This article was downloaded by [UQ Library]On 14 November 2014 At 0014Publisher Taylor amp FrancisInforma Ltd Registered in England and Wales Registered Number 1072954 Registered office Mortimer House37-41 Mortimer Street London W1T 3JH UK

Desalination and Water TreatmentPublication details including instructions for authors and subscription informationhttpwwwtandfonlinecomloitdwt20

Experimental study of high-performance hybrid NF-MSF desalination pilot test unit driven by renewableenergyAbdel Nasser A Mabrouka amp Hassan El-banna S Fathb

a Faculty of Petroleum amp Mining Engineering Department of Engineering Science SuezUniversity Suez Egyptb Faculty of Engineering Department of Mechanical Engineering Alexandria UniversityAlexandria EgyptPublished online 15 Mar 2013

To cite this article Abdel Nasser A Mabrouk amp Hassan El-banna S Fath (2013) Experimental study of high-performancehybrid NF-MSF desalination pilot test unit driven by renewable energy Desalination and Water Treatment 5137-396895-6904 DOI 101080194439942013773860

To link to this article httpdxdoiorg101080194439942013773860

PLEASE SCROLL DOWN FOR ARTICLE

Taylor amp Francis makes every effort to ensure the accuracy of all the information (the ldquoContentrdquo) containedin the publications on our platform However Taylor amp Francis our agents and our licensors make norepresentations or warranties whatsoever as to the accuracy completeness or suitability for any purpose of theContent Any opinions and views expressed in this publication are the opinions and views of the authors andare not the views of or endorsed by Taylor amp Francis The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information Taylor and Francis shall not be liable forany losses actions claims proceedings demands costs expenses damages and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with in relation to or arising out of the use ofthe Content

This article may be used for research teaching and private study purposes Any substantial or systematicreproduction redistribution reselling loan sub-licensing systematic supply or distribution in anyform to anyone is expressly forbidden Terms amp Conditions of access and use can be found at httpwwwtandfonlinecompageterms-and-conditions

Experimental study of high-performance hybrid NF-MSFdesalination pilot test unit driven by renewable energy

Abdel Nasser A Mabrouka Hassan El-banna S Fathb

aFaculty of Petroleum amp Mining Engineering Department of Engineering Science Suez University Suez EgyptEmail abdelnaser_mabrouks-petrolsuezeduegbFaculty of Engineering Department of Mechanical Engineering Alexandria University Alexandria Egypt

Received 23 September 2012 Accepted 2 February 2013

ABSTRACT

This study presents the experimental results of a renewable energy (RE) driven high-perfor-mance multi-stage flash (MSF) desalination unit integrated with nanofiltration (NF) mem-brane The desalination pilot test is built to evaluate the performance of the novel De-aeration and brine Mix (MSF-DBM) configuration The NF pilot is build in order to enableMSF desalination unit the operation at high top brine temperature (TBT) The capacity of thedesalination pilot plant is 10m3day of water The whole pilot plant is installed at Wadi El-Natroun (Egypt) This study presents the first phase of the project including the pilot testdescription and experiments of concentrated solar trough steam generator NF and MSFunits Comparison between the simulation and the experimental results of the pilot unitrsquo sub-systems are relatively satisfactory The newly developed NF-MSF-DBM (de-aerator and brinemix) configuration is tested at TBT= 100˚C and the gain output ratio (GOR) is calculated as15 which almost twice of traditional MSF under the same operating conditions The newlyhigh-performance NF-MSF-DBM and unitrsquos input thermal energy which make the integrationwith (the relatively expensive) RE as a desalination plant driver is a viable option

Keywords Desalination Renewable energy MSF plant NF membrane

1 Introduction

Desalination has been favored as a method toincrease the water supply in many counties of theworld On the other hand renewable energy (RE) asdriver to desalination technologies is best suited toremote off-grid communities Use of RE is morepopular in academic circles and at test sites but hasnot been as popular with decision makers whoconcentrate on large fossil-fueled desalination plant toprovide potable water to large populations

RE driven desalination systems fall into two cate-gories The first includes distillation processes drivenby heat produced directly by the RE system (RES)while the second includes membrane and distillationprocesses driven by electricity or mechanical energyproduced by RES Attention has been directed towardimproving the efficiencies of the solar energy conver-sions desalination technologies and their optimalcoupling to make them economically viable for small-and medium-scale applications

Various small-scale RE desalination projects havebeen conducted in the Middle East and North Africa(MENA) region over the past twenty years A hybrid

Corresponding author

1944-39941944-3986 2013 Balaban Desalination Publications All rights reserved

Desalination and Water Treatmentwwwdeswatercom

doi 101080194439942013773860

51 (2013) 6895ndash6904

November

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

multi-stage flash-reverse osmosis (MSF-RO) systemdriven by a dual purpose solar plant was installed inKuwait which produces 25m3 of water per day [1]Pilot projects for solar pond collectors have been car-ried out in addition to multiple effect distillation(MED) desalination plant operated for 18 years atUmm El-Nar (Abu Dhabi) with a 120m3day plant[1] In Saudi Arabia a PV-RO brackish water plantwas installed and connected to a solar still with a pro-duction of 5m3day [1] These were small-scale pro-jects that have contributed to the general consensusthat renewable desalination is a feasible option in theMiddle East The most popular combination is eitherMED with thermal collectors or RO with photovoltaics(PV) technology

The energy demand of seawater reverse osmosis(SWRO) plant located in Canary Islands of a capacityof 25000m3d is provided by a combination of PVcells (rooftop) with minor share of the grid which con-sist of energy mix including wind energy [2]

There are various concentrated solar power (CSP)technologies In 2011 13GW of CSP were operatingand a further 23GW were under construction [3]Currently base-load electricity generated by CSPplants costs two to three times that from existing fos-sil-based technologies without carbon capture andstorage CSP generation costs are competitive withphotovoltaics and offshore wind but are almost twotimes of onshore wind [3] A reduction of 50ndash60 inCSP generating cost may reasonably expected overthe next 10ndash15 years provided that research develop-ment and demonstration programs are sustaining thecommercial deployments of CSP plants It is antici-pated that CSP should become cost competitive withfossil-based load generation at some point between2020 and 2030 [3]

The energy cost of solar desalination is reported asan equivalent to 5ndash15 kWh of electricity per cubicmeter of water either directly by reverse osmosis orindirectly as pump losses and decreased efficiency inbackpressure turbines [4]

Multi stage flash (MSF) technology has proven tobe a mature thermal technology for large-scale capac-ity production with high-quality desalinated waterespecially for high severe feed water quality How-ever at high top brine temperature (TBT) scaledeposits of high seawater brine concentration presentsa real problem in MSFMED plants as it directlyaffect the heat transfer rates on the heating surface Athigher temperature greater than 120˚C nonalkalinecalcium sulfate (CaSO4) precipitates if saturation limitsare exceeded due to the inverse solubility of thesesalts with temperature Formation of alkaline scale(CaCO3 amp MgOH) can be controlled by lowering pH

(acid additives) or by adding antiscalant Nonalkaline(hard) scale (as CaSO4) is only controlled nowadaysby limiting TBT below 120˚C

Increasing TBT with hard scale control can be car-ried out by (1) introduce high-temperature antisca-lant (2) reducing hard scale ions to avoid it reachingsaturation conditions The first approach currently isunder experimental evaluation to check antiscalantperformance under different operating conditionsbefore introducing to the market [5] The secondapproach is considered in the literature [6ndash12]through the use of nanofiltration (NF) membrane sys-tem for make-up feed water pretreatment

The application of NF membrane in seawater desa-lination has recently gained significant attention in thedesalination industry due to selective removal of diva-lent ions NF is originally applied to reject electrolytesand obtain ultrapure water with high-volume flux atlow operating pressure because most membraneshave either positive or negative charge due to theircompositions

Extensive experiments have been carried out onMSF pilot test unit with NF as pretreatment [6ndash9] NFpressure was 24 bars and its recovery ratio rangedfrom 60 to 65 Almost 90 of the total concentrationof the sulfate and calcium ions is rejected with NFbrine while the MSF make-up entirely formed fromNF permeate was below their solubility limits Thisresult indicated the possibility to operate the MSFplant safely and without any scaling problem at TBTat or higher than 130˚C

A partial pretreatment of the MSF feed (25 ofmake-up) by introducing NF membrane was proposedand presented in [1011] A test has been carried outon one evaporator of Layyah MSF plant located inUAE The results showed an increase in the TBT from105 to 110˚C and the production from 1044 to 1253 th (199) The maximum production was 1260 th(206) at 117˚C with product conductivity of 454 lScm However based on personal communication theNF plant was shut down due to operational problemsin the pretreatment section

The NF membrane possesses molecular weightcutoff of about hundreds to a few thousands which isintermediate between RO membranes and ultrafiltra-tion (UF) The pore radii and fixed charge density ofpractical membranes were evaluated from permeationexperiments of different neutral solutes of sodiumchloride The pore radii of these NF membranes wereestimated to range from 04 to 08 nm [12]

Simulation of a novel integrating NF system withMSF-de-aerator and brine mix (DBM) desalinationplant at the TBT= 130˚C was carried out [13] Theprocess calculations showed that the gain output ratio

6896 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

(GOR) of the novel configuration is tow times of theconvention MSF-BR [13] The levelized water cost ofNF-MSF-DM (at TBT= 130˚C) was found to be 14lower than conventional MSF (at 110˚C) at the currentoil price of 104 $bbl [13] The novel NF-MSF-DBMconfiguration significantly reduces the unitrsquos inputthermal energy cost which suits integration with (therelatively expensive) RE

The objective of this work is to present the pilottest description and experimental performance resultsof CSP steam generator NF and MSF units Thewell-developed process design program (VDS) [13ndash16]is upgraded to consider the mathematical model ofCSP and steam generator and used as a processdesign tool for MSF-DBM NF steam generator andCSP unitsThe goal of MSF desalination pilot test unitis to verify the new concept of (DBM-MSF) at highTBT and high performance

2 Pilot test unit CSP-NF-MSF-DBM overview

The process design and simulation for the testpilot is developed to prepare specifications of differentcomponents Some units are manufactured by Egyp-tian contractor while other are purchased from ven-dors The site is prepared where civil work andfoundation are constructed The test pilot componentsare installed assembled and finally individual com-missioning for each component is carried out The siteis located at lsquoWadi El-Natrounrsquo remote area almost150 km southwest of Alexandria city (Egypt) The sitebelongs to Alexandria University

