unit means it can be designed tofeed more accurately.Accuracy can be improved byusing double dispensers (multistage), as is in somerobots.

16.3 Feed controlThe appetite of the fish is affected by externalfactors, such as variation in water temperature,water quality, waves (in cages) and light conditions.With a normal feed control, the amount of feed todistribute during a given period is fixed. If there isvariation in fish appetite there are no possibilitiesfor controlling this. This requires one of two solu-tions: the use of restrictive feeding with no feed loss,(really underfeeding), or acceptance of a certainfeed loss which is expensive and damaging to theenvironment.

Hand feeding is an old-fashioned system for regu-lating feed supply. The person who is feedingobserves the appetite of the fish visually, and in thisway regulates the amount of feed supplied accord-ing to fish appetite. One way to improve the feedcontrol and utilization of feed is to use a feeder forbasic feeding and hand feeding for topping up; this,however, requires manpower. Based on this,systems have been developed for automatic feed

control which are of special interest in productionunits where large amounts of feed are used, such aslarge sea cages.

16.4 Feed control systemsFeed control systems can be divided into manualand automatic systems. In tanks, manual systemsare used. The dual drain system with a particle traprepresents such a system (see Chapter 13). If thissystem is correctly designed, it is easy to observeany feed loss in the screen or separation unit for theparticle outlet. When screening the total outletwater from each tank, the feed loss may also beobserved. This, however, requires a large screen oneach tank, which is more expensive.

In sea cages a number of methods have beenintroduced. A manually operated method is to usea submersible video camera under the cage andwatch randomly for feed loss when the feeder isrunning. Another manual method is to use a stock-ing (a small net bag in the shape of a tube) in thelower part and under the net bag. Here feed lossand dead fish are collected. An airlift pump bringsthe feed loss and dead fish to the surface and intoa collecting bucket, where feed loss can be visuallycontrolled and the daily feeding amount may beregulated (Fig. 16.10). Equipment to detect thegathering behaviour, such as infrared photoelectricsensors, has also been used to control feeding.19

Figure 16.10 By using an airlift pump the feed loss onthe bottom of the cage is collected and brought to thesurface and into a tray where it can be seen. The systemcan also be used for collecting dead fish.

Hydro-acoustic sensors, photocells, linear videoand Doppler signals are all used for automaticallymeasuring feed loss.20–25 Either the sensor can beplaced under the cage or inside it to measure thefeed loss over a sample area. The sensor sends asignal to the feed controller to stop when feed lossexceeds a certain level. In tanks ultrasonic deviceshave been used to control the waste feed.20 Asystem observing the gathering behaviour in rela-tion to feeding has also been used for automaticfeed control.26

16.5 Dynamic feeding systemsDynamic feeding systems go even further and usethe feed loss to control the amount of feed to bedelivered. In one system a collector for feed loss is placed inside the cage.27 When feeding starts,the pump takes the collected food lost from the previous feeding and pumps this through a pipe that delivers it to the top of the cage bag

(Fig. 16.11). In this pipe circuit an infrared devicedetects eventually uneaten feed. If there are nofeed particles left, the feeder starts to add a newportion of feed. The same procedure happens when starting the next feeding. The system followsthe appetite of the fish in a dynamic way, withincreasing appetite increasing the amount of feedand decreasing appetite decreasing the amount of feed. Such systems have no feed loss to the environment, and it is possible to get early warning of eventual abnormal behaviour of the fish.

References1. Goodard, S. (1996) Feed management in intensive

aquaculture. Chapman and Hall.2. Swift, D. (1993) Aquaculture training manual. Fishing

News Books. Blackwell Science.3. Hochheimer, J. (1999) Equipments and controls. In:

Wheaton, F. (ed.) CIGR handbook of agriculturalengineering, part II aquaculture engineering (ed. F.

Feeding Systems 225

Figure 16.11 In a dynamic feeding system feed is delivered according to the appetite of the fish. (Adapted fromref. 27.)


Wheaton), pp. 281–307. American Society of Agricul-tural Engineers.

4. Woodcock, C.R., Mason, J.S. (1998) Bulk solids handl-ing: an introduction to the practice and technology.Springer Verlag.

5. Levy, A., Kalman, H. (2001) Handbook of conveyingand handling of particulate solids. Elsevier Science.

6. Klinzing, G.E., Marcus, R., Rizk, F., Leung, L.S. (1997)Pneumatic conveying of solids: a theoretical and prac-tical approach. Springer-Verlag.


8. Alanärä, A. (1996) The use of self-feeders in rainbowtrout (Oncorhynchus mykiss) production. Aquacul-ture, 145: 1–20.

9. Covès, D., Gasset, E., Lemarié, G., Dutto, G. (1998) Asimple way of avoiding feed wastage in Europeanseabass, Dicentrarchus labrax, under self-feeding con-ditions. Aquatic Living Resources, 11: 395–401.

10. Gélineau, A., Corraze, G., Boujard, T. (1998) Effectsof restricted ratio, time-restricted access and rewardlevel on voluntary food intake, growth and growthheterogeneity of rainbow trout (Oncorhynchusmykiss) fed on demand self-feeders. Aquaculture, 167:247–258.

11. Alänäre, A., Kadri, S., Paspatis, M. (2001) Feedingmanagment. In: Food intake in fish (eds D. Houlihan,T. Boujard, M. Jobling). Blackwell Science.

12. Rubio, V.C., Vivas, M., Sánchez-Mut, A., Sánchez-Vázquez, F.J., Covès, D., Butto, G., Madrid, J.A. (2004)Self feeding of European sea bass (Dicentrarchuslabrax, L.) under laboratory and farming conditionsusing a string sensor. Aquaculture, 233: 393–403.

13. Fullerton, B., Robinson Swift, M., Boduch, S.,Eroshkin, O., Rice, G. (2004) Design and analysis ofan automated feed-buoy for submerged cages. Aqua-cultural Engineering, 32: 95–111.

14. Larsson, K. (1978) Transport och portionering avkraftfôder vid mekanisk utfôdring. Medelande nr 374,Jordbrukstekniske instituttet (in Swedish).

15. Sørlin, S. (1985) Teknik för mängdbästemning. Ned-delande nr 407. Jordbrukstekniska Institutten (inSwedish).

16. Thomassen, J.M., Lekang, O.I. (1993) Optimal distri-bution of feed in sea cages. In: Fish farming tech-

nology (eds H. Reinertsen, L.A. Dahle, L. Jørgensen,K. Tvinnereim) pp. 439–442. A.A. Balkema.

17. Guajardo, M. (2004) Relation between feed qualityand handling in a feeding system. Master Thesis.Norwegian University of Life Science.

18. Norambuena, F. (2005) Aquaculture’s feeding system,optimisation of pick up velocity based on feed rate andpipeline length. Master Thesis. Norwegian Universityof Life Science.

19. Chang, C.M., Fang, W., Jao, R.C., Shyu, C.Z., Lia, I.C.(2005) Development of an intelligent feeding con-troller for indoor intensive culturing of eel. Aquacul-tural Engineering, 32: 343–353.

20. Blyth, P.J., Pursher, G.J., Russel, J.F. (1993) Detectionof feeding rhythms in sea cage rearing of Atlanticsalmon. In: Fish farming technology (edsH. Reinertsen, L.A. Dahle, L. Jørgensen,K. Tvinnereim), pp. 209–216. A.A. Balkema.

21. Dunn, M., Dallard, K. (1993) Observing behaviourand growth using the Simrad FCM 160 fish cagesystem. In: Fish farming technology (edsH. Reinertsen, L.A. Dahle, L. Jørgensen,K. Tvinnereim), pp. 269–274. A.A. Balkema.

22. Juell, J.E., Furevik, D.M., Bjordal, Å. (1993) Demandfeeding in salmon farming by hydro acoustic fooddetection. Aquacultural Engineering, 12: 155–167.

23. Foster, M., Petrell, R., Ito, M.R., Ward, R. (1995)Detection and counting of uneaten food pellets in seacage using image analysis. Aquacultural Engineering,14: 251–269.

24. Kevin, D.P., Rayann, J.P. (2003) Accuracy of amachine-vision pellet detection system. AquaculturalEngineering, 29: 109–123.

25. Summerfelt, S.T., Holland, K.H., Hankin, J.A.,Durant, M.D. (1995) Hydro acoustic waste feed con-troller for tank systems. Water Science and Tech-nology, 31: 123–129.

26. Chang, C.M., Wang, W., Jao, R.C., Shyu, C.Z., Liao,I.C. (2005) Development of an intelligent feedingcontroller for indoor intensive culturing of eel. Aqua-cultural Engineering, 32: 343–353.

27. Skjervold, P.O. (1993) Fish feeding station. In: Fishfarming technology (eds H. Reinertsen, L.A. Dahle,L. Jørgensen, K. Tvinnereim), pp. 443–445. A.A.Balkema.

17Internal Transport and Size Grading

ports the fish out of the farm. Several handlingoperations may be necessary throughout theseprocesses. One example of a handling line for inter-nal transport of fish on a farm with tanks can be:

• Crowding in tank• Dip net for lifting fish out of the tank• Bucket for internal transport of the fish• Dip net for lifting of the fish from the bucket an

into the new tank.

It is important that the separate handling methodsand equipment used in the handling line are com-patible. The handling methods may also be an inte-gral part of the farm construction,1 and thereforebe designed and the equipment selected beforeplanning and building the farm. Use of ‘alternativemodels’, where alternative methods for performingthe different handling operations in the line areincluded, can be an effective tool for selecting han-dling methods (see Chapter 22).

17.2 The importance of fish handling

17.2.1 Why move the fish?

The amount of fish or shellfish that can be producedon a farm depends on the fish density in the pro-duction units. Land-based fish farms equipped forintensive farming require high investment per unitfarming volume. Continuous high fish density is therefore necessary to attain good productioneconomy;2–4 this will require frequent transport orreallocation of fish as they grow (Fig. 17.1).

How often the fish must be moved is mainlydecided by their growth rate, but input density andmaximum allowed density in the production units

17.1 IntroductionVarious forms of handling are necessary in all aqua-culture activity. In extensive farming the fish arehandled very few times, but frequency increaseswhen the farming becomes more intensive. Thereare several reasons for handling fish and otheraquatic organisms. They are transported within orbetween farms, or from farms to slaughterhouse;examples include the transport of fry to on-growingfarms, and adult fish from on-growing farms toslaughterhouse (see Chapter 18). The need for thistype of transport, of course, depends on the pro-duction strategy of the farm; however, most is performed inside the farm area. Internal transportis performed for various reasons in connection with other handling activities, such as division offish groups, size grading, weight sampling and vaccination. In this chapter a description of differ-ent methods and equipment for handling of aquaticorganisms inside the farm is given, mainly for fish, although some of the methods and equipmentmay also be used on shellfish and other aquaticorganisms. A brief description of the advantagesand disadvantages of fish handling is also included.

Independent of the handling procedures andequipment used, it is important that the operationis performed by trained personnel and in a way thatminimizes the possibility for injury and stress to thefish.

When deciding on fish handling systems, it isimportant to consider the entire operation startingfrom when the fish are moved from the productionunit and ending when the fish have been returnedto the production unit or are in the unit that trans-

227


are also important.An example can be used to illus-trate this.

ExampleOn a farm for ongrowing fish, the fish will not bemoved during production (from input to delivery).Average stocking density is set at 45kg/m3, and themaximum density to avoid growth reduction is set at100kg/m3. Calculations show that the input densitycannot therefore exceed 2kg/m3 (determined byexponential growth in kg/m3). This shows that poorutilization of the production units will result if thefish are not moved. The length of productiondepends on input size, harvesting size and specificgrowth rate (SGR) in relation to fish size (dailygrowth rate expressed as percentage of body weight).

If higher average density is to be achieved, fre-quent moving of the fish is necessary, as illustratedby the following example. Maximum density mustnot exceed 100kg/m3, and required average fishdensity in the production units on the farm is 70kg/m3.The SGR is set at 0.9.The intervals betweenthe movements for dividing/splitting of the fishgroup are calculated to be 3–4 months. The table

illustrates the interval between handling (months) inrelation to SGR, input density, average density andmaximum density where the input weight of the fishis 100g and the harvesting weight 4kg. This clearlyshows that increased growth rate increases the needfor handling.

Interval between handling Fish density (kg/m3) (months) for different SGR

min avg max 0.3 0.5 0.7 0.9 1.1 1.3

40 70 100 10–11 ca. 6 4–5 3–4 ca. 3 2–340 60 80 ca. 8 ca. 5 3–4 2–3 ca. 2 1–2

17.2.2 Why size grade?

In a fish group there are several reason for sizegrading of fish.

Improved growth

In a large group of fish in tanks or cages, there isindividual variation in the growth rate. Some indi-

Figure 17.1 To keep high average fishdensity in the production unit, frequent handling is necessary.

viduals grow faster than others so differences in theindividual weights for the fish in the group willdevelop over a period of time, even if all the fish inthe group have exactly the same weight when thefeeding period starts. This is a sub-optimal situationfor several reasons.

An ordinary biological population will typicallyhave a normal (Gaussian) distribution of weightresulting from genetic variation in growth rate.Under farming conditions with artificial feeding,this normal distribution may develop to form ahierarchy preventing smaller, less dominant indi-viduals gaining access to feed. Because of this, thevariation in size might become so large, the biggerfish will eat the smaller ones. How fast cannibalismdevelops in the population is species dependent.For instance, cannibalism will develop very rapidlyif bass are not graded.5 The distribution in the fishgroup may develop from normal distribution to atwo group distribution, with typical winners andlosers; this may reduce the total growth in a fishgroup6,7 (Table 17.1).

The coefficient of variation (CV) can be used todescribe the weight variation in the fish group. TheCV is the standard deviation (d) expressed as a per-centage of the mean value

CV = (d/Xmean) × 100.

If the mean weight of fish in group is 2kg and thestandard deviation is 0.5kg, the CV will be (0.5/2)× 100 = 25%. In a fish group the CV varies withspecies, size, age and farming conditions. The CV in a fish group is related to the growth rate;faster development occurs with faster growth. Forexample, in salmon smolt production the CV can beup to 100%, while in on-growing production in sea-water the CV is seldom above 30%.

When size grading a fish group into severalweight groups, the individual weight variation ineach of the new groups will be less than in the start-ing group. When dividing a fish group into two, fishsmaller than the mean size are put in one group andthose that are larger in the other.When grading intothree or four groups, the size variation in eachgroup is of coarse even less. On a typical salmonsmolt farm between three and five gradings peryear are normal.8,9 For on-growing in cages, onegrading is normally enough plus a possible gradingin connection with harvesting.

Production control

To ensure good production control in an intensivedrifted farm, it is necessary to maintain a small sizevariation, i.e. a low CV.10 Production must be con-trolled to maintain a satisfactory growth rate on thefarm in relation to budget. This requires regularweight sampling to ascertain average fish size in thegroup. In a tank or cage with tens of thousands offish it is impossible to weigh all the fish individually,so only a sample of the fish is withdrawn for weightsampling; this sample must be representative of the whole group. However, this is difficult, espe-cially if there is a large size variation in the group.Several methods, including the use of monitoringinstruments (see Chapter 19), can be used for thispurpose.

One manual method of weight sampling uses adip net to take fish from a group, for instance, froma tank of juvenile fish.11 Three samples of at least50 fish each are taken out from the total fish groupin the production unit using the dip net. A require-ment of this method is that the CV of the threesamples shall vary by less than 2%, otherwise moresamples are needed. Below an example is used toillustrate how this functions in a population withsome size variation.

ExampleThree samples taken from a large fish group in atank of juvenile fish give the following results:

Internal Transport and Size Grading 229

Sample 1 2 3

No. of fish 103 107 98

Total weight (g) 1102.3 1155.6 1009.4

Av. weight of each fish 10.5 10.8 10.3(g)

The average fish weight based on the average of thethree samples is 10.53g. To simplify the calculationsan equal number of fish between the samples isassumed. Between the three samples the standarddeviation (d) is 0.252. This gives a (coefficient ofvariation) (CV) of:

(0.252/10.53) ¥ 100 = 2.39%


Tab

le 1

7.1

Res

ult f

rom

an

expe

rimen

t whe

re th

e gr

owth

and

coe

ffici

ent o

f var

iatio

n in

a g

roup

of j

uven

ile A

tlant

ic s

alm

on w

ere

stud

ied

for

non-

grad

edan

d gr

oups

gra

ded

into

tw

o or

thr

ee w

eigh

t cl

asse

s.7

20 O

ct21

Nov

19 D

ec18

Jan

16 F

eb

Xm

ean

dC

VX

mea

nd

CV

Xm

ean

dC

VX

mea

nd

CV

Xm

ean

dC

V

Not

gra

ded

1A5.

573.

804.

588.

385.

9614

.22

11.2

57.

254.

9612

.90

9.77

2.10

16.1

013

.69

11.3

21B

7.78

3.98

4.98

10.1

06.

058.

0712

.63

7.72

5.68

16.9

310

.75

9.63

23.4

213

.75

4.52

Div

ided

into

22A

a4.

131.

935.

965.

593.

313.

697.

044.

8210

.95

9.02

7.15

13.9

211

.81

9.58

7.75

wei

ght

clas

ses

2Ba

4.17

2.01

4.16

5.59

3.71

8.25

8.25

5.60

2.68

10.8

66.

950.

9612

.42

10.0

615

.02

2Ab

11.2

12.

592.

515.

933.

602.

5621

.20

6.81

1.15

28.9

85.

744.

2634

.00

9.09

3.19

2Bb

10.8

72.

662.

8316

.71

3.52

2.33

22.3

95.

682.

5628

.20

6.16

2.77

34.5

09.

970.

44

Div

ided

into

33A

a3.

051.

310.

193.

811.

382.

774.

262.

313.

326.

174.

256.

076.

585.

278.

69w

eigh

t cl

asse

s3B

a3.

151.

262.

173.

971.

623.

184.

652.

365.

096.

554.

096.

476.

775.

0715

.38

3Ab

6.34

1.73

1.53

10.2

53.

312.

4313

.00

5.61

4.99

18.7

07.

361.

8122

.26

9.42

1.21

3Bb

7.15

1.88

2.10

9.46

3.43

1.42

12.7

05.

814.

9617

.52

7.57

6.29

22.9

411

.07

1.92

3Ac

11.3

82.

801.

9617

.55

2.97

1.13

13.0

13.

541.

4230

.41

6.30

4.59

37.7

17.

555.

403B

c12

.91

2.57

0.20

16.9

82.

972.

2113

.04

4.20

2.77

27.9

15.

986.

1736

.63

6.80

1.69

Key

:Xm

ean

=av

erag

e va

lue;d

=st

anda

rd d

evia

tion

;CV

=co

effic

ient

of

vari

atio

n.

This does not fulfil our requirements for a CV below2% and another sample needs to be taken.

Sample 4: 102 fish of total weight 1071.0g, so averageweight of each fish 10.50g.

New average weight = 10.525g new standard devi-ation (δ) = 0.206; new CV = 1.96%. This is satisfac-tory, below 2%.

This shows the difficulties of taking representativeweight samples from a fish group. It is not normallypossible to obtain an acceptable (CV) without sizegrading the fish.

Manual weight samples also include othersources for mistakes. Before taking out the fish forsampling it is important to mix them. Experience intanks has shown that there is a tendency for thelargest fish to be near the bottom; in cages thelargest fish are always in the deepest layers whenfish are collected for weight sampling (Fig. 17.2).Further, all the fish that have been withdrawn in thesample in the dip net must be weighed, not just thefirst individuals that are removed (see examplebelow). If the fish are small it is quite easy to getfar more than 50 fish in the dip net.

ExampleBelow the results from the author’s experimentwhere the weight (g) of different fish withdrawnfrom a fish group collected in a deep net are shown.

The results are given for the first 10 fish, the middle10 fish and the last 10 fish withdrawn from a sample of 50 fish, for an ungraded and a graded fish group.

Ungraded fish group


Sample 1 2 3

Weight First 10 16.1 15.7 13.6(g)

Middle 10 8.7 12.2 9.6

Last 10 5.2 4.7 4.5

Xmean 9.99 10.93 9.31

Sample 1 2 3

Weight First 10 38.4 37.2 37.3(g)

Middle 10 38.1 37.0 40.2

Last 10 32.1 33.5 33.7

Xmean 36.92 35.93 37.07

Figure 17.2 If the large fish are notincluded in a weight sample the wrongaverage weight of the fish group will beobtained.

Total Xmean = 10.10g, d = 6.05, CV = 8.07%

Graded fish group

Total Xmean = 36.63g, d = 6.8, CV = 1.69%


These results show the largest fish are withdrawn firstand the smallest at the end.

Harvesting of fish

When a pond, tank or cage is to be harvested, it isa great advantage if the fish in it have been sizegraded. If not, the slaughterhouse must deal withseveral sizes of fish which must go to the consumerin different weight classes. It is also more difficultto estimate precisely the amount of fish in eachweight class and therefore to achieve a good price.An example is given below where the fish areungraded (adapted from ref.10). Another solutionis to size grade in connection with harvesting andsend the fish that are too small back to the pro-duction unit.

ExampleA fish population with 10000 fish (n) is to be har-vested. The average fish size (x) is 4kg and the standard deviation (d) is measured as 1kg giving a coefficient of variation (CV) of 25%. Find thenumber of the different weight classes represented,and the assumed number in each weight class of 1kg if the fish group distribution is normal.

In this case a normal distribution table can be used,from which the value of d is found to be 0.841. Thismeans that 84.1% of the fish have a weight less thanx + d (5kg). 15.9% of the fish in the group musttherefore be larger than this, so of the 10000 fish,1590 are over 5kg.

Similarly the number of fish under x − d = 4 − 1 =3kg can be calculated and is 1590.Therefore between3 and 5kg there are 10000 − (1590 + 1590) = 6820fish. These are equally divided between 3–4kg and4–5kg weight classes, each of which comprise 3410fish.

Next the numbers of fish larger than 6kg and smallerthan 2kg are calculated; this represents x + 2d. Againthe normal distribution table is used and 2δ = 2 givesthe value 0.977, meaning that 97.7% of the fish aresmaller than 6kg and larger than 2kg, so 2.3% arelarger than 6kg and 2.3% smaller than 2kg. 2.3% of10000 = 230 fish. Therefore there are 1360 fishbetween 2 and 3kg and also between 5 and 6kg. Thenumbers of fish below 1kg and over 7kg are verylow (13 fish in each case) and can be ignored.

As this shows, there are six different weight classesthat have to be sent to the slaughterhouse. Thismakes management quite difficult.

17.3 Negative effects of handling the fishEven if handling is necessary, especially in intensivefarming, it includes a number of possible adverseeffects. Before selecting handling routines andequipment, this must be taken into consideration.Handling creates a stress response in the fish, whichmay affect the production results negatively. Whenthe fish become stressed, the primary and sec-ondary effects will not normally be discoveredunless special measurements of heart rate, oxygenconsumption or blood characteristics (for instancecortisol or glucose) are taken.12–17 The farmer nor-mally registers the secondary or tertiary effects ofstress manifested by reduced growth and reducedimmune defence,18–21 which again may directlyreduce productivity.

It is also important to consider the possible stressresponse involved in pre-harvest handling. Thismay increase the consumption of glycogen storedin the muscle (part of the stress response). Theresults of this may be an earlier occurrence and ashorter duration of rigor mortis after slaughtering,which again will reduce the fish quality.22

How much the fish is affected by handling isspecies dependent: some species are more tolerantof handling than others. Results also show that fishmay adapt to handling procedures, and the stressresponses will gradually be reduced. This can beseen, for example, when the fish tanks are washed.14

The first time the tanks are washed it is possible tomeasure a high stress response, but this will gradu-ally decrease as the fish begin to tolerate this pro-cedure. Breeding programmes may also be used toadapt the fish to more and more of the normal han-dling operations in fish farming.23–25 When startingto rear a new species, it is collected from naturalwild stocks and put into farming conditions. Thebehaviour of such stocks differ from that of wildstocks that have been farmed and bred for genera-tions as is clearly seen when looking over the edgeof the tanks containing farmed and wild stock; dif-ference in behaviour is also shown by the numberof involuntary collisions between the fish and thetank walls.

Fish may also suffer physical damage if handledtoo roughly. Tolerance here, of course, also dependson species and life stage. The fish may be wounded,by rough handling leading to fungal attack. It isespecially important to avoid physical damage inpre-slaughter handling because it may reduce theflesh quality and hence the price of the product.

All handling includes some kind of human work,which requires time and creates costs.The total eco-nomic cost and possible negative effects of handlingmust therefore be compared to the positive effectshandling will have on the production. For thisreason it is very important to use effective handlingprocedures and handling lines which affect the fishas little as possible.

17.4 Methods and equipment forinternal transportTwo different principles may be used to move fishinside the farm:

(1) With a supply of energy(2) By the use of signals or stimuli to get the fish

to move voluntarily.

The first method is totally dominant, while thesecond is mainly the subject of research. The mostcommon methods within each group are described

below. For more information about integration ofhandling methods in farms see Chapter 21.

17.4.1 Moving fish with a supply of externalenergy

With supplied energy, the total internal transportprocess may again be divided into three phases:

(1) Crowding of the fish inside the production unit

(2) Vertical transport where the fish are liftedbetween the levels

(3) Horizontal transport of the fish between theunits.

When moving the fish to a lower level no verticaltransport is necessary; stored potential energy isused here as a source for the process. When movingfish between two equal levels, crowding can be usedto force them to move.

Crowding

In almost all methods for vertical transport, crowd-ing of the fish in a restricted volume is necessary.Too much crowding may, however, result inunwanted stress (Fig. 17.3). During the crowdingprocess fish behaviour should be observed; if


Figure 17.3 Fish that are overcrowded.


odd behaviour occurs, further crowding must beavoided.Two methods are commonly used to crowdfish:

• Reduction of water level (tanks, ponds)• Reduction of available volume (ponds, tanks and

cages)

If reduction of water level is to be used to crowdthe fish, the outlet of the tank or pond must bedesigned in a way that makes drainage possible (seeChapter 13). If there is no drain, a drainage pumpcan be put inside the tank or pond, but this makesthe handling operation more difficult. Lack ofoxygen in the water may occur during this opera-tion and supplementary oxygen may have to besupplied; the oxygen level must therefore be monitored.

In cages, the net bag can be lifted to reduce thevolume available to the fish, and through this crowdthe fish.A seine net may also be used in ponds, largetanks or cages, but is less effective. In tanks, fixedor removable grids may be used to crowd the fish.A combination of collection grid and a decrease inwater level can also be employed (Fig. 17.4).

Vertical transport

Dip net: The dip net is constructed with a round orrectangular frame with a net or tarpaulin bag inside(Fig. 17.5). Round frames are suitable for net bagsto be used in cages, while rectangular frames aremore suitable for use inside tanks. For a dry net,small mesh knotless netting should be used to avoidwounds. Plastic net have also been tried to reducethe possibilities for wounding.26

If a tarpaulin bag is used the fish will always bein water; this is also called a wet net. A hydraulic ormechanical crane must be used to operate a wet netbecause of the weight. A mechanism opens thebottom to empty the wet net of fish and water.

Use of nets is labour-intensive, especially for han-dling larger fish. Normally wet nets are between 100and 500 l capacity. Fish densities above 50–70% fishcompared to water are normally avoided to reducethe possibility of wounding the fish.

Pumps: A pump supplies energy to the water so itis either set under pressure or vacuum which causesthe water and hence the fish to move. The fish aretherefore in water throughout the entire handling

operation. Several systems for pumping fish areavailable, including centrifugal, vacuum, ejector andairlift pumps; crucially these have an open con-struction that does not injure the fish.

Centrifugal pump: A centrifugal pump used forpumping fish utilizes the same principle as a cen-trifugal pump used for pumping of water (seeChapter 2). To avoid injuring the fish it does,however, have an open impeller with large channelsand no narrow passages (Fig. 17.6). Because of itsconstruction, it is not commonly used on fish largerthan 1kg. If such a pump were to be made for 4–5kg fish the required dimensions of the impellerwould be very large. It is normal to use submergedpumps or at least pumps with a supply pressure.Fish pumps can be made self-sucking with specialadaptors. The impeller may be driven by hydraulicpressure.This makes it quite easy to move the pumparound in the farm area, because only the impellerunit is moved. The pump may also be drivendirectly via a shaft from an electric motor. Cen-trifugal pumps are commonly used both in produc-tion farms and in well boats; they have been usedfor many years in traditional fishery well boats,among others, for pumping herring.

Vacuum–pressure pump: A vacuum–pressurepump consists of a tank to which inlet and outlettubes are connected via valves (Fig. 17.7); a smallpump is also attached. This pump can either pres-surize the larger tank or withdraw the air from it,causing a partial vacuum. The function of the pumpis first to evacuate the tank; then the valve to theinlet pipe is opened and water and fish are suckedinto the tank; after this the inlet valve is closed andthe tank is pressurized; lastly the outlet valve isopened and the fish are forced out through theoutlet tube. The operation is repeated, and a newbatch is pumped through.

The pump does not deliver water and fish continuously, because it operates in two phases:vacuum and pressure. However, two pumps can beused alternately to obtain more equal delivery offish and water. A vacuum head of more than 5mH2O is normally avoided to prevent injuries to thefish; use of less than 40% water relative to fishshould also be avoided for the same reason. Herethe manufacturer’s recommendations must be followed.


Figure 17.4 A grid can be used tocrowd the fish.


Pumps of different sizes are required for han-dling small fish and harvesting large fish. The dif-ference is the size of the tank, the pipes and thepump used to evacuate and pressurized the tank.Pumps are used on farms, in well boats and inslaughterhouses.

Ejector pump: In an ejector pump, a high velocity,high pressure part flow creates a region of low pres-sure (suck) in the larger main stream (Fig. 17.8).The fish travel with the water in the main stream.When the water flows past the ejector it will gofrom low to higher pressure. The pump can

Figure 17.5 Various types of dip net can be used to lift the fish.

therefore deliver fish in a continuous flow of water.The pump has no moveable parts that can injure thefish. A suction head that is too high must beavoided; it is better to take a larger part of thelifting head on the pressure side. When fish havealready lost scales, as occurs to young salmonduring smoltification, they can easily lose morescales in these pumps, especially if the ejector isbadly adjusted.

Different sized ejectors are available, adjusted tofry and to on-growing fish. The smallest size isportable and easy to move around the farm. Har-vesting of mussels has also been performed withthis type of pump. Some pumps have ejectors atboth ends of the pipes so it is possible to change theflow direction; these are of interest on well boatsfor pumping fish in and out of the well.

Airlift pump: An airlift pump can be used forpumping fish (Fig. 17.8). Air is added to createbubbles that rise to the surface through a vertical,water-filled pipe suspended in the fish cage or tank.The bubbles cause drag on the water particles nearthem and create an upward flow of water inside thepipe. This moves the fish that are in the water upthrough the pipe. The capacity and the lift height insuch a pump depend on the depth of the water, atwhich depth the air is supplied and the amount ofair supplied. If using airlift pumps for fish transport,there are always possibilities for supersaturation ofthe water with nitrogen gas. Because of the short

fish retention time in the pump, this will not normally result in any problems, but if the fish stayin the water for a longer period, problems mayoccur.

More generally, fish pumps can be used for dif-ferent fish sizes from juvenile through to fish readyfor harvesting.The difference is the size of the tubesor pipes. Pumps may also be used for harvestingshells and mussels from the bottom culture, espe-cially airlift and ejector pumps.

It is very important that pumps are used accord-ing to the supplier’s recommendations. Severalproblems have occurred when transporting fish dueto incorrect pump use; examples include fish thathave been cut by valves, eyes that come out becauseof incorrect pressure conditions, and scale loss;they all show the importance of correct pump use.Awareness of the correct suction head is especiallyimportant.

The great advantage of pumps over other han-dling methods is their large capacity. Today pumpsare being used increasingly in intensive aquacultureas well as in slaughterhouses.

Fish screw: The fish screw, or pescalator, is basedon the Archimedes screw which was used for liftingwater in ancient times (Fig. 17.9). A screw is fixedinside a pipe which has a belt around its circumfer-ence. The belt is further connected to a small elec-tric motor; when the motor starts the belt will rotateas will the pipe and therefore the screw relative to


Figure 17.6 A centrifugal pump for pumping fish.


Figure 17.7 A vacuum–pressurepump for pumping fish.


Figure 17.8 Airlift (top) and ejector (bottom) pumps for pumping fish.


it. This is not a traditional screw which rotates; thepipe in which the screw is fixed rotates. Rotation of the pipe will result in lifting of water and the fish within it. The fish appear to be lying with waterin small basins. Fish must be crowded around theinlet of the pipe for the screw to be filled with fish.Hence the screw may be placed in a well where thefish are automatically crowded. Another solution isto use a special perforated tank around the inlet,which again stands inside the fish basin or pond. Adip net may be used to lift the fish up and into aperforated tank. The lift height achieved dependson the length of the screw. To avoid damage to thefish it is not recommended to have an angle towardsthe horizontal plane of more than 40°. Usually thescrew length is between 3 and 6m, while the diam-

eter of the pipe varies between 30 and 45cm. Anarea around the tank or pond where the screw is tobe used is required, so that the screw can be repo-sitioned. Such screws are commonly used in tankswith low water levels and in ponds.

Transport tanks: The transport tank is speciallydesigned with smooth surfaces and angles to avoidwounding the fish during transport (Fig. 17.10). Thefish have first to be transferred from their ordinaryproduction unit to the transport tank, for instanceby use of a dip net or by pumping the fish. It is nec-essary to have mechanical equipment, such as aforklift truck or a tractor with a front loader, to liftthe transport tank because of its heavy load ofwater and fish. The fish can stay in the transport

Figure 17.9 A fish screw used tomove fish.

tank at quite a high density; it is not usual to addextra oxygen into the tank during internal transporton the farm. The period that the fish can stay in thetank without a supply of oxygen is, however,limited.

The size of tank selected depends on the size ofthe fish to be transported. Tanks of 300–800 l arenormal. Fibreglass or aluminium are the usual con-struction materials for transport tanks, and thedesign is very similar to that of tanks used for ordi-nary fish transport (Chapter 18), but smaller. Tanksfor internal transport may therefore be used for

transporting small amounts of fish out from thefarm (external transport). The tanks are not,however, insulated like many of the long distancetransport tanks. To enable the fish to be tapped outof the transport tank there may be a hatch close tothe bottom which makes emptying of the tankquick and easy. The tank may also be used for ver-tical transport at the farm, being lifted with eithera forklift truck or a tractor.

If tanks are to be used for fish transport, quite alarge transport area is required. This must be aslevel as possible, to prevent the tractor or forklifttruck tilting when lifting the tank. Use of a trans-port tank to move juvenile fish is physically unde-manding. The capacity of this internal transportmethod depends on the size of the tank, the fishdensity and the duration. Fish density is speciesrelated; for example, over 500kg in 1000 l of watercan, easily result in lack of oxygen. Transport withsuch high densities must therefore only last for ashort period of time.

Horizontal transport between units

Pipes are commonly used for internal transport offish. The method is also often combined withpumping, for instance in cage farming. Here thepump transports the fish vertically and the hori-zontal transport is done through pipes. A certainexcess pressure (head) is necessary for fish trans-port in pipes. If not using a pump to create this pressure, one possibility is to create a magazine offish and water from which the fish can be tapped bygravity; alternatively, the pipe must slope down-wards and use gravity.

If there is a natural slope in the building area, afish farm may be built in terraces with start feedingin the upper part and on-growing in the lower parts(Fig. 17.11). Fish transport can then be performed


Figure 17.10 A tank used for internal transport of fish.

Figure 17.11 The farm may be ter-raced to utilize gravity for internaltransport.


through pipes utilizing the natural slope. Anothermethod is to use a common tapping centre to whichtapping pipes from the fish tanks are connected.This can either be a common tapping pipe system,or there can be individual pipes from the separatetanks. From the tapping centre, vertical transportcan be performed by some of the methods men-tioned previously, a fish pump or a fish screw, forexample. It must be possible to reduce the waterlevel in the tapping centre to crowd the fish beforevertical transport (Fig. 17.12); this can, for instance,be used to transport the fish back to other tanks orto a grader.

The same principle may also be applied in ponds,where the fish are tapped out to the harvestingtank. This requires the pond to be constructed toallow this, with a self-tapping pipe through the pondlevee.

When selecting the diameter of the tapping pipe,experiments have shown that it should be large

Figure 17.12 A common tappingcentre may be used for collecting fishfrom several tanks.

enough to enable the fish to turn inside.27 A dia-meter of at least half a fish length is a good start,but this will of coarse also be species dependent.The flow velocity inside the pipe must be suffi-ciently high that the fish understand it is not possi-ble to fight against the water flow and will only bedragged along with the water flow. If the velocity istoo low the fish will fight against the water flow andbe exposed to more stress. The correct velocity will,of course, depend on the species and its swimmingperformance; for salmonids an appropriate velocityin the pipe is three to four fish lengths per second27

(Fig. 17.13).The capacity of the system depends on the velo-

city of the water, the diameter of the pipe and thesize of the fish; the distribution of fish through thepipe depends on the diameter of the pipe and the size of the fish. Normally, completely waterfilled pipes, such as from pumps, are recommended;otherwise quite large slopes on the pipes are nec-

Figure 17.13 Use of pipes for fishtransport.

essary. If using partially filled pipes, after-flushingwith additional water must be carried out to emptythe pipe completely of fish. If after-flushing isstarted before the pipes are empty of fish a moreeven distribution of fish through the pipes willresult. If the farm is constructed with a good systemfor tapping of fish there is minimal requirement forhuman work to operate the handling system. Thecapacity is large both in total and per man-hour.

One question may be raised regarding the use ofa common tapping centre where pipes from severaltanks are connected, because there is the possibil-ity of transferring diseases between fish tanksthrough the pipes and the common tapping centre.When a farm is going to be constructed or recon-structed, much effort must be put into the disinfec-tion systems, drainage and insulation of singletanks, when a tapping centre is to be used.

If the fish are to be tapped through pipelines, allthe valves in the system must be of the ball or throt-tle type; with these valves the pipe diameter can becompletely open through the total cross section.This is important to avoid narrowing in the pipesand by this possible damage to the fish.

17.4.2 Methods for moving of fish without the need for external energy

Today there are no commercial systems availablebased only on voluntary movement of the fish.Some experiments have, however, been performedwhere the aim has been to move fish without sup-plying energy. If stress, wounds and labour are con-sidered, such methods are, of course, of interest.

To get the fish to move voluntarily, they need asignal or a stimulus that tells them to go. This caneither be a positive signal that will attract the fishor it can be a negative signal from which the fishwill swim away. The fish may also be trained so thata signal gives a positive or negative response. Whattype of signal or stimuli that can be used dependson the species and age. Practical observation hasshown that the fish will keep away from areas withlow oxygen levels. It is, however, important thatenvironmental conditions are varied in a positiveand not a negative way, which again may stress thefish. Flow of water is interpreted as positive stimu-lus for species that naturally prefer to stay in awater flow, such as salmon. Here juvenile and on-growing fish will swim against the water flow while

smolts will drift with the flow.28 One way to achievevoluntary transport is to equip the tanks withhatches, and have a channel where the fish can swimthrough and into the next tank (see Chapter 21).Water flow and light conditions can be manipulatedto improve voluntary transport.29,30

The addition of chemical substances to the watercan be interpreted as a positive stimulus by thefish.31–33 Experiments have shown that certain typesof amino acids can attract fish34 (Fig. 17.14). Mani-pulation of the light conditions may also have a pos-itive effect on the voluntary movement of fish; darkzones may attract them.35

The fish may also be trained to respond to a stim-ulus voluntarily.The simplest way is to teach the fishto associate a signal with a positive stimulus, i.e. tocondition their behaviour.The Russian physiologistPavlov demonstrated this with his famous dogtrials: every time the dog was fed a bell rang andafter a while the dog salivated in anticipation offood merely when it heard the bell ring. In fishfarming, similar experiments have been carried outin connection with collecting fish.36,37 In the author’sexperiments fish were trained to associate a flash-ing light with feeding38,39 (Fig. 17.15); after a periodthe fish crowded together around the flashing light,even if they were not fed. Similar principles havebeen used in the sea. Sound has been used as asignal when feeding fish in a cage. After a period oftraining the net bag was removed and the fish


Figure 17.14 Chemicals can be used to attract thefish.


A

B

C

Figure 17.15 (A–C) If fish aretrained to associate feeding with aflashing light, they will come to thelight when it is flashed. (C) A rig wasdeveloped for doing experiments thatallowed the fish to move betweentanks.

released;40 they could now swim and search fornatural feed over a larger area. By making a soundagain the fish crowded around the feeder and col-lection for harvesting was possible. In tanks a soundsignal is difficult to use because of echoes from thewalls which make it impossible for the fish to locatethe sound source exactly.

The main problem with voluntary transport is itseffectiveness and the time needed when the fish areto be moved. Not all the fish respond to the chosenstimulus, which also represents a problem.

17.5 Methods and equipment for sizegrading of fishSimilar to the methods and equipment for internaltransport, the size grading equipment may also bedivided into systems that do and do not requireaddition of extra energy. It may not be necessary tosupply energy to the grader, but the equipment maybe of a design that requires potential energy; thefish must be lifted to a higher level before gradingand sent from there into the grader.

All size grading will stress the fish, even if there are variations from species to species. For this reason, it is important that the equipment used and methods employed are implemented correctly to minimize the stress response of the fish. Wounds may also occur as a result of incor-rectly adjusted grading equipment. Thereforegraders must be used according to the supplier’srecommendations.

Several methods are used for grading of fish.Equipment can be separated into that needing asupply of energy and equipment where the fish vol-untarily grade themselves. The first method istotally dominant. The effectiveness of the differentmethods is to some degree dependent on thespecies to be graded. The latter method is mainlyused in research activities. A brief survey of themethods most used follows.

17.5.1 Equipment for grading that requires anenergy supply

Methods where the fish are taken out of the water

Manual: Fish can, of course, be size graded manu-ally. Each fish is taken on a table, visually graded by

hand and sent to the different size classes. Thismethod can be used to grade very small amounts offish, but it is labour intensive. Therefore some typesof automatic grader are used.

Fish cradle: A grading cradle is simple in construc-tion and cheap to buy (Fig. 17.16). It is quitecommon to use a cradle, especially for the firstgrading, or for small species. The same is the casefor smaller farms or on more extensive drift farmsthat seldom grade their fish. A cradle is basically abox with ribs or bars in the bottom. When using atraditional cradle, fish of different sizes are crowdedinto the cradle which is placed inside the fish tank.The cradle is then lifted up and shaken. Small fishwill now fall between the ribs in the bottom of thecradle and remain in the fish tank. The cradle


Figure 17.16 Using a fish cradle to grade fish.


containing the larger fish is then lifted out of thetank and swung into another tank where the largefish are released. In relation to capacity the use ofcradle for grading is labour intensive. A cradle thatis lifted manually will impose large loads on theoperator’s back. The advantage with using a cradleis that it is simple, inexpensive and does not requireany additional equipment.

Grading box: A grading box is based on the sameprinciple as the grading cradle. The fish are liftedout of the fish tank, for instance using a dip net, andpoured into a box with a grading grid in the bottom.The smallest fish will fall through the bottom of thegrading box under gravity. The bottom consists of agrid made of bars; the distance between the bars isadjustable.The box normally slides on a rail system,and when the small fish have fallen out the box ismoved over a new tank using the rail system. Herethe distance between the bars in the bottom of thebox is increased and fish of a new determined sizethen fall through the grid. Afterwards the slidingbox can be moved above yet another tank and thedistance between the bars increased once more, sothe largest fish will fall out. The equipment has lowcapacity and is only recommended for gradingsmall fish.

Tilt grader: A tilt grader is based on a similar prin-ciple. Fish are poured over a grid system, normallywith two or more sections on top of each other (Fig.17.17). The fish are poured into the middle section

and the smallest fish fall through all the grids andinto a tank below. Then the top grading grid, whichhas the largest distance between the bars, is tiltedto one side and the largest fish will follow and fallinto a tank. Then the grader tilts the intermediategrade to the other side and the medium sized fishwill fall into another tank. If a small amount of fish is being graded, the grid may be tilted manu-ally. For larger fish and larger quantities, hydrauliccylinders may be used to tilt the grids. This type ofgrader normally divides the fish into three sizeclasses. It has quite low capacity and is also labourintensive.

Another method based on the same principle is also used on large fish. The fish are crowded into a grading box fixed to a sledge. When a reasonable number of fish have come into the box, hydraulic cylinders lift it. The smallest fish then fall through the grids in the bottom and downinto the tank fixed below. When the sledge reachesthe top of the rail system, the fish are dropped intodifferent size groups separate from each other. Inthis system the vertical transport is part of thegrader.

Grading gridsDesign of a grading grid: There are a number ofways to design a grading grid, which may be placedhorizontally or vertically. The grading grid caneither have a fixed or variable distance between thebars. If the distance between the bars is fixed sepa-rate sizes of grading grid must be available on thefarm according to the size of the fish to be graded.To achieve a variable distance between the barsseveral methods are available (Fig. 17.18). Onemethod is to set the bars in a frame where the twoopposite sides can be displaced parallel to eachother, so changing the distance between the ribs.The scissor principle may also be used. Here theseparate bars are placed in the centre of a scissorconstruction; by opening this out, the distancebetween the bars is changed. It is also possible toplace removable knots of various sizes between the single bars to obtain a grading grid with variable distances between the bars, but changingthe spacing between the bars will require moretime.

The same bar construction may also be used ingrading grids which are to be placed vertically inthe water, in a net cage, pond or sea cage. The gridFigure 17.17 A tilt grader for grading fish.


A

B

C

Fig

ure

17.

18S

ever

al m

etho

ds a

re u

sed

to a

chie

ve a

var

iabl

e di

stan

ce b

etw

een

the

bars

in a

gra

ding

grid

: (A

) gr

id c

onst

ruct

ion;

(B

, C

) th

e gr

idis

rot

ated

and

the

sm

alle

r fis

h sw

im t

hrou

gh w

hile

the

larg

er fi

sh a

re r

etai

ned.


may either stay fixed in one place or draggedthrough the production unit as an integral part ofthe seine net.

Distance between the bars: The distance betweenthe grading bars determines the size of fish to beplaced in the different classes. This size, or actuallythe thickness of the fish, is related to its weight. Itis, however, difficult to give exact values for the dis-tances that should be used to grade out fixed sizesof fish. This of course varies with the species,because they have different body shapes. However,it also varies with the condition of the fish withinthe same species. A fish in good condition will bethicker than one in poor condition. Normally a fishis thickest just behind the gills, but this may varyfrom species to species.

To obtain a rough estimate of the distancebetween the bars, salmonids can be taken as anexample. A rough estimate says that The width ofthe fish i.e. the thickness (T), is around 1/10 of thelength of the fish. The condition factor of the fish(C) and the weight of the fish (W) in grams mayalso be used to estimate in mm (and hence the dis-tance between the bars). The following formulamay be employed:

C = (W × 100)/T 3

T = C/(W × 100)1/3

where:

T = thickness of the fish (mm)C = condition factor of the fishW = weight of the fish (g).

Grading machines (graders): A number of princi-ples are used to determine the design of gradingmachines or grader. For all machines described thefish must be lifted out of the water; it is also neces-sary to lift the fish up to the grader, and the liftingheight depends on the principle used for grading.Because the fish are graded in air, it is usual to spraythem with water to prevent them drying out. Aftergrading, the different fish sizes are delivered intodifferent tanks through a pipe system, each housinga different size of fish, or the different sizes canreturn directly to the different fish tanks, depend-ing on the total handling system. To get effectivegrading the grader must be fed continuously; thefish must not come in batches. Depending on con-

struction, the grading machines are to variousdegrees adapted to take fish with different bodyshapes, for instance flatfish. Before choosing agrader it is therefore important to ensure that thegrader is appropriate for the species.

Bar graders: In a bar grader the fish slide down ona slightly sloped ‘rib table’ constructed of beams orbars (Fig. 17.19). The distance between the bars issmallest close to where the fish enter the table andthen gradually increases.Therefore the smallest fishwill fall through the bars first into the tank under-neath. Since the distance between the bars gradu-ally increases larger fish will fall through when theyhave advanced some distance from the entry point.This type of grader is usually used to divide the fish

Figure 17.19 A bar grader.

group into two or three size classes. The advantagewith the bar grader is that there are no movableparts, but it has limited capacity compared to theother graders. What decides the capacity is theslope angle on the grading board. However, if the slope is too large sub-optimal grading willresult; many fish will enter the wrong weight groups.In each case there will be an optimal slope, and thismust be tested on the site.

Roller grader: The design of the roller grader issimilar to the bar grader, but the bars are replaced

by rotating rollers. This system can be utilized forall size classes; the size of the graders is the only dif-ference. Roller graders are normally installed onland-based farms, but may also be installed on boatsor rafts for grading fish in sea cages. A machinelocated on a raft can be partly submerged to reducethe required lifting height.

The principle of ‘dry placed’ roller graders forjuvenile and on-growing fish involves lifting andpouring the fish, so that they flow over the ‘table’with the rotating rollers (Fig. 17.20). Two pairs ofrollers rotate away from each other so that the fish


Figure 17.20 A grader equipped with rotating rollers,where two pairs of rollers rotate away from each other.The rollers are driven by electric motors.


are not squeezed between the rollers, but lifted upso that they fall through the grading table in thecorrect place. Between the pairs of rollers there isa ridge. The distance between the rollers increasesfrom the start to the end of the grading table. Twodifferent types of rollers are used: the first type ofroller is the same diameter along the whole lengthand is installed with a fixed angle between the pairof rollers; in this way an increasing distancebetween the two rollers is achieved. The other typeof roller has a diameter that decreases in stepsalong the direction of movement of the fish,because of which there will be an increase the dis-tance between the rollers; this ensures separation ofthe fish based on thickness. A roller grader has anelectric motor to drive the rollers via gear wheels.Normally this grader will divide the fish group intothree to four different sizes. Roller graders havehigher capacity than bar graders. The capacitydepends on the number of rollers, or actually the

width of the machine, the length and the slope ofthe rollers.

To use this type of machine, the fish must be liftedup to the grading table. The head loss over themachine is quite low (about 50cm H2O). Normallyit is therefore possible for the fish to fall directlyfrom the machine back into ordinary fish tanksthrough pipelines. This type of machine is normallyequipped with wheels so can be easily movedaround the farm and stored when not in use.

Belt grader: In a belt grader the fish slide betweentwo rotating belts that are positioned obliquely toform a V-shaped channel with no bottom (Fig.17.21). The rotation of the belts helps to drive thefish forward in the channel. To rotate, each belt isequipped with its own electric motor. From thepoint where the fish are poured into the machine,the distance between the belts gradually increases.When the distance between the belts is large

Figure 17.21 In a belt grader thedistance between two rotating beltsgradually increases.

enough the fish will fall through into tanks under-neath, the smallest fish first followed by the othersizes. Normally this grader is used to grade intothree size classes or more. These graders have avery low head loss, so only a low lifting height isrequired for the fish to enter the grader. This typeof grader is also equipped with wheels for easymovement inside the farm; it is long and narrow(3–5m) and so requires a long space.

It is important not to overload the grader with atoo many fish. Sub-optimal grading will result,because the small fish may stay on top of the largerfish and in this way drop into the wrong size class.This problem may also occur with roller and beamgraders, but in these cases a wider grading table canbe used to increase the capacity and in this wayreduce the problem.

Band grader: This is a fairly new grading principlewhich combines the belt and the roller grader (Fig.17.22). A tilted rotating belt or band of PVC intowhich the fish are lead is used. The fish lie on theirside and move in the direction of rotation of theband. A roller is sited above the belt. The distancebetween the band and the roller gradually increasesin the direction of transport. When the distance is

large enough the fish will slide under the roller. Inthis way division into groups is achieved. It isnormal to grade in groups of three sizes with thismachine.The machine has also proved adequate forgrading flatfish such as turbot.

Level grader: On a level grader the fish are pouredonto the top and gradually slide down throughtilted grids that form a ‘grading table’ similar to thebar grader (Fig. 17.23). The larger fish will not gothrough the grid and will therefore be removed onthe first grid. Smaller fish will continue to fallthrough onto a new tilted grid where the next fishsize is removed, and then the same process maycontinue with new grids. Normally this grader willalso divide fish into three size groups, but themachine may quite simply be adjusted to gradeseveral sizes. The advantage with this machine isthat there are no moveable parts and that thelargest fish are removed first. The machine has aquite a high head loss and the fish must be lifted inorder to enter the machine. Such graders are notnormally mobile, but lightweight versions may beproduced so that moving is possible. This type ofmachine is recommended to stay centrally placed inthe farm. If the system for getting fish in and out


Figure 17.22 A band grader utilizes aprinciple that is also adequate for gradingflat fish species.


Figure 17.23 In a level grader thelargest fish are separated away first,which prevents the smaller fishstaying on the back of larger fish.

from the grader is well designed, a large gradingcapacity can be achieved with this system, forinstance in connection with a tapping centre.

Other types of grader: There are also a number ofother grading systems and principles that can beused. These are not described here where focus ison the most general types and basic principles. Forinstance, many fish farmers have constructed theirown graders adapted to their specific needs.

Methods for grading the fish in the water

Methods where the fish are maintained in the waterthroughout the entire grading process have beendeveloped for raceways, ponds, sea cages andgrading channels.5,41,42 In raceways this is possibleby using movable vertical grading grids (Fig. 17.24).When using the grid it is moved towards the fishwhich are forced to swim through. It is often usedin combination with a water flow towards the fish.The smallest fish will pass through whilst the largerfish will not be able to do so and will gradually becrowded together.

Pulling two cages together and placing thegrading grid in the middle gives a similar system.

Alternatively, a seine net with a grading gridincluded can be dragged through a cage or pond(Fig. 17.24). The smaller fish will swim through thegrid and remain in the cage or pond, while thelarger fish will be crowded together and can beremoved. In circular tanks a similar method can beused. A vertically placed grading grid is draggedthrough the tank like the hand on a watch. Anotherdense grid is fixed in the tank. Both the grading gridand the dense grid are fixed on a fitting in the centreof the tank; this may, for instance, be a centre drainin the tank.43

17.5.2 Methods for voluntary grading (self grading)

The same stimuli that are used for voluntary fishtransport have been used for self grading34,42,44 (Fig.17.25). Water flow towards the fish, scented sub-stances, manipulation of light conditions andbehaviour training have all had some effect, but fullsize grading of an entire fish group has been shown


A

B

Figure 17.24 Moveable grids can beused for grading in the water in: (A) seacages and (B) raceways.


to be impossible. In addition, it is necessary to havea slightly larger distance between the ribs whenusing stimuli compared to traditional methods forgrading because the fish will not go through smallpassages voluntarily. In particular, when the fishfeel the ribs on their sides they will stop swimmingvoluntarily. In practice, where time is limited,stimuli are not useful for voluntary grading of fish.

References1. Lekang, O.I., Fjæra S.O. (1993) Fish handling in land

based fish farms. In: Fish farming technology. Pro-ceedings of the first international conference on fishfarming technology (eds H. Reinertsen, L.A. Dahle,L. Jørgensen, K. Tvinnereim). A.A. Balkema.

2. Kjartansson, H., Fivelstad, S., Thomassen, J., Smith,M.J. (1988) Effects of different stocking densities onphysiological parameters and growth of adultAtlantic salmon (Salmo salar L.) reared in circulartanks. Aquaculture, 73: 261–274.

3. Lekang, O.I., Fjæra, S.O., Skjervold, P.O. (1992) Enlønnsomhetsvurdering av interntransport og flyttingav fisk. Norsk Fiskeoppdrett, 1a: 22–23 (in Norwe-gian).

4. Lekang, O.I. (1994) Logistikksystemer i landbaserteoppdrettsanlegg. ITF rapport nr 53. Norwegian Uni-versity of Life Science (in Norwegian, Englishsummary).

5. Parker, R. (2002) Aquaculture science. ThomsonLearning.

6. Gunnes, K. (1976) Effect of size grading youngAtlantic salmon (Salmo salar) on subsequent growth.Aquaculture, 9: 381–386.

7. Fjæra, S.O., Lekang, O.I. (1991) Betydning av stør-relsessortering av lakseyngel på tilvekst, dødelighet og

størrelsesvariasjon. ITF-rapport nr 22. NorwegianUniversity of Life Science (in Norwegian, Englishsummary).

8. Lekang, O.I., Kittelsen, A. (1989) Spørreundersøkelseblant settefiskprodusenter om utstyr og rutiner forhandtering. LTI-trykk nr. 101. Norwegian Universityof Life Science (in Norwegian).

9. Willougby, S. (1999) Manual of salmonid farming.Fishing News Books, Blackwell Science.

10. Lekang, O.I., Fjæra, S.O., Skjervold, P.O. (1991) Stør-relsessorter ofte, unngå størrelsesspredning. NorskFiskeoppdrett, 7: 24–25 (in Norwegian).

11. Lekang, O.I., Fjæra, S.O. (1992) Erfaringer med uttakav fisk til veieprøver. Norsk Fiskeoppdrett, 1: 26–27(in Norwegian).

12. Mazeaud, M.M., Mazeaud, F., Donaldson, E.M.(1977) Primary and secondary effects of stress in fish:some new data with a general review. Transactions ofthe American Fisheries Society, 106: 201–212.

13. Pickering, A.D. (1981) Stress and fish. AcademicPress.

14. Lekang, O.I. (1989) Effects of handling on juvenilefish. PhD thesis, Norwegian University of LifeScience.

15. Fjæra, S.O., Lekang, O.I. (1991) Oksygenforbruk etterhandtering for tre ulike fiskestørrelser. ITF-rapport nr.20. Norwegian University of Life Science (in Norwe-gian, English summary).

16. Fjæra, S.O., Lekang, O.I. (1991) Additiv stressresponsved trenging, transport i fiskeskrue, sortering ogvaksinering av smolt. ITF-rapport nr. 21. NorwegianUniversity of Life Science (in Norwegian, Englishsummary).

17. Barton, B.A. (2000) Stress. In: Encyclopedia of aquaculture (ed. R.R. Sickney). John Wiley & Sons.

18. Sniesko, S.O. (1974) The effects of environmentalstress on outbreaks of infectious diseases of fishes.Journal of Fish Biology, 6: 197–208.

19. Pickering, A.D. (1993) Growth and stress in fish pro-duction. Aquaculture, 111: 51–63.

20. Wendelaar Bonga, S.E. (1997) The stress response infish. Physiological Review, 77: 591–625.

21. Jentoft, S., Aastvit, A.H., Torjesen, P.A., Andersen, Ø.(2005) Effects of stress on growth, cortisol andglucose levels in non-domesticated Eurasian perch(Perca fluviatilis) and domesticated rainbow trout(Oncorhynchus mykiss). Comparative Biochemistryand Physiology, 141: 353–358.