Fig 1 shows the pilot test of solar energy anddesalination units The concentrated solar parabolictrough with thermal energy storage facility providesthe necessary heating in order to generate requiredsteam of the MSF desalination unit The system isequipped also with a backup boiler for steam compen-sation Solar PV and wind turbine (not appeared inFig 1) are installed and run separately However in

this phase diesel engine is used to provide the pumpselectricity until finalize the match and synchronizationbetween PV and wind turbine

21 MSF desalination unit

The MSF pilot unit consists of 28 flashing cham-bers with 28 connected condensers as shown inFigs 2a and b The stages are arranged in double deckin order to reduce the foot print There are four setseach set consists of 7 stages MSF chambers areequipped with glass windows for monitoring theflashing process The shell material of MSF is fabri-cated from 2ndashmm-thick stainless steel 316 L The flashchamber is 05m length and 05m width while theheight is 10m The condenser tube is 8mm diameterand 6m long and made of stainless steel (07mmthick) The number of tubes is 2 per condenser andarranged in multipass inside 05m shell length Theunit is manufactured in Egyptian and assembled atthe project site

The orifice opening area controlled using gatevalve which is located between the flash chambersThe interstages valves controls the interstages flowrates to guarantee the brine flashing at each stage Thesplash plate is designed just above of the inlet open-ing to reduce carry over Demister is fixed near thevapour outlet vapor pipes to reject brine carry overbefore going to the condenser The shells are insulatedto minimize energy losses In addition to the brineheater different supporting systems are added includ-ing vacuum system and chemical injection systemsThe vacuum system has control valves at each stageto adjust the venting rate of noncondensable gases(NCG) and the stage pressure

The MSF is the main subsystem where distillationis produced using the flashing process Differentinstrumentations are installed to measure and recordthe temperatures pressures and flow rates as shownin Fig 1 In the heating section steam input and out-put temperatures in addition to pressure and flow are

Fig 1 Basic flow diagram of CSP-NF-MSF-DBM pilot plant

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6897

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

measured using proper transducers All chambers areequipped with temperature and pressure indicatorsFirst and last chambers are equipped with tempera-ture transmitter (TT) and pressure transmitter (PT)and additional two movable PT and TT are suppliedto be inserted in the chambers of the operator choiceInput seawater flow and output brine and distilledwater flow rates are measured using flow transmit-ters

22 Concentrated solar power (CSP) system

Four modules of solar concentrator (parabolictrough) are purchased and assembled in series at thesite of the project as shown in Fig 3 Each module is

36m length and 1524m width The collector area permodule is 56m2 while the collector reflective area is5m2 The assembled collector length becomes 146mwhile the total area is 224m2 The receiver absorptiv-ity is 092 the mirror reflectivity is 091 while thereceiver emittance is 023 The black-coated pipes are10 in diameter and placed in 20 in diameter glasspipes to minimize convection losses The concentratorshave a tracking system and were placed EastndashWestand facing south

The CSP system contains a steam generator to sup-ply the MSF brine heater with the required heatingsteam Thermal oil is circulated through the collectingpipes gains the solar thermal energy and flowthrough steam generator and energy storage tank

The steam generator consists of shell and tube andhas a separate vapor header The shell diameter is10 in and 2m length The hot oil passes throughtubes while the water flows through the shell Thetube length is 4m and diameter is 6mm the numberof tubes is 24 and arranged in 2 passes

The CSP system instrumented with TT flowmeters (FT) and PT as shown in Fig 1 to monitorthe temperatures flow rates and pressure in bothsteam and oil loops

23 NF pretreatment

Fig 4(a) shows the P and I diagram of the NF sys-tem The system consisting of a dual media filter acartridge filter a high-pressure pump chemical injec-tion pumps and NF membrane One dual media filtervessel is installed with a specified feed flow rate of 15and 35 tonhr for back wash The vessel contains

Photo of double deck MSF desalination testpilot unit

(a) Process flow diagram of Double deck MSF(b)

Fig 2 MSF desalination unit with double deck

Fig 3 Four modules of concentrated solar collector inseries

6898 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

sand gravel and anthracite The cartridge filter of5micron is installed after high-pressure pump andjust before membrane section The membrane sectionconsists of 4 pressure vessels running in parallel eachvessel contains one membrane element of NF270 4040type The whole NF system except feed permeateand brine tanks is placed inside one container with itscontrol panel as shown in Fig 4(b) For the watersalinity samples are collected periodically to measureits conductivity using a mobile conductivity meter

3 Pilot test performance

31 NF test performance

The NF system testing is carried out using the sitebrackish water (TDS= 2000 ppm) On the other handa mathematical model of NF membrane is developedand verified against typical operating NF unit datausing the Visual design and simulation (VDS) soft-ware developed by the authors [13ndash16] The VDS sim-ulation results of NF system at different feed pressureof 8 and 10 bar and compared with experimentalresults as shown in Table 1 The NF performance wascarried out and assessed by the recovery ratio andsalts rejection The recovery ratio (permeatefeed)increases as the feed pressure increases The salt rejec-tion (1-(permeate salinityfeed salinity)) is calculatedas shown in Table 1 The salt rejection decreases asthe feed pressure increase due to increase in permeatesalinity The measured recovery ratio is slightly lowerthan the simulation results although the salt rejectionof experimental is lower than that of simulation Thedifference between measured values of permeate flow

salinity recovery ratio salt rejection and simulationresults is within acceptable range

32 Concentrated solar power (CSP) test performance

CPS system including solar collector and steamgenerator is simulated using VDS program The massand heat balance equations of solar collector steamgenerator and pumps are developed The oil andwater thermo physical properties correlations at dif-ferent temperature are considered in the program Thecharacteristic surface of collector reflectivity receiveremission and absorptivity and glass tube materialtransmittance are specified in the VDS program Thespecifications of the solar collector and steam genera-tor are defined and fed to the program The measuredweather conditions (solar intensity ambient tempera-ture and wind velocity) at each hour are fed to theprogram The duration time starts at 700 AM to 800PM with 1-hr step

Fig 5(a) shows the interface of VDS programresults at 400 PM The oil mass flow rate and temper-ature at inlet and outlet for both solar collector andthe steam generator are presented The collectedenergy is transferred to the steam generator to gener-ate 38 kghr saturated steam at 1138˚C The solarintensity (I) and calculated absorbed energy by thereceiver is shown in Fig 5(b) The difference is notice-able at mid-day time

Fig 6(a) shows comparison between the VDS sim-ulation and experimental results of oil temperaturerise through the solar collector along the day timeThe oil temperature difference increases as the solar

Fig 4 NF system

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6899

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

intensity increase while the maximum difference inmid-day reach to 25˚C Fig 6(b) shows a comparisonbetween the simulation and experimental results of

the oil temperature drop in the steam generator unitThe maximum heat transfer occurs at the mid-dayand the maximum temperature drop is 14˚C It is

Table 1Typical NF pretreatment experimental results compared with VDS results

VDS simulation Experimental diff VDS simulation Experimental diff

Feed pressure bar 8 ndash 10 ndash

Feed flow rate tonhr 1325 ndash 1375 ndash

Feed salinity ppm 2000 ndash 2000 ndash

Permeate flow rate tonhr 067 06 10 085 076 11

Permeate salinity ppm 648 600 7 761 650 15

Brine flow rate tonhr 065 0725 12 052 06 15

Brine salinity ppm 3395 3158 7 4027 3760 7

Recovery ratio 51 45 12 62 55 11

Salt rejection 676 70 4 62 68 10

Fig 5 (a) VDS simulation results of CSP system at 400 PM June 2012 and (b) Measured solar intensity and absorbedenergy by the receiver

Fig 6 Oil temperature variation through CSP trough amp steam generator

6900 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

noticed that at the same day time the temperaturedrop in steam generator unit is less than the tempera-ture rise in the solar collector This means that part ofgained energy in the collector is absorbed in the steamgenerator and the remaining is maintained with theoutlet oil stream from steam generator and come backthe collector This explains the increase in oil tempera-ture at the concentrated solar collector inlet throughthe day time

The CSP system average efficiency (gCSP) is calcu-lated as the average useful gained poweraveragesolar input power

gCSP frac14 mCp

oilethTooil TioilTHORN

I ACSP

eth1THORN

Fig 7 shows the simulation and experimentalresults of the collector efficiency through the day timeThe collector efficiency decreases along the day timedue to the increase in the average oil temperaturewhich increases the energy loss to the ambient Theexperimental collector efficiency shows relatively lowvalue than that of the simulation collector efficiencydue to (1) inaccurate tracking system that could notfollow the sun movements accurately and (2) the inef-ficient concentrated tubes location in the CSP focusand possible convection loss

Fig 8 shows comparison between the simulatedand the measured generated steam temperature Thewater inlet and steam exit valve remain closed whilethe oil valves are open to allow energy transfer fromoil to heat the enclosed water in the boiler The waterfeed and steam valves are opened when the watertemperature reaches 77˚C The generated steam tem-perature increases as the solar intensity increase andthe maximum temperature reached is 110˚C at mid-day time

The steam valve is opened at 100 PM at steamflow rate of 43 kghr The generated steam is directedto MSF desalination unit as a heating source The con-densate steam in the brine heater of MSF is feed backto the steam generator The amount of generatedsteam flow rate decreased linearly as shown in Fig 9this due to the decreasing in the solar collector effi-ciency The measured generated steam flow rateshows lower values than the simulation results due tothe thermal losses encountered through insufficientinsulation of steam generator and throughout the con-nection pipe between CSP outlet and steam generatorAs shown in Fig 9 the operation of the steam genera-tion extends up 11 PM this is due to heat storage inthe CSP system

Fig 10 shows the simulation and experimental val-ues of TBT variation through day time As the CSPsteam condenses in the brine heater of MSF the TBTraise up due to gained energy of the latent heatUnder the ambient and operating condition of June2012 the mid-day TBT reaches up to 100˚C while theCSP steam condenses at 106˚C that is 6˚C tempera-ture differences is maintained