22. Skjervold, P.O. (2002) Live chilling and pre-rigor fil-leting of salmonids: technology affecting physiologyand product quality. Dr.Agric. thesis. Norwegian Uni-versity of Life Science.

23. Pottinger, T.G., Moran, T.A., Morgan, J.A.W. (1994)Primary and secondary indices of stress in theprogeny of rainbow trout (Onchorhynchus mykiss)selected for high and low responsiveness to stress.Journal of Fish Biology, 44: 149–163.

24. Pottinger, T.G., Carrick, T.R. (1999) Modification ofplasma cortisol response to stress in rainbow trout by

Figure 17.25 Use of stimuli to attract the fish throughthe grading grid.

selctive breeding. General and ComparativeEndocrinology, 116: 122–132.

25. Gjedrem, T. (2005) Selection and breeding programsin aquaculture. Springer-Verlag.

26. Fjæra, S.O., Lekang, O.I. (1991) Effekt av ulikt håv-materiale. ITF-rapport nr. 19, Norwegian Universityof Life Science (in Norwegian, English summary).

27. Ravn Larsen, H.I. (1990) Rørtransport av fisk. Nor-wegian University of Life Science (in Norwegian).

28. Holm, M., Knutsson, S. (1977) Sorteringsforsøk medlaksesmolt. Norsk Fiskeoppdrett, 5: 4–6 (in Norwe-gian).

29. Lekang, O.I., Fjæra, S.O. (1995) Effect of light condi-tions on voluntary fish transport. Aquacultural Engi-neering, 14: 101–106.

30. Lekang, O.I., Fjæra, S.O., Thommassen, J.M. (1996)Voluntary fish transport in land based fish farms.Aquacultural Engineering, 15: 13–25.

31. Hara, T.J. (1973) Olfactory responses to amino acidsin Rainbow trout (Salmo gairneri). Comparative Bio-chemistry and Physiology, 44: 407–416.

32. Døving, K.B., Selset, R., Thomesen, G. (1980) Olfac-tory sensitivity to bile acids in salmonid fishes. ActaPhysiological Scandinavica, 108: 23–31.

33. Mearns, K.J. (1986) Sensitivity of brown trout (Salmotrutta L.) and Atlantic salmon (Salmo salar L.) fry toamino acids at the start of exogenous feeding. Aqua-culture, 55: 191–200.

34. Ness, G., Lekang, O.I., Mearns, C., Skjervold, P.O.(1991) Som man lokker på fisken får man svar. NorskFiskeoppdrett, 9: 24–25 (in Norwegian).

35. Claussen, O. (1992) Kanalforhold i landbaserte

anlegg. Norwegian University of Life Science (in Nor-wegian).

36. Øiestad, V., Pedersen, T., Folkvord, A., Bjordal, Å.,Kvenseth, P.G. (1986) Automatic feeding and har-vesting of juvenile Atlantic cod (Gadus morhua L.)in a pond. Modeling, Identification and Control, 8:39–46.

37. Ivanov, V. (1988) Tame trout: Soviet scientist developnew method. Freshwater Catch, 35: 12–13.

38. Lekang, O.I., Skjervold, P.O., Fjæra, S.O. (1991) Nymetode for interntransport av fisk. Norsk Fiskeopp-drett, 2: 30–31 (in Norwegian).

39. Ness, G. (1990) Selvsortering av fisk ved hjelp avlokemidler. Norwegian University of Life Science (inNorwegian).

40. Midling, K.Ø., Øiestad, V. (1987) Fjordranching withconditioned cod, ICES-CM F: 29. InternationalCouncil for Exploration of the Sea.

41. Gessel, M.H., Farr, W.E., Long, C.V. (1985) Under-water separation of juvenile salmonids by size.Marine Fish Review, 3: 38–42.

42. Fjæra, S.O., Skogesal, G. (1993) Sub surface size-grading of fish. In: Fish farming technology. Proceed-ings of the first international confernce on fish farmingtechnology (eds H. Reinertsen, L.A. Dahle, L.Jørgensen, K. Tvinnereim). A.A. Balkema.

43. Hovda, J. (1992) Effekt av økende tetthet på sorter-ing av laks. Master thesis. Norwegian University ofLife Science (in Norwegian).

44. Lekang, O.I., Skjervold, P.O. (1990) Selvsorteringsme-toder for fisk. ITF notat nr 28/90. Norwegian Univer-sity of Life Science (in Norwegian).


18Transport of Live Fish

hand. Several methods are employed for transportof fish, and a survey of those used follows.

18.2 Preparation for transportSince the duration of transport is normally quitelong, several hours at least, it is important toprepare the fish beforehand. The fish must be ingood conditions and should be starved before beingtransported to empty the stomach and digestivesystem and hence reduce the release of waste meta-bolic products that cause the quality of the trans-port water to deteriorate. Starving the fish will alsoreduce the metabolic rate and hence oxygen con-sumption and secretion of ammonia and carbondioxide during transport. The length of the starva-tion period needed depends on the water tempera-ture and the fish species, but is 24 hours or more.

Internal transport, size grading and other han-dling must be carried out in anticipation of trans-port; the fish must be fully recovered from the stressthat these actions involve before being transported.Loading the fish into the transport container mustbe performed in a manner that affects the fish aslittle as possible and minimizes stress.After loadingit is also important to expose the fish to as littlestress as possible to keep recovery times short.

The water used in the transport equipment musthave similar characteristics to the water quality thatthe fish are used to so that they are not exposed toa new stressor. Variations between the temperatureof the farming water and that of the transport watermust also be avoided to minimize stress to the fish.3

Normally a reduced water temperature during fishtransport is beneficial because the metabolic rate,and hence oxygen consumption and the release of

18.1 IntroductionBecause juvenile farms, on-growing farms andslaughterhouses can be located in different places,it is necessary to transport live fish and otheraquatic organisms. Live fish may be transported asfry or juveniles to on-growing farms, and the adultfish may be transported to the slaughterhouse.There is also some transport of fry and juvenile fishassociated with restocking in the wild. Transport offish can be classified as external transport (normallyknown as transport) and internal transport of fishinside the farm area (see Chapter 17). The dif-ferences are in the distance and duration of thetransport.

The equipment used for transporting fry/juvenileand adult fish is similar in design. The main differ-ence concerns the size of the tank, which must beof sufficient volume, and fitted with large enoughhatches and/or valves for filling and tapping out the fish.

All procedures will vary depending on thespecies to be transported. However, all transportwill result in extra stress for the fish, possiblyleading to death;1,2 this will not necessarily occurduring transport, but can do so over several daysafter transport. Good preparation before transportand good routines during transport and receptionare therefore important. There may also be gov-ernment regulations concerning the transport oflive fish and other aquatic animals, based on animalwelfare needs. Acceptable fish densities andrequirements for adding new water, or waterexchange, serve as examples.These regulations mayalso include requirements for design of the trans-port equipment, so this must be checked before-

256

metabolic products such as ammonia and carbondioxide, will be reduced. Furthermore, the contentof available oxygen in the water will increase.However, this may stress the fish unnecessarily,and some warm water species will also die if thetemperature is too low. Whether it is appropriatefor the transport water to be chilled by adding ice,for instance, must be checked with regard to thetransported species.

Fish may be more vulnerable during some lifestages than others; for instance, transport of salmonsmolt during smoltification, when the scales areloose. Special care must be taken when transport-ing fish in such situations.

18.3 Land transportLand transport with trucks or smaller vehicles iscommonly used for live fish and also for otheraquatic organisms. The method is especially suit-able for fry, juvenile and small sized species (<1kg),because the weight is limited. When moving largeadult fish, the total cargo weight of the vehicle is alimiting factor. The same is the case for the size ofthe fish transport tanks.

18.3.1 Land vehicles

There are several types of land vehicle used for fishtransport, including vans, wagons and trucks. Nor-mally the vans and trucks have two axles, but largertrucks with several axles and trucks with a trailer orsemi trailer may also be used. Calculations haveshown that the unit transport cost decreases withincreasing size of truck: in going from a two to threeaxle truck, and a tank volume of 4m3 to one of 8m3, the trucking costs will increase by 30% whilethe transport volume is doubled.4 A good journeyon the truck is an advantage for both the fish andthe equipment.

Fish transport trucks can be completely special-ized. Combined trucks that are also used for otherpurposes, such as general cargo, may also be used;the advantage with the second option is that itmight be difficult to keep the truck fully booked allyear round with fish transport alone.

18.3.2 The tank

Two main types of tank are used for fish transport(Fig. 18.1): either a separate removable tank is

placed on a platform on the truck, or the tank canbe attached directly to the truck chassis. Removabletanks are normally prefabricated in sizes rangingfrom 200 to 4000 l capacity. Fibreglass is commonlyused for tanks, but stainless steel and aluminiummay also be used. If the truck is to be used for otherpurposes the fish tanks can easily be removed. If thetank is specially designed to be fixed directly to thetruck chassis more time will be needed to removeit before the truck can be used for other purposes.Fixed tanks are normally made of stainless steel oraluminium; inside they are divided into separatecompartments. The principle is the same as thatused for tanks for transporting other liquids such asmilk or fuel. These tanks are also divided into sep-arate compartments inside, even if this is not appar-ent from the outside.

Transport of Live Fish 257

Figure 18.1 Trucks for fish transport either haveremovable tanks placed on a platform body or largertanks fixed directly to the chassis frame of the truck(lower photograph).


The fish are loaded through hatches on the topof the tanks and unloaded through special tappingvalves in the bottom. It is important that there aresmooth connections so the fish are not wounded.Toavoid temperature variations during transporttanks may be constructed with a double wall withinsulation in between. However, if insulated cabinettrucks are used for transport of single tanks, this isnot necessary.

Regular cleaning and disinfection of the trans-port equipment is necessary; this requires the tankto be designed in a way that makes it easy to clean,with smooth surfaces for instance. ‘Dead areas’where cleaning is difficult must be avoided. Theinner tank surface area must tolerate standard dis-infection liquids. The material must be of a closedstructure to avoid water being forced into it so thatlater disinfection is difficult.

18.3.3 Supply of oxygen

The fish consume oxygen during transportation and additional oxygen must be supplied if the transport lasts for some time. The amount ofoxygen needed can easily be calculated by lookingat the available oxygen in relation to the consump-tion by the fish. Oxygen is usually supplied frombottles attached to the truck and is added to thewater in the fish tanks through diffusers which lieon the tank bottom. Air may also be used as anoxygen source for the fish; in this case the truckmust be equipped with an air blower. If transport-ing fish at high density and only adding air throughdiffusers in the bottom of the tank, large numbersof bubbles in the water can result, which is not rec-ommended because it will stress the fish. Whenadding air, supersaturation with nitrogen must beavoided (see Chapter 8). It is important to get agood distribution of the added air or oxygenthroughout the tank;5 an airlift pump may be usedto create a flow inside the tank and hence animproved distribution.

When fish are transported for long periods oftime (>12h), depending on the density, problemsmay result from accumulation of carbon dioxide inthe tanks. Airlift pumps may be used to aerate thewater and will also remove some of the excesscarbon dioxide. They will also cause the water tocirculate inside the tank (Fig. 18.2). When usingsuch a pump, the air is supplied via a perpendicu-

lar closed pipe which runs from a few centimetresabove the tank bottom to a few centimetres abovethe water surface. The addition of air to the waterinside the pipe creates bubbles which will drag thewater towards the surface and in this way generatea water flow inside the tank. The air blower is fixedto the frame of the truck. To get proper water cir-culation, an airlift pump can be placed near thecentre of the tank; however, it must not inhibitfilling of the tanks with fish through the hatches onthe top.

Near the bottom of the tank the airlift pump willcreate an area of low pressure and the fish may get stuck to the inlet of the perpendicular pipe.To avoid this, a wide perforated screen should be

Figure 18.2 An airlift pump may be used to addoxygen and at the same time remove excess carbondioxide. The photograph shows the installation in atransport tank.

placed around the inlet to reduce the water veloc-ity and possibilities for the fish to get stuck.

When using the airlift pump it is important to beaware that when carbon dioxide is removed fromthe water there will be an increase in pH and hencean increase in the concentration of ammonia rela-tive to ammonium ion. Ammonia is rather moretoxic to the fish than ionic ammonium and couldcause problems in high concentrations. The reasonfor this is that the equilibrium NH3 ª NH4

+ is pHdependent (see Chapter 9).

18.3.4 Changing the water

In addition to high concentrations of carbondioxide it is possible to get high concentrations of ammonia (total ammonia nitrogen, TAN)depending on fish density, water temperature, trans-ported species and duration of transport. To avoidfish mortalities it is therefore necessary to have the following:

• Proper water exchange routines• A system for cleaning and re-use of the transport

water or dosing with additives.

Few trucks are equipped with cleaning and waterre-use systems, and few operators use additives (seebelow), so water exchange is therefore the mostcommon method. Because of this fish transportmust be properly planned, taking into considera-tion the duration of the trip, the oxygen require-ments of the fish and the necessity of waterexchange. When exchanging water, the incomingwater must be of satisfactory quality, so that a stablewater environment is maintained for the fish. Toavoid additional stress it must have a similar qualityto that of the production water. When exchangingwater, mixed zone problems must be taken intoconsideration,6 In addition the water must notcontain any micro-organisms which are harmful to the transported fish. The outlet water must betreated and released in a secure place, because itmight contain micro-organisms that could harm thenatural fish stocks in the area. It is therefore impor-tant to know the government regulations regardingintake and release of transport water.

The length of time during which fish can be trans-ported without water exchange depends on watertemperature, fish density and tolerance of the fishspecies to the decreasing water quality. The water

quality gradually deteriorates as a result of fishmetabolism until it is toxic to the fish. Normal waterexchange intervals are 10–16h, but it is possible tohave longer intervals between water exchanges: lowwater temperature will reduce the metabolic rateand therefore the fish may be transported for alonger period without water exchange.

18.3.5 Density

The density of the fish during transport is of impor-tance for transport economy. Hence the amount ofwater transported must be as low as possible inrelation to the amount of fish. An accurate limit ofthe correct density for transport is difficult to cal-culate and may vary greatly among species, size anddevelopment stage. For instance, Tilapia can betransported at higher densities than rainbow trout.If the water temperature increases, the density mustnormally be reduced. For salmon, which is a speciesthat has very high requirements regarding waterquality, a rough estimate says that for temperaturesbelow 8°C a density of up to 90–100kg/m3 may besupported;7,8 if the temperature is above 8°C, adensity of up to 50kg/m3 is recommended. If thetemperature is very high, transport is difficult,because both the oxygen requirements of the fishwill increase and the oxygen content in the waterwill decrease. Another way to express density is asthe percentage of fish in the total transport watervolume. At a water temperature of 18°C, densitiesfor catfish of up to 38%, Tilapia 32%, carp 15% anddrum 4% have been reported for transports ofduration 8–16h for fish of various sizes.9

Some insurance companies have their own guide-lines regarding density of fish under transport.These guidelines must be known before trans-portation, if the fish are to be insured. There mayalso be government regulations governing fishdensity.

18.3.6 Instrumentation and stopping procedures

The temperature and oxygen level should be mon-itored during fish transportation to control andavoid critical situations. Therefore the truck mustpossess the necessary instruments such as a ther-mometer and an oxygen meter. These may besimple portable devices or fixed so that they can beread straight from the driver’s cab.An alarm system



may also be installed to detect and respond to vari-ations in these parameters.

When transporting fish one should not, however,depend entirely on the instruments installed in thetank. Visual control is also necessary. Programmedstops are recommended for observation of fishbehaviour. The frequency of these stops may vary:a stop quite early in the trip is recommended, forinstance 15min after the start, and subsequentlyeach hour or every two hours thereafter.

It is necessary to keep a logbook during thejourney in which parameter values and observa-tions are registered. It is also necessary to recordfish behaviour and any special occurrences duringtransportation. If something happens to the fishduring or directly after transportation, the logbookcould be used to find a possible explanation and toclaim on the insurance policy. The economic valueof the transported fish can be quite high: if trans-porting 100g salmon smolt, each valued at 1 €, in a10m3 tank with a fish density of 50kg/m3, the totalvalue of the cargo will be 5000 €.

18.4 Sea transportLarge quantities of fish can be effectively trans-ported by sea. This may be carried out by specialboats, well boats or by hauling the sea cages orother floating installations with boats. Compared to

trucks on shore, the advantage with using well boatsis that the well is much larger than the truck tanksso the amount of transported fish can be increasedand therefore the costs per fish transportedreduced. Larger tanks are also better for larger fish,for instance for transporting fish to the slaughter-house because the available volume for the fish toswim in is greater.

The fish can also be transported by hauling thenet cage or another closed construction. However,the recommended drag velocity is quite low; this isof course also species dependent, but is never rec-ommended to be above 1 or 2 knots. Such veloci-ties will also impose large forces on the draggedconstructions.

18.4.1 Well boats

Boats used for live fish transport are equipped withwells (internal tanks) for holding the fish (Fig. 18.3).Well boats have for many decades been used fortransporting various species, such as herring andpollack, from traditional fisheries. The same boatsmay today be used for transporting farmed fish.Special systems for loading and unloading the fish may be installed. The normal size of well boatsvaries from 20 to 60m and they may accommodateup to 150t of live fish.10 Traditional V-shaped hullsare usually used, but catamaran types that are faster

Figure 18.3 Well boat for live fishtransport.

have also been tried. The wheel house is placedeither in the bow or the stern. On new boats it istypically placed in the bow.

18.4.2 The well

In the well boat there is a constant flow of newwater through the well to ensure enough oxygen issupplied to the fish and the waste products areremoved. Special well valves are placed in the frontand in the back of the well (Fig. 18.4). If the boat ismoving these valves are opened and new water willflow through the well from front to back. The wellvalves are operated directly from the bridge. If thewell boat is in dock, large circulation pumps ensurethe renewal of water in the well and by this thesupply of necessary oxygen to the fish and removalof waste products.

When the well boat is transporting fish it may benecessary to close the well valves in some areas(special zones) and continue with closed valves.Reasons for this are to avoid possible transfer ofdisease either from surroundings to the fish in thewell, or from the fish in the well to the surround-ings. Special zones can be areas near fish farms orimportant natural fish stocks. On these occasions it is necessary to ensure that the fish get enoughoxygen even if the circulation pumps are inopera-tive. An oxygen supply system or air blower istherefore necessary, together with the necessarydistribution system, for example diffusers located inthe bottom of the well. Most modern well boats are

equipped with systems for adding oxygen and maycarry the fish for a long period of time withoutopening the well valves.The wells function similarlyto those used on truck transport. Similarly the tankscan be equipped with systems for individual circu-lation of the water and also water purification (re-use) systems. The well boat may also beequipped with a refrigerated seawater (RSW)system for cooling the seawater, to keep a low tem-perature in the well and also to chill the live fishbefore slaughter.

Well boat size commonly varies from 50 to 1000m3. Inside the well is normally divided intoseveral tanks separated by fixed walls. The con-struction and surface of the sidewalls and bottommust be designed not to damage the fish. Use of nets for dividing the well into additional com-partments is therefore not satisfactory. It is alsoimportant that the well is designed in a way thatensures good water distribution and exchange ofthe total water volume. Well boats with circulartanks have also been constructed.The well will thenfunction in the same way as a circular tank with thesame flow pattern, and a very good distribution ofthe incoming water is ensured.

As for tanks on trucks, it is important that thewells are easy to clean and disinfect. It is alsoimportant to have control of the real well volumewhen adding disinfectant.11

18.4.3 Density

As for truck transport, the transport density willvary with the species, but here it is also importantfrom an economic perspective. For Atlantic salmon,the density is normally around 35–50kg/m3, and foradult fish to slaughterhouse densities of between150 and 180kg/m3 have been reported.12 Whentransporting large amounts of fish to the processingplant, the cargoes are very valuable. A well capac-ity of 1000m3 and a fish density of 150kg/m3 meanthat 150t of fish is being transported. If the price is3 €/kg the total value of the cargo is 450000 €. Thisalso shows the necessity of control to avoid accidents.

18.4.4 Instrumentation

Instrumentation to control water temperature,oxygen content and salinity, when transporting in


Figure 18.4 Valves that are open in the front and backof the wells ensure water flow through the well duringtransport.


seawater, is essential in well boats. In all boat trans-port it is important to have proper records and tokeep a logbook. This can either be done manuallyor in a computer log with automatic data-logging ofthe parameters.

When transporting fish by well boat, specialattention must be given to variations in tempera-ture and salinity in the sea because these changesmay be very stressful for the fish. Large variationsin temperature may occur when there are large sup-plies of freshwater, for example close to the outletof large rivers. Variation in water quality duringtransport is not good; for instance, there may bereduced salinity just outside large rivers and alsolarge organic burdens so that a mixing zone withnegative water quality may occur.6,13

Boat transport in rough weather must also beavoided. The fish in the well boat may suffer fromsome kind of seasickness. Some species will attemptto compensate for the wave pressure and try to stayat the same depth in the well by taking air in andout of the swim bladder. This may exhaust the fishtotally and they will sink to the bottom of the wellwhere they will lie, sliding about on the bottom ofthe well so incurring wounds and losing scales.Therefore it is important to listen to the weatherforecast before transport, and record the weatherconditions and approximate wave height.

All well boats have some type of equipment forloading and unloading the fish. Some sort of fishpump, for instance an ejector or vacuum–pressurepump, is normally used today, but a wet net mayalso be employed (Fig. 18.5). There have also beensome specially designed boats that allow the fish tomove in and out of the well voluntarily. This is pos-sible by lifting the stern so that the well is opened.Sliding walls can be used to crowd the fish withoutpumping the water out from the well and reducingthe water level. It is thus possible to reduce thevolume in the well by moving one wall.

Many well boats are also equipped with gradingequipment so that they can be used to grade fish insea cages in addition to transporting fish.

18.5 Air transportLive fish may also be transported by helicopter orby aircraft. Because of the high cost this option is seldom used. Reasons for using air transport are mainly remote geographical conditions or bad

infrastructure (roads, bridges), that make sea orland transport difficult and expensive. For restock-ing of lakes transport of live fish with small seaplanes is quite normal, but because of the lowloading capacity (some hundreds of kilograms) thisis not very useful in commercial aquaculture.Larger multi-engine aircraft can be used; in suchcases seaplanes are equipped with large internaltanks for storing the fish, and pure oxygen gas isused. However, it is fry or small fish that are mostoften transported by air.

The use of helicopters is quite a new method forlive fish transport. In one method, a large barrel of400–500 l capacity filled with fish and water hangson a wire below the helicopter.14,15 This is basicallythe same equipment as when using helicopters tofight fire. To fill the barrel, the helicopter lowers itand the fish and water are lifted in; to empty thebarrel, the helicopter lowers it into the water and

Figure 18.5 Detail of the deck on a well boat with thewell hatch and wet net for lifting the fish.

the fish are released by tilting the barrel. The heli-copter is maintained in flight throughout the entireloading and unloading procedure.This method maybe a solution for fry/juvenile transport in areas withlow quality roads and where the geographical con-ditions made long distance sea transport necessary.

18.6 Other transport methodsPlastic bags may be used to transport smallamounts of fish, for instance to transport fish forlake cultivation and for transport of ornamentalfish. A common size of transport bags is 85cm ×60cm. Thick polyethylene (PE) is used to make thebags resistant to puncturing. The bag is normallyfilled with between one-quarter and one-third of itsvolume with water before the fish are loaded. Airis then removed from the bag by pressing ittogether, which is then filled with pure oxygen untilit assumes a balloon like form. The atmosphereover the water in the bag will then be pure oxygen.Cooling of the plastic bag minimizes the oxygenconsumption of the fish and maximizes the durationof oxygen supply. Using ice on top of the bag, andnot exposing the bag to direct sunlight may helpwith this. During transportation it is recommendedthat the bag is shaken from time to time to mix theoxygen that is in the atmosphere above into thewater. The duration of the journey using thismethod depends directly on the fish density andtheir oxygen consumption. For example, a densityof 1kg salmon smolt per 3–4 l of water and 10 l ofpure oxygen in plastic bags at 5°C has been usedfor air transportation for a couple of hours.16

An alternative to plastic bags is plastic contain-ers. Here also the containers are filled with approx-imately one-third of water and fish, above whichthere is a pure oxygen atmosphere. Special equip-ment is required to remove the air from the can andreplace it with pure oxygen.

Special types of tractor-trailers designed for fishtransport can be used for short journeys (Fig. 18.6).Goods wagons equipped with tanks for live fishtransport have also been tried for railway transport.17

18.7 Cleaning and re-use of waterIn some situations it can be necessary to transportlive fish over large distances and for longer periods

of time (12–16h). In these circumstances there mustbe either a water exchange or continuous flow-through of water as in well boats, depending of thespecies being transported. It is, however, also pos-sible to design a transport tank system that includesa circuit for cleaning and re-using the water (Fig.18.7) so that high concentrations of carbon dioxideand ammonia in the water can be avoided. Suchtransport tanks really function as a recycling plantwith 100% re-use of water. Complete removal ofmetabolic waste products is impossible from aneconomic view, so even here the water quality willgradually decrease, but over a much longer timeperiod. Compared to a re-use plant, the degenera-tion of the water quality will be slower because fishrespiration is reduced due to prior starvation.

Airlift pumps may be used for removal of carbondioxide in addition to adding oxygen and creatingwater flow. Biofilters are not efficient systems forremoval of TAN during transport because these arebiodynamic systems that require a start-up time toestablish a culture of bacteria before use. However,an ion exchanger can be used to remove ammonia.When the water passes through the ion exchangerthe ammonia is removed (NH4

+ exchanged withNa+); after use the ion exchanger must be regener-ated, and this can be done between journeys.

Long journeys and high degrees of water circu-lation may also cause the water temperature toincrease as a result of the pumping required forwater re-use and fish metabolism. Installation of acooling system on the transport tank can thereforebe advantageous. The oxygen supplied to the water


Figure 18.6 Using a tractor trailer for fish transport.


may be either produced by generators or by bottlesof oxygen gas attached to the truck.

In Norway salmon smolt and juvenile turbothave been transported for up to 5 days withoutwater exchange using a specially designed tank witha water re-use system.18

18.8 Use of additivesMany additives are available that can be added tothe transport water to improve the transport resultsand allow fish transport over longer periods.

Instead of using filters and water re-use systems,additives can be applied to the water to give similareffects.Addition of seawater to freshwater will havea positive pH regulating effect because of its buffer-ing capacity. The same will be the case with addi-tion of other buffering substances. Sodium chloridecan be added to freshwater as a defoaming agent.Addition of clinoptilolite to the water (as used inan ion exchanger) will reduce the concentration ofammonia.

Addition of salt may reduce osmoregulatory dis-turbance, and by this the total stress when trans-porting freshwater fish.19,20 Antibiotic can be addedto reduce the development of bacteria in the trans-ported water.21

Addition of sedatives, such as MS222 and cloveoil, to the transport water may also be done to calmthe fish down and reduce the metabolic rate.22–26

This will reduce oxygen consumption and decreasethe excess of carbon dioxide and ammonium ion;the fish will also tolerate more stress.

All use of additives will, however, require accu-rate control of the water quality during transport;they may not always function or the effect may beminor. It is best therefore normally to avoid using additives. Adding of antibiotics and sedativesought to be avoided unless there are very specialconditions pertaining to the live transport. Thesemay not be allowed before the fish go to process-ing plants because they may leave residues in theflesh.

References1. Barton, B.A., Haukenes, A.H., Parsons, B.G., Reed,

J.R. (2003) Plasma cortisol and chloride stressresponse in juvenile walleyes during capture, trans-port, and stocking procedures. North AmericanJournal of Aquaculture, 65: 210–219.

2. Iversen, M., Finstad, B., McKinley, R.S., Eliassen,R.A., Carlsen, C.T., Evjen, T. (2005) Stress responsesin Atlantic salmon (Salmo salar L.) during commer-cial well boat transport, and effects on survival aftertransfer to sea. Aquaculture, 243: 373–382.

3. Strange, R.J., Schreck, C.B., Golden, J.T. (1977) Cor-ticoid stress responses to handling and temperaturein salmonids. Transactions of the American FisheriesSociety, 106: 213–218.

4. Wahlberg, B. (1977) Aktuelle fisktransportsfrågor.Rapport Vattenfall (in Swedish).

5. Børjesson, H. (1987) Synspunkter for trans-portutrustning for fisk. In: Fisktransport handbok.Mittnordekommitten før vattenbruk (in Swedish).

6. Krogelund, F., Teien, H-C., Rosseland, B.O., Salbu, B.(2001) Time and pH-dependent detoxification of aluminium in mixing zones between acid and non-acid rivers. Water, Air and Soil Pollution, 130:905–910.

Figure 18.7 It is possible to design atransport tank system for long journeysthat includes a circuit for cleaning andre-using the water.

7. Mittnordenkomitten før vattenbruk (1987) Fiske-transporthandboken. Mittnordenkomitten før vatten-bruk (in Swedish).

8. Pursher, J., Forteath, N. (2003) Salmonids. In: Aqua-culture, farming aquatic animals and plants (eds J.S.Lucas, P.C. Southgate). Fishing News Books, Black-well Publishing.

9. Johnsen, S.K. (2000) Live transport. In: Encyclopediaof aquaculture (ed. R.R. Sickney). John Wiley & Sons.

10. Willougby. S. (1999) Manual of salmonid farming.Fishing News Books, Blackwell Publishing.

11. Johnsen, S., Simolin, P. (2002) Volumbergning av van-nmengde i brønn og rørsystem i brønnbåt ved bruk avsporstoff. VESO rapport (in Norwegian).