Fig 7 Solar collector efficiency

Fig 8 Generated steam temperature

Fig 9 Generated steam flow rate

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6901

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

33 New MSF with De-aerator and Brine Mix (DBM)

The permeate water of NF system is directed tothe deaeration and brine mix tower where the feed issprayed for oxygen removal The deaerated water ismixed with part of the brine blow down then itpumped to the MSF condensers The brine mix feedabsorbs latent heat energy in condnesers before passthough brine heater where the brine reaches its TBTThe brine is then directed to the first flash chamberwhere flashing process occurs and vapor releases Thereleased vapor condenses to form a product waterThe flashing process occurs in the successive stagesuntil reach the last stage where the unflashed brineexit as brine blown down The condensate of all stagesis collected and directed to the water product tankThe brine level is adjusted above the interconnectingpipes (interstage gates) to guarantee the flash cham-bers sealing

Under the same feed saline water flow rate (NFpermeate) of 370 kghr and feed temperature of 27˚Cand controlling the brine mix ratio at 20ndash70 of theMSF brine blow down the distillate water is mea-sured and recorded at different TBTs as shown inFig 11 as compared with design calculated valuesThe pressure of saline water before the first chamberis controlled and fixed value of 15 bar absolute (abovesaturation conditions) by partially closing the valveAlso the orifices among chambers are controlled bypartially close the valve between each two successivechambers The in tube water velocity is controlled at2ms

The rate of both the design and the measured dis-tilled water increase as a result of TBT increases asshown in Fig 11 The amount of distillate is lowerthan expected this may be due to the partial loss offlashed vapor through vacuum system and the

irreversibility of flashing process occur within the ori-fices and weirs

Fig 12 shows the design and experimental GOR ofMSF variation with TBT This is defined as the ratiobetween distillate flow rate to the heating steam con-sumption GOR frac14 _m distillate

_m steam The average value of theunit design GOR is 17 which is almost twice the con-ventional MSF GOR The average measured GOR is15 as shown in Fig 12 The small difference betweenthe measured and designed value of GOR is due tothe lower distillate productivity under fixed amountof heating steam flow rate

The MSF specific power consumption (SPC) whichis defined as the ratio between the pumping powerconsumption (kW) and the rate of water distillate

(m3hr) SPC frac14 PumpingPower_m distillate Fig 13 shows that the

SPC decreases as the TBT increase mainly due to theincreases in the water productivity The experimentalSPC is calculated based on the measured distillateflow rate and the rated power consumption The

Fig 10 TBT variation Fig 11 MSF distillate productivity with TBT variation

Fig 12 The GOR of MSF system variation with TBT

6902 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

exprimental SPC is higher however than the designvalue is mainly due to lower experimental distillatefor the same saline water feed and may be due topressure drop in piping and valves that were not con-sidered properly in the design stage The SPC of testpilot unit is relatively higher than that of commercialvalue of large-scale MSF desalination plant due to thevery small test pilot unit productivity

In the next phase of the project it is expected torun MSF system at TBT= 130˚C using heating steam at140˚C from CSP system or auxiliary boiler for highTBT MSF Completing insulation for all piping con-nected CSP and MSF and oil tank need to be finalizedproper operation of tracking system of CSP needed tobe fix also completing installation of the storage tankand filling with molten salt Complete installation ofback up boiler of 10 kghr capacity in case of insuffi-cient steam from CSP at high TBT requirements Alsomatching and synchronize among PV and wind tur-bine electricity in 3 phases instead of 2 phase as perrequired for high-pressure pump of NF system andfeed pump of MSF desalination unit

4 Conclusion

bull A pilot test unit of CSP-NF-MSF-DBM water desali-nation of 10m3day capacity is designed andinstalled at Wadi AL-Natron Egypt

bull The VDS [13ndash16] is upgraded to consider the math-ematical model of CSP and steam generator andused as a process design tool for MSF-DBM NFsteam generator and CSP units

bull Comparison between the numerical simulation andthe experimental performance results of the pilotunitrsquo subsystems are relatively satisfactory

bull The newly developed NF-MSF-DBM configurationis tested under TBT= 100˚C and the GOR is esti-

mated as 15 which almost twice of traditional MSFunder the same operating conditions

bull There is ongoing adjustment to operate the plant atTBT= 130˚C to improve the overall plant perfor-mance with expected GOR=18

bull The newly high-performance NF-MSF-DBM desali-nation unit significantly reduces the unitrsquosinput thermal energy to suits integration with (therelatively expensive) RE as a desalination plantdriver

Acknowledgement

The authors acknowledge the financial support ofresearch development and innovation (RDI) programproject RE-NF-MSF ( C2-S1-148)

Nomenclature

ACSP mdash area of solar collector m2

I mdash solar intensity Wm2

m mdash mass flow rate kgs

MSF mdash multistage flash

MED mdash multieffect distillation

RO mdash reverse osmosis

NF mdash nano filtration

CSP mdash concentrated solar power

TBT mdash top brine temperature ˚C

GOR mdash gain output ratio

SPC mdash specific power consumption kWhm3

Cp mdash specific heat capacity kJkg k

TT mdash transmitter transducer

FT mdash flow meter transducer

PT mdash pressure transducer

References

[1] A Fried Water Industry Segment Report World Trade Cen-ter San Diego 2011 wwwwtcsdorg

[2] JJ Sadhwani JM Veza Desalination and energy consump-tion in Canary Islands Desalination 221 (2008) 143ndash150

[3] The European Academies Science Advisory Council (EASAC)report 2011 wwweasaceu

[4] G Fiorenza V Sharma G Braccio Technoeconomic evalua-tion of a solar powered water desalination plant Energy Con-ver Manage 44 (2003) 2217ndash2240

[5] Amr Mahmoud J Robert Abdelnasser Mabrouk A ImteyazA Nafey JS Choi JK Park S Nied J Detering New anti-sca-lant performance evaluation for MSF technology DesalinWater Treat 1944-39941944-3986 2012 httpdxdoiorg101080194439942012704701

[6] O Hamed A Hassan K Al-Shail K Bamardouf S Al-Sulami Ali Hamza M Farooque A Al-Rubaian HigherTBT more product less scaling Desalin Water Reuse Vol133 IDA NovDec 2003

[7] A Hassan Fully Integrated NF-Thermal Seawater Desalina-tion Process and Equipment US Patents No 20060157410A1 July 20 (2006)

Fig 13 SPC of MSF with TBT variation

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6903

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

multi-stage flash-reverse osmosis (MSF-RO) systemdriven by a dual purpose solar plant was installed inKuwait which produces 25m3 of water per day [1]Pilot projects for solar pond collectors have been car-ried out in addition to multiple effect distillation(MED) desalination plant operated for 18 years atUmm El-Nar (Abu Dhabi) with a 120m3day plant[1] In Saudi Arabia a PV-RO brackish water plantwas installed and connected to a solar still with a pro-duction of 5m3day [1] These were small-scale pro-jects that have contributed to the general consensusthat renewable desalination is a feasible option in theMiddle East The most popular combination is eitherMED with thermal collectors or RO with photovoltaics(PV) technology

The energy demand of seawater reverse osmosis(SWRO) plant located in Canary Islands of a capacityof 25000m3d is provided by a combination of PVcells (rooftop) with minor share of the grid which con-sist of energy mix including wind energy [2]

There are various concentrated solar power (CSP)technologies In 2011 13GW of CSP were operatingand a further 23GW were under construction [3]Currently base-load electricity generated by CSPplants costs two to three times that from existing fos-sil-based technologies without carbon capture andstorage CSP generation costs are competitive withphotovoltaics and offshore wind but are almost twotimes of onshore wind [3] A reduction of 50ndash60 inCSP generating cost may reasonably expected overthe next 10ndash15 years provided that research develop-ment and demonstration programs are sustaining thecommercial deployments of CSP plants It is antici-pated that CSP should become cost competitive withfossil-based load generation at some point between2020 and 2030 [3]

The energy cost of solar desalination is reported asan equivalent to 5ndash15 kWh of electricity per cubicmeter of water either directly by reverse osmosis orindirectly as pump losses and decreased efficiency inbackpressure turbines [4]

Multi stage flash (MSF) technology has proven tobe a mature thermal technology for large-scale capac-ity production with high-quality desalinated waterespecially for high severe feed water quality How-ever at high top brine temperature (TBT) scaledeposits of high seawater brine concentration presentsa real problem in MSFMED plants as it directlyaffect the heat transfer rates on the heating surface Athigher temperature greater than 120˚C nonalkalinecalcium sulfate (CaSO4) precipitates if saturation limitsare exceeded due to the inverse solubility of thesesalts with temperature Formation of alkaline scale(CaCO3 amp MgOH) can be controlled by lowering pH

(acid additives) or by adding antiscalant Nonalkaline(hard) scale (as CaSO4) is only controlled nowadaysby limiting TBT below 120˚C

Increasing TBT with hard scale control can be car-ried out by (1) introduce high-temperature antisca-lant (2) reducing hard scale ions to avoid it reachingsaturation conditions The first approach currently isunder experimental evaluation to check antiscalantperformance under different operating conditionsbefore introducing to the market [5] The secondapproach is considered in the literature [6ndash12]through the use of nanofiltration (NF) membrane sys-tem for make-up feed water pretreatment

The application of NF membrane in seawater desa-lination has recently gained significant attention in thedesalination industry due to selective removal of diva-lent ions NF is originally applied to reject electrolytesand obtain ultrapure water with high-volume flux atlow operating pressure because most membraneshave either positive or negative charge due to theircompositions

Extensive experiments have been carried out onMSF pilot test unit with NF as pretreatment [6ndash9] NFpressure was 24 bars and its recovery ratio rangedfrom 60 to 65 Almost 90 of the total concentrationof the sulfate and calcium ions is rejected with NFbrine while the MSF make-up entirely formed fromNF permeate was below their solubility limits Thisresult indicated the possibility to operate the MSFplant safely and without any scaling problem at TBTat or higher than 130˚C

A partial pretreatment of the MSF feed (25 ofmake-up) by introducing NF membrane was proposedand presented in [1011] A test has been carried outon one evaporator of Layyah MSF plant located inUAE The results showed an increase in the TBT from105 to 110˚C and the production from 1044 to 1253 th (199) The maximum production was 1260 th(206) at 117˚C with product conductivity of 454 lScm However based on personal communication theNF plant was shut down due to operational problemsin the pretreatment section

The NF membrane possesses molecular weightcutoff of about hundreds to a few thousands which isintermediate between RO membranes and ultrafiltra-tion (UF) The pore radii and fixed charge density ofpractical membranes were evaluated from permeationexperiments of different neutral solutes of sodiumchloride The pore radii of these NF membranes wereestimated to range from 04 to 08 nm [12]