12. NIVA (2003) Transportkvalitet av fisk I brønnbåt.Prosjektfakta, NIVA (in Norwegian).

13. Rosseland, B.O., Blakar, I.A., Bulger, A., Kroglund,F., Kvellestad,A., Lydersen, E., Oughton, D.H., Salbu,B., Staurnes, M., Vogt, R. (1992) The mixing zonesbetween limed and acidic river waters: complex alu-minium chemistry and extreme toxicity for salmonids.Environmental Pollution, 78: 3–8.

14. Sheperd, B.G., Bérézay, G.F. (1987) Fish transporttechniques in common use at salmonid enhanchementfacilities in British Columbia. Canadian manuscriptreport of fisheries and aquatic sciences 1946. Cana-dian Department of Fisheries and Oceans.

15. Anon. (1988) Up, up and away. Air-lifting smolts is ajob for helicopter specialists. Fish Farmer, 6: 37.

16. Gjedrem, T. (1986) Fiskeoppdrett med framtid. Land-bruksforlaget (in Norwegian).

17. Berka, R. (1986) The transport of live fish: a review.EIFAC Technical Report 48, FAO.

18. Lekang, O.I. (1992) Avansert levendefisktransport.Nordisk Akvakultur, 6: 32–33 (in Norwegian).

19. Carneiro, P.C.F., Urbinati, E.C. (2001) Salt as a stressresponse mitigator of matrinxã; Brucon cephalus(Günter) during transport. Aquaculture Research, 32:297–304.

20. Johnson, D.L., Metcalf, M.T. (1982) Causes and con-trols of freshwater drum mortality during transport.Transactions of the American Fisheries Society, 111:58–62.

21. Amed, D.F., Croy, T.R., Beverly, A.G., Johnson, K.A.,McCarthy, D.H. (1982) Transportation of fish inclosed systems. Transactions of the American FisheriesSociety, 111: 603–611.

22. Mishra, B.K., Kumar, D., Mishra, R. (1983) Observa-tions on the use of carbon acid anaesthesia in fish frytransport. Aquaculture, 33: 405–408.

23. Prinsloo, J.F., Schoonebee, H.J. (1985) Note on pro-cedures for the large scale transportation of juvenilefish. Water SA, 11: 215–218.


25. Swanson, C., Mager, R.C., Doroshow, S.I., Cech, Jr.,J.J. (1996) Use of salt, anaesthetics and polymers tominimize handling and transport mortality in deltasmelt. Transactions of the American Fisheries Society,125: 326–329.

26. Cooke, S.J., Suski, C.D., Ostrand, K.G., Tufts, B.L.,Wahl, D.H. (2004) Behaviour and physiologicalassessment of low concentration of clove oil anaes-thetic for handling and transport of largemouth bass (Micropterus salmoides). Aquaculture, 239:509–529.


19Instrumentation and Monitoring

can be more damaging than if no measuring equip-ment were used at all. This implies that mainte-nance and running costs must be included in theprice of an instrument, not just the purchase cost.Extra effort must be given to maintenance ofinstruments used to monitor water quality.This alsoincludes frequent calibration according to the man-ufacturer’s instructions so that the values shown arereliable. Depending on the type of instrument, thesensors may have a limited duration, so must beexchanged at fixed intervals.

Measurement of biological performance has alsoincreased during the past few years as a result ofthe increased focus on profitability in intensive fishfarming. By automatically measuring developmentin terms of weight and total fish biomass, it is pos-sible to control the development and intervene ifsomething does not correspond to the productionplans.

Due to the large expense involved and theamount of technical equipment that can fail, it isincreasingly common to have a total monitoringsystem on the farm, which also includes a significantuse of computer tools.1–4 On land-based farms usingpumps for the water supply or in farms with re-useof water such systems are essential.5

There is much general literature availabledescribing measurements, instruments and sensors(see, for example, refs 6–8). This chapter gives ageneral description of the construction of instru-ments, with some thorough investigation of thoseused in aquaculture. In addition there is a briefreview of methods used to count fish, measure fishsize and estimate biomass. The chapter ends with adescription of monitoring systems of which mea-suring instruments constitute a major part.

19.1 Introduction

Equipment for measuring and recording of variousparameters is more and more commonly used inaquaculture, especially in intensive aquaculture.Such equipment controls and adjusts the environ-mental conditions to obtain optimal productionresults. Until now several of the measurementshave been taken manually, which is normally moretime consuming and labour intensive, and thereforeresults in fewer measurements. During the past fewyears, there has been rapid development in theautomation of instruments and monitoring systemsthat can also be used in the aquaculture industry,mainly based on developments in electronics andcomputer science. Therefore many of the trivialmanual measurements are now carried out by spe-cially designed instruments, releasing manpowerfor more important intellectual tasks and toimprove the production results, especially in inten-sive aquaculture.

One reason for using instruments is to automatethe management of fish farming as much as possi-ble. For example, video cameras and image analy-sis can be used to monitor fish and give alarmsignals if odd behaviour is observed. The biologicalprocesses underlying fish production are, however,both complex and difficult, unlike the production ofnails. Even with today’s knowledge, it is only adream to believe that it is possible to fully replacethe fish farmer with instruments and robots.

When buying and installing instruments, therequirements for maintenance and calibration,adjusted for special circumstances must be takeninto account. The values read from the measuringequipment must be reliable; otherwise the result

266

19.2 Construction of measuring instrumentsThe construction of measuring instrumentsdepends on the measuring principles used and thesignal transfer. One classification is mechanical,hydraulic, pneumatic, electrical and electronic,where the last is being increasingly used.9

A measuring instrument often comprises threemajor parts (Fig. 19.1):

(1) A sensor or probe(2) A transmitter to transfer the signal(3) A display or another type of indicator (con-

nected to the transmitter).

In some instruments the three major parts are con-nected within the same unit, while in other instru-ments the parts are separate and connected viacables for electric signals or another principle fortransfer of the measured values. Measuring equip-ment can either give continuous signals (ana-logues), or on/off signals (digital). An example ofthe first case is an oxygen meter that shows the con-centration of oxygen. Flow indicators that registerif there is water flow or not (on/off) are an exampleof the latter case.

The sensor in the unit is used to record the phys-ical conditions in the medium, such as the probe inan oxygen meter. The transmitter can either beelectrical or mechanical and translates the signalcoming from the sensor to a scaled signal that as isfurther transported to the display or indicatorwhere the results are shown in an understandableway. In the display the physical conditions of themedium are shown. Signals may also go directly to a recording unit such as a computer for

storing the results, or can be used to control a regulator.

A short description of the measuring instrumentsmost used in aquaculture facilities is given below.Equipment is separated into that used for measur-ing water quality and that used to measure physi-cal conditions.

19.3 Instruments for measuring water qualityBoth chemical and physical parameters are used tocharacterize water quality, and many instrumentsbased on different principles are used. Instrumentsfor measuring water quality can be divided into on-line and off-line instruments based on their con-struction. On-line instruments that can stay in thewater carry out continuous monitoring and arealways a type of electrode or sensor. The sameinstrument may, however, also be used off-line forindividual measurements. Off-line instruments aremore closely related to those in laboratories, andnormally more work is needed to perform an analy-sis, together with experience in performing labora-tory work in some cases.

On-line instruments or chemical analysers can beseparated into those that use sensors but noreagents, and wet-chemistry analysers.10 Use ofinstruments with a sensor is advised, because theyreduce the manual input. Wet-chemistry analysiscan be classified on the principle of the analyses:typical examples are colorimetric, titrimetric and ion-selective electrodes. Colorimetric measure-ments are much used, and function by adding areagent to the water sample and monitoring thecolour change with a detector. A typical detector isa spectrophotometer that measures the lightabsorption of the sample. The great disadvantagewith wet-chemistry analysers is that a sample fromthe water flow needs to be taken to which a reagentis added. Stopped flow/batch mode analysers andflow injection analysers are available.10 In the latterpumps and valves ensure that samples are auto-matically taken from the water flow and trans-ported to the instrument for monitoring beforebeing released so that a new sample can be taken.Some common devices for water analyses aredescribed briefly below; more information can beobtained from the literature, for example, refs 2 and 9.

Instrumentation and Monitoring 267

Figure 19.1 An electronic instrument normally com-prises three major components either linked by cablesor integrated.


19.3.1 Measuring temperature

In fish farming, temperature needs to be measuredin several situations, for instance, the farming wateror inside slaughtered fish. One principle utilized tomeasure temperature is the expansion of certainsubstances with temperature as in a mercury ther-mometer. The construction of the thermometer issuch that even quite a small expansion of themercury causes a noticeable change in the level inthe thin pipe where the reading is taken. Ther-mometer manufactures calibrate the scale accord-ing to known values: a bath with freshwater and iceis 0°C at normal atmospheric pressure and is theboiling point of freshwater is 100°C under the sameconditions. The Celsius scale divides the intervalbetween 0 and 100 into equal degrees. Other sub-stances besides mercury can be used in thermome-ters using this principle.

That the electrical resistance in materials changeswith temperature may also be used to measuringtemperature. What actually is measured is the elec-tric current going through a circuit of which the resis-tance material is an integral part; if such material isexposed to temperature variation its resistance willchange together with the electric current passingthrough the circuit. Platinum is often used in digitalthermometers which measure electrical resistance.

The difference in voltage occurring between twodifferent materials may also be used for measuringtemperature. If two different metals are solderedtogether, for instance copper and constantan, acurrent will flow, the size of which will be propor-tional to the temperature. If the point where thetwo materials are soldered together is affected by atemperature gradient, the electrical current willalso vary. This equipment is called a thermocouple;it is simple but not as accurate as other devices.

A thermistor is also a temperature dependantresistance, but is more complex and includes semi-conductor technology. Higher accuracy is achievedwith this principle than with the other methodsdescribed. The device is simple and the price rea-sonable; thermistors are therefore widely used fortemperature measurement.

19.3.2 Measuring oxygen content of the water

The oxygen content may be measured either chem-ically or electronically. The normal chemical

method is the so-called Winkler method which is atitration method, consisting of adding certain chem-icals to the water and observing a colour changewhich will be directly related to the oxygen concentration.

When measuring oxygen concentration electron-ically, a sensor or probe is used. This can be constructed using a positively charged conductor(anode) and a negatively charged conductor(cathode) separated by an insulator (Fig. 19.2). Inone design, an electrolyte is located around theelectrodes (anode and cathode). A special mem-brane covers the electrolyte, keeping it in place andprotecting it. The membrane is permeable tooxygen molecules. As there is a positive and a neg-ative electrode, electrolysis will occur, and electronswill pass between the two electrodes. The magni-tude of the electron transport is affected by theamount of oxygen in the electrolyte; this again

Figure 19.2 Measuring oxygen content: detail of thesensor for measuring oxygen content.

depends on the amount of oxygen transportedthrough the membrane, which is related to theamount of dissolved oxygen in the water.

It is important to calibrate this type of instrumentaccurately. For some instruments it is necessary tocalibrate for air pressure and temperature, while inothers this is integrated. The electrolyte and the membrane or the complete sensor, must bechanged at defined intervals. Since the membraneis very thin it is easy to break, care must be takenwhen handling it. Ageing of the membrane can bedetected, because it becomes difficult to get stableresults and the values fluctuate constantly. Themembrane is also exposed to fouling, and must reg-ularly be visually inspected and cleaned. If foulingoccurs the oxygen values will normally drop in rela-tion to the correct values.

19.3.3 Measuring pH

Several methods are used to measure pH to preventit dropping to critical values. A simple but inaccu-rate way is to observe the colour change of pHpaper, also known as litmus paper, when it is dippedinto water. This paper contains a chemical, thecolour of which is pH dependent. Similarly, a smallwater sample can be taken, a specially preparedliquid added to it and the colour change observed.

The pH can also be measured by a standardchemical analysis with titration and observing thechange in colour, which is pH dependent. Thissystem may be implemented for daily measure-ments on a fish farm and has the advantages ofrequiring low maintenance and being reliable. Thedisadvantages are, however, that this takes sometime and experience of working with chemicalequipment, and measurements in a laboratory arenecessary.

A pH meter is constructed on the same principleas the oxygen meter. Here the probe is constructedwith two electrodes, one for measuring the concen-tration of H+ and the other as a reference electrode.In this case the pH electrode consists of a mem-brane permeable to hydrogen ions. Here also themembrane creates a cell that encloses the elec-trolyte. Between the electrodes a current will passthat depends on the concentration of H+. Thevoltage between the pH electrode and the reference electrode is then measured, and the pHcalculated.

Instruments for measuring pH must be calibratedbefore use. This is performed by placing the probein different solutions of known pH and then adjust-ing the instrument. The pH probe is compact, incontrast to the O2 probe; therefore it must be com-pletely replaced if it ceases to function. A disad-vantage of the pH probe is its limited duration,normally from 3 months up to 1 year in specialcases. Practical experience with the use of pHmeters in fish farming shows that maintenance is ofextreme importance. Both calibration and changingof electrodes must be done regularly to achieve reli-able measurements.

19.3.4 Measuring conductivity and salinity

Conductivity is a measure of the ability of water toconduct an electric current. In fish farms this isimportant in order to evaluate the ability of thewater to inhibit pH fluctuations, i.e. the bufferingcapacity. In seawater, Na+ and Cl− ions dominateand here the instrument is used to measure thesalinity.

The probe consists of two electrodes and islowered into the water. A small electric potential(voltage) is applied across the electrodes. An elec-tric current will occur between the electrodes, thesize of which depends on the ion concentration inthe water. To prevent the establishment of a layeron the electrodes which affects the current, it is nec-essary to use an alternating current as pre-voltage.

Conductivity is affected by temperature, so it isimportant to compensate for this parameter whentaking measurements. Each instrument should havea special table setting out the effect of temperatureon the conductivity. Advanced instruments incor-porate automatic temperature compensation.

19.3.5 Measuring total gas pressure and nitrogen saturation

The total gas pressure in the water is measuredmainly to find not only the total pressure, but alsothe amount and saturation of dissolved nitrogen gas(N2). If the saturation of nitrogen in the water isabove 100%, the fish may suffer from gas bubbledisease. This is more critical in fry stage fish than in adult fish. In salmonids problems have beenobserved when saturation is over 102%, but it isrecommended that saturation be maintained below



100.5%. Marine fish fry has been shown to be verysensitive for supersaturation of nitrogen. Problemsmay also occur if the total gas pressure is too highand there are some indications that above 100%total pressure may be detrimental.

One method to measure the total gas pressure inthe water is to use a saturometer (saturation meter)(Fig. 19.3). The main part of the instrument is asmall silicon tube into which the dissolved gasesfrom the water pass (the silicon acts as a mem-brane) and become enclosed. A pressure meter isattached to the tube and the measurement iscarried out by determining the difference in pres-sure between the local atmospheric pressure andthe pressure inside the silicon tube.

Before using the saturometer, the pressure insidethe cylinder must be equalized to the surrounding

environmental pressure. This is carried out byplacing the probe in the air for some minutes beforeplacing it in the water. A perforated cover sur-rounds the cylinders and a manual pump is used toensure water flow past the cylinder. The instrumenthas to stay in the water for 5–10 minutes before areading can be taken.

In order to determine the total gas pressure whenusing a saturometer the following equation must beemployed:

Where:

TGP = total gas pressure of dissolved gasesBP = local barometric pressure (normally read in

mmHg (mercury))ΔP = pressure difference (read from the saturom-

eter) between total gas pressure in the waterand local barometric pressure (mmHg).

When calculating the nitrogen pressure, which isa critical factor for avoiding gas bubble disease, thefollowing equation may be used:

N2 (%) = [BP + ΔP − ((O2/bO2) × 0.5318− PH2O)/((BP − PH2O) × 0.7902)] × 100

Where:

N2 = partial pressure of nitrogen gas in the water(percentage nitrogen saturation)

BP = local barometric pressure (mmHg)ΔP = difference between total gas pressure and

local barometric pressure, measured by asaturometer (mmHg)

O2 = oxygen concentration in the water measuredby an oxygen meter (mg/L)

bO2 = Bunsen’s coefficient for oxygen (see Appendix 8.2)

PH2O = partial pressure of the water vapour (mmHg)

The two numbers in the equation are conversionfactors.

There are also other instruments available formonitoring the nitrogen gas saturation which aresimpler to use.

19.3.6 Other

As a result of the rapid development of instrumentsfor online measurements, sensors for online mea-

TGP %BP

BP( ) = +⎛

⎝⎞⎠ ×ΔP

100

Figure 19.3 A saturometer used for measuring thetotal gas pressure in the water.

suring of carbon dioxide, ammonia and nitrate areavailable. The sensors can either be single, or multiinstruments comprising several sensors connectedin a multiprobe (Fig. 19.4). The sensors in the multiprobe can be changed with sensors for otherquality parameters, but the instrument remains thesame.

Instruments using traditional chemical labora-tory methods have also been developed for useunder field conditions; for instance there are specialtypes for aquaculture facilities. One commonlyused instrument is based on a spectrophotometer,in which the sample is illuminated by light of a specific wavelength and the amount of light passing through is monitored. Before putting thesample in the spectrophotometer a chemical isadded and so that a colour change will occur,the size of which will depend on the amount of substance to be measured in the sample. To avoidmuch work with weighing out of chemicals speciallyprepared ampoules, each with the correct amountof chemical for one water sample, are deliveredwith the instrument. Many chemical analyses can berun in such instruments which are quite simple to use.

19.4 Instruments for measuringphysical conditionsThere are many places were there is a need tocontrol physical conditions. In aquaculture, thewater condition is of major importance, particularlythe following factors: water flow, water level andwater pressure. Many methods can be used tomeasure these parameters; some can also be usedto measure more than one. There are differencesboth in price and accuracy of available instruments.Methods used are reviewed below; for more infor-mation see, for instance, refs 9 and 11.

19.4.1 Measuring the water flow

It is a common practice to measure water flow tobe sure that it is constant and that the correctamount of water passes through the pipes. Waterflow measurements can be carried out at variousplaces in a fish farm. Flow meters may be locatedin the main inlet pipe, in part flow pipes or in theinlet pipe to a single tank. Several principles areactually used, such water as velocity and head loss.The methods used to measured the flow in openchannels and pipes are different. The main empha-sis here is on measurement in pipes that are full ofwater, since this is the most common situation infish farming. Some of the methods presented may,however, also be used in open channels.

Measuring water velocity

The water velocity is proportional to water flow(see Chapter 2). If the water velocity and the inter-nal diameter of the pipe are known, it is easy to cal-culate the water flow.

A simple way to measure the water velocity in apipe is to use a propeller, paddle wheel or turbine(Fig. 19.5). Propellers with a variety of designs areused. The working principle is that the water willmove the propeller and the rotational velocity ofthe propeller will reflect the water velocity in thepipe. Either the propeller can be installed in anexisting pipe system, known as an inset meter, or itcan be a completely separate system in which thepropeller and instruments are connected, forming a complete unit adapted to the pipe. Propellersystems are simple and inexpensive, but one of thedisadvantages is that the head loss will increase.


Figure 19.4 Sensors connected in a multiprobe.


However, the main disadvantage is that the pro-peller will be exposed to fouling because it is withinthe water flow. Normally, fouling will reduce themeasured velocity compared with the actual veloc-ity. Correct maintenance is therefore important insuch a system.A further disadvantage is wear of thecontinuously moving propeller.

An electromagnetic flow meter can also be usedto measure water velocity (Fig. 19.5). The principleutilized here is that an electromagnetic fieldchanges when the water velocity varies. A conduct-

ing fluid, such as water, in a magnetic field willinduce an electric voltage, which correlates with thewater velocity. The water must have some electricalconductivity if this type of instrument is to function;in fish farming this will always be the case.

Ultrasound waves may also be used for measur-ing the water flow, as in an ultrasonic flow meter(Fig. 19.5). Two combined ultrasound source andreceiver units are placed, one on each side of thepipe and crossed diagonally. Sound waves are sentfrom one side and travel with the water flow; from

A

C BB

Figure 19.5 Instruments for measuring water velocity: (A) propeller; (B) ultrasound; (C) electromagnetic.

the opposite side of the pipe another sound wave issent against direction of the water flow. The timestaken for the sound waves to reach the receivers arecompared; the water velocity can be calculatedbecause the water flow deflects the sound wavesand alters their velocity. A fixed unidirectionalultrasonic system can also be used which calculatesthe time difference during which the sound wavestravel. The advantage of this system is that theinstrument may be attached to the exterior of exist-ing pipes; if the type of pipe is known the water flowinside can be found. Flow meters utilizing this prin-ciple will not give any head loss and are very accu-rate; however, they are quite expensive. Theinstruments may also be affected by air bubbles orhigh particle concentrations.

The various types of flow meter have differentaccuracies, and this must be checked before selec-tion. Generally, greater accuracy incurs greaterexpenditure, so the meter must be fit for purpose.

Measuring head loss

Head loss in a pipe will occur when the water passesan obstruction in the pipe or any other object thatreduce the water flow. Measurement of the headloss may therefore be used to calculate the watervelocity, since the head loss depends on the veloc-ity; an increase in velocity will increase the headloss. When this principle is used for flow measure-ments, a known obstruction is set inside the pipe. Avery accurate plate with a hole in the centre, slightlysmaller than the internal pipe diameter, known asan orifice plate, is usually used (Fig. 19.6). The dif-ference in the pressure before and after the plategives an indirect measurement of the water flow.This instrument is also called a differential pressureflow meter. The physical relation used when mea-suring the head loss is defined by the Bernoulliequation:

Where:

p = pressurer = densityv = velocityh = elevation (height)g = acceleration due to gravity.

p v gh+ + =12

2r r constant

The disadvantage with this method is that a headloss occurs, especially with the orifice plate, soinstead a venturi can be used which functions in thesame way, but the head loss is reduced.

Another instrument based on measuring the dif-ferential pressure (actually the head loss) to calcu-late water flow is the pitot tube which measures the dynamic head (total head) and the static head(Fig. 19.6). Based on the difference between thesetwo values, the water velocity can be calculated(from the Bernoulli equation). Correct use of this is important because there is a velocity pro-file inside the pipe that must be taken into consideration when locating the pitot tube.

A rotameter or hover velocity meter also utilizesthe head loss measurement, but with a constanthead loss. The construction of a rotameter includesa conical glass or plastic pipe, with a floating deviceinside (Fig. 19.6); this device is lifted when the waterstarts to flow. Since the pipe is conical the height bywhich the floating device is lifted will depend on themagnitude of the water flow, because the areawhere the waters is flowing is increased. A normalrotameter must be placed vertically to function. Ina similar instrument the device is set on a springwith known characteristics, and this is compressedwhen the water flows by an amount that dependson the water flow. Vertical installation is not neces-sary for this flow meter.

19.4.2 Measuring water pressure

Water pressure is measured to control water levelsin tanks or the pressure in pipes. If the pressure istoo high or low a warning signal can be given.

Diaphragm manometers are often used tomeasure the water pressure. The principle is thatpressure can alter the shape of a diaphragm (Fig.19.7). An increase in pressure will change the formof the diaphragm and this change can be used tocontrol an indicator. The manometer is fixeddirectly to the pipe where the measurements aretaken.

A bourdon tube manometer may also be used forpressure measurements (Fig. 19.7).This system con-sists of a hollow bent pipe that is connected directlyto the place where the pressure is measured. Whenthe pressure increases, the bent pipe will try tostraighten as the pressure is considerably higher onthe outside of the bend than on the inside because



the area where the force acts is larger; this move-ment will be registered on a scale calibrated toshow the pressure. There are also other devicesusing the same principle, such as pressure capsulesand bellows.

Today, however, the most commonly used instru-ments are electronic load cells or pressure trans-ducers. They function because there is a smallchange in the electric resistance in an electric circuitrelated to the pressure. A pressure transducer maybe installed on the bottom of a tank or inside a pipe.

19.4.3 Measuring water level

Measuring the water level is necessary at variousplaces to avoid overflows and water shortages. This

can be carried out in a head tank or directly in theindividual fish tanks. Water level sensors are nor-mally digital, sending a signal if the level of thewater is below or above defined values. The systemmay also be analogue; this generates a signal that isproportional to the water level.

Different types of electronic floats may be usedfor controlling the water level. The float will lie onthe surface and change position if the water levelvaries; this variation can be utilized to control thelevel. When this happens a signal will be sent andaction taken, for instance switching the watersupply on and off. A commonly used float is thelevel rocking sensor (Fig. 19.8) which floats on thesurface; if the water level decreases it will graduallybecome more upright. The sensor is located in the

A B

CFigure 19.6 Instruments for measuring head loss: (A) orifice plate; (B) pitottube; (C) rotameter.

Figure 19.7 Membrane and bourdon manometersused to measure water pressure.

end of a cable and when it is in the floating posi-tion there may or may not be contact between thetwo conductors located inside the float. If thesensor is hanging vertically an automatic switch willtake the opposite position (opened or closed as thecase may be) to make or break the electric circuit.This can then be used to control the water level. Onsubmerged pumps this is a common method forstarting and stopping the pumping, depending onthe water level.

Other methods are also used to measure thewater level in fish farms, of which pressure sensors,as described above, are the most common.

Capacitance sensors are becoming popular formonitoring water levels (Fig. 19.9). These sensorsdetect how well a material keeps its electricalcharge. Water may keep 80 times as much electricalcharge compared to air and the sensor can there-fore register changes in electrical charges associ-ated with changes in water level. If the tank wall isthin, the sensor may be placed on the outside of thetank wall and will be capable of sensing whether thetank is filled with water or not. If the water leveldecreases to beneath the position of the sensor, anelectrical circuit will be either opened or closed anda signal given.

Water level can be controlled very accurately by ultrasound devices A transmitter and receiverare placed above the water surface. By transmittinga sound wave and measuring the time of reflectionwith the receiver, it is possible to calculate the distance from the transmitter to the water surfaceand hence the water level. If the water leveldecreases, the time taken by the sound wave totravel between the transmitter and receiver willincrease.

19.5 Equipment for counting fish,measuring fish size and estimation oftotal biomass

19.5.1 Counting fish

To count fish manually on large aquaculture farms is very labour intensive; for example, a juve-nile production plant delivering 1 million juvenilesto an on-growing plant. Therefore it is beneficial todo this automatically. The problem when countingfish automatically is to separate the individuals and to avoid two or more fish coming together and


Figure 19.8 A floating sensor used for monitoring andcontrolling the water level.


Figure 19.9 Capacitance sensors used for measuringwater level (A) in a pipe and (B) in a tank.

being counted as one. One challenge when devel-oping equipment for counting fish is therefore tofind a way to separate the fish on an individualbasis.

Counting can be done when the fish are in or outof the water. Self-counting of the fish when they areput into water has been very difficult in practicebecause the fish are reluctant to pass any obstaclevoluntarily (see Chapter 17). Some of the equip-ment used for counting fish can also be used for measuring the size of the fish (see section19.5.2).

A suitable place to count the fish is in connectionwith internal transport or grading. Here the fish arenormally taken out of the water and it is quite easyto count them and carry out the other operations,most of which will also require counting to controlthe number of fish.

One fairly cheap and simple method used to sepa-rate the individual fish is to let them slide in aconvex V-shaped channel (Fig. 19.10). By having aconvex channel the velocity of the sliding fish willgradually increase. This will separate the individ-uals and it is now quite easy to count the single fish,for instance with the use of light sensitive cells. Thelight diodes create a beam that is broken by thesliding fish, which is thus counted.

Counting of the fish in water is normally done inconnection with pumping or transport throughpipes. Some kind of camera is commonly employed,either a video camera or a linear camera.

During the past few years camera technology hasdeveloped rapidly together with the use of imageanalysis, which has also been utilized for countingfish (Fig. 19.11). The challenge when using camerasis to achieve good pictures, with sufficient contrastto the background because black and white imagesare commonly used. A chamber with proper lightconditions is therefore required. This may beincluded on the pump pipe as a separate unit. Toavoid two fish coming together and being countedas one, image analysis, either with two cameras orone camera and a mirror to give two pictures, isemployed. This system may also be used to calcu-late the size of the fish (see section 19.5.2) but thenother algorithms are used.

The difference in conductivity between waterand fish may also be used to count them, but here also it will be difficult to separate the individuals.

A

B

19.5.2 Measuring fish size and total fish biomass

It is no longer necessary to lift fish out of the waterand weigh them to obtain a traditional weightsample, or to lift them into a calibrated tank anduse Archimedes’ law to calculate the weight of thefish group.Archimedes law can be used because thefish will displace a volume of water equal to theirown volume. If it is assumed that 1 l of water weighs1kg of water the volume displaced in litres will beequal to the weight of the fish in kilograms. Byhaving a scale on the tank wall the increase in waterlevel can be monitored and in this way the biomassfound. Equipment for measuring fish size and totalbiomass is useful for production planning and

control, and also for deciding when to size gradeand for planning harvesting.

Today, various equipment is available for mea-suring fish size automatically whilst they remain inthe water.The equipment used for automatic count-ing and weighing of the fish can be separated in twodepending on whether: the fish stay in or are takenout of the production unit. The first category com-prises equipment that is lowered into the produc-tion unit and does not disturb the behaviour of the fish; the second normally comprises measuringequipment connected to a fish pump.

Measurements where the fish stay in the production unit

Several systems have been developed, mainly forsea cages, to measure either the single fish size orthe total biomass of the fish in the cage, or a com-bination of the two. The reason for the interest inmeasuring in sea cages is due to the large amountof fish and high feed consumption. Good control of fish weight and development are therefore


Figure 19.10 Counting fish with a convex V-shapedchannel and light cells.

Figure 19.11 A video camera in a chamber used forcounting fish and estimating their size.


important. In a cage it is impossible to controltotally environmental factors that affect theappetite of the fish. Neither is it possible to foreseeexactly changes in environmental factors (i.e. theweather) for prognosis.

Two principal methods dominate12:

• Remote sensing by hydroacoustics or submergedcameras

• Detailed measurements obtained by leading thefish through a sensor for individual registration,e.g. by optical, impedance, capacitance or ultra-sounds techniques

One commonly used technique based on thesecond method is a measuring frame13. This is a rec-tangular frame that is lowered into the cage. In thewalls of the frame there are a number of lightdiodes placed in rows so they create a linear camera(Fig. 19.12). When the fish swim through the framethe light diodes create a shadow picture. Using acomputer, the approximate size of the fish is calcu-lated, based on image analysis of the shadowpicture. The system is based on a random sample offish voluntarily swimming through the frame duringa given period. This will occur in cages because thefish are swimming voluntarily around in the cage.The number of passages through the frame dependson the fish density. At fish densities between 10 and20kg/m3 a typical swimming frequency is 30–200fish per hour13. If two fish pass through the frametogether, the image analyses will remove the

picture as not being useful. Only the passageswhere a single fish passes through the frame and agood picture is achieved are used to calculate theaverage fish size. This is done based on a side viewof the fish when it passes through the frame. Theside area of the fish is highly correlated with its size.Based on the large number of measurements onsingle fish, the average fish weight with a standarddeviation can be calculated for the total cage.Comparison of these data with the actual weightachieved at the slaughter house has shown goodcorrelation.The standard deviation gives importantinformation about the size variation of the fish inthe cage and can be used to tell when size gradingis necessary. If the average size of fish is known, itis possible to calculate the total biomass providedthat the number of fish in the cage is known. Whenusing such a system for calculating the totalbiomass, it is of course very important to accountfor the dead fish, and there must be no unregisteredescape of fish from the cage. To use the frame it isnecessary that the fish swim voluntarily around inthe production unit; it is not possible to use thesystem in fish tanks with a water flow, because herethe fish will stay still in the water flow.

Another method based on camera technology forfinding the size of fish in a cage utilizes a stereo-video system14,15. Two video cameras sitting aboveeach other in a rack with a known distance betweenthem are used (Fig. 19.13). By taking pictures of thefish, with both cameras, it is possible to calculate

Figure 19.12 A measuring frameused to determine fish size in seacages13.

the distance from each camera to every point on thefish; image analysis will give the length and heightof the fish; there is a high correlation between thesemeasurements and the weight of the fish. Experi-ments using three cameras simultaneously to countfish in cages have also been performed.16

Video cameras like those used for counting fishcan also be used for biomass estimation.17,18 Thissystem is quite large for use in cages and it may bedifficult to get the fish to swim through the unit orchannel voluntarily. When using such a unit in con-nection with pumping, this does not present anyproblem because here the fish are moved by force.By taking a picture from the side when it passes thevideo camera and calculating the area of the fish,its weight can be found.

Hydroacoustics is a totally different principleused for measuring total fish biomass in a sea cagewhich employs an echo sounder.19 This is actuallythe same method as used in traditional fishing tofind where and at what depth the fish are swimming.It also gives some information about the amount offish. An echo sounder contains a source that emitssound pulses at fixed intervals and a transducerwhich is normally placed below the cage. Thesource sends the sound pulses up to the watersurface in the cage; this is the opposite direction tothe echo sounders used for fishery which sendsound pulses down to the bottom. When the soundpulse hits a fish it is reflected. A receiver fixed nearthe sound source beyond the cage receives thisreflected signal (echo), which is travelling in theopposite direction to the sound pulse that was sentout (Fig. 19.14). The sound pulses are very short (1ms is typical) and separated by longer intervals

during which the receiver is listening for echoesbefore the next sound pulse is sent out. Thereforeseveral echoes are received over the course of aminute. The size of the echo depends on the size ofthe individual fish or fish shoal. By using an echosounder it is also possible to obtain informationabout the behaviour of the fish and the depth in thecage at which they are swimming under variousenvironmental conditions. It is often the swimbladder of the fish that is the source of the echoesregistered by the echo sounder; a good echo is


Figure 19.13 Use of two cameras and stereoscopy forcalculating fish size.

Figure 19.14 The principle used in an echo sounderfor finding fish in the sea may also be used for findingthe total biomass in a sea cage, but it is used in theopposite way with the receiver being placed under thecage.


generated here because the swim bladder is filledwith air. Echo sounders are not used to measure thesize of single fish with high accuracy because themagnitude of the echo is highly dependent on thedirection of the swim bladder in relation to thesound source and will vary if the fish is swimmingupwards, downwards or horizontally. For the wholefish group it will, however, give quite a good estimate. Echo sounders cannot so easily be used in tanks because sound reflections from the tank bottom and walls will interfere with themeasurements.

Measurements where the fish are removed fromthe production unit

It is possible to install a video camera in a pipe usedfor pumping fish or an ordinary pipe.This takes pic-tures from the side and eventually above, and cal-culates the fish size from image analysis of thepictures (Fig. 19.15). The system is the same asdescribed earlier for counting fish, the only differ-ence being the image analysis program used to cal-culate fish size.

19.6 Monitoring systemsOn an aquaculture facility it is necessary to controlall the different factors that are involved in themanagement of the farm. This can for instance be awater quality factor such as the concentration of

oxygen, ensuring sufficient water flow, or thecorrect water level in a head basin. Of course thiscan be done with manual measurements and obser-vations, but is very time-consuming if done withhigh regularity. How regularly measurements mustbe taken depends on the economic consequences ifsomething fails. In intensive fish farming with highstock density and much use of technical installa-tions that can fail, the need for regular measure-ment and control is obvious. The time available for remedial action when something fails is alsolimited. Systems for automatic control are there-fore used based on economic and practical conse-quences, especially in intensive fish farming. Toillustrate this a land-based marine fish farm thatrequires pumping to supply water to the fish can beused as an example. The importance of continuouscontrol of the water flow into the fish tanks isobvious.

A monitoring system comprises three majorcomponents (Fig. 19.16):

• Sensors and measuring equipment which controlthe conditions

• Monitoring centre which receives signals fromthe sensors and measuring equipment, interpretsthem and eventually sends out alarm signals orsignals to regulators

• Equipment for warning when something is failingand emergency equipment that is started andstopped by regulators.

Figure 19.15 A well boat equippedwith a system for automatically measuring fish size when they arepumped.

Signals are transferred between the components.These are normally electrical signals via cables;wireless connections however, may also be used. Amonitoring centre is always equipped with a batteryback-up so that it can always function indepen-dently whether or not there is an electricity supply.

19.6.1 Sensors and measuring equipment

To control the conditions a number of sensorswhich function as monitoring points in the alarmsystem can be used.20 Sensors can be used to controlwater level, water flows or water quality parame-ters, all described earlier in this chapter. Sensors aremuch used also to indicate whether the electricitysupply is connected or not. The number of factorsto be controlled determines the size of the moni-toring system. From the sensors or measuringequipment electrical signals are sent to the moni-toring centre: these can be digital signals, i.e. currentor not current, for example from a level sensorwhere an electric circuit is broken and the signal tothe monitoring centre changes from current to nocurrent. The signal may also be analogues whichmeans that the current coming into the monitoringcentre varies depending on the value read from thesensor; an example of such an instrument is anoxygen meter.A current corresponding to the meas-ured oxygen level is sent to the monitoring centre;

thus if the oxygen concentration decreases, lesscurrent flows. Normally this standardizes lowcurrent signals of between 4 and 20mA.

19.6.2 Monitoring centre

The construction of the monitoring centre thatreceives the signals, interprets them and eventuallysends out signals to the warning equipment, dependson the complexity of the system. Even though themonitoring centre is usually built especially foraquaculture facilities, it is based either on a pro-grammable logic controller (PLC) or a computer.

A PLC can be visualized as an electronicallyoperated switch system. It includes a number ofinput channels and a number of output channels.Electric signals from the sensors come through theinput channels. The electrical output signals (i.e.how the switches are functioning) can be pro-grammed based on the input signals. Programmingof the PLC is done using an external keyboard con-nected to the ‘brain’ of the PLC. Whether outputsignals will be sent, and through which of the outputchannels, depends on the input signals and the loca-tion of the switch system. In and out signals can bedigital or analogue, or most commonly both.

To illustrate the functioning of a PLC, take asimple monitoring system on a farm as an example.The farm equipment includes a pump that deliverswater to a head tank from which it flows undergravity to the fish tanks. Pure oxygen gas from anoxygen bottle can be added to the single fish tanksthrough diffusers at the bottom of the tanks. Thefollowing measuring instruments are included: alevel sensor in the head tank and an oxygen sensorin each of the fish tanks. In addition there is areserve pump on stand-by, a magnetic valve thatcontrols the emergency oxygen supply, and a sirenavailable on the farm.

The programme for the PLC might be as follows.If the water level in the head tank is too low a signalwill go from the level sensor to the PLC. The PLCis then programmed to send an output signal to thereserve pump to start it. If the oxygen level in thetank is too low, the value input to the PLC is undera programmed value and a signal is sent outthrough the channel that controls the magneticvalve on the oxygen gas bottle. This will changefrom the closed to the open position. In addition, asignal is sent to the channel that starts the siren.


Figure 19.16 A monitoring system comprises threemajor parts: the sensors, the monitoring centre and thewarning equipment.


If a computer is used as a monitoring centre it isnormally equipped with special cards known as I/Ocards or data acquisition cards (DAQ) that make it possible to import signals, including analoguesignals, and send out the same types of signals asthe PLC. By using a computer the input signals caneasily be recorded and stored. It is also possible toget a hard copy of the time points for alarms, includ-ing which of the sensors registered the failure. It isalso possible to get a picture of the farm, includingthe sensors and their individual status, on the com-puter screen, which makes visual control of opera-tions easier (Fig. 19.17). Today, both PLC and PCare used in combined systems.

19.6.3 Warning equipment

From the monitoring centre electric signals are sentthrough the output channels to external warningequipment. The warning equipment is installed onthe farm and normally includes both equipmentthat creates light and sound, such as warning lightsand sirens. A telephone may also be included in the system. This dials fixed programmed phonenumbers, among others to the watchman’s mobilephone. On more advanced systems, a messagesaying what is wrong appears on the telephonedisplay.

Figure 19.17 By using a PC as themonitoring centre it is easy to obtaina picture of the farm and the status ofeach sensor on the computer screen.

Figure 19.18 A generator for providing electrical powerin the event of a power failure on the farm.

19.6.4 Regulation equipment

The signals may also be used to start emergencyequipment or to regulate the functioning of equip-ment. For instance, in the event of a power failure,a generator will start (Fig. 19.8). The equipmentmay also open emergency oxygen supplies to thefish tanks or start stand-by pumps. Often there is acombination of starting emergency equipment andstarting external warning equipment.

19.6.5 Maintenance and control

For a monitoring system to be reliable it is veryimportant to carry out proper maintenance andtesting. The maintenance normally includes testingand calibration of the sensors. It is advisable to havea fixed routine for testing the monitoring system,for instance once a week. The sensors are thentested to ensure that the signals being emitted andalso that the signals are being sent from the centreto the warning equipment. Emergency systems such as electric generators and systems for addingoxygen must also be tested at fixed intervals, to besure that they will function if a real emergency sit-uation occurs.

References1. Lee, P.G. (1995) A review of automatic control

systems for aquaculture and design criteria for their implementation. Aquacultural Engineering, 14:205–227.

2. Hochheimer, J. (1999) Equipments and controls. In:Wheaton, F. (ed.) CIGR handbook of agriculturalengineering, part II. Aquaculture engineering, pp.281–307.American Society of Agricultural Engineers.

3. Ernest, D., Nath, S. (2000) Computer tools for siting,designing and managing aquaculture facilities. Aqua-cultural Engineering, 23: 1–78.

4. Lee, P.G. (2000) Process control and artificial intelli-gence software for aquaculture. Aquacultural Engi-neering, 23: 13–36.

5. Ebeling, J.M. (1994) Monitoring and control. In:Aquaculture water reuse systems: engineering designand management (eds M.B. Timmons, T.M. Losorod).Elsevier Science.

6. Webster, J.G. (1998) The measurement, instrumenta-tion and sensors handbook. CRC Press.

7. Anderson, N.A. (1997) Instrumentation for processmeasurement and control. CRC Press.

8. Morris, A.S. (2001) Measurement and instrumentationprinciples. Butterworth-Heinemann.

9. Montgomery, J.M. (1985) Water treatment, principlesand design. John Wiley & Sons.

10. Gibbs, R. (1990) Advances in on-line monitoring. In:Aquaculture and water quality (eds D.E. Brune, J.R.Thomasso) Advances in world aquaculture, vol. 3.World Aquaculture Society.

11. Wheaton, F. (1977) Aquacultural engineering. JohnWiley and Sons.

12. Heyerdahl, P.H. (1995) Biomass estimation, a power-ful tool in fish farming management. In: Quality inaquaculture, Special publication no. 23. EuropeanAquaculture Society.

13. Heyerdahl, P.H. (1995) Optical biomass estimation, atechnological approach. In: Quality in aquaculture,Special publication no. 23. European AquacultureSociety.

14. Nailberg, A., Petrell, R.J., Savage, C.R., Neufeld, T.P.(1993) A non-invasive fish assessment method for tanks and sea cages using stereo video. In: Pro-ceedings of an Aquaculture Engineering Conference,Spokane, Washington. American Society of Agricul-tural Engineers.

15. Harvey, E., Cappo, M., Shortis, M., Robson, S.,Buchanan, J., Speare, P. (2003) The accuracy and pre-cision of underwater measurement of length andmaximum body depth of southern bluefin tuna(Thunnus maccoyii) with a stereo-video camerasystem. Aquacultural Engineering, 63: 315–326.

16. Petrell, R.J., Neufeld, T.P., Savage, C.R. (1993) Avideo method for noninvasively counting fish in seacages. In: Proceedings of an Aquaculture EngineeringConference, Spokane, Washington. American Societyof Agricultural Engineers.

17. Boyle, W.A., Ásgeirsson, Á., Pigott, G.M. (1993)Advances in the development of computer vision fishbiomass measurement procedure for use in aquacul-ture. In: Proceedings of an Aquaculture EngineeringConference, Spokane, Washington. American Societyof Agricultural Engineers.

18. Eide, T. (1997) Videoregistrering av fisk i elv. Intern-publikasjon, Norwegian University of Life Science,(in Norwegian).

19. Dunn, M., Dallard, K. (1993) Observing behaviourand growth using the Simrad FCM 160 fish cagesystem. In: Fish farming technology (eds H. Reinert-sen, L.A. Dahle, L. Jørgensen, K. Tvinnereim),pp. 269–274. A.A. Balkema.



20Buildings and Superstructures

insulated or uninsulated, and can be built forvarious lifetimes, either short or long.

The simplest superstructure is a shadow net. Thisis used to prevent the fish getting sunburned and toinhibit solar heating of the water. A slightly moreadvanced construction results when plastic sheet-ing or tarpaulin is substituted for the shadow netand a tent is created. Normally a framework of steelpipes keeps this upright. The tent can either be forone tank or larger for the complete farm. On singletanks a hemispherical shaped construction is com-monly used. It is also possible to have an insulatedtent made of two layers of plastic with insulation inbetween. A bowed shape is a cheap construction touse here.

A typical building is, however, based on weather-tight walls and a roof made of fixed materials.The duration of such buildings is much longer thanof the simpler tent constructions. Buildings on thefarm may have one or several storeys. If there arepossibilities for utilizing the terrain and hencegravity for internal transport of fish and feed, it willbe advantageous. For instance, if small fish can begrown on the first floor and on-growing takes placein the ground floor, tapping of the fish to the on-growing department is possible, which gives a simplesystem for internal transport of fish. The feedersmay also be filled from the first floor if they hang inthe roof of the ground floor over the fish tanks.

20.2.2 Shape

Buildings may have different shapes. When think-ing of economy, the best is to have a square or espe-cially a rectangular shape for the building. Anglesin the ‘body of’ the building will increase the costs.

20.1 Why use buildings?Buildings or superstructure are used in aquaculturefor several reasons. It will shelter the productionfacilities from environmental factors such as wind,sun, rain or snow and the working environment forthe farmers will be improved. It also makes possiblephotomanipulation of the grown organisms, whichis an important part of optimizing the farming conditions for different species, especially in the fryand juvenile stages and during maturation.

The buildings or superstructure can be used tocover the entire farm, part of the farm, or only theproduction units. Buildings are used mostly onland-based fish farms; on sea-based farms they canbe used for protection of feed storage on a bargeand for housing the mess room and WC. In a land-based farm there will always be a building housingthe toilet, mess room and office; it might, however,be an integral part of a larger building that alsocovers the production area.

Building design and construction is a large spe-cialist field and here only a very brief survey isgiven with special attention to aquaculture facili-ties. Many textbooks are available on the subject;see, for example, refs 1–4.

20.2 Types, shape and roof design

20.2.1 Types

Several types of buildings can be used for aquacul-ture facilities. On the basis of their constructionthey can be categorized into different types rangingfrom simple superstructures with or without wallsto complete buildings (Fig. 20.1). Buildings can be

284

The reason for having angles must therefore be to improve the functionality of the building, forexample, by improving the feed or fish handling.

20.2.3 Roof design

The roof may also have different constructions. Aridge roof is the most usual, but a sloping roof mayalso be used. Normally a ridge roof will be the mosteconomic. A flat roof may also be used, but is muchmore difficult to build correctly and is especiallydifficult to get completely watertight. In areas withsnow a flat roof is not recommended; it may be nec-essary to clear the snow, which is difficult. In addi-tion the weight of the snow will press the roofdown, so a large and costly frame is necessary. Thechosen shape and roof of the building will alsodepend on available materials.

20.3 Load-carrying systemsA load-carrying system keeps the walls upright andthe roof in position; it also takes the weight of the roof which, in colder regions also includes theload from snow. The cost of the load-carryingsystem is an important part of the total cost of thebuilding.

Various constructions are used as load-carryingsystems; these are based, among other things, on thematerial used (Fig. 20.2). Either separate construc-tions can be used for the wall and the roof, or com-bined systems can be employed, where the wallsand roof are integral parts of the load-carryingsystem. In the first case the wall system carries theroof; different types are explained later.

When designing the load-carrying system, thenecessary free span, i.e. the available floor area with

Buildings and Superstructures 285

Figure 20.1 Different types of buildings used in aquaculture.

Figure 20.2 Different load-carrying systems can beused to keep the walls upright and the roof in position.

no pillars, is of great interest. The size of the freespan determines the size of the construction; alarger free span results in a larger construction. Toreduce the size of the free span, pillars or internalwalls are used to carry the weight of the roof atseveral points.The use of pillars will reduce the costof the load-carrying system, but inhibit the use ofthe building, constraining internal handling andspace utilization.

Roof constructions that cover a large free span(above 10–15m) are expensive. Long distancesbetween the outside walls, without pillars or inter-nal walls, will increase the building costs becausethe size of the roof construction is increased.

The timber lattice roof (W) truss is quite a com-monly used roof load-carrying system. It may bebuilt on site, but is normally prefabricated. A freespan distance of up to 12–13m width is easilycovered with such a construction. If the building iswider, it is recommended that pillars be used totake the weight of the roof, if the aim is to keep thecost of construction low; a load carrying internalwall will have the same effect. This is a cheap andsimple building construction.

Steel, concrete or wood beams or trussed beamsare also much used. They can be used on large freespans in excess of 20m, and even up to 100m if thebeams are bowed, as in a velodrome.Trussed beamsare favoured for large constructions. However, forthis type of construction also, the costs increaseexponentially with the free span. Longitudinalbeams of wood or steel in the building give quite asimple construction, especially when used in freespans below 20–25m; however, pillars are recom-mended if the span is longer.

In combined systems, the load-carrying systemfor the roof and walls is integrated. A much-usedconstruction here consists of large frames, either insteel or wood and often of a bowed shape, that areset into the foundations. The distance between theframes varies, but is normally in the range 3–5m, orequal to a section of the building. The rest of thewall and roof structure is also built of steel or woodand connected directly to the frames. A greatadvantage of this construction is that it is quite fastto erect, and can be used on large buildings with alarge free span. The method is fairly expensive andwill normally require cranes to put up the building.

If the tanks are large, individual superstructuresfor separate tanks can be used. A cheap construc-

tion uses a hemisphere created with a frame of steelpipes and a cover of tarpaulin.

20.4 MaterialsVarious materials are used for constructing build-ings (Fig. 20.3). Wood is simple to work with andsimple to join together. Normally it is cheap, at leastin smaller buildings and in areas with timber. It isused in walls, roofs and load-carrying constructions.To increase the strength and the length, gluedbeams can be used. Wood may also be part of con-structions, for example in chipboard panels.

Metals such as steel or aluminium are also quiteeasy to work with; pieces are either welded orscrewed together. Metals are much used in load-carrying constructions, such as beams or frame-work. Metal plates are used for covering interiorand exterior surfaces on walls and roofs.

Concrete is a widely used building material. It ismade of a mixture of sand and gravel with cementthat functions as a glue, and water.After mixing fol-lowed by some hardening time this makes a per-manent construction. The method of mixing andproportions of materials used will give concrete ofdifferent strengths.

Concrete has good compressive strength but poortensile strength. Therefore iron is used, either as rods or mats, as reinforcement in concrete toenable it to withstand tension, while the concretecan withstand compressive forces. In small con-structions, concrete is quite simple to handle andwork with, and in addition it is fairly cheap. It caneither be mixed on site or in a factory and deliveredin special trucks ready mixed but not hardened.This latter method is most normal in larger constructions.

Concrete can also be delivered as prefabricatedelements which are finished and hardened. Ele-ments, such as beams, may also be pre-stressed toincrease tolerance to higher forces without increas-ing the weight too much. Sizes of components vary from small blocks to bars to complete parts ofbuildings, such as wall elements or roof elements.Concrete may also be used for beams in load-carrying constructions.

A lightweight version of concrete is also avail-able as blocks and bars. Here the gravel is replacedwith a light material, for example, expanded clayproducts.



20.5 Prefabricate or build on site?Buildings can be built on site from the foundationsup or can be delivered as prefabricated parts thatare put together on site. Various sizes of prefabri-cated elements can be delivered; options includethe degree of finishing. These elements could becomplete parts of the walls or the roof, whereeverything is finished. The time to establish a pre-fabricated building is, of course, much less than forbuildings constructed on site. Whether prefabri-cated buildings are to be used or not depends onthe price which will vary from case to case depend-ing, among other factors, on the freight costs. Theavailable building period will also influence thefinal choice.

20.6 Insulated or not?In a cold climate, the building should be insulatedif the air inside is to be temperate. The greatestamount of heat is lost through the roof; it is also lostthrough the walls and the floor. In addition, thereare high heat losses through the windows anddoors. The thickness of the insulation depends onthe climate and winter temperature; lower wintertemperatures mean that more insulation isrequired. In walls and roofs mineral wool made of glass or rock is used for insulation; expandedpolystyrene (PS) is commonly used in the floorwhen this is of concrete. All these materials have alow k value (see Chapter 7). PS can also be used in walls or roofs providing it is covered by concrete.

A B

C D

Figure 20.3 Different materials used in buildings on aquaculture plants: (A) wood panelling; (B) metal sheets; (C)concrete; (D) lightweight concrete blocks.

This is because it can produce toxic gases if itcatches fire. Taking Norway as an example, it isnormal to use 15cm mineral wool in the walls and30cm in the roof.The PS in the floor is 5–10cm thick(equal to 20–25cm of mineral wool which has ahigher k value.

Whether insulation is to be used or not dependson the rooms and how they are utilized. The office,mess room and toilet are normal insulated in coldclimates. Whether the production rooms are insu-lated or not depends on the desire to improve theworking environment.

If the building is insulated, it will be of sealedconstruction so it will be necessary to use a venti-lation system to ensure exchange of the air inside.If the walls are not insulated and airtight, the wallconstruction can be open so that a ventilationsystem is not required. Then the wall can be simplymade with split panels or only a plastic grating;natural air exchange is thus ensured and there is noneed for ventilation.

20.7 Foundations and groundconditionsWhen starting to build, proper foundations are veryimportant to prevent part of the constructionmoving after the building is finished. Importantcomponents may break if the foundations moveunder the load from the building.

The ground conditions must be suitable for erect-ing a building. Rock, stone and sand/gravel formgood building ground, while clay and silt are not assuitable because they are less stable. In colderregions, where the ground freezes in the winterseason, clay and silt are not recommended forbuilding ground, because the frozen ground willcreate movement in the building during freezingand thawing. The ground must carry the weight ofthe building, which makes marshy areas unfit fornormal building constructions unless special pre-cautions are taken, such as having raft constructionas the foundation for the building.

Foundations are normally laid as concrete slabs.In areas with frozen ground, insulation under theslabs is necessary to avoid problems with heave.Alternatively, a ring foundation wall that goesdown to frost-free ground must be used to ensurethat the frozen ground not does affect the building.

Normally this is recommended to go below 1.5mdepth, but this varies with the depth of the frozenground. It is important to use drainage pipes toensure that water is removed from the proximity ofthe walls to avoid possible movement.

20.8 Design of major parts

20.8.1 Floors

In aquaculture facilities concrete is normally usedfor the floor (Fig. 20.4). Iron mats or rods are usedto reinforce the concrete.The normal slab thicknessis 10–15cm, but it might be greater under the wallsand pillars because the floor needs to withstand theadditional weight transferred from the wall orpillars. In cold climates is it necessary to use PS asinsulation under the floor. To avoid frost heaveeither flank insulation or a ring foundation can beused for the building. If using flank insulation, it islaid under the surface, extending 1.5–2m outsidethe flanks of the building. In this way frost inhibitfrom going down in the ground close to the build-ing; it will go some distance out, so does not reachinto the building and the problem with frozenground is eliminated. If using a ring foundation thisgoes down to below the frozen ground and ensuresthat the building remains independent of the frozenground. The thickness of a ring foundation wall is10–20cm, and typical depths are 1.5–2m belowground level.

In the production room, at least, it is necessary to have a smooth surface on the floor that is easyto clean. It is normal to seal the floor with some


Figure 20.4 Normal method for constructing a con-crete floor in a production room.


type of epoxy two component paint, especially suitable for concrete surfaces, to increase thesmoothness of the floor so that it is easier to clean. However, it must not be too slippery; in thewalkways, sand grains can be included in the epoxypaint.

To keep the floor clean and dry, it is important tohave sufficient slope on the floor and enough gulliesto collect the water; the recommended slope is from0.1 to 1%. It is better to have too great a slope thantoo little. Farms have also used heat cables in theconcrete floor to keep the floor dry, but this is veryexpensive.

20.8.2 Walls

The walls must be constructed to stay upright andkeep the roof in position, keep the building stable,and prevent wind damage to or deformation of thebuilding. How the wall is constructed dependswhether it is insulated or not, the material used, andthe load carrying system employed.

If the wall is not insulated, an open constructionis recommended, because then no ventilationsystem is necessary and the air inside will be thesame as the air around the building. It is, however,an advantage to have some walls to shield the insideof the building from the wind.

A closed or insulated wall construction, if this isnecessary, consists of the following three majorparts:

• Exterior covering to shelter and protect the load-carrying system

• Wall construction including the load-carryingsystem and insulation layer

• Internal covering to screen the inside and protectthe load-carrying system.

These parts can be separate, or all the parts can beincluded in one construction, depending on thematerial used.

Several materials can be used for construction ofthe wall (Fig. 20.5). One major material could beused or the wall can be a mix of different ma-terials. The major materials are wood, steel andconcrete. A normal weathertight wall (the sameprinciple is also used in the roof) of wood or steelcan be constructed as described below.

The exterior covering that protects the wall con-struction and prevents ingress of rain and wind can

be wooden boarding or metal plates (steel or alu-minium). Under these boards or plates, sheathingcardboard or plates can be used to stop strong wind and driving rain penetrating into the wall construction.

In the farming room, plastic-covered chipboardsheets or metal plates can be used for the internalwall covering. To avoid humidity from the produc-tion room penetrating into the wall, a damp courseis used under the plates and on the wall construc-tion. This is a thin clear plastic sheet, and is of great

Figure 20.5 Construction of a wall.

importance in protecting the wall against highhumidity and possible decay.

Timber framework is widely used; typical planksizes are 5cm × 10cm, or 5cm × 15cm, which giveswalls of thickness 10cm or 15cm, respectively. Theplanks are joined to form a frame that is torsionstable in every direction. The frame is typicallycovered on both sides with metal plates or woodpanelling.

If using concrete, the thickness of the walls is10–30cm, and iron bars are used as reinforcement.If the building is insulated, PS can be an integralpart of the wall. Because the PS is enclosed in con-crete, problems with toxic gases from the PS in theevent of fire are avoided. When building concretewalls, formwork must be used on both sides beforethe concrete is poured in and the hardening processcan take place. Concrete blocks or light concreteblocks may also be used. This is a simple way tobuild a wall, where no formwork is needed. Theblocks are laid with cement mortar. Blocks are alsodelivered with integral insulation. The concrete orconcrete blocks represent the total wall construc-tion, and will keep the roof in position, so they func-tion as the load-carrying system.