Simulation of a novel integrating NF system withMSF-de-aerator and brine mix (DBM) desalinationplant at the TBT= 130˚C was carried out [13] Theprocess calculations showed that the gain output ratio

6896 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

(GOR) of the novel configuration is tow times of theconvention MSF-BR [13] The levelized water cost ofNF-MSF-DM (at TBT= 130˚C) was found to be 14lower than conventional MSF (at 110˚C) at the currentoil price of 104 $bbl [13] The novel NF-MSF-DBMconfiguration significantly reduces the unitrsquos inputthermal energy cost which suits integration with (therelatively expensive) RE

The objective of this work is to present the pilottest description and experimental performance resultsof CSP steam generator NF and MSF units Thewell-developed process design program (VDS) [13ndash16]is upgraded to consider the mathematical model ofCSP and steam generator and used as a processdesign tool for MSF-DBM NF steam generator andCSP unitsThe goal of MSF desalination pilot test unitis to verify the new concept of (DBM-MSF) at highTBT and high performance

2 Pilot test unit CSP-NF-MSF-DBM overview

The process design and simulation for the testpilot is developed to prepare specifications of differentcomponents Some units are manufactured by Egyp-tian contractor while other are purchased from ven-dors The site is prepared where civil work andfoundation are constructed The test pilot componentsare installed assembled and finally individual com-missioning for each component is carried out The siteis located at lsquoWadi El-Natrounrsquo remote area almost150 km southwest of Alexandria city (Egypt) The sitebelongs to Alexandria University

Fig 1 shows the pilot test of solar energy anddesalination units The concentrated solar parabolictrough with thermal energy storage facility providesthe necessary heating in order to generate requiredsteam of the MSF desalination unit The system isequipped also with a backup boiler for steam compen-sation Solar PV and wind turbine (not appeared inFig 1) are installed and run separately However in

this phase diesel engine is used to provide the pumpselectricity until finalize the match and synchronizationbetween PV and wind turbine

21 MSF desalination unit

The MSF pilot unit consists of 28 flashing cham-bers with 28 connected condensers as shown inFigs 2a and b The stages are arranged in double deckin order to reduce the foot print There are four setseach set consists of 7 stages MSF chambers areequipped with glass windows for monitoring theflashing process The shell material of MSF is fabri-cated from 2ndashmm-thick stainless steel 316 L The flashchamber is 05m length and 05m width while theheight is 10m The condenser tube is 8mm diameterand 6m long and made of stainless steel (07mmthick) The number of tubes is 2 per condenser andarranged in multipass inside 05m shell length Theunit is manufactured in Egyptian and assembled atthe project site

The orifice opening area controlled using gatevalve which is located between the flash chambersThe interstages valves controls the interstages flowrates to guarantee the brine flashing at each stage Thesplash plate is designed just above of the inlet open-ing to reduce carry over Demister is fixed near thevapour outlet vapor pipes to reject brine carry overbefore going to the condenser The shells are insulatedto minimize energy losses In addition to the brineheater different supporting systems are added includ-ing vacuum system and chemical injection systemsThe vacuum system has control valves at each stageto adjust the venting rate of noncondensable gases(NCG) and the stage pressure

The MSF is the main subsystem where distillationis produced using the flashing process Differentinstrumentations are installed to measure and recordthe temperatures pressures and flow rates as shownin Fig 1 In the heating section steam input and out-put temperatures in addition to pressure and flow are

Fig 1 Basic flow diagram of CSP-NF-MSF-DBM pilot plant

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6897

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

measured using proper transducers All chambers areequipped with temperature and pressure indicatorsFirst and last chambers are equipped with tempera-ture transmitter (TT) and pressure transmitter (PT)and additional two movable PT and TT are suppliedto be inserted in the chambers of the operator choiceInput seawater flow and output brine and distilledwater flow rates are measured using flow transmit-ters

22 Concentrated solar power (CSP) system

Four modules of solar concentrator (parabolictrough) are purchased and assembled in series at thesite of the project as shown in Fig 3 Each module is

36m length and 1524m width The collector area permodule is 56m2 while the collector reflective area is5m2 The assembled collector length becomes 146mwhile the total area is 224m2 The receiver absorptiv-ity is 092 the mirror reflectivity is 091 while thereceiver emittance is 023 The black-coated pipes are10 in diameter and placed in 20 in diameter glasspipes to minimize convection losses The concentratorshave a tracking system and were placed EastndashWestand facing south

The CSP system contains a steam generator to sup-ply the MSF brine heater with the required heatingsteam Thermal oil is circulated through the collectingpipes gains the solar thermal energy and flowthrough steam generator and energy storage tank

The steam generator consists of shell and tube andhas a separate vapor header The shell diameter is10 in and 2m length The hot oil passes throughtubes while the water flows through the shell Thetube length is 4m and diameter is 6mm the numberof tubes is 24 and arranged in 2 passes

The CSP system instrumented with TT flowmeters (FT) and PT as shown in Fig 1 to monitorthe temperatures flow rates and pressure in bothsteam and oil loops

23 NF pretreatment

Fig 4(a) shows the P and I diagram of the NF sys-tem The system consisting of a dual media filter acartridge filter a high-pressure pump chemical injec-tion pumps and NF membrane One dual media filtervessel is installed with a specified feed flow rate of 15and 35 tonhr for back wash The vessel contains

Photo of double deck MSF desalination testpilot unit

(a) Process flow diagram of Double deck MSF(b)

Fig 2 MSF desalination unit with double deck

Fig 3 Four modules of concentrated solar collector inseries

6898 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

sand gravel and anthracite The cartridge filter of5micron is installed after high-pressure pump andjust before membrane section The membrane sectionconsists of 4 pressure vessels running in parallel eachvessel contains one membrane element of NF270 4040type The whole NF system except feed permeateand brine tanks is placed inside one container with itscontrol panel as shown in Fig 4(b) For the watersalinity samples are collected periodically to measureits conductivity using a mobile conductivity meter

3 Pilot test performance

31 NF test performance

The NF system testing is carried out using the sitebrackish water (TDS= 2000 ppm) On the other handa mathematical model of NF membrane is developedand verified against typical operating NF unit datausing the Visual design and simulation (VDS) soft-ware developed by the authors [13ndash16] The VDS sim-ulation results of NF system at different feed pressureof 8 and 10 bar and compared with experimentalresults as shown in Table 1 The NF performance wascarried out and assessed by the recovery ratio andsalts rejection The recovery ratio (permeatefeed)increases as the feed pressure increases The salt rejec-tion (1-(permeate salinityfeed salinity)) is calculatedas shown in Table 1 The salt rejection decreases asthe feed pressure increase due to increase in permeatesalinity The measured recovery ratio is slightly lowerthan the simulation results although the salt rejectionof experimental is lower than that of simulation Thedifference between measured values of permeate flow

salinity recovery ratio salt rejection and simulationresults is within acceptable range

32 Concentrated solar power (CSP) test performance

CPS system including solar collector and steamgenerator is simulated using VDS program The massand heat balance equations of solar collector steamgenerator and pumps are developed The oil andwater thermo physical properties correlations at dif-ferent temperature are considered in the program Thecharacteristic surface of collector reflectivity receiveremission and absorptivity and glass tube materialtransmittance are specified in the VDS program Thespecifications of the solar collector and steam genera-tor are defined and fed to the program The measuredweather conditions (solar intensity ambient tempera-ture and wind velocity) at each hour are fed to theprogram The duration time starts at 700 AM to 800PM with 1-hr step

Fig 5(a) shows the interface of VDS programresults at 400 PM The oil mass flow rate and temper-ature at inlet and outlet for both solar collector andthe steam generator are presented The collectedenergy is transferred to the steam generator to gener-ate 38 kghr saturated steam at 1138˚C The solarintensity (I) and calculated absorbed energy by thereceiver is shown in Fig 5(b) The difference is notice-able at mid-day time

Fig 6(a) shows comparison between the VDS sim-ulation and experimental results of oil temperaturerise through the solar collector along the day timeThe oil temperature difference increases as the solar

Fig 4 NF system

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6899

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

intensity increase while the maximum difference inmid-day reach to 25˚C Fig 6(b) shows a comparisonbetween the simulation and experimental results of

the oil temperature drop in the steam generator unitThe maximum heat transfer occurs at the mid-dayand the maximum temperature drop is 14˚C It is

Table 1Typical NF pretreatment experimental results compared with VDS results

VDS simulation Experimental diff VDS simulation Experimental diff

Feed pressure bar 8 ndash 10 ndash

Feed flow rate tonhr 1325 ndash 1375 ndash

Feed salinity ppm 2000 ndash 2000 ndash

Permeate flow rate tonhr 067 06 10 085 076 11

Permeate salinity ppm 648 600 7 761 650 15

Brine flow rate tonhr 065 0725 12 052 06 15

Brine salinity ppm 3395 3158 7 4027 3760 7

Recovery ratio 51 45 12 62 55 11

Salt rejection 676 70 4 62 68 10

Fig 5 (a) VDS simulation results of CSP system at 400 PM June 2012 and (b) Measured solar intensity and absorbedenergy by the receiver

Fig 6 Oil temperature variation through CSP trough amp steam generator

6900 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

noticed that at the same day time the temperaturedrop in steam generator unit is less than the tempera-ture rise in the solar collector This means that part ofgained energy in the collector is absorbed in the steamgenerator and the remaining is maintained with theoutlet oil stream from steam generator and come backthe collector This explains the increase in oil tempera-ture at the concentrated solar collector inlet throughthe day time

The CSP system average efficiency (gCSP) is calcu-lated as the average useful gained poweraveragesolar input power

gCSP frac14 mCp

oilethTooil TioilTHORN

I ACSP

eth1THORN

Fig 7 shows the simulation and experimentalresults of the collector efficiency through the day timeThe collector efficiency decreases along the day timedue to the increase in the average oil temperaturewhich increases the energy loss to the ambient Theexperimental collector efficiency shows relatively lowvalue than that of the simulation collector efficiencydue to (1) inaccurate tracking system that could notfollow the sun movements accurately and (2) the inef-ficient concentrated tubes location in the CSP focusand possible convection loss

Fig 8 shows comparison between the simulatedand the measured generated steam temperature Thewater inlet and steam exit valve remain closed whilethe oil valves are open to allow energy transfer fromoil to heat the enclosed water in the boiler The waterfeed and steam valves are opened when the watertemperature reaches 77˚C The generated steam tem-perature increases as the solar intensity increase andthe maximum temperature reached is 110˚C at mid-day time

The steam valve is opened at 100 PM at steamflow rate of 43 kghr The generated steam is directedto MSF desalination unit as a heating source The con-densate steam in the brine heater of MSF is feed backto the steam generator The amount of generatedsteam flow rate decreased linearly as shown in Fig 9this due to the decreasing in the solar collector effi-ciency The measured generated steam flow rateshows lower values than the simulation results due tothe thermal losses encountered through insufficientinsulation of steam generator and throughout the con-nection pipe between CSP outlet and steam generatorAs shown in Fig 9 the operation of the steam genera-tion extends up 11 PM this is due to heat storage inthe CSP system

Fig 10 shows the simulation and experimental val-ues of TBT variation through day time As the CSPsteam condenses in the brine heater of MSF the TBTraise up due to gained energy of the latent heatUnder the ambient and operating condition of June2012 the mid-day TBT reaches up to 100˚C while theCSP steam condenses at 106˚C that is 6˚C tempera-ture differences is maintained

Fig 7 Solar collector efficiency

Fig 8 Generated steam temperature

Fig 9 Generated steam flow rate

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6901

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

33 New MSF with De-aerator and Brine Mix (DBM)

The permeate water of NF system is directed tothe deaeration and brine mix tower where the feed issprayed for oxygen removal The deaerated water ismixed with part of the brine blow down then itpumped to the MSF condensers The brine mix feedabsorbs latent heat energy in condnesers before passthough brine heater where the brine reaches its TBTThe brine is then directed to the first flash chamberwhere flashing process occurs and vapor releases Thereleased vapor condenses to form a product waterThe flashing process occurs in the successive stagesuntil reach the last stage where the unflashed brineexit as brine blown down The condensate of all stagesis collected and directed to the water product tankThe brine level is adjusted above the interconnectingpipes (interstage gates) to guarantee the flash cham-bers sealing

Under the same feed saline water flow rate (NFpermeate) of 370 kghr and feed temperature of 27˚Cand controlling the brine mix ratio at 20ndash70 of theMSF brine blow down the distillate water is mea-sured and recorded at different TBTs as shown inFig 11 as compared with design calculated valuesThe pressure of saline water before the first chamberis controlled and fixed value of 15 bar absolute (abovesaturation conditions) by partially closing the valveAlso the orifices among chambers are controlled bypartially close the valve between each two successivechambers The in tube water velocity is controlled at2ms

The rate of both the design and the measured dis-tilled water increase as a result of TBT increases asshown in Fig 11 The amount of distillate is lowerthan expected this may be due to the partial loss offlashed vapor through vacuum system and the

irreversibility of flashing process occur within the ori-fices and weirs

Fig 12 shows the design and experimental GOR ofMSF variation with TBT This is defined as the ratiobetween distillate flow rate to the heating steam con-sumption GOR frac14 _m distillate

_m steam The average value of theunit design GOR is 17 which is almost twice the con-ventional MSF GOR The average measured GOR is15 as shown in Fig 12 The small difference betweenthe measured and designed value of GOR is due tothe lower distillate productivity under fixed amountof heating steam flow rate

The MSF specific power consumption (SPC) whichis defined as the ratio between the pumping powerconsumption (kW) and the rate of water distillate

(m3hr) SPC frac14 PumpingPower_m distillate Fig 13 shows that the

SPC decreases as the TBT increase mainly due to theincreases in the water productivity The experimentalSPC is calculated based on the measured distillateflow rate and the rated power consumption The

Fig 10 TBT variation Fig 11 MSF distillate productivity with TBT variation

Fig 12 The GOR of MSF system variation with TBT

6902 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

exprimental SPC is higher however than the designvalue is mainly due to lower experimental distillatefor the same saline water feed and may be due topressure drop in piping and valves that were not con-sidered properly in the design stage The SPC of testpilot unit is relatively higher than that of commercialvalue of large-scale MSF desalination plant due to thevery small test pilot unit productivity

In the next phase of the project it is expected torun MSF system at TBT= 130˚C using heating steam at140˚C from CSP system or auxiliary boiler for highTBT MSF Completing insulation for all piping con-nected CSP and MSF and oil tank need to be finalizedproper operation of tracking system of CSP needed tobe fix also completing installation of the storage tankand filling with molten salt Complete installation ofback up boiler of 10 kghr capacity in case of insuffi-cient steam from CSP at high TBT requirements Alsomatching and synchronize among PV and wind tur-bine electricity in 3 phases instead of 2 phase as perrequired for high-pressure pump of NF system andfeed pump of MSF desalination unit

4 Conclusion

bull A pilot test unit of CSP-NF-MSF-DBM water desali-nation of 10m3day capacity is designed andinstalled at Wadi AL-Natron Egypt

bull The VDS [13ndash16] is upgraded to consider the math-ematical model of CSP and steam generator andused as a process design tool for MSF-DBM NFsteam generator and CSP units

bull Comparison between the numerical simulation andthe experimental performance results of the pilotunitrsquo subsystems are relatively satisfactory

bull The newly developed NF-MSF-DBM configurationis tested under TBT= 100˚C and the GOR is esti-

mated as 15 which almost twice of traditional MSFunder the same operating conditions

bull There is ongoing adjustment to operate the plant atTBT= 130˚C to improve the overall plant perfor-mance with expected GOR=18

bull The newly high-performance NF-MSF-DBM desali-nation unit significantly reduces the unitrsquosinput thermal energy to suits integration with (therelatively expensive) RE as a desalination plantdriver

Acknowledgement

The authors acknowledge the financial support ofresearch development and innovation (RDI) programproject RE-NF-MSF ( C2-S1-148)

Nomenclature

ACSP mdash area of solar collector m2

I mdash solar intensity Wm2

m mdash mass flow rate kgs

MSF mdash multistage flash

MED mdash multieffect distillation

RO mdash reverse osmosis

NF mdash nano filtration

CSP mdash concentrated solar power

TBT mdash top brine temperature ˚C

GOR mdash gain output ratio

SPC mdash specific power consumption kWhm3

Cp mdash specific heat capacity kJkg k

TT mdash transmitter transducer

FT mdash flow meter transducer

PT mdash pressure transducer

References

[1] A Fried Water Industry Segment Report World Trade Cen-ter San Diego 2011 wwwwtcsdorg

[2] JJ Sadhwani JM Veza Desalination and energy consump-tion in Canary Islands Desalination 221 (2008) 143ndash150

[3] The European Academies Science Advisory Council (EASAC)report 2011 wwweasaceu

[4] G Fiorenza V Sharma G Braccio Technoeconomic evalua-tion of a solar powered water desalination plant Energy Con-ver Manage 44 (2003) 2217ndash2240

[5] Amr Mahmoud J Robert Abdelnasser Mabrouk A ImteyazA Nafey JS Choi JK Park S Nied J Detering New anti-sca-lant performance evaluation for MSF technology DesalinWater Treat 1944-39941944-3986 2012 httpdxdoiorg101080194439942012704701

[6] O Hamed A Hassan K Al-Shail K Bamardouf S Al-Sulami Ali Hamza M Farooque A Al-Rubaian HigherTBT more product less scaling Desalin Water Reuse Vol133 IDA NovDec 2003

[7] A Hassan Fully Integrated NF-Thermal Seawater Desalina-tion Process and Equipment US Patents No 20060157410A1 July 20 (2006)

Fig 13 SPC of MSF with TBT variation

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6903

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

(GOR) of the novel configuration is tow times of theconvention MSF-BR [13] The levelized water cost ofNF-MSF-DM (at TBT= 130˚C) was found to be 14lower than conventional MSF (at 110˚C) at the currentoil price of 104 $bbl [13] The novel NF-MSF-DBMconfiguration significantly reduces the unitrsquos inputthermal energy cost which suits integration with (therelatively expensive) RE

The objective of this work is to present the pilottest description and experimental performance resultsof CSP steam generator NF and MSF units Thewell-developed process design program (VDS) [13ndash16]is upgraded to consider the mathematical model ofCSP and steam generator and used as a processdesign tool for MSF-DBM NF steam generator andCSP unitsThe goal of MSF desalination pilot test unitis to verify the new concept of (DBM-MSF) at highTBT and high performance

2 Pilot test unit CSP-NF-MSF-DBM overview

The process design and simulation for the testpilot is developed to prepare specifications of differentcomponents Some units are manufactured by Egyp-tian contractor while other are purchased from ven-dors The site is prepared where civil work andfoundation are constructed The test pilot componentsare installed assembled and finally individual com-missioning for each component is carried out The siteis located at lsquoWadi El-Natrounrsquo remote area almost150 km southwest of Alexandria city (Egypt) The sitebelongs to Alexandria University

Fig 1 shows the pilot test of solar energy anddesalination units The concentrated solar parabolictrough with thermal energy storage facility providesthe necessary heating in order to generate requiredsteam of the MSF desalination unit The system isequipped also with a backup boiler for steam compen-sation Solar PV and wind turbine (not appeared inFig 1) are installed and run separately However in

this phase diesel engine is used to provide the pumpselectricity until finalize the match and synchronizationbetween PV and wind turbine

21 MSF desalination unit

The MSF pilot unit consists of 28 flashing cham-bers with 28 connected condensers as shown inFigs 2a and b The stages are arranged in double deckin order to reduce the foot print There are four setseach set consists of 7 stages MSF chambers areequipped with glass windows for monitoring theflashing process The shell material of MSF is fabri-cated from 2ndashmm-thick stainless steel 316 L The flashchamber is 05m length and 05m width while theheight is 10m The condenser tube is 8mm diameterand 6m long and made of stainless steel (07mmthick) The number of tubes is 2 per condenser andarranged in multipass inside 05m shell length Theunit is manufactured in Egyptian and assembled atthe project site

The orifice opening area controlled using gatevalve which is located between the flash chambersThe interstages valves controls the interstages flowrates to guarantee the brine flashing at each stage Thesplash plate is designed just above of the inlet open-ing to reduce carry over Demister is fixed near thevapour outlet vapor pipes to reject brine carry overbefore going to the condenser The shells are insulatedto minimize energy losses In addition to the brineheater different supporting systems are added includ-ing vacuum system and chemical injection systemsThe vacuum system has control valves at each stageto adjust the venting rate of noncondensable gases(NCG) and the stage pressure