The exterior and internal coverings are easy toapply to concrete walls. The concrete can be plas-tered and painted so a smooth surface that is easyto clean can be established on both sides (Fig. 20.6).

When using concrete and light concrete, a weath-ertight construction is achieved. Some kind of ven-tilation is therefore necessary, even if the walls arenot insulated, to avoid excessive humidity.

20.9 Ventilation and climatizationIn a building with a weathertight wall construction,there is no exchange of air. After working there fora period the air quality will gradually declinebecause, no new oxygen is added and no carbondioxide removed. Exchange of air (ventilation) istherefore necessary.

Both in areas for water treatment and in the fishproduction rooms there are large free water sur-faces. This will increase the humidity, which meansthat there are large amounts of water vapour in theair. The material and equipment that is going to beused in the room must be able to withstand highhumidity.

The amount of water that the air can take updepends on the air temperature (Fig. 20.7); if thisdrops the amount of water that can be kept in theair is reduced and condensation will result. Typicalsurfaces that have a lower temperature and wherecondensation will accumulate are exterior walls,windows and floors. Excessive humidity can causeproblems with the feed. Especially problematic iswhen the feed is taken from a cold room; because


Figure 20.6 It is important that sur-faces are easy to clean.


the feed is colder than the new surroundings, waterwill condense on its surface. Problems with humid-ity occur when the water temperature is high and itis cold outside the building. To avoid these, thehumidity of the air must be reduced; this is knownas air conditioning.

Two things must therefore be done: exchange ofair to improve the air quality and reduce the humid-ity.These problems can be solved in combination orseparately, and a number of methods and solutionsare available.

One method of reducing the humidity in the airis to use a dehumidifier. This functions in the sameway as a heat pump or a refrigerator. A fan blowsthe air from the room over a cold surface (the evap-orator); the water in the air condenses on thesurface and is collected in a tray. Since the temper-ature of the air is reduced the amount of water thatthe air can contain will be reduced. Afterwards theair is transported over a heated plate (the con-denser) and its temperature is increased; because ofthis the humidity is decreased even more.

Another method to reduce the humidity is toincrease the air temperature 2–3°C above that ofthe water. It is, however, difficult to achieve a highertemperature in all parts of the building; this shows

Figure 20.7 The amount of water that can be taken upby air depends on the temperature.

Figure 20.8 A ventilation system containing an air-to-air heat exchanger.

the importance of good insulation in all cold areasof the building. In practice, this is difficult, becausethere will always be parts of the building where thetemperature is lower and condensation will occur.Otherwise the room temperature must be veryhigh, which is very expensive. Good circulation ofair in a room is also very important when using suchmethods to smooth out temperature differences.

To combine reduction of humidity with ventila-tion is also a solution. Air from outside normallyhas a lower humidity than air inside a building.By bringing air in from outside to replace the air inside, humidity is reduced and ventilationenhanced.

The simplest way is to achieve ventilation is tocreate a small vacuum inside the building. By using

vacuum in the building and new cold air can betaken in from outside through valves in the wall. Ofcourse, a combination of the different methods mayalso be used.

To avoid the need for ventilation, as mentionedpreviously, an open wall and roof constructions canbe used. This requires free circulation of air fromthe outside to the inside. Split panels or plastic grat-ings are examples of this. In addition an open roofridge can be used, so that hot air can escape. This isa specially designed roof part.

References1. Ching, F.D.K., Ching, F.D. (1996) Architecture: form,

space and order. John Wiley & Sons.2. Ching, F.D.K., Adams, C. (2000) Building construction

illustrated. John Wiley & Sons.3. Allen, E., Iano, J. (2003) Fundamentals of building con-

struction: materials and methods. John Wiley & Sons.4. Chudly, R., Greeno, R. (2004) Building construction

handbook. Elsevier Science.


a fan that blows air from inside the building reducesthe air pressure. The air from outside will then flowin through air valves on the walls due to the partialvacuum inside the building. In this way good airexchange can be achieved. Alter-natively, air fromoutside can be forced into the building with a fanand inside air forced out through valves in the walls.This is not recommended, because humid air will be forced into equipment and the fabric of thebuilding.

To recover the energy from the air that is takenout, heat exchange with the inflowing air can beused. An air-to-air heat exchanger is shown in Fig.20.8. Air is withdrawn with a fan, and an equalamount of new air is dragged in via a pipe system.A neutral pressure is then created inside the room.

Another method is to use a so-called breathingintermediate ceiling. The phenomenon utilized isthat hot air will go upwards because it is less densethan cold air. A special open ceiling through whichthe hot air can go is then used to create a small

21Design and Construction of

Aquaculture Facilities

reduce the amount of new inlet water necessary, butthe need for equipment for water treatmentincreases. The design will also depend on whetherthe re-use units are placed on single tanks, orwhether one re-use unit is centrally placed to serveseveral tanks. In a plant with a water re-use system,the amount and size of the equipment for watertreatment is increased, and the same will be the casefor the required area.

A land-based farm for hatching and juvenile pro-duction based on intensive production and a flow-through system can be separated into the followingsections (Fig. 21.1):

• Water intake and water transfer to the farm• Water treatment• Hatchery• First feeding• On-growing• Feed storage• Workshop, staff room• Wastewater treatment• Equipment for feeding and handling that could

be integral parts of the farm construction.

21.2.2 Water intake and transfer

The design of the water intake and transfer systemdepends on whether the water source is at a higheraltitude than the farm, so the water flows undergravity into the farm, or whether the source is onthe same or at a lower level than the farm, so thatpumping of the inlet water is necessary. If possible,the first alternative is of course recommended (Fig. 21.2).

21.1 IntroductionThe design and construction of a production plantfor aquaculture depends on a number of factors,including the intensity and size of the production.Production plants can be classified according towhether the production is based on freshwater orseawater, and if it is onshore or offshore. The development stage of the organism may also beused as a basis for classification. A farm mayproduce eggs, fry, juvenile or on-growing fish readyfor market, or it can have a complete productionsystem with all life stages, from eggs to harvesting,on the farm.

In this chapter some general examples of farmdesign are given.The focus is on intensive fish farmsbecause such farms use technology to the greatestextent.A land-based farm for hatching and juvenileproduction, and a sea-based farm for on-growingare used as examples.

21.2 Land-based hatchery, juvenileand on-growing production plant

21.2.1 General

Land-based production may be classified accordingto the production units used: tanks, ponds or netpens in lakes. The tanks can either have circularwater flow or be raceways. Here the focus is onintensive farms having tanks with circular flow,because this system can give the largest productionper unit surface area.

The design of an intensive land-based farmdepends on whether a flow-through system or a re-use system is used. Use of a re-use system will

294

Transfer pipes can be quite large, and thereforeexpensive, so it is recommended to have as short adistance from the water source to the farm as pos-sible. If the distance is more than 500m the cost ofthe inlet pipe will be considerable. Hence land-based fish farms should be located as close to

the water source as possible. The same is of coursethe case for the distance between the farm and therecipient water body, if the water is to be trans-ported through a pipeline.

If the water is to be pumped into the farm, a highlift head is not recommended because this increasesthe pumping costs; heads 10–15m will result inquite large costs. If such conditions prevail, the casefor increased oxygenation and use of recyclingsystems must be evaluated, to reduce the amountof water needed and hence reduce the operatingcosts of the farm.

Water inlet

The inlet design depends on the source: lake, river,groundwater or the sea.

Lake: In lakes deeper than 10–20m there is nor-mally thermal stratification. The ratio between thewater temperature and the water density causesthis stratification. Because water is most dense at4°C, water at this temperature will sink to thebottom and warmer water will float on top. In areaswhere the water temperature falls below 4°C in thewinter, there will be mixing of water during springand autumn. In the autumn, the surface tempera-ture is reduced; when it reaches 4°C mixing willoccur because all the water will have the samedensity and only a slight breeze on the surface will ensure mixing. Strong winds may also result inmixing of water in the column at other times of theyear.

It is normal to divide the water column into threelayers, the surface layer (epilimnion), a layer with asteep temperature gradient where a large differ-ence in temperature occurs (metalimnion), and thebottom layer (hypolimnion) (Fig. 21.3). The depthat which the steepest temperature increase/decrease occurs is known as a thermocline. Watercollected from below this depth is said to have beencollected below the thermocline. If there is such alayer in the lake it is recommended that the inletbe sited below the thermocline to avoid tempera-ture variations that are large and difficult tocontrol.Two intakes can be installed, one above andone below the thermocline, so that the warmsurface water can be used to increase fish growth,but then possible problems with fouling must betaken into account.

Design and Construction of Aquaculture Facilities 295

Figure 21.1 Main parts in an intensive drifted land-based farm for hatching and juvenile production.


Figure 21.2 Water sources supplyingthe farm under gravity are recom-mended to avoid pumping.

A

B

Figure 21.3 (A) The water in a deeplake can be divided into three layers,but a shallow lake may show no stratification. (B) Circulation of waterby season due to water temperaturegradients (highest density at 4°C).

In a lake the inlet pipe is normally laid on thebottom. The pipe must be correctly aligned, andlarge stones and deep cracks avoided; regularsloping of the bottom is advantageous. Underwatercameras or an echo sounder should be usedto align the pipe before it is fixed in position withweights. These are normally concrete blocksclamped to the pipe; the distance between themdepends on the pipe size and water flow.

The actual inlet ought to be placed some distancefrom the bottom to avoid mud and small stonesbeing sucked into the inlet pipe. This can beachieved in different ways, for instance by adding afloat at the end of the pipe or by using an elbowthat directs the pipe upwards in the water column(Fig. 21.4). To avoid fish and other large substancesbeing dragged into the inlet pipe, a screen is usedat the orifice. The water velocity through the screenmust not be too large to prevent small fish andother objects getting sucked onto the surface of thescreen and blocking the inlet. Water inlet velocitiesof less than 0.1m/s are recommended, while in theinlet pipe itself flows of 1–1.5m/s are used. Poly-ethylene (PE) has proved to be a suitable materialfor inlet pipelines, because it is both reasonablypriced and to some extent will follow the contoursof the terrain.

Sea: In the sea the same principles for an inlet in alake are used. However, the thermocline is nor-mally located deeper in the sea, normally from 30to 70m. Here also it is advantageous to have theinlet below the thermocline. The water tempera-tures will then be predictable, at least when thefarm has been in use for some years, or if much his-torical data are available. For future productionplanning predictable temperatures are a great

advantage but predicting surface water tempera-ture is impossible. Another great disadvantage withhaving the water inlet above the thermocline andclose to the surface is the problem of fouling insidethe pipe. With high water temperatures, the inletpipe will become totally fouled quite quickly andthe water flow will be dramatically reduced. Facili-ties to clean the inlet pipe must, in such cases, be anintegral part of the construction.

For water inlets at great depths, it will be difficultto clean the grating if it becomes blocked, so self-cleaning inlet gratings should be used (Fig. 21.5).Another, and probably a better solution, is to place the grating in the pumping station near thesurface instead of end of the inlet pipe. The inlet pipe is now completely open at the bottom;objects will be swept in with the water but arestopped by the grating in the pumping station. Afunnel at the start of the inlet pipe may also beused. This will reduce the water velocity and at thesame time the resistance head (see Chapter 2). Inthe pumping station it will be quite easy to inspectthe grating visually and remove accumulateddebris.

An infiltration intake may also be used if groundconditions are suitable. As for infiltration inletsfrom rivers described below (Fig. 21.6), variousdesigns of the inlet are possible.1 For example, awell can be built on the beach but the beach mate-rial must be sufficiently permeable.

The cost of inlet pipes in the sea can be consid-erable, because they must be long enough to reachadequate depth and be below the thermocline; agood site therefore has a short distance to theacceptable depth, so deep water close to the shoreis advantageous. A site on a long shallow coastlineis normally poor.


Figure 21.4 Methods for installingthe inlet pipe to avoid bottom mudbeing dragged into the inlet pipe.


River: When the inlet source is a river, the locationof the inlet is very important. Rivers normallytransport a lot of suspended particles and largerobjects that are unwanted in the inlet water to thefish farm. The autumn period is especially critical

when a lot of fallen leaves are transported in thewater. Flood situations will also be critical becausethe amount of suspended particles in the waterincreases due to erosion. The inlet ought thereforeto be laid in an area were the velocity of the river

A

B

Figure 21.5 The inlet grating can (A) be made self-cleaning, or (B) placed within the pumping station so that it isclose to the surface and easily available for cleaning.


Figure 21.6 An infiltration intake ensures that the inletwater is purified before entering the farm.

laid as spokes in a wheel; thus more water is drainedinto the vertical well, because a larger area istapped.

Pumping station

If pumps are used to transfer water to the farm apumping station is needed. The pumps are either

dry placed or submerged. Special equipment is nec-essary to make dry-placed pumps self-priming if thepump is installed at a higher level than the seawa-ter. If such pumping stations are used, special caremust be taken to avoid sucking of ‘false air’. Dry-placed pumps may be installed in a well wherethere is pressure from the water, so circumventingthis problem; this specially designed well must, in


A

B

C

Figure 21.8 Types of groundwaterwell: (A) drilled well; (B) excavatedwell; (C) horizontal well with drainagepipes to increase the area that drainsto the well.


these cases, be excavated and cast. Submergedpumps will always be placed in some type of well(Fig. 21.9); both propeller and centrifugal pumpscan be used. Propeller pumps are recommendedwith large water flows and low lift height (below 10m). If submerged pumps are used, the shorewhere the pumping station is to be placed must beshielded from the waves. It is also recommended

that areas with large waves are avoided when sitingpipes to dry-placed pumping stations. There arealso experiences with drilling the inlet pipe in rockand siting the entire pumping station inside amountain.2 The pumping station can also be placedon the seashore and water is lead into the stationthrough pipes. This requires appropriate groundconditions, sand or gravel for example.

Figure 21.9 Pumping station with submerged pumps forpumping seawater to a land-based fish farm.

For emergency reasons, the water should be sup-plied to a farm using at least two pumps. It is impor-tant that the total efficiency of the pumps is high toreduce the cost per cubic metre of water. If therequirements for water delivery vary, it could beadvantageous to use several pumps of differentsize; however standardization of the pumps so thatit is possible to change spare parts and have acommon stock of the most needed spares, will bedifficult in such cases. There should be at least onestand-by pump in the pumping station. A pro-gramme that alternates between the pumps shouldbe used so that all pumps will be run for someperiod. In this way the stand-by pump is always inuse and not found to have seized up with rust whenneeded.

The pumps are installed in a pumping station.

Dry-placed pumps are placed in a building onshore. For submerged or dry-placed pumps in awell, the pumping station is placed below the lowtide mark in the sea. The pumping station is oftenmade of prefabricated concrete or fibreglass. Sub-merged pumps should be placed on guide rails sothat is easy to take them up for maintenance.

Transfer pipeline

As mentioned earlier, transfer pipes should be asshort as possible to avoid unnecessary expense. Ifonly a low head is available (<5m) or if the wateris pumped, it is especially important to use pipingthat is as smooth as possible. Single resistances withhigh friction coefficients must be avoided; 90°elbows should be substituted with shallower bends,for instance three 30° elbows, or one long smoothelbow. Reduced velocity through the pipelinescould also be used to reduce the friction loss, but itmust be above 0.5m/s to avoid settling.

If the water source is at a sufficiently higher alti-tude than the farm (>30m) the water velocity in thepipeline could be very high; velocities above 3m/sshould be avoided to prevent breakage in the watercolumn and vacuum effects. Increased velocitieswill also increase the requirements for anchoringthe pipes. In such circumstances a small generatorcan be installed to utilize the energy in the inletwater (Fig. 21.10); it is important to choose onemade of material that is not toxic to the fish. Afterthe turbine, it is important to aerate the water toavoid possible supersaturation with nitrogen gas.

Care must be taken if the pipe is laid over a hillor ridge (high crest) from a lake and is functioningas a siphon (Fig. 21.11) because a vacuum effect will


Figure 21.10 A power generator may be included onthe transfer pipeline if there is large height differencebetween the water source and the fish farm.

Figure 21.11 On a high crest there maybe a vacuum in the inlet pipe.


occur on the top crest and the pipe may collapse.To avoid this, it is important to use a pipe of a suf-ficiently high pressure class which has a thick wall.Another problem that may occur when the pipelineis not evenly sloped in one direction, is the creationof gas bubbles at the crests. Normally there willalways be some gas bubbles in the water and theywill collect at the crests; when numerous bubblescollect at this point the effective cross-sectionalarea of the pipe is reduced. This will again reducethe water flowing through the pipe and in the worstcase the water flow can be totally blocked by thegas bubble. The use of a siphon on the inlet pipe istherefore not optimal, and if possible should beavoided. Instead of laying the pipe across the shoreof a lake it may be advantageous to excavate achannel out of the lake in which the pipe is placed.It is also important that the pipe is laid with a goodand even slope.

If there are possibilities for gas collection atpoints in the transfer pipeline or pipelines insidethe farm, special degassing valves can be used.These can be automatic and discharge gas whennecessary. Degassing can also be done manually byinstalling a valve at the critical point in the pipelineand open it at fixed intervals. It should be remem-bered that there must be pressure in the systemwhen the valve is opened to force the gas out. Thevalve can be closed as soon as the water comes out.In a siphon construction a special valve must beused to avoid breaking the siphon.

If fouling occurs in the inlet pipeline, it can becleaned out by sending plugs through the inlet pipe,which is known as plugging. The cleaning plug canbe inserted at the start of the inlet pipe and with-drawn at a point inside the farm (Fig. 21.12). Theplug can, for instance, be made of some type offoam rubber. It is important to have enough waterpressure to force the plug through the transferpipeline.

21.2.3 Water treatment department

In this department the equipment to control andeventually improve the water quality is installed. Itis a very difficult department to design well, espe-cially when much equipment is needed (Fig. 21.13).In several established farms this department looksa mess. When planning, it is important to allow sufficient space. The department will, in all proba-

bility, be changed and modified several times. It isadvantageous to include several valves in thesystem so that water flows can easily be stoppedand sent in different directions. It must be possibleto remove all the individual pieces of equipmentwithout having to shut off the inlet water supply tothe farm.

The amount of equipment needed in this depart-ment varies with the quality of inlet water andtherefore the need to treat it. In this departmentequipment is typically installed for:

• Aeration• Disinfection• Oxygenation• pH control• Removal of suspended solids• Heating and cooling.

Figure 21.12 A cleaning plug for sending through thepipeline to remove fouling.

If using a central re-use system, the equipment forammonia removal and the re-use pumps may alsobe placed here.

Before starting to plan a water-treatment depart-ment on a new farm, it is always recommended thata flow chart be drawn that includes the differentfree water surfaces to prevent mistakes (seeChapter 22). It is quite normal to site the watertreatment department in two rooms, a machineroom and a water treatment room. Equipment foroxygen production together with equipment forheating and cooling can be placed in the machineroom. In the water treatment room there are largefree water surfaces and therefore high humidity, soproper ventilation is necessary here. Examples ofequipment placed in this room include that for aeration, ammonia removal and solids removal.By having two rooms, the expensive mechanicalequipment can be placed in a separate room withlower humidity.

It is advantageous to locate some equipmentclose to the water inlet or where the water transferpipe to the farms starts.This ensures some exposuretime before the inlet water reaches the farm. Whenusing ozone as a disinfectant, or when adding chem-icals for changing the pH, it can be done in the inletand the need for a large retention basin inside thefarm avoided.

There are advantages in having a feeder tank asa last step before the water reaches the productionsystem. This will ensure equal pressure in the inter-

nal pipelines.At the same time, the pressure will notbe too high. High pressure in the internal pipelineswill create a lot of noise in the pipes and valves inthe production hall. In addition, it may be necessaryto use pipes and parts of a higher pressure class,which is more expensive. A feeder tank is also asuitable place to install the alarm sensors, becausethe level will immediately drop when there istrouble with the water supply (Fig. 21.14); somereaction time is also achieved if the water flow


Figure 21.13 Good planning of thewater treatment department is a chal-lenge because there are several com-ponents and it is easy to arrangethem poorly.

Figure 21.14 If possible, it is advisable to have afeeder tank in the water supply before the productionroom to ensure equal water pressure in the pipes to theproduction tanks. The feeder tank is also a suitableplace for alarm sensors.


drops, depending on the volume of the header tank.The disadvantages of using a feeder tank are that itis necessary to lift the water to a higher level, if nopressurized water is available. For this reason sea-water is sometimes sent directly through channelsor pipes into the production room.

21.2.4 Production rooms

The production rooms are normally divided into ahatchery, a first feeding department and an on-growing department. The hatchery can be sepa-rated into an incubation room and a hatching roomdepending on the species grown. Reasons for divi-sion could be to inhibit disease transfer, or toimprove the control and the possibilities for differ-ent light regimes for different production units.It may also be that the on-growing department isoutdoors.

Hatchery

Hatchery designs are based on the species farmed.For some species separate rooms are used for theincubation of fertilized eggs and for the hatching ofthe eggs, because different production units arerequired; water temperature and light conditionsmay also be different. In a hatchery for salmonids,for example, it is common to use trays with 40cm ×40cm troughs (Fig. 21.15).Two to four trays are used

in a stack,and are commonly 2.1m or 3.6m long.Thewater inlet is at one end of the tray and the outlet isat the other.To control the eggs easy accessibility toall units is important. Because of this it is rather dif-ficult with four tray stacks. If producing eye eggs forsalmonids, cylinders may be used; this will, however,require another arrangement. Two cylinders placedat different heights on both sides of the walkwaymight be used.

The main inlet pipeline to the hatchery comesfrom the feeder tank. It is recommended that thereis a separate supply to the hatchery to providewater at the correct temperature and quality; thiscan also include removal of smaller particles, use ofprotein skimmers and disinfection of the water.Theflow rate to the hatchery will be quite low. Inletpipes are either run along the wall or in the roofwith valves to the separate trays. The main outletpipeline is laid in the floor (see below).

To avoid disease transfer when personnel enterthe hatchery, they must first pass through a disin-fection area. If selling eggs, there should be a roomspecially for disinfection and packaging of eggs.Personnel movement between these rooms shouldbe avoided. A hatch in the wall that connects thetwo rooms might be a solution. Between the hatch-ery and the first feeding department the use of ahatch in the wall to avoid movement of personneland possible disease transfer is also common (Fig.21.16).

Figure 21.15 Typical arrangementof trays in a hatchery for salmonidproduction.

Normally no windows are used in the hatcheries.Sunshine may injure the eggs of some species. Somelight sources must also be avoided because they candamage the eggs; for example, blue–violet lighttubes. Photomanipulation is also necessary forsome species, and having no windows makes thispossible. Some species will also require total dark-ness, or the use of only red light during someperiods in the hatchery; for example, halibut.

The hatchery must be easy to clean and disinfectbetween the hatching seasons. It is important toremember this when designing the roof and walls,and installing the equipment.

If a combined hatching and start feeding is used,a mixed department is built and there is no sepa-rate hatchery. In smaller farms this might be a goodsolution, but of course, this has disadvantages con-cerning disease control.

First feeding department

The first feeding department is normally a separateroom, but can also be combined with the hatcheryor be a part of the on-growing department. Tanksizes also vary with the species; this is also the casewith the method of first feeding, for instance if livefeed is used. Normal tanks sizes are in the range 1–8m2 surface area.

It is important to have good accessibility, becausefirst feeding is normally the most difficult stage inthe production cycle and where control is mostimportant (Fig. 21.17). The lighting conditions over

every tank must therefore ensure good visibility.The foundations of the production results are laidin this department.

Two storey tanks could be used, but this arrange-ment can inhibit the accessibility and control of thetanks and is a challenge in the planning process(Fig. 21.17).The water can be supplied directly froma general feeder tank or warm water can be sup-plied from a separate feeder tank. Photomanipula-tion can be used to improve the first feeding results,so no windows are required in this department. Thetanks can be placed along the walls or in the centre,or both with wider buildings. Walkways betweenthe tank rows must be at least 1m wide, or wider,depending on the systems used for transport of feedand fish.

On-growing department

The on-growing department can be inside oroutside. It has become quite common to have partsof it inside, because of possibilities for photoma-nipulation, increased growth and better diseasecontrol, even if this increases the costs. If the on-growing department is outside it is normal to havea bird net above to protect the fish from birds. Thismay also be combined with a shadow net thatshades the fish from the sun and protects somespecies against sunburn.

The tanks are normally quite large with a diam-eter of 5–20m and water height up to 5m. Becauseof their size it can be cheaper to have individualsuperstructures for the tanks. The recommendedtank shape is circular, or square with cut corners.Circular tanks optimize material utilization, whilesquare tanks optimize utilization of the farmingarea.

Depending on the arrangement of the tanks, theplan can be similar to that in the first feedingdepartment. Because the tanks are normally higher,visual control of the fish in them can be a problem.The following solutions can be used (Fig. 21.18):

• Raised walkways for individual tanks• Raised walkways for several tanks• Submerged tanks

Raised walkways are simply constructed. If using abuilding, all installations can theoretically be on thefloor and the building will then have a high secondhand value. Submerged tanks are best for fish


Figure 21.16 Eggs and yolk sac fry can be transportedfrom hatchery to first feeding department through ahatch in the wall without the need for staff to movedirectly between the departments.


farming purposes from the point of view of hygiene,and all the area around the tanks is easy to clean.This requires a concrete floor, located on the upperpart of the tanks. On very large tanks (>10m) awalkway across, like a bridge, can be used toincrease the general view of the fish in the tanks.

Since the fish are now larger, the methods chosenfor fish handling are more important. More feed isalso used, and the feed handling method will there-fore also influence the design.

Water supply, inlet pipelines

The water supply can be routed on the walls, in theroof or in the middle of the room. Normally, thewater is supplied in pipes, but open channels canalso be used although they create more noise andwater under pressure cannot be used. A continuous

fall from the feeder tank to all the production tanksis required.

It is important that the main inlet pipe is largeenough to avoid significant variations in the totalflow if the water flow to one tank is increased.Alternatively, the main pipeline can be installed asa ring connected at both ends (Fig. 21.19). Thisequalizes the pressure at each point in the pipeline,and ensures enough water in the tanks at the endof the pipeline. Use of a ring pipeline will also eliminate the risk of stagnant water staying in the pipeline, when some length of the pipe is not inuse. In seawater this can be a particular problembecause of the biological processes that occur in thestagnant water.

The water velocity is normally between 1 and 1.5m/s; flows over 1.5m/s create high head loss. Lowervelocities are advantageous in achieving a larger

Figure 21.17 Different layouts in a first feedingdepartment.

reservoir and reducing the variation in flow.However, settling in the pipes may now be aproblem and it must be possible to flush the pipesout to remove this and prevent blockages. Thevalves through which the water flows to the indi-vidual tanks are set in the sides of the inlet pipe, notthe bottom, to avoid settling in closed or partiallyopen valves. One valve is set in the bottom to flushand clean the main inlet pipe. It is best to have aslope from the inlet reservoir to the valves, andfrom the valves to the fish tank. The main pipelinecan also be close to the floor, or below the floor withvertical pipes up to the single tanks. When usingsuch systems, care must be taken to avoid air locksin the pipelines; degassing valves can be used atcritical points in the pipe system. Degassing of theinlet pipe is especially critical when using oxy-genated water.

Outlet pipelines

The outlet pipeline starts in the tank. To avoidfouling and blockages the use of 90° bends in thispipeline is not recommended; bends should be asshallow as possible. This will also result in a moregentle treatment of the faeces, so crushing tosmaller particles can be avoided. The head lossthrough the outlet system will also be reduced.

It is easy to get sedimentation in and blockage ofthe outlet pipes.To prevent this, a slope on the pipesof a least 0.5%/m is recommended. In addition itmust be possible to flush and plug the pipes. Whendesigning the system these requirements must bemet.

For laying the outlet pipes inside the farm,several systems are employed (Fig. 21.20):

(1) Pipes under the floor(2) Open channels or culverts in the floor(3) Pipes in culverts in the floor, either covered

with a grating or concrete block(4) Pipes laid upon the floor.

All the alternatives have advantages and disadvan-tages. Method (1) is best for avoiding smell fromthe outlet and getting a surface inside the farm thatis easy to clean. A major disadvantage is that lackof access prevents remedial measures being takenwhen something happens. Later reconstructions ofthe piping system are also difficult. Open channelscreate much noise and are a source of unwanted


Figure 21.18 Methods of access to higher tanks forinspection, feeding, etc. include use of submergedtanks, a raised walkway to a series of tanks, or raisedwalkways to individual tanks.

Figure 21.19 If possible, a ring pipeline should beused to equalize the pressure and avoid flow variationin the connected tanks.


smell, but are reasonably cheap. Method (4) is notrecommended, because it will prevent adequatecleaning of the floor. Its great advantage is,however, that no special arrangement is necessaryand that re-use of the buildings for other purposesis easy. Instead of laying pipes directly on the floor,it can be possible to hang them in the tanks,depending on the tanks used, and then use racks tobring the tanks to a correct working height.

Outlet pipes are not always filled with water; twophases can flow, with air above the water.The veloc-ity ought to be above 0.3m/s to avoid settling ofsolids; otherwise a good slope and easy flushing andplugging must be possible.

21.2.5 Feed storage

Feed storage depends on the type of feed and howit is packed. If using dry feed, as is most common inintensive aquaculture, the feed can be delivered in

bulk and then stored in silos. However, this requiresthe use of large quantities. Big sacks are used formedium quantities and small sacks for small quan-tities. Whatever their size, sacks should be stored inbuildings, or at least in a sheltered area to protectthe contents from animals and birds. The feed sacksmust also be protected against direct sunlight toreduce heating and possible destruction of the feed.This is not necessary if insulated buildings are usedfor storage of dry feed. In the feed storage houseconcrete floors are typically used because spilledfeed is easily cleaned up.