The MSF is the main subsystem where distillationis produced using the flashing process Differentinstrumentations are installed to measure and recordthe temperatures pressures and flow rates as shownin Fig 1 In the heating section steam input and out-put temperatures in addition to pressure and flow are

Fig 1 Basic flow diagram of CSP-NF-MSF-DBM pilot plant

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6897

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

measured using proper transducers All chambers areequipped with temperature and pressure indicatorsFirst and last chambers are equipped with tempera-ture transmitter (TT) and pressure transmitter (PT)and additional two movable PT and TT are suppliedto be inserted in the chambers of the operator choiceInput seawater flow and output brine and distilledwater flow rates are measured using flow transmit-ters

22 Concentrated solar power (CSP) system

Four modules of solar concentrator (parabolictrough) are purchased and assembled in series at thesite of the project as shown in Fig 3 Each module is

36m length and 1524m width The collector area permodule is 56m2 while the collector reflective area is5m2 The assembled collector length becomes 146mwhile the total area is 224m2 The receiver absorptiv-ity is 092 the mirror reflectivity is 091 while thereceiver emittance is 023 The black-coated pipes are10 in diameter and placed in 20 in diameter glasspipes to minimize convection losses The concentratorshave a tracking system and were placed EastndashWestand facing south

The CSP system contains a steam generator to sup-ply the MSF brine heater with the required heatingsteam Thermal oil is circulated through the collectingpipes gains the solar thermal energy and flowthrough steam generator and energy storage tank

The steam generator consists of shell and tube andhas a separate vapor header The shell diameter is10 in and 2m length The hot oil passes throughtubes while the water flows through the shell Thetube length is 4m and diameter is 6mm the numberof tubes is 24 and arranged in 2 passes

The CSP system instrumented with TT flowmeters (FT) and PT as shown in Fig 1 to monitorthe temperatures flow rates and pressure in bothsteam and oil loops

23 NF pretreatment

Fig 4(a) shows the P and I diagram of the NF sys-tem The system consisting of a dual media filter acartridge filter a high-pressure pump chemical injec-tion pumps and NF membrane One dual media filtervessel is installed with a specified feed flow rate of 15and 35 tonhr for back wash The vessel contains

Photo of double deck MSF desalination testpilot unit

(a) Process flow diagram of Double deck MSF(b)

Fig 2 MSF desalination unit with double deck

Fig 3 Four modules of concentrated solar collector inseries

6898 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

sand gravel and anthracite The cartridge filter of5micron is installed after high-pressure pump andjust before membrane section The membrane sectionconsists of 4 pressure vessels running in parallel eachvessel contains one membrane element of NF270 4040type The whole NF system except feed permeateand brine tanks is placed inside one container with itscontrol panel as shown in Fig 4(b) For the watersalinity samples are collected periodically to measureits conductivity using a mobile conductivity meter

3 Pilot test performance

31 NF test performance

The NF system testing is carried out using the sitebrackish water (TDS= 2000 ppm) On the other handa mathematical model of NF membrane is developedand verified against typical operating NF unit datausing the Visual design and simulation (VDS) soft-ware developed by the authors [13ndash16] The VDS sim-ulation results of NF system at different feed pressureof 8 and 10 bar and compared with experimentalresults as shown in Table 1 The NF performance wascarried out and assessed by the recovery ratio andsalts rejection The recovery ratio (permeatefeed)increases as the feed pressure increases The salt rejec-tion (1-(permeate salinityfeed salinity)) is calculatedas shown in Table 1 The salt rejection decreases asthe feed pressure increase due to increase in permeatesalinity The measured recovery ratio is slightly lowerthan the simulation results although the salt rejectionof experimental is lower than that of simulation Thedifference between measured values of permeate flow

salinity recovery ratio salt rejection and simulationresults is within acceptable range

32 Concentrated solar power (CSP) test performance

CPS system including solar collector and steamgenerator is simulated using VDS program The massand heat balance equations of solar collector steamgenerator and pumps are developed The oil andwater thermo physical properties correlations at dif-ferent temperature are considered in the program Thecharacteristic surface of collector reflectivity receiveremission and absorptivity and glass tube materialtransmittance are specified in the VDS program Thespecifications of the solar collector and steam genera-tor are defined and fed to the program The measuredweather conditions (solar intensity ambient tempera-ture and wind velocity) at each hour are fed to theprogram The duration time starts at 700 AM to 800PM with 1-hr step

Fig 5(a) shows the interface of VDS programresults at 400 PM The oil mass flow rate and temper-ature at inlet and outlet for both solar collector andthe steam generator are presented The collectedenergy is transferred to the steam generator to gener-ate 38 kghr saturated steam at 1138˚C The solarintensity (I) and calculated absorbed energy by thereceiver is shown in Fig 5(b) The difference is notice-able at mid-day time

Fig 6(a) shows comparison between the VDS sim-ulation and experimental results of oil temperaturerise through the solar collector along the day timeThe oil temperature difference increases as the solar

Fig 4 NF system

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6899

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

intensity increase while the maximum difference inmid-day reach to 25˚C Fig 6(b) shows a comparisonbetween the simulation and experimental results of

the oil temperature drop in the steam generator unitThe maximum heat transfer occurs at the mid-dayand the maximum temperature drop is 14˚C It is

Table 1Typical NF pretreatment experimental results compared with VDS results

VDS simulation Experimental diff VDS simulation Experimental diff

Feed pressure bar 8 ndash 10 ndash

Feed flow rate tonhr 1325 ndash 1375 ndash

Feed salinity ppm 2000 ndash 2000 ndash

Permeate flow rate tonhr 067 06 10 085 076 11

Permeate salinity ppm 648 600 7 761 650 15

Brine flow rate tonhr 065 0725 12 052 06 15

Brine salinity ppm 3395 3158 7 4027 3760 7

Recovery ratio 51 45 12 62 55 11

Salt rejection 676 70 4 62 68 10

Fig 5 (a) VDS simulation results of CSP system at 400 PM June 2012 and (b) Measured solar intensity and absorbedenergy by the receiver

Fig 6 Oil temperature variation through CSP trough amp steam generator

6900 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

noticed that at the same day time the temperaturedrop in steam generator unit is less than the tempera-ture rise in the solar collector This means that part ofgained energy in the collector is absorbed in the steamgenerator and the remaining is maintained with theoutlet oil stream from steam generator and come backthe collector This explains the increase in oil tempera-ture at the concentrated solar collector inlet throughthe day time

The CSP system average efficiency (gCSP) is calcu-lated as the average useful gained poweraveragesolar input power

gCSP frac14 mCp

oilethTooil TioilTHORN

I ACSP

eth1THORN

Fig 7 shows the simulation and experimentalresults of the collector efficiency through the day timeThe collector efficiency decreases along the day timedue to the increase in the average oil temperaturewhich increases the energy loss to the ambient Theexperimental collector efficiency shows relatively lowvalue than that of the simulation collector efficiencydue to (1) inaccurate tracking system that could notfollow the sun movements accurately and (2) the inef-ficient concentrated tubes location in the CSP focusand possible convection loss

Fig 8 shows comparison between the simulatedand the measured generated steam temperature Thewater inlet and steam exit valve remain closed whilethe oil valves are open to allow energy transfer fromoil to heat the enclosed water in the boiler The waterfeed and steam valves are opened when the watertemperature reaches 77˚C The generated steam tem-perature increases as the solar intensity increase andthe maximum temperature reached is 110˚C at mid-day time

The steam valve is opened at 100 PM at steamflow rate of 43 kghr The generated steam is directedto MSF desalination unit as a heating source The con-densate steam in the brine heater of MSF is feed backto the steam generator The amount of generatedsteam flow rate decreased linearly as shown in Fig 9this due to the decreasing in the solar collector effi-ciency The measured generated steam flow rateshows lower values than the simulation results due tothe thermal losses encountered through insufficientinsulation of steam generator and throughout the con-nection pipe between CSP outlet and steam generatorAs shown in Fig 9 the operation of the steam genera-tion extends up 11 PM this is due to heat storage inthe CSP system

Fig 10 shows the simulation and experimental val-ues of TBT variation through day time As the CSPsteam condenses in the brine heater of MSF the TBTraise up due to gained energy of the latent heatUnder the ambient and operating condition of June2012 the mid-day TBT reaches up to 100˚C while theCSP steam condenses at 106˚C that is 6˚C tempera-ture differences is maintained

Fig 7 Solar collector efficiency

Fig 8 Generated steam temperature

Fig 9 Generated steam flow rate

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6901

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

33 New MSF with De-aerator and Brine Mix (DBM)

The permeate water of NF system is directed tothe deaeration and brine mix tower where the feed issprayed for oxygen removal The deaerated water ismixed with part of the brine blow down then itpumped to the MSF condensers The brine mix feedabsorbs latent heat energy in condnesers before passthough brine heater where the brine reaches its TBTThe brine is then directed to the first flash chamberwhere flashing process occurs and vapor releases Thereleased vapor condenses to form a product waterThe flashing process occurs in the successive stagesuntil reach the last stage where the unflashed brineexit as brine blown down The condensate of all stagesis collected and directed to the water product tankThe brine level is adjusted above the interconnectingpipes (interstage gates) to guarantee the flash cham-bers sealing

Under the same feed saline water flow rate (NFpermeate) of 370 kghr and feed temperature of 27˚Cand controlling the brine mix ratio at 20ndash70 of theMSF brine blow down the distillate water is mea-sured and recorded at different TBTs as shown inFig 11 as compared with design calculated valuesThe pressure of saline water before the first chamberis controlled and fixed value of 15 bar absolute (abovesaturation conditions) by partially closing the valveAlso the orifices among chambers are controlled bypartially close the valve between each two successivechambers The in tube water velocity is controlled at2ms

The rate of both the design and the measured dis-tilled water increase as a result of TBT increases asshown in Fig 11 The amount of distillate is lowerthan expected this may be due to the partial loss offlashed vapor through vacuum system and the

irreversibility of flashing process occur within the ori-fices and weirs

Fig 12 shows the design and experimental GOR ofMSF variation with TBT This is defined as the ratiobetween distillate flow rate to the heating steam con-sumption GOR frac14 _m distillate

_m steam The average value of theunit design GOR is 17 which is almost twice the con-ventional MSF GOR The average measured GOR is15 as shown in Fig 12 The small difference betweenthe measured and designed value of GOR is due tothe lower distillate productivity under fixed amountof heating steam flow rate

The MSF specific power consumption (SPC) whichis defined as the ratio between the pumping powerconsumption (kW) and the rate of water distillate

(m3hr) SPC frac14 PumpingPower_m distillate Fig 13 shows that the

SPC decreases as the TBT increase mainly due to theincreases in the water productivity The experimentalSPC is calculated based on the measured distillateflow rate and the rated power consumption The

Fig 10 TBT variation Fig 11 MSF distillate productivity with TBT variation

Fig 12 The GOR of MSF system variation with TBT

6902 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

exprimental SPC is higher however than the designvalue is mainly due to lower experimental distillatefor the same saline water feed and may be due topressure drop in piping and valves that were not con-sidered properly in the design stage The SPC of testpilot unit is relatively higher than that of commercialvalue of large-scale MSF desalination plant due to thevery small test pilot unit productivity

In the next phase of the project it is expected torun MSF system at TBT= 130˚C using heating steam at140˚C from CSP system or auxiliary boiler for highTBT MSF Completing insulation for all piping con-nected CSP and MSF and oil tank need to be finalizedproper operation of tracking system of CSP needed tobe fix also completing installation of the storage tankand filling with molten salt Complete installation ofback up boiler of 10 kghr capacity in case of insuffi-cient steam from CSP at high TBT requirements Alsomatching and synchronize among PV and wind tur-bine electricity in 3 phases instead of 2 phase as perrequired for high-pressure pump of NF system andfeed pump of MSF desalination unit

4 Conclusion

bull A pilot test unit of CSP-NF-MSF-DBM water desali-nation of 10m3day capacity is designed andinstalled at Wadi AL-Natron Egypt

bull The VDS [13ndash16] is upgraded to consider the math-ematical model of CSP and steam generator andused as a process design tool for MSF-DBM NFsteam generator and CSP units

bull Comparison between the numerical simulation andthe experimental performance results of the pilotunitrsquo subsystems are relatively satisfactory

bull The newly developed NF-MSF-DBM configurationis tested under TBT= 100˚C and the GOR is esti-

mated as 15 which almost twice of traditional MSFunder the same operating conditions

bull There is ongoing adjustment to operate the plant atTBT= 130˚C to improve the overall plant perfor-mance with expected GOR=18

bull The newly high-performance NF-MSF-DBM desali-nation unit significantly reduces the unitrsquosinput thermal energy to suits integration with (therelatively expensive) RE as a desalination plantdriver

Acknowledgement

The authors acknowledge the financial support ofresearch development and innovation (RDI) programproject RE-NF-MSF ( C2-S1-148)

Nomenclature

ACSP mdash area of solar collector m2

I mdash solar intensity Wm2

m mdash mass flow rate kgs

MSF mdash multistage flash

MED mdash multieffect distillation

RO mdash reverse osmosis

NF mdash nano filtration

CSP mdash concentrated solar power

TBT mdash top brine temperature ˚C

GOR mdash gain output ratio

SPC mdash specific power consumption kWhm3

Cp mdash specific heat capacity kJkg k

TT mdash transmitter transducer

FT mdash flow meter transducer

PT mdash pressure transducer

References

[1] A Fried Water Industry Segment Report World Trade Cen-ter San Diego 2011 wwwwtcsdorg

[2] JJ Sadhwani JM Veza Desalination and energy consump-tion in Canary Islands Desalination 221 (2008) 143ndash150

[3] The European Academies Science Advisory Council (EASAC)report 2011 wwweasaceu

[4] G Fiorenza V Sharma G Braccio Technoeconomic evalua-tion of a solar powered water desalination plant Energy Con-ver Manage 44 (2003) 2217ndash2240

[5] Amr Mahmoud J Robert Abdelnasser Mabrouk A ImteyazA Nafey JS Choi JK Park S Nied J Detering New anti-sca-lant performance evaluation for MSF technology DesalinWater Treat 1944-39941944-3986 2012 httpdxdoiorg101080194439942012704701

[6] O Hamed A Hassan K Al-Shail K Bamardouf S Al-Sulami Ali Hamza M Farooque A Al-Rubaian HigherTBT more product less scaling Desalin Water Reuse Vol133 IDA NovDec 2003

[7] A Hassan Fully Integrated NF-Thermal Seawater Desalina-tion Process and Equipment US Patents No 20060157410A1 July 20 (2006)

Fig 13 SPC of MSF with TBT variation

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6903

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

measured using proper transducers All chambers areequipped with temperature and pressure indicatorsFirst and last chambers are equipped with tempera-ture transmitter (TT) and pressure transmitter (PT)and additional two movable PT and TT are suppliedto be inserted in the chambers of the operator choiceInput seawater flow and output brine and distilledwater flow rates are measured using flow transmit-ters

22 Concentrated solar power (CSP) system

Four modules of solar concentrator (parabolictrough) are purchased and assembled in series at thesite of the project as shown in Fig 3 Each module is

36m length and 1524m width The collector area permodule is 56m2 while the collector reflective area is5m2 The assembled collector length becomes 146mwhile the total area is 224m2 The receiver absorptiv-ity is 092 the mirror reflectivity is 091 while thereceiver emittance is 023 The black-coated pipes are10 in diameter and placed in 20 in diameter glasspipes to minimize convection losses The concentratorshave a tracking system and were placed EastndashWestand facing south

The CSP system contains a steam generator to sup-ply the MSF brine heater with the required heatingsteam Thermal oil is circulated through the collectingpipes gains the solar thermal energy and flowthrough steam generator and energy storage tank

The steam generator consists of shell and tube andhas a separate vapor header The shell diameter is10 in and 2m length The hot oil passes throughtubes while the water flows through the shell Thetube length is 4m and diameter is 6mm the numberof tubes is 24 and arranged in 2 passes

The CSP system instrumented with TT flowmeters (FT) and PT as shown in Fig 1 to monitorthe temperatures flow rates and pressure in bothsteam and oil loops

23 NF pretreatment

Fig 4(a) shows the P and I diagram of the NF sys-tem The system consisting of a dual media filter acartridge filter a high-pressure pump chemical injec-tion pumps and NF membrane One dual media filtervessel is installed with a specified feed flow rate of 15and 35 tonhr for back wash The vessel contains

Photo of double deck MSF desalination testpilot unit

(a) Process flow diagram of Double deck MSF(b)

Fig 2 MSF desalination unit with double deck

Fig 3 Four modules of concentrated solar collector inseries

6898 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

sand gravel and anthracite The cartridge filter of5micron is installed after high-pressure pump andjust before membrane section The membrane sectionconsists of 4 pressure vessels running in parallel eachvessel contains one membrane element of NF270 4040type The whole NF system except feed permeateand brine tanks is placed inside one container with itscontrol panel as shown in Fig 4(b) For the watersalinity samples are collected periodically to measureits conductivity using a mobile conductivity meter

3 Pilot test performance

31 NF test performance

The NF system testing is carried out using the sitebrackish water (TDS= 2000 ppm) On the other handa mathematical model of NF membrane is developedand verified against typical operating NF unit datausing the Visual design and simulation (VDS) soft-ware developed by the authors [13ndash16] The VDS sim-ulation results of NF system at different feed pressureof 8 and 10 bar and compared with experimentalresults as shown in Table 1 The NF performance wascarried out and assessed by the recovery ratio andsalts rejection The recovery ratio (permeatefeed)increases as the feed pressure increases The salt rejec-tion (1-(permeate salinityfeed salinity)) is calculatedas shown in Table 1 The salt rejection decreases asthe feed pressure increase due to increase in permeatesalinity The measured recovery ratio is slightly lowerthan the simulation results although the salt rejectionof experimental is lower than that of simulation Thedifference between measured values of permeate flow

salinity recovery ratio salt rejection and simulationresults is within acceptable range

32 Concentrated solar power (CSP) test performance

CPS system including solar collector and steamgenerator is simulated using VDS program The massand heat balance equations of solar collector steamgenerator and pumps are developed The oil andwater thermo physical properties correlations at dif-ferent temperature are considered in the program Thecharacteristic surface of collector reflectivity receiveremission and absorptivity and glass tube materialtransmittance are specified in the VDS program Thespecifications of the solar collector and steam genera-tor are defined and fed to the program The measuredweather conditions (solar intensity ambient tempera-ture and wind velocity) at each hour are fed to theprogram The duration time starts at 700 AM to 800PM with 1-hr step

Fig 5(a) shows the interface of VDS programresults at 400 PM The oil mass flow rate and temper-ature at inlet and outlet for both solar collector andthe steam generator are presented The collectedenergy is transferred to the steam generator to gener-ate 38 kghr saturated steam at 1138˚C The solarintensity (I) and calculated absorbed energy by thereceiver is shown in Fig 5(b) The difference is notice-able at mid-day time

Fig 6(a) shows comparison between the VDS sim-ulation and experimental results of oil temperaturerise through the solar collector along the day timeThe oil temperature difference increases as the solar

Fig 4 NF system

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6899

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

intensity increase while the maximum difference inmid-day reach to 25˚C Fig 6(b) shows a comparisonbetween the simulation and experimental results of

the oil temperature drop in the steam generator unitThe maximum heat transfer occurs at the mid-dayand the maximum temperature drop is 14˚C It is

Table 1Typical NF pretreatment experimental results compared with VDS results

VDS simulation Experimental diff VDS simulation Experimental diff

Feed pressure bar 8 ndash 10 ndash

Feed flow rate tonhr 1325 ndash 1375 ndash

Feed salinity ppm 2000 ndash 2000 ndash

Permeate flow rate tonhr 067 06 10 085 076 11

Permeate salinity ppm 648 600 7 761 650 15

Brine flow rate tonhr 065 0725 12 052 06 15

Brine salinity ppm 3395 3158 7 4027 3760 7

Recovery ratio 51 45 12 62 55 11

Salt rejection 676 70 4 62 68 10

Fig 5 (a) VDS simulation results of CSP system at 400 PM June 2012 and (b) Measured solar intensity and absorbedenergy by the receiver

Fig 6 Oil temperature variation through CSP trough amp steam generator

6900 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

noticed that at the same day time the temperaturedrop in steam generator unit is less than the tempera-ture rise in the solar collector This means that part ofgained energy in the collector is absorbed in the steamgenerator and the remaining is maintained with theoutlet oil stream from steam generator and come backthe collector This explains the increase in oil tempera-ture at the concentrated solar collector inlet throughthe day time