The size of the feed store depends on the feedconsumption, types and sizes of feed, and the shelflife. If much feed is bought at the same time, theprice for feed and transport is reduced, but thisrequires a larger feed store. Problems will occur ifthe feed is stored for too long. The shelf lifedepends on the composition and the tempera-ture in the feed store, and is given by the feed supplier.

It is important to be aware of the feed handlinglines when designing the feed store. If using bigsacks, how is the feed going to be transported in andout of the feed store? For instance, if using bigsacks, they can be hung up for manual tapping intoa wheelbarrow. It is then important to have equip-ment for lifting the sacks and doors that are wideenough.

21.2.6 Disinfection barrier

To protect the farm against transfer of disease,disinfection is recommended. This includes a dis-infection barrier in front of the total farm area andothers before entering the different productiondepartments on the farm. A simple method is toutilize a disinfection mat/bath to which a disinfec-tant is added. The shoes are disinfected when step-ping into this bath. It is, however, preferable to usea barrier where the shoes are changed. Having aclean zone, where people only walk in their socksis a possible solution here (Fig. 21.21). Clothes mayalso be changed in this zone. It is advantageous touse different colours for shoes and working clothesin the different departments. How much sectioningis to be done into separate departments with priordisinfection will always be a problem because it iscostly, and more time is needed for changing, espe-cially of clothes.

Figure 21.20 The outlet pipes can be laid below thefloor, in open or closed channels/culverts or upon thefloor.

21.2.7 Other rooms

In a fish farm it is normal to have a small workshop,or at least a place to store the tools, because thereis quite a large amount of technical equipment onan intensive fish farm.

Depending on the size of the farm it is normal tohave rooms for personnel including a mess room,wardrobe, toilet and bathroom. An office is alsonecessary on a fish farm; on large farms this couldeven be an administration building.

21.2.8 Outlet water treatment

If the outlet water is to be treated, this normallyonly includes equipment for removing particles andstoring the sludge.This equipment should be placedas close to the production units as possible to avoiddamaging the particles. The water must be treatedas gently as possible before it enters this equip-ment. Normally the effluent water is treated in aseparate department or building. It is important tohave sufficient slope on the pipes from the produc-tion unit to the treatment plant to avoid having topump because this breaks up the particles. The par-ticle filter is normally of the rotating screen typewith a mesh size of 90–100μm. Outdoor settlingponds may also be used. However, phosphorus canbe released from settling ponds.

Contamination of the inlet water by the outletwater must be avoided to reduce the possibilitiesfor disease transfer. Therefore there should be no

possibilities for direct movement of personnelbetween these departments. Neither should therebe any possibilities for short circuits between theinlet and outlet pipes. If the inlet water is pumpedfrom a lake or from the sea and the outlet water issent back to the same source, it is important thatshort cuts and/or cross contamination are avoided.The inlet and outlet pipes must be spaced far apartboth in the vertical and horizontal directions.Transport of the outlet water directly to where theinlet is placed by the main current must also beavoided.

21.2.9 Important equipment

Equipment for feeding and feed handling, andsystems for handling fish are important on intensivefish farms because they are used so much. A greatdeal of equipment is available here (see Chapters16 and 17). This equipment can to some degree bean integral part of the farm construction, or it canbe portable. Therefore the choice of equipmentinfluences the design of the farm, so if a farm isestablished using a particular system, much recon-struction can be necessary if it is later decided tochange that system.

Feed handling

Some types of feeding equipment are commonlyused on all intensive land-based farms. This canrange from simple feeders on a single productionunit, to larger automated central feeding systems orfeeding robots. Automation of the feeding systemdepends on the amount of feed used. It is impor-tant that the complete handling process for the feedfrom delivery to end use is well thought out. Thenecessary components in the handling line will, ofcourse, depend on the chosen feeding system; if afeeding robot or a complete feeding system is used,the feed must be stored in silos.

If using traditional feeders, the feed must betaken from the store to the fish tanks. Alternativesfor performing this operation can be to carry sacks,or to use a trolley or wheelbarrow. Trolleys orwheelbarrows set requirements for the width of thewalkways. The feed is normally lifted from floorlevel up into the hoppers of the feeders manually.The hoppers must therefore be easy to fill.The farmmay also be designed with two storeys, a first floor


Figure 21.21 A disinfection barrier before the produc-tion rooms is recommended.


in addition to the ground floor. Feed is stored onthe first floor; this is also where hoppers to the tankson the ground floor are filled, so gravity is utilizedand heavy lifting avoided.

Fish handling

Fish handling is an important part of the work in anintensive farm, and equipment for doing this isimportant, especially when the number of fish andsize increase. Reasons for handlings are various andinclude moving between departments, dividinggroups to avoid excessive density, size grading andwhen delivering fish. Also here it is important tothink in complete handling lines, where the fish aretaken from the production unit and sent back to thetank. A handling line for moving fish may, forinstance, include:

• Crowding in the tank• Vertical transport out of the tank• Horizontal transport between tanks.

The first operation, crowding, can be brought aboutby (1) reducing the water level (it must then be pos-sible to reduce the water level in the tank), (2) usinga rotating grid, (3) having removable tank bottom.For the second operation, vertical transport out ofthe tank, the following methods can be used: (1)net, (2) pumps of various types, such as centrifugalthat is lowered into the tank or the handling centre,vacuum (pressure) that also sucks up fish, ejector orairlift that require deep tanks, (3) transport tankslifted with a forklift truck, (4) fish screw (needs alarge area). For horizontal transport the followingmethods are used: (1) dip net and/or buckets, (2)transport tank, (3) pipelines.

The handling system may also be an integral partof the farm (Fig. 21.22). Examples of such systemsare: (1) a centrally placed pumping chamber towhere the fish are tapped through in pipelines fromall tanks, (2) by having pump adaptors close to thebottom in all tanks and a centrally placed pump, (3)pipelines near the top of all tanks, into which thefish are poured after using a dip net to take themout the tank.

For larger fish the tanks may be equipped withhatches through which the fish either swim volun-tarily or are forced to go when the water level isreduced. The first system only requires one channelbetween the tanks, and there is a combination of

voluntary and forced movement by movable grids(Fig. 21.23).The second system is based on channelson two floors and an elevator that lifts the fishbetween the two levels (Fig. 21.24). This systemfunctions as follows. The water level in the tank isreduced and the fish are forced to leave via thehatch in the tank wall. There is a slope on the

Figure 21.22 Farm in which the handling system is anintegral part.

Figure 21.23 Fish farm equipped with channels fortransport of fish.

bottom in the lower channel and by continuouslyreducing the water level the fish are forced into theelevator at the end of the channel system. One tankinside the silo constitutes the elevator.The fish fromone tank are collected in the elevator that is actu-ally an ordinary fish tank. By closing the inlet hatchto the elevator and supplying water to the silo thewater level will increase and the tank that floatsinside will move upwards like an elevator.When the

tank reaches the upper channel level the hatch isopened and the fish will start to swim out; by con-tinuing the elevating process they will gradually beforced to leave. After this the water level in theupper channel will be gradually reduced and thefish forced back into their tank. Then fish from anew tank can be moved in a stepwise process.

The solutions for size grading must also bedecided before the farm is established. If the farm


Figure 21.24 Fish farm equipped with channels for volun-tary fish movement. Between the two channel levels there isan elevator for vertical transport of fish.


is small and small amounts of fish are going to begraded, a cradle used directly in the productiontanks might be a solution. Otherwise a machine canbe used, such as a roller, belt or band type. If usinga level grader the high head loss must be remem-bered. If thinking about the complete system, thegrading machine can be placed in a grading centre,or it can be portable and moved to the tank that isto be graded. In the latter case it is important tohave sufficient space in the walkways to place thegrader.

21.3 On-growing production, sea cage farms

21.3.1 General

An on-growing farm bases its production on buyingfry or juvenile fish and producing fish ready for slaughtering. An inexpensive system for on-growing production is to use cages, in lakes, calmrivers or in the sea. Later in this chapter a briefdescription of the design of a complete cage farmis given, focusing on a sea-based farm.

A total sea cage farm includes the following com-ponents (Fig. 21.25):

• The cage farm with fixed equipment• An operations base

• A boat• Net handling equipment.

The cage farm includes the cages with the mooringsystem. Equipment fixed to the cage can be feedingequipment, equipment for collection of dead fishand feed loss, and a lighting system.

21.3.2 Site selection

When selecting the sites for sea cage farms severalcriteria are important. Some species independentcriteria used when evaluating a site are as follows:

• Stable water quality (the actual quality require-ments will be species dependent)

• Good water exchange, but not too high a veloc-ity (below 1m/s is recommended)

• Minimum depth under the cages of 5m• Good infrastructure• Temperature above 0°C to avoid icing; otherwise

temperature appropriate for the species• Not close to potential sources of water

contamination• For wave heights above 3m the equipment costs

are rather high.

When having a sea cage farm the use of severalsites is recommended. If having species that needmore than one year to reach the marked size orwhen having several inputs of juveniles every yearit is advisable to use different sites (Fig. 21.26) toreduce the possibility of disease transfer betweenthe inputs/generations. In addition, is it advisable tolet the sites rest for a year after some years in pro-duction. This has been shown to improve produc-tion on the site. If this production model is used, thesites must not be too close together. The distancedepends on the current conditions between thesites, but typically at least 1.5km is recommended.

21.3.3 The cages and the fixed equipment

Cages and net bags

The number, type and size of the cages depends onthe production regime and the farmed species. Theexposure of the site to waves is also important.If the site is very exposed, special offshore cages are used. Large cages have become increasinglypopular because the production costs can bereduced.3 In the Norwegian salmon industry, circu-Figure 21.25 Plan of a typical sea cage farm.

lar cages with a circumference of 90 or 110m arebecoming quite common. However, extra demandsare made on the equipment used for handling thenets.Another trend is to utilize more exposed waterfor farming, which sets extra requirements for thecages, mooring systems and operating boats.

Sea cage farms may also be established close tothe shore, so that a walkway can be constructedfrom the land directly to the cages (Fig. 21.27). Insuch conditions it is normal to use a system farm,or at least a farm with a floating walkway, to wherethe cages are moored.

A dead fish collector may be used in the cage: thiscan either be in the form of a stocking below theordinary net bag, or a basket inside the cage. Thebasket can easily be lifted to the surface with a rope,and the dead fish collected.

Lighting system

To increase the growth and reduce early matura-tion, additional and continuous lights can be usedin the sea cages.4,5 In particular, during the winter

season in high latitudes, this has been shown toimprove the production results for salmoinds by20–30%.6 Thus in almost every sea cage in Norwayextra light is used to increase growth and reducematuration. The light source can stay above orbeneath the surface. When having the light abovethe water surface it must be quite strong because ithas to go through the water surface (Fig. 21.28).Light above the surface may be detrimental to thesurroundings; for instance, there have been com-plaints from ships. For example, on a cage 15m × 15m four 500W floodlights have been used; the nec-essary brightness is around 180 lux, the same as rec-ommended for rooms not in continuous use, suchas storerooms.

When planning a light installation, the followingfactors are important:

• Type of light source• Placement of light source above or beneath the

surface• Number of lights• When in the year it shall be used• For how large a part of the day it shall be used• Source of energy• How to get the energy to the cage; where shall

electrical cables be laid or shall generators beused?

Today, underwater lights have become much morecommonly used, because they are closer to what isto be illuminated. Also, they do not cause the prob-lems associated with non-submerged lights, e.g. nav-igational confusion.

Feeding system

Several methods are used for feeding fish in cages,ranging from hand feeding where no additionalequipment is required to automatic feeding wherethe additional installations depend on the chosenfeeders (see Chapter 16). If traditional feeders areused, they are placed on a platform that could beintegrated in the collar. The hoppers can be quitelarge (>1m3) and additional buoyancy is necessary.If central feeding systems are used it is only thepipes with the feed that enter the cages and noadditional equipment is necessary. It can be possi-ble to use additional equipment inside the cagesuch as detectors for uneaten feed and biomass esti-mation. Such equipment will, however, be portable


Figure 21.26 The use of several sites is recom-mended, both to inhibit to transfer of disease and to letthe sites rest.

Figure 21.27 Cages can be individ-ual, or part of a system with andwithout a walkway to land.

and have no influence on the construction of thecages.

21.3.4 The base station

A base station is necessary for a sea cage farm (Fig.21.29). This can be land based or sea based (lyingon a raft), or it can be a combination of the twoalternatives.The size and content of the base stationdepends on the size of the production unit and themanagement of the farm, including what servicesare bought from subcontractors.A base station maycontain various pieces of equipment and storagefacilities.

It is normal to have at least a wardrobe, toilet andmess room for the workers on the base. In additionthere is a small workshop, or at least a place to havesome tools.

Every farm also needs a system for taking careof the dead fish; to prevent them becoming an envi-ronmental problem, they must not be dropped intothe water. A tank to which acid is added can be used, and the fish are ensiled (Fig. 21.30); theproduce can, for example, be used to feed fur-bearing animals. The dead fish may also be storedin a cold-storage room, for later collection and utilization.

It is also normal to have a feed store on the base.A central feeding system may also be installed onthe base. Feeding barges have become quitepopular because they can be moved between sites


Figure 21.28 Lighting system usedon a cage farm to improve fishgrowth.

if the sites are fallowed. Common constructionmaterials are steel or concrete; old ferries haveproved quite popular as feeding barges.

The base, whether it is on land or sea, must be ofa design that makes it easy for boats to dock. On asea-based barge this is quite simple; if the base ison land a quay is necessary.

The base may or may not include net storage,net washing equipment and impregnation equip-ment. However, today it is quite common to employsubcontractors to do all the net handling andstorage.

21.3.5 Net handling

Net handling represents a major part of the totalworkload on a sea cage farm, and requires addi-tional equipment. For many sites fouling is a largeproblem for nets in the sea and to reduce thedegree of fouling the mesh size needs to be as largeas possible. This means that it is an advantage tochange the nets according to fish growth, so as largea mesh size as possible can be maintained. One or two sizes are typically used per year, but thisdepends on growth and species; nets can be muchmore frequently changed on exposed sites.7 Netexchange is heavy work, especially on larger cages;large cranes are required to handle the nets. This isalso one reason for using subcontractors, becausethey have large equipment specially adapted for thepurpose.


If there is much fouling on the site cleaning orwashing of the nets is necessary. This ensuresenough water passes through the net panel tosupply oxygen and remove waste products. Onexposed fouling sites, the nets need to be washedseveral times a year, up to once a month. Washingmay either be done when the net bags are standingin the sea, or they can be removed and taken toshore for washing. Special washing equipment isused by divers to wash nets in the sea. If the netsare taken to shore, large washing machines are

used. These machines are similar to a traditionaldomestic washing machine but have a larger druminto which the nets are loaded (Fig. 21.31). Theeffluent from the washing machines has a highcontent of fouling materials. This is discharged atone place through the outlet pipe and the pointoutlet may be too high. Today there are calls forpurification of wastewater from such washingmachines; there is a particular problem with dis-charge of antifouling agents, because normally atleast 20% of the antifouling agents remain on thenets at the start of the washing process.

The nets will also need regular repair, becausethe mesh will break. This is normally done in con-nection with washing and before the nets are set outagain.

It is also usual to treat the net with an antifoul-ing agent to inhibit fouling so that they can stay inthe sea for longer before removal for washing.Copper-based antifouling agents are widely usedand quite effective. However, there are environ-mental concerns regarding the use of copper, but noother antifouling agents have so far achieved thesame efficiency. Different types of biocides may beused, but there are also environmental concerns

Figure 21.29 Normally a cage farm will be equippedwith a base station which is either land or sea based.

Figure 21.30 A tank for ensiling dead fish.

about these. Much research is being undertaken tofind alternatives. Before setting out the impreg-nated nets they are dried so that the antifoulingstays on the net.

21.3.6 Boat

All sea-based on-growing farms need a boat (Fig. 21.32). In the past, small boats of polyethylene

or fibreglass with an outboard motor were oftenused. Today, larger boats of steel or aluminium with a working deck are becoming increasinglycommon as a result of the trend towards usinglarger cages and sites with more feed and heavierequipment. Faster boats are more common thanpreviously, and speeds of up to 15–20 knots are normal. One reason for this is that the distance from the coast is increasing and


A

C

B

Figure 21.31 Equipment for washing net bags: (A) and(B) when the net is on shore with the use of largewashing machines; (C) when the net is in the sea.


more exposed water is being used for farming.Catamarans with wide decks and lengths in excessof 10m are much used today. It is common to have a crane on the boats for handling nets andfeed sacks.

References1. Huguenin, J.E., Colt, J. (2002) Design and operating

guide for aquaculture seawater systems. ElsevierScience.

2. Lekang, O.I. (1991) Lukkede produksjonsanlegg for laksefisk i Norge. ITF-rapport nr. 18, NorwegianUniversity of Life Science (in Norwegian, Englishsummary).

3. Guldberg, B., Kittelsen, A., Rye, M., Åsgård, T. (1993)Improved salmon production in large cage systems. In:Fish farming technology. Proceedings of the first inter-national confernce on fish farming technology (edsH., Reinertsen, L.A., Dahle, L. Jørgensen, K.Tvinnereim). A.A. Balkema.

4. Kråkness, R., Hansen, T., Stefansson, S.O., Taranger,G.L. (1991) Continuous light increase growth rate ofAtlantic salmon (Salmo salar L.) post-smolts in seacages. Aquaculture, 95: 281–287.

5. Hansen,T., Stefanson, S.,Tarnager, G.L. (1992) Growthand sexual maturation in Atlantic salmon, Salmo salarL., reared in sea cages at two light regimes. Aquacul-ture and Fisheries Management, 23: 275–280.

6. Willougby. S. (1999) Manual of salmonid farming.Fishing News Books, Blackwell Publishing.

7. Lucas, J.S., Southgate, P.C. (2003) Aquaculture, farmingaquatic animals and plants. Fishing News Books,Blackwell Publishing.

Figure 21.32 Different types of boat used on seacages farms. Today there is a trend towards larger andfaster boats.

22Planning Aquaculture Facilities

facilities.1,2 Despite the biological aspects involvedin the planning of aquaculture facilities, some ofthese basic theories and methods can be employedin addition to the important matters regardingaquaculture.

In this chapter important elements in the planning process of an aquaculture facility aredescribed. This is done by presenting one simplifiedmethod for performing the planning process.This isutilized as a tool to avoid missing important ele-ments during the planning process which, becauseof the complexity of the planning process for inten-sive fish farms is very easy, at least for personnelwho are not highly trained. It is important toremember that this is only one of several methods,and planners will often have developed their ownbased on experience.

22.2 The planning processWhen the planning process starts there is always aninitiative to alter the prevailing situation. Forexample, someone may want to establish a com-pletely new farm, or only be a minor reconstructionof the farm is wanted. Both require a planningprocess to have been completed before buildingcommences. For the planner, it is important toensure that they really understand the needs of theapplicant to be able to execute the planning processoptimally.

The planning process may be separated into thefollowing parts from the choice of site, to when thefacility is finished and in production:

(1) Site evaluation and selection(2) Production plan

22.1 IntroductionPlanning of aquaculture facilities, of whatever type,is a complicated process that requires much knowl-edge to achieve a good result. It is, for example,more difficult than planning a typical industrial pro-duction plant, such as for manufacturing metalparts. Aquaculture facilities involve living individu-als. The production result depends, for instance, onthe suitability of the tanks, water flow conditionsand whether water quality meets the requirementsof the individuals, whether fish or shellfish. Planningfaults will reduce the performance of the individu-als, which can be manifest as reduced growth ormore frequent disease problems, for example.

The requirements for planning will vary accord-ing to the type of facility. The planning of a farmwith one or a few excavated ponds is fairly simple.A rather more complex situation occurs when plan-ning a land-based fish farm for indoor juvenile pro-duction. The planning will be even more complex ifthe farm is to include water re-use technology inaddition to flow-through technology. Such complexplanning tasks involve several fields of competence:for instance, sanitary, electrical, building and archi-tectural. These are all technological, but becausethe production involves living organisms it is also necessary to have biological knowledge, forexample of the optimal environmental growth con-dition for the fish. Since so many subjects areinvolved, planning is not simple to perform and anumber of specialists will normally be involved, atleast when planning larger farms.

A number of theories and methods have beenintroduced to optimize the planning process,especially for planning buildings and industrial

321


(3) Room programme(4) Necessary analyses, such as function, form,

technology, environmental impact andeconomy

(5) Development of alternative solutions basedon the analysis

(6) Evaluation and synthesis of the alternativesolutions

(7) Actual design, making the necessary drawingsand description, calculation of costs

(8) Drawing up invitations to tender, choice ofcontractor, starting building

(9) Function test of the plant, with and withoutfish

(10) Project review.

If there is only one site available, it will only need to be evaluated. It is, however, important toshow the limitations of the site. If an extension of an existing plant is wanted, the same will be the case. Independent of this, it is always importantto carry out the site evaluation and control to ascertain whether the site really can tolerate theextension and what problems may occur. This mayalso set additional requirements in the planningprocess, for instance that there will be a limitationin the water supply and that re-use technology isrequired.

A production target or a given production planmay also be the starting point of the planningprocess. If this is the case, a proper site has to bechosen based on these requirements.

22.3 Site selectionTo choose a good site is of course of major impor-tance for future production results and possibleproblems, so proper investigations about site per-formance must be carried out. In relation to plan-ning this description will give the criteria for thefurther planning process. If several sites could beused, the description will give the necessary basisfor evaluation and selection of the site.

The site chosen will, of course, depend on thetype of farm that is being planned – a hatchery oron-growing, land- or sea-based. An extremelyimportant selection criterion when talking aboutaquaculture facilities is, of course, the amount andquality of available water. There are many storiesof land-based freshwater aquaculture facilities

suffering from lack of water after some years inproduction.

For cage farming, the water quality and currentare of great interest. The depth and bottom condi-tions are also important because of the mooringrequirements (see Chapter 15). For land-basedfarms the water quality will also be of major impor-tance, but here the amount of water available is alsoof great interest. It is important to remember thatwhen a farm is planned it is designed for a givenwater flow. In almost every case, after a period inproduction there will be a desire to increase pro-duction and therefore the need for water willincrease. It is therefore advantageous to includethis possibility in the planning process.When check-ing the possible amounts of water that could bewithdrawn from the water source and used in thefarm, it is important to find values for the possiblewater supply for every month all year round. Amonitoring programme before establishing a farmmust be implemented and surprises resulting fromdry seasons must be avoided. Therefore it is impor-tant to look up as much historical data as possibleregarding the water source. Possibilities for regula-tion of the water level in lakes, or damming uprivers must also be evaluated.

Good water quality will always be the best,regardless of species farmed. Some species, such ascarps, will not have such high requirements for thewater quality, while others such as salmonids aremore stringent. Water of poor quality can be usedon a salmonid farm but requires additional treat-ment before use, so resulting in increased costs.

Available infrastructure is also important whenselecting a site. To have easy access to electricity,good roads and telephone lines reduces the costs ofestablishing the farm.

22.4 Production planIn the production plan an estimate of the futureproduction is given, for instance how much fish andof what size is going to be in the farm at differenttimes of the year, normally every month at least,to ensure a given production (Table 22.1). Thisincludes the requirements for oxygen and water, eventually requirements for heating orcooling of water, and necessary units for storing ofthe fish.

Planning Aquaculture Facilities 323

Tab

le 2

2.1

Exa

mpl

e of

a p

rodu

ctio

n pl

an f

or a

land

-bas

ed f

arm

.

Mon

th

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

Wat

er t

empe

ratu

re (

°C)

11

12

810

1212

86

42

Egg

wat

er t

empe

ratu

re (

°C)

88

Tem

pera

ture

0+

(°C

)8

8.0/

12.0

1212

1212

1212

86

42

Tem

pera

ture

1+

(°C

)1

11

28

10

Wei

ght

0+(g

)0.

20.

180.

270.

691.

724.

319.

8722

.548

.569

.388

.797

.3W

eigh

t 1+

(g)

103

107

110

114

117

158

Num

ber

of e

ggs

683

971

674

419

Num

ber

of 0

+fis

h65

734

362

958

359

003

155

473

255

063

154

656

154

252

153

851

053

452

953

057

852

665

652

276

2N

umbe

r of

1+

fish

520

581

518

898

511

254

507

475

503

725

500

000

Tota

l num

ber

of fi

sh1

177

624

114

379

01

101

285

106

218

91

050

285

104

656

154

252

153

851

053

452

953

057

852

665

652

276

2

Bio

mas

s eg

gs (

l)1

371

137

1B

iom

ass

0+fis

h (k

g)13

111

315

838

394

92

355

535

312

135

2589

736

766

4673

050

887

Bio

mas

s 1+

fish

(kg)

5373

955

020

5633

257

675

5899

178

924

Tota

l fish

bio

mas

s (k

g)53

870

5513

356

490

5805

059

940

8127

95

353

1213

525

897

3676

646

730

5088

7

Wat

er r

equi

rem

ent

eggs

(l/m

in)

120

120

Wat

er r

equi

rem

ent

0+fis

h (l

/min

)22

136

450

71

228

223

04

239

802

916

989

1502

013

604

1121

37

633

Wat

er r

equi

rem

ent

1+fis

h (l

/min

)6

986

715

37

323

865

129

496

5840

4To

tal w

ater

for

wat

er (

l/min

)7

207

751

77

830

987

931

726

6264

38

029

1698

915

020

1360

411

213

763

3

Hat

chin

g tr

ays

(No.

of)

6969

Tank

vol

ume

0+(m

3 )69

tra

ys63

m2

59m

238

9523

653

560

71

295

183

81

557

169

6Ta

nk v

olum

e 1+

(m3 )

179

11

834

187

81

923

196

62

631

Tota

l tan

k vo

lum

e (m

3 )1

791

183

41

878

196

12

061

286

753

560

71

295

183

81

557

169

6


A well prepared and detailed production plan isthe basis for the rest of the planning process. It isimportant to allow sufficient time when developingthe production plan; mistakes here will be trans-ferred to mistakes in the planned facility.

22.5 Room programmeThe real planning process can now start, and thefirst task is to get a survey of the main installations,including the buildings with necessary rooms andheavy fixed installations. This is done through aroom programme (Table 22.2). Here a first estima-tion of the need for size and area is also given. A

Table 22.2 A room programme for a small fish pro-duction plant.

Area Size

Rearing sectionbroodstock 2 tanks, >8 m2

hatchery 14 hatching traysfirst feeding 12 m2

on-growing 800 m3

Water treatment 50 m2

Variousfeed storage 15 m2

workshop/storage 20 m2

office/mess room 15 m2

wardrobe 7 m2

WC/shower 5 m2

disinfection barrier 4 m2

Figure 22.1 Different ways to prepare a connectionanalysis.

Figure 22.2 Drawing process diagrams can be a toolin the planning process.


Figure 22.3 Various methods forfish handling illustrated in an alterna-tives chart.

Figure 22.4 Blasting for establish-ing aquaculture facilities in the coastalarea may result in unsightly scars inthe landscape.

given production will, for instance, need a tankvolume based on the requirements of the fish. If theproduction of fish is known, some assumptionsabout the space for feed storage can be made. Somerooms, such as a changing room, bathroom, messroom and office may also be necessary. The aim ofthe room programme is to obtain some idea aboutthe size of the different rooms and installationswhich can be used in the further analyses, not tomake a complete list of what the facility is going tolook like when it is finished. By summing all thesecomponents an estimate of the total size of theplant is obtained.

22.6 Necessary analysesPart of the aim with analysis is to remember tothink through the different possible solutions.

Advantages and disadvantages of the differentsolutions are to be discussed, which is really helpfulin the planning process. Whilst analysis can be per-formed on many topics, it is important to performthe main analyses. This part must not be confusedwith the description of the chosen solution; it mustbe an analysis.

One necessary analysis is that concerning areaconnections; which areas in the plant are or are notto have connections. This can be illustrated with anexample: a farm is fenced in and the only entranceis through a disinfection barrier where the shoes aredisinfected in a bath; there should be no possibili-ties for direct entrance to the plant in other ways.During the planning process an area connectionanalysis will identify such relations.

One method of performing the analysis is tospread out the rooms and areas from the room pro-


Fig

ure

22.

5A

ltern

ativ

e pl

ans

shou

ld b

e de

velo

ped

befo

re c

hoos

ing

the

layo

ut o

f th

e fis

h fa

rm.


Fig

ure

22.

5C

ontin

ued.

gramme like pieces of a puzzle and draw linesbetween the areas were connections are wanted(Fig. 22.1).The same can also be done with an arrowdiagram, where connections and the reason for connections are illustrated (Fig. 22.1).

To remember the different process that musttake place process diagrams can used as a tool (Fig.22.2). The technical analysis includes a survey ofways of solving technical problems with theiradvantages and disadvantages. For instance, if thewater is to be aerated, what types of aerators are tobe used and what are the advantages and disad-vantages of the different types; another analysis canbe whether or not to use oxygen. Analysis of dif-ferent materials includes the advantages and disad-vantages of each. Process diagrams and alternativescharts are also examples of assistance tools; forinstance, alternatives charts are helpful for showingvarious handling methods (Fig. 22.3).

Form and situation analysis includes where in theterrain the farm can be located, with advantagesand disadvantages; for example, should it be in theground or on top. Aesthetic considerations mustalso be included.

Analysis of environmental impact is becomingincreasingly important for aquaculture facilities.How to reduce the discharge is an important analy-sis. To establish aquaculture facilities near beachzones may result in large impacts in the landscape,caused, for instance, by blasting operations thatcreate large ‘scars’ in the landscape (Fig. 22.4). Theneed for proper analysis is necessary in such cases.

An area function analysis of the different areasis also commonly included where the requirementsand their function are discussed. Taking the feedstorage as an example, this could include the fol-lowing analysis: will there be possibilities for expan-sion or not; will there be possibilities for drainingthe floor or not; are there any special requirementsfor the surface of the floor or not?

22.7 Drawing up alternative solutionsBased on the analyses, the development and plan-ning of alternative solutions may start.This includessimple sketches of the different possible options.The reason for stressing development of alternativeoptions to meet planning requirements is that thisfunctions as a tool to develop optimal solutions.Theplans can, with advantage, be as different as possi-

ble from each other. This stresses the variability,which is important in improving creativity. At leasttwo or three alternatives should be developed (Fig.22.5); these can be discussed with the owner of thefarm, to involve them in the planning process andto make sure that the developed solution meetstheir requirements. The water levels are extremelyimportant, and when planning land-based fishfarms it is important to have control of the freewater levels; therefore it is helpful to prepare dia-grams showing this.

22.8 Evaluation of and choosingbetween the alternative solutionsThe next step in the planning process is to evaluatethe alternatives and choose from among the devel-oped solutions. On this basis the chosen plan can befurther developed. Hopefully the analysis and con-sideration of different solutions have improved theplan compared to first proposals. All the developedsolutions will of course have advantages and disad-vantages, and these must be weighed when devel-oping the final plan which will often be a mix of thealternatives.

22.9 Finishing plans, detailed planningAfter choosing a solution, this can be further devel-oped, by preparing more detailed plans and draw-ings of constructions and/or buildings. Here moredetailed design of the necessary components is alsoincluded and, based on this, a more detailed calcu-lation of the costs. This is labour-intensive. It mayalso be a two-step planning process with a pilotplanning project followed by the detailed planningprocess.

The next step is usually to draw up invitations totender with the necessary descriptions. When atender is accepted, the building process can start.During the building process is it important to checkprogress regularly.

22.10 Function test of the plantAfter finishing building the plant or part of theplant, a period of function testing is necessary, start-ing with single components and ending with theentire farm. First is it performed without fish in the plant and when everything is functioning the



testing can be continued with fish in the system. Suf-ficient time must be taken at this important stage,and when establishing advanced facilities can takeup to several months.This is an important stage thatis often underestimated. To put the fish into thefacilities too early may end in disaster if somethingfails. If contractors build the entire project, theowner of the farm must not take it over beforeoperational testing of components and the wholefarm has been carried out with satisfactory results.

22.11 Project reviewIt is important to undertake a post-hoc review ofthe building process and of the chosen options. The

major object of doing this is to optimise the processin later planning, and to create future optimal solu-tions. Post-hoc project reviews are mainly for thebenefit of the planner.

References1. Muther, R. (2000) Systematic planning of industrial

facilities (SPIF). Management of Industrial ResearchPublication, Kansas City.

2. Svardal, S. (1994) Planning of rural buildings. Theoryand method. Lecture notes. Norwegian University ofLife Sciences (in Norwegian).

additives, live fish transport, 264aeration, 97–120

see also oxygen/oxygenationaerators, 101–106cascade aerators, 105–106construction, 101–106design, 101–106equilibrium, 99–100evaluation, 102–103gas theory, 99–101gas transfer, 100–101gases in water, 97–9gravity aerators, 103–106Inka aerators, 106, 107methods, 101–102packed column aerators, 103–106paddle wheel aerators, 106, 107ponds, 179principles, 101–102propeller aerators, 106, 107purpose, 97saturation, 97–9subsurface aerators, 106, 107surface aerators, 106, 107

afterevaluation, planning, 329air transport, live fish transport, 262–3airlift pumps, 237, 239

oxygen/oxygenation, 258–9alternatives, planning, 325–8aluminium, pH, 38ammonia

monitoring, 270–71water quality, 33

ammonia removal, 121–32bacteria, 121–3biodrum, 125–6biofilters, 123–9biological removal of ammonium ion, 121chemical removal, 129–30

Index

330

denitrification, 128–9filters, 123–9flow-through system, 123–5fluid bed/active sludge, 126–7ion exchangers, 129–30nitrification, 121–3Nitrobacter bacteria, 121–3Nitrosomonas bacteria, 121–3oxidizing, 121–3pH, 121–2rotating biofilter (biodrum), 125–6

analyses, planning, 325–8anchors, mooring systems, 203–204angle seat valves, pipes, 10aquaculture, classification, 1–2aquaculture facilities

construction, 294–320design, 294–320hatcheries, 294–314juvenile production, 294–314land-based, 294–314planning, 321–9

artificial substrate, egg storage/hatching,154–5

automationfeeding systems, 215, 218–22instrumentation, 266

back-flushingdepth filtration, 50–51screens, 46–9

bacteria, ammonia removal, 121–3ball valves, pipes, 10band graders, size grading, 251bar graders, size grading, 248–9base station, sea cages, 5, 317, 318basins, see tanksbead filters, ammonia removal, 127belt graders, size grading, 250–51

Bernoulli equation, water transport, 16biodrum, ammonia removal, 125–6biofilters, ammonia removal, 123–9biological removal of ammonium ion, 121biomass estimation, 277–80boats, sea cages, 5, 319–20breakwaters, sea cages, 197buildings, 284–93

cleaning, 290–91climatization, 291–3design, 285, 289–91environmental factors, 291–3floors, 289–90foundations, 289ground conditions, 289insulation, 288–9load-carrying systems, 285–7materials, 287–8prefabricated, 288reasons, 284roof design, 285shapes, 284–5types, 284–5ventilation, 291–3walls, 290–91

Bunsen coefficient, oxygen/oxygenation, 99,119

buoys, mooring systems, 202–203butterfly valves, pipes, 10–11

cage collars, sea cages, 193–5cameras, fish size, 278–80carbon dioxide, monitoring, 270–71cascade aerators, 105–106cavitation, pumps, 21centrifugal pumps, 23–7, 234, 237characteristics curves, pumps, 25–7characterization of the water, particles, 45check valves, pipes, 11Chick’s law, disinfections, 64chlorine, disinfections, 73classification

aquaculture, 1–2pipes, 9–10production units, 144–9sea cages, 183

cleaningbuildings, 290–91closed production units, 165–6live fish transport, 258, 263–4self-cleaning, 165–6transfer pipeline, 304water inlet, 304

climatization, buildings, 291–3

closed production units, 145–7, 158–73cleaning, 165–6components, 158–60dead zones, 162design, 162–5drains, 169–72flow pattern, 165–6materials, 163–5mixing, 162self-cleaning, 165–6types, 158–60velocity profile, 165–6water exchange rate, 161–2water flow, 165–6water inlet, 167–9water outlet, 169–72water quantity, 160–61

closed sea cages, 145–7, 158–73combination units, egg storage/hatching, 157components

closed production units, 158–60farms, 2–5instrumentation, 267land-based farms, 2–4mooring systems, 198–9re-use, 139–40sea cages, 183–4

conductivity, monitoring, 269connections

pipes, 12–13planning, 325–8pumps, 28

constructionsee also buildingsaeration, 101–106aquaculture facilities, 294–320heat pumps, 87–9instrumentation, 267ponds, 176–8re-use systems, 136–9sea cages, 193–8

conveyor belt feeders, 216–17cooling

chilling of water, 94–5heat exchangers, 80, 94–5reasons, 75

counting fish, instrumentation, 275–7crowding, internal transport, 233–4, 235current

measuring, 193oceanic, 193sea cages, 191–3, 204–210tidal, 192–3wind-generated, 191–2

Index 331

332 Index

DE, see diatomite filtersdead fish, tanks, 318dead zones, closed production units, 162demand feeders, 217–18denitrification, ammonia removal, 128–9density, fish

live fish transport, 259, 261re-use, 136

density, water, production units, 148–9depth filtration

back-flushing, 50–51filters, 49–52

designaeration, 101–106aquaculture facilities, 294–320buildings, 285, 289–91closed production units, 162–5mooring systems, 198–201re-use systems, 141–3tanks, 162–5ultraviolet light, 65–7water inlet, 167–9water intake/transfer, 294–5water outlet, 169–71

diatomite (DE) filters, 51–2diffusers, oxygenation, 111dip nets, internal transport, 234, 236disc feeders, 216–17disinfection barriers, production rooms, 310disinfections, 63–74

basis, 64–5Chick’s law, 64chlorine, 73dose-response curve, 65ground filtration, 73heat treatment, 72–3methods, 63natural methods, 73oxidizing, 63–4ozone, 68–72pH, 73photozone, 72ultraviolet light, 65–8Watson’s law, 64–5wetlands, 73

ditches, pipes, 14–15drainable/non-drainable ponds, 177–8drains/drainage

closed production units, 169–72dual drain tanks, 57, 58, 171–2ponds, 177–8, 180–81

dual drain tanks, 171–2particles, 57, 58

echo sounding, biomass estimation, 279–80ecosystem, ponds, 174

effluent, water quality, 33–5egg storage/hatching, 150–57

artificial substrate, 154–5bottom-lying eggs, 153–7combination units, 157hatching cabinets, 155–7hatching cylinders, 155–7hatching troughs, 153–4incubators, 151–2intensive/extensive production units, 150–51pelagic eggs, 151–3water flow, 152–3

ejector pumps, 236–7, 239embankment ponds, 176–8energy loss, water transport, 16–18energy, pumps, 22–3energy requirement, heating, 75–6environmental factors

buildings, 291–3sea cages, 185–93

environmental forcesmooring systems, 210–13sea cages, 204–210

environmental impact, production units, 149equilibrium

aeration, 99–100oxygenation, 108

escaped fish, water quality, 35evaluation

aeration, 102–103afterevaluation, 329oxygen/oxygenation, 110–11planning, 328

excavated ponds, 176–8extensive/intensive production units, 144–7

facilities, aquaculture, see aquaculture facilitiesfaeces, particles, 33, 34, 45feed blowers, 216feed dispensers, 216–17feed storage, production rooms, 310feeding equipment

feed handling, 311–12land-based farms, 4sea cages, 4–5

feeding systems, 215–26automatic feeders, 218–22automation, 215cell wheel, 219central, 222–3control units, 221conveyor belt feeders, 216–17demand feeders, 217–18disc feeders, 216–17distribution mechanisms, 218–20dynamic, 225

electric current, 221–2feed blowers, 216feed control, 224–5feed control systems, 224–5feed dispensers, 216–17feed hopper, 220feeding robots, 223–4requirements, 215–16screws, 219sea cages, 315–17selection, 215spreading of feed, 220–21types, 216–24vibrators, 219

filtersammonia removal, 123–9back-flushing, 46–9bead, 127biofilters, 123–9depth filtration, 49–52diatomite (DE), 51–2efficiency, 56–7examples, 128fuller’s earth, 51–2granular, 127granular medium, 49–52hydraulic loads, 56hydrocyclones, 53–4integrated treatment systems, 55–6management, 127–8mechanical, 45–9media, biofilters, 125mesh sizes, 49nitrification, 123–8particles, 45–54purification efficiency, 56–7settling/gravitation, 52–3swirl separators, 53–4vacuuming, 46–9

fish feeding department, production rooms, 307fish counting, instrumentation, 275–7fish cradles, size grading, 245–6fish density, re-use, 136fish handling, see internal transportfish screws, 237–40fish size, instrumentation, 277–80fish transport, see live fish transportfittings

head loss, 18, 19pipes, 12, 13

fixing point, mooring systems, 201flat outlets, water outlet, 169–71floors, buildings, 289–90flow pattern

closed production units, 165–6heat exchangers, 85–6

flow-through system, ammonia removal, 123–5flow, water, see water flowfluid bed/active sludge, ammonia removal, 126–7forces calculations, sea cages, 204–210frameworks, sea cages, 193–5freshwater/salt water, production units, 148–9fry production ponds, 174–6fuller’s earth, filters, 51–2function test, planning, 328–9future trends, 6

gas concentrations, water quality, 33gas/oil burners, 79gas pressure, total, see total gas pressuregases in water, 97–101

see also aeration; oxygen/oxygenationTGP, 269–70

grading boxes, size grading, 246grading grids, size grading, 246–8, 253–4grading machines (graders), size grading,

248–53grading, size, see size gradinggranular filters, ammonia removal, 127granular medium filters, 49–52gravitation/settling filters, 52–3gravity aerators, 103–106ground conditions, buildings, 289ground filtration, disinfections, 73groundwater, water inlet, 300–301growth, size grading, 228–9

handling fish, see internal transportharvesting fish, size grading, 232hatcheries

land-based, 294–314production rooms, 306–307

hatching cabinets, egg storage/hatching, 155–7hatching cylinders, egg storage/hatching, 155–7hatching troughs, egg storage/hatching, 153–4head loss

monitoring, 273, 274water transport, 16–18, 19

heat exchangerscooling, 80, 94–5flow pattern, 85–6fouling, 86–7heat transfer, 80–81materials, 86NTU, 81–2pipes, 83–5plate exchangers, 83several-stroke exchangers, 83–4shell and tube exchangers, 83, 85size, 81–3specific pressure drop, 82types, 83–5

Index 333

334 Index

heat pumps, 87–91coefficient of performance, 89–91construction, 87–9function, 87–9installation, 90–91log pressure–enthalpy (p–H) diagram, 89maintenance, 91management, 91reasons, 87

heat treatment, disinfections, 72–3heating, 75–94

coefficient of performance, 89–94composite heating systems, 91–4energy requirement, 75–6heat exchangers, 79–87heat pumps, 87–91heaters, 77–9immersion heaters, 77–9methods, 76–7oil/gas burners, 79reasons, 75

Henry’s law, oxygenation, 108high pressure pumps, 28–9horizontal transport, 241–3hydrocyclones, filters, 53–4hydroxides, pH adjustment, 41–2

ice, sea cages, 193immersion heaters, 77–9impellers, pumps, 27–9importance, aquaculture engineering, 6incubators, egg storage/hatching, 151–2injection systems, oxygenation, 109–115Inka aerators, 106, 107inlet, water, see water inletinset layout ponds, 181–2instrumentation, 266–83

see also monitoringautomation, 266biomass estimation, 277–80components, 267conductivity, 269construction, 267fish counting, 275–7fish size, 277–80head loss, 273, 274land-based farms, 4live fish transport, 259–60, 261–2nitrogen saturation, 269–70oxygen content, 268–9pH, 269physical conditions measuring, 271–5salinity, 269saturometer, 269–70temperature, 268

TGP, 269–70water flow, 271–3water level, 274–5water pressure, 273–4, 275water quality, 267–71

insulation, buildings, 288–9integrated treatment systems, particles, 55–6intensive/extensive production units, 144–7

egg storage/hatching, 150–51interactions, water quality, 33internal transport, 4, 227–45

see also size gradingcrowding, 233–4, 235dip nets, 234, 236equipment, 233–45external energy, 233fish handling, 227–33, 312–14, 324–5horizontal transport, 241–3, 312methods, 233–45negative effects, 232–3pipes, 241–3planning, 324–5pumps, 234–9reasons, 227–8, 312transport tanks, 240–41, 242vertical transport, 234–41, 312–14voluntary movement, 243–5, 312–14

ion exchangers, ammonia removal, 129–30

jointing, pipes, 12–13juvenile production, aquaculture facilities, 294–314

lakes, water inlet, 295–7land-based aquaculture facilities, 294–314land-based farms

components, 2–4site selection, 322

land transport, live fish, 257–60layout

planning, 325–8ponds, 181–2

legal issues, sea cages, 185levee ponds, 176–8level graders, size grading, 251–3lighting systems, sea cages, 315, 317lime

pH adjustment, 39–41sludge production/utilization, 60

live fish transport, 256–65additives, 264air transport, 262–3bags/cans, 263changing water, 259cleaning, 258, 263–4density, fish, 259, 261

instrumentation, 259–60, 261–2land transport, 257–60oxygen/oxygenation, 258–9preparation, 256–7sea transport, 260–62stopping procedures, 259–60tanks, 257–8vehicles, 257well boats, 260–61

load-carrying systems, buildings, 285–7lye, pH adjustment, 41–2

materialsbuildings, 287–8closed production units, 163–5heat exchangers, 86net bags, 195–7pipes, 7–9sea cages, 194–5

mesh sizes, filters, 49metal ions, pH, 33, 38–9micro-organisms, water quality, 33–5micro screens, particles, 45–9monitoring, 266–83

see also instrumentationammonia, 270–71carbon dioxide, 270–71components, 280–81conductivity, 269control, 283head loss, 273, 274land-based farms, 4maintenance, 283nitrate, 270–71nitrogen saturation, 269–70oxygen content, 268–9pH, 269physical conditions, 271–5PLC, 281–2regulation equipment, 283salinity, 269sensors, 280–82systems, 280–83temperature, 268TGP, 269–70warning equipment, 282water flow, 271–3water level, 274–5water pressure, 273–4, 275water quality, 267–71water velocity, 271–3

Moody diagram, water transport, 17mooring, pipes, 13–14mooring systems

anchors, 203–204

buoys, 202–203calculations, 210–13components, 198–9control, 213design, 198–201environmental forces, 210–13fixing point, 201mooring lines, 201–202, 211–13sea cages, 198–204size, 210–13types, 198–201

negative effects, handling fish, 232–3net bags

materials, 195–7sea cages, 195–7, 314–15, 317–19

net handling, sea cages, 317–19net positive suction head (NPSH), pumps, 21–2, 27nitrate, monitoring, 270–71nitrification

ammonia removal, 121–3filters, 123–8

Nitrobacter bacteria, ammonia removal, 121–3nitrogen saturation, monitoring, 269–70Nitrosomonas bacteria, ammonia removal, 121–3NPSH, see net positive suction headnumber of transfer units (NTU), heat exchangers,

81–2nutrients, water quality, 34

ocean sea cages, 198oceanic current, 193oil/gas burners, 79on-growing production, 314–20

ponds, 174–6production rooms, 307–308

outlet water, see water outletoxidizing

ammonia removal, 121–3disinfections, 63–4

oxygen/oxygenation, 106–120see also aerationairlift pumps, 258–9Bunsen coefficient, 99, 119compressed oxygen gas, 115–16diffusers, 111equilibrium, 108evaluation, 110–11examples, 111–15gas transfer, 108Henry’s law, 108increasing equilibrium concentration,

108injection systems, 109–115liquid oxygen (LOX), 116–17

Index 335

336 Index

live fish transport, 258–9monitoring oxygen content, 268–9on-site oxygen production, 117–19oxygen cones, 113, 114oxygen gas characteristics, 115oxygen sources, 115–19oxygen wells, 113, 114packed column, 111, 112principles, 109PSA, 117–19purpose, 97saturation, 97–9sea cages, 113–15solubility, 119sources, 115–19supply, 258–9systems, 109–115theory, 108water quality, 33

ozonecontent measuring, 71–2design/dimensioning, 70disinfections, 68–72dose, 70–71function, 68mode of action, 68–70problems, 71

packed column aerators, 103–106packed column oxygenation, 111, 112paddle wheel aerators, 106, 107parallel layout ponds, 181–2particles

characterization of the water, 45definitions, 44–5dual drain tanks, 57, 58faeces, 33, 34, 45filters, 45–54integrated treatment systems, 55–6local ecological solutions, 60–61micro screens, 45–9removal, 44–62removal methods, 45–56screens, 45–9sludge production/utilization, 57–60TS, 44TSS, 44water quality, 33–5wave calculations, 186–7

pathogens, water quality, 33–5pelagic eggs, egg storage/hatching, 151–3pH

adjustment, 37–43adjustment examples, 39–42

aluminium, 38ammonia removal, 121–2definitions, 37–8disinfections, 73hydroxides, 41–2lime, 39–41low, 38lye, 41–2metal ions, 33, 38–9monitoring, 269problems, 38seawater, 41sodium hydroxide, 41–2water quality, 33water sources, 38–9

photozone, disinfections, 72physical conditions, monitoring, 271–5pipes

classification, 9–10connections, 12–13ditches, 14–15fittings, 12, 13head loss, 16–18, 19heat exchangers, 83–5internal transport, 241–3jointing, 12–13materials, 7–9mooring, 13–14pressure class, 9production rooms, 308–310transfer pipeline, 303–304vacuum, 9valves, 10–12water flow, 15–16water hammer, 9water inlet, 303–304water transport, 7–15

planningafterevaluation, 329alternatives, 325–8analyses, 325–8aquaculture facilities, 321–9connections, 325–8detailed, 328evaluation, 328function test, 328–9internal transport, 324–5layout, 325–8process, 321–2production plan, 322–4room programme, 324–5site selection, 322size grading, 324–5

plastic sea cages, 197

PLC, see programmable logic controllerponds, 174–82

aeration, 179construction, 176–8drainable/non-drainable, 177–8drainage, 180–81ecosystem, 174embankment, 176–8excavated, 176–8fry production, 174–6inset layout, 181–2layout, 181–2levee, 176–8on-growing production, 174–6parallel layout, 181–2production units, 144–7, 158–73, 174–82radial layout, 181–2series layout, 181–2site selection, 178–9size, 178types, 176–8water inlet, 179–80water outlet, 180–81water supply, 179watershed, 176–8

pressure class, pipes, 9pressure, pumps, 27–9pressure swing adsorption (PSA), oxygen/oxygenation,

117–19production control, size grading, 229–32production plan, 322–4production rooms, 306–310production units

see also closed production unitsaims, 144classification, 144–9closed, 145–7, 158–73closed sea cages, 145–7, 158–73design, 144–8environmental impact, 149freshwater/salt water, 148–9fully controlled/semi-controlled, 147intensive/extensive, 144–7land-based farms, 4ponds, 144–7, 158–73, 174–82raceways, 144–7, 158–73sea cages, 4, 145–9tanks, 145–7, 158–73tidal basin, 144–8water density, 148–9

programmable logic controller (PLC), monitoring,281–2

propeller aerators, 106, 107propeller pumps, 24–5

PSA, see pressure swing adsorptionpumping stations, water inlet, 301–303pumps

airlift, 237, 239, 258–9cavitation, 21centrifugal, 23–7, 234, 237characteristics curves, 25–7connections, 28costs, 23definitions, 21–2ejector, 236–7, 239energy, 22–3fish screws, 237–40high pressure, 28–9impellers, 27–9internal transport, 234–9NPSH, 21–2, 27performance, 25–7pressure, 27–9propeller, 24–5pump height, 21re-use, 134, 141–3regulation, water flow, 29–31RPM, 29–30throttling, 30–31types, 19–20vacuum-pressure, 234–6, 238water flow, 27–31water inlet, 301–303water transport, 18–31working point, 27

purification efficiencyfilters, 56–7re-use, 136

quality, water, see water quality

racewaysproduction units, 144–7, 158–73size grading, 253

radial layout ponds, 181–2re-use, 133–43

advantages, 133–4centralized, 141–3components, 139–40construction, systems, 136–9definitions, 134–6degree of, 134–5, 139–40density, fish, 136design, systems, 141–3disadvantages, 134effectiveness, 138–9mass flow, 136–7pumps, 134, 141–3

Index 337

338 Index

purification efficiency, 136theoretical models, 136–9waste handling, 137–8water flow, 136–8water, live fish transport, 263–4water requirements, 137

recirculation, see re-userecycling, see re-useReynolds number, water transport, 17river, water inlet, 298–300roller graders, size grading, 249–50room programme, planning, 324–5rotating biofilter (biodrum), ammonia removal,

125–6RPM, pumps, 29–30

salinity, monitoring, 269salt water/freshwater, production units, 148–9saturation, aeration/oxygenation, 97–9saturometer, instrumentation, 269–70screens, particles, 45–9screws

feeding systems, 219fish screws, internal transport, 237–40

sea cages, 4–5, 183–214base station, 5, 317, 318boats, 319–20breakwaters, 197cage collars, 193–5classification, 183components, 183–4conditions, 184construction, 193–8current, 191–3, 204–210environmental factors, 185–93environmental forces, 204–210examples, 197–8feeding systems, 315–17forces calculations, 204–210frameworks, 193–5ice, 193legal issues, 185lighting systems, 315, 317materials, 194–5mooring systems, 198–204net bags, 195–7, 314–15, 317–19net handling, 317–19ocean, 198on-growing production, 314–20plastic, 197production units, 145–9site selection, 184–5, 314, 322size grading, 253steel, 198water quality, 184–5

waves, 185–91, 204–210wind, 204–210

sea transportdensity, fish, 261instrumentation, 261–2live fish transport, 260–62well boats, 260–61

sea, water inlet, 297–8seawater, pH adjustment, 41self-cleaning, closed production units, 165–6self grading, size grading, 253–4sensors, monitoring, 280–82series layout ponds, 181–2settling/gravitation filters, 52–3site selection

land-based farms, 322planning, 322ponds, 178–9sea cages, 184–5, 314, 322

size gradingsee also internal transportband graders, 251bar graders, 248–9belt graders, 250–51energy supply, 245–53equipment, 245–54fish cradles, 245–6grading boxes, 246grading grids, 246–8, 253–4grading machines (graders), 248–53growth, 228–9harvesting fish, 232land-based farms, 4level graders, 251–3manual, 245methods, 245–54planning, 324–5production control, 229–32raceways, 253reasons, 228–32roller graders, 249–50sea cages, 5, 253self grading, 253–4tilt graders, 246voluntary grading, 253–4in water, 253

sludge production/utilizationlime, 60particles, 57–60wet composting reactor, 59–60

SMB method, see Sverdrup–Munk–Bretsneidermethod

sodium hydroxide, pH adjustment, 41–2steel sea cages, 198stopping procedures, live fish transport, 259–60

subsurface aerators, 106, 107superstructures, see buildingssurface aerators, 106, 107Sverdrup–Munk–Bretsneider (SMB) method, waves,

189–90swell, waves, 191swirl separators, filters, 53–4

tanks, 158–73dead fish, 318design, 162–5dual drain, 57, 58, 171–2internal transport, 240–41, 242live fish transport, 257–8production units, 145–7, 158–73transport, 240–41, 242

temperature, monitoring, 268TGP, see total gas pressurethrottling, pumps, 30–31tidal basin, production units, 144–8tidal current, 192–3tilt graders, size grading, 246total gas pressure (TGP), monitoring, 269–70total solids (TS), particles, 44total suspended solids (TSS), particles, 44tower outlets, water outlet, 169–71transfer pipeline, water inlet, 303–304transport, see internal transport; live fish transport;

water transporttreatment, water, see water treatmenttrends, future, 6triple way valves, pipes, 11–12TS, see total solidsTSS, see total suspended solids

ultraviolet lightdesign, 65–7dimensioning, 67–8disinfections, 65–8dose, 68function, 65mode of action, 65problems, 68

vacuum, pipes, 9vacuum-pressure pumps, 234–6, 238vacuuming, screens, 46–9valves, pipes, 10–12velocity profile, closed production units, 165–6velocity, water, monitoring, 271–3ventilation, buildings, 291–3video cameras, fish size, 278–80voluntary grading, size grading, 253–4voluntary movement, internal transport, 243–5,

312–14

walls, buildings, 290–91waste handling

land-based farms, 4re-use, 137–8

water density, production units, 148–9water flow

closed production units, 165–6egg storage/hatching, 152–3monitoring, 271–3pumps, 27–31re-use, 136–8water transport, 15–16

water hammer, pipes, 9water inlet

aquaculture facilities, 295–301cleaning, 304closed production units, 167–9groundwater, 300–301lakes, 295–7land-based farms, 2–3pipes, 303–304ponds, 179–80production rooms, 308–309pumping stations, 301–303pumps, 301–303river, 298–300sea, 297–8transfer pipeline, 303–304water quality, 32–3wells, 300–301

water intake/transfer, design, 294–5water level, monitoring, 274–5water outlet

closed production units, 169–72flat outlets, 169–71ponds, 180–81production rooms, 309–310tower outlets, 169–71treatment, 311water quality, 33–5

water pressure, monitoring, 273–4, 275water purification, see particles; water

treatmentwater quality, 32–6

ammonia, 33effluent, 33–5escaped fish, 35gas concentrations, 33inlet water, 32–3instrumentation, 267–71interactions, 33micro-organisms, 33–5monitoring, 267–71nutrients, 34outlet water, 33–5

Index 339

340 Index

oxygen, 33particles, 33–5pathogens, 33–5pH, 33sea cages, 184–5

water supplyponds, 179production rooms, 308–309

water transport, 7–31energy loss, 16–18head loss, 16–18, 19pipes, 7–15pumps, 18–31water flow, 15–16

water treatment, 35–6, 304–306see also particlesland-based farms, 2–3production rooms, 311

water velocity, monitoring, 271–3watershed ponds, 176–8

Watson’s law, disinfections, 64–5waves

breaking, 187–8calculations, 186–7, 210creating, 188–91diffraction, 187–8reflecting, 187–8sea cages, 185–91, 204–210SMB method, 189–90swell, 191terminology, 185–6wind, 188–91

wells, water inlet, 300–301wetlands, disinfections, 73wind

calculations, 210current, 191–2sea cages, 204–210waves, 188–91

working point, pumps, 27

aquaculture engineering

Science

water quality

inlet water

water treatment

water transport

pumping of water

change of water

blackwell publishing

pipe parts ttings