The CSP system average efficiency (gCSP) is calcu-lated as the average useful gained poweraveragesolar input power

gCSP frac14 mCp

oilethTooil TioilTHORN

I ACSP

eth1THORN

Fig 7 shows the simulation and experimentalresults of the collector efficiency through the day timeThe collector efficiency decreases along the day timedue to the increase in the average oil temperaturewhich increases the energy loss to the ambient Theexperimental collector efficiency shows relatively lowvalue than that of the simulation collector efficiencydue to (1) inaccurate tracking system that could notfollow the sun movements accurately and (2) the inef-ficient concentrated tubes location in the CSP focusand possible convection loss

Fig 8 shows comparison between the simulatedand the measured generated steam temperature Thewater inlet and steam exit valve remain closed whilethe oil valves are open to allow energy transfer fromoil to heat the enclosed water in the boiler The waterfeed and steam valves are opened when the watertemperature reaches 77˚C The generated steam tem-perature increases as the solar intensity increase andthe maximum temperature reached is 110˚C at mid-day time

The steam valve is opened at 100 PM at steamflow rate of 43 kghr The generated steam is directedto MSF desalination unit as a heating source The con-densate steam in the brine heater of MSF is feed backto the steam generator The amount of generatedsteam flow rate decreased linearly as shown in Fig 9this due to the decreasing in the solar collector effi-ciency The measured generated steam flow rateshows lower values than the simulation results due tothe thermal losses encountered through insufficientinsulation of steam generator and throughout the con-nection pipe between CSP outlet and steam generatorAs shown in Fig 9 the operation of the steam genera-tion extends up 11 PM this is due to heat storage inthe CSP system

Fig 10 shows the simulation and experimental val-ues of TBT variation through day time As the CSPsteam condenses in the brine heater of MSF the TBTraise up due to gained energy of the latent heatUnder the ambient and operating condition of June2012 the mid-day TBT reaches up to 100˚C while theCSP steam condenses at 106˚C that is 6˚C tempera-ture differences is maintained

Fig 7 Solar collector efficiency

Fig 8 Generated steam temperature

Fig 9 Generated steam flow rate

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6901

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

33 New MSF with De-aerator and Brine Mix (DBM)

The permeate water of NF system is directed tothe deaeration and brine mix tower where the feed issprayed for oxygen removal The deaerated water ismixed with part of the brine blow down then itpumped to the MSF condensers The brine mix feedabsorbs latent heat energy in condnesers before passthough brine heater where the brine reaches its TBTThe brine is then directed to the first flash chamberwhere flashing process occurs and vapor releases Thereleased vapor condenses to form a product waterThe flashing process occurs in the successive stagesuntil reach the last stage where the unflashed brineexit as brine blown down The condensate of all stagesis collected and directed to the water product tankThe brine level is adjusted above the interconnectingpipes (interstage gates) to guarantee the flash cham-bers sealing

Under the same feed saline water flow rate (NFpermeate) of 370 kghr and feed temperature of 27˚Cand controlling the brine mix ratio at 20ndash70 of theMSF brine blow down the distillate water is mea-sured and recorded at different TBTs as shown inFig 11 as compared with design calculated valuesThe pressure of saline water before the first chamberis controlled and fixed value of 15 bar absolute (abovesaturation conditions) by partially closing the valveAlso the orifices among chambers are controlled bypartially close the valve between each two successivechambers The in tube water velocity is controlled at2ms

The rate of both the design and the measured dis-tilled water increase as a result of TBT increases asshown in Fig 11 The amount of distillate is lowerthan expected this may be due to the partial loss offlashed vapor through vacuum system and the

irreversibility of flashing process occur within the ori-fices and weirs

Fig 12 shows the design and experimental GOR ofMSF variation with TBT This is defined as the ratiobetween distillate flow rate to the heating steam con-sumption GOR frac14 _m distillate

_m steam The average value of theunit design GOR is 17 which is almost twice the con-ventional MSF GOR The average measured GOR is15 as shown in Fig 12 The small difference betweenthe measured and designed value of GOR is due tothe lower distillate productivity under fixed amountof heating steam flow rate

The MSF specific power consumption (SPC) whichis defined as the ratio between the pumping powerconsumption (kW) and the rate of water distillate

(m3hr) SPC frac14 PumpingPower_m distillate Fig 13 shows that the

SPC decreases as the TBT increase mainly due to theincreases in the water productivity The experimentalSPC is calculated based on the measured distillateflow rate and the rated power consumption The

Fig 10 TBT variation Fig 11 MSF distillate productivity with TBT variation

Fig 12 The GOR of MSF system variation with TBT

6902 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

exprimental SPC is higher however than the designvalue is mainly due to lower experimental distillatefor the same saline water feed and may be due topressure drop in piping and valves that were not con-sidered properly in the design stage The SPC of testpilot unit is relatively higher than that of commercialvalue of large-scale MSF desalination plant due to thevery small test pilot unit productivity

In the next phase of the project it is expected torun MSF system at TBT= 130˚C using heating steam at140˚C from CSP system or auxiliary boiler for highTBT MSF Completing insulation for all piping con-nected CSP and MSF and oil tank need to be finalizedproper operation of tracking system of CSP needed tobe fix also completing installation of the storage tankand filling with molten salt Complete installation ofback up boiler of 10 kghr capacity in case of insuffi-cient steam from CSP at high TBT requirements Alsomatching and synchronize among PV and wind tur-bine electricity in 3 phases instead of 2 phase as perrequired for high-pressure pump of NF system andfeed pump of MSF desalination unit

4 Conclusion

bull A pilot test unit of CSP-NF-MSF-DBM water desali-nation of 10m3day capacity is designed andinstalled at Wadi AL-Natron Egypt

bull The VDS [13ndash16] is upgraded to consider the math-ematical model of CSP and steam generator andused as a process design tool for MSF-DBM NFsteam generator and CSP units

bull Comparison between the numerical simulation andthe experimental performance results of the pilotunitrsquo subsystems are relatively satisfactory

bull The newly developed NF-MSF-DBM configurationis tested under TBT= 100˚C and the GOR is esti-

mated as 15 which almost twice of traditional MSFunder the same operating conditions

bull There is ongoing adjustment to operate the plant atTBT= 130˚C to improve the overall plant perfor-mance with expected GOR=18

bull The newly high-performance NF-MSF-DBM desali-nation unit significantly reduces the unitrsquosinput thermal energy to suits integration with (therelatively expensive) RE as a desalination plantdriver

Acknowledgement

The authors acknowledge the financial support ofresearch development and innovation (RDI) programproject RE-NF-MSF ( C2-S1-148)

Nomenclature

ACSP mdash area of solar collector m2

I mdash solar intensity Wm2

m mdash mass flow rate kgs

MSF mdash multistage flash

MED mdash multieffect distillation

RO mdash reverse osmosis

NF mdash nano filtration

CSP mdash concentrated solar power

TBT mdash top brine temperature ˚C

GOR mdash gain output ratio

SPC mdash specific power consumption kWhm3

Cp mdash specific heat capacity kJkg k

TT mdash transmitter transducer

FT mdash flow meter transducer

PT mdash pressure transducer

References

[1] A Fried Water Industry Segment Report World Trade Cen-ter San Diego 2011 wwwwtcsdorg

[2] JJ Sadhwani JM Veza Desalination and energy consump-tion in Canary Islands Desalination 221 (2008) 143ndash150

[3] The European Academies Science Advisory Council (EASAC)report 2011 wwweasaceu

[4] G Fiorenza V Sharma G Braccio Technoeconomic evalua-tion of a solar powered water desalination plant Energy Con-ver Manage 44 (2003) 2217ndash2240

[5] Amr Mahmoud J Robert Abdelnasser Mabrouk A ImteyazA Nafey JS Choi JK Park S Nied J Detering New anti-sca-lant performance evaluation for MSF technology DesalinWater Treat 1944-39941944-3986 2012 httpdxdoiorg101080194439942012704701

[6] O Hamed A Hassan K Al-Shail K Bamardouf S Al-Sulami Ali Hamza M Farooque A Al-Rubaian HigherTBT more product less scaling Desalin Water Reuse Vol133 IDA NovDec 2003

[7] A Hassan Fully Integrated NF-Thermal Seawater Desalina-tion Process and Equipment US Patents No 20060157410A1 July 20 (2006)

Fig 13 SPC of MSF with TBT variation

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6903

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

sand gravel and anthracite The cartridge filter of5micron is installed after high-pressure pump andjust before membrane section The membrane sectionconsists of 4 pressure vessels running in parallel eachvessel contains one membrane element of NF270 4040type The whole NF system except feed permeateand brine tanks is placed inside one container with itscontrol panel as shown in Fig 4(b) For the watersalinity samples are collected periodically to measureits conductivity using a mobile conductivity meter

3 Pilot test performance

31 NF test performance

The NF system testing is carried out using the sitebrackish water (TDS= 2000 ppm) On the other handa mathematical model of NF membrane is developedand verified against typical operating NF unit datausing the Visual design and simulation (VDS) soft-ware developed by the authors [13ndash16] The VDS sim-ulation results of NF system at different feed pressureof 8 and 10 bar and compared with experimentalresults as shown in Table 1 The NF performance wascarried out and assessed by the recovery ratio andsalts rejection The recovery ratio (permeatefeed)increases as the feed pressure increases The salt rejec-tion (1-(permeate salinityfeed salinity)) is calculatedas shown in Table 1 The salt rejection decreases asthe feed pressure increase due to increase in permeatesalinity The measured recovery ratio is slightly lowerthan the simulation results although the salt rejectionof experimental is lower than that of simulation Thedifference between measured values of permeate flow

salinity recovery ratio salt rejection and simulationresults is within acceptable range

32 Concentrated solar power (CSP) test performance

CPS system including solar collector and steamgenerator is simulated using VDS program The massand heat balance equations of solar collector steamgenerator and pumps are developed The oil andwater thermo physical properties correlations at dif-ferent temperature are considered in the program Thecharacteristic surface of collector reflectivity receiveremission and absorptivity and glass tube materialtransmittance are specified in the VDS program Thespecifications of the solar collector and steam genera-tor are defined and fed to the program The measuredweather conditions (solar intensity ambient tempera-ture and wind velocity) at each hour are fed to theprogram The duration time starts at 700 AM to 800PM with 1-hr step

Fig 5(a) shows the interface of VDS programresults at 400 PM The oil mass flow rate and temper-ature at inlet and outlet for both solar collector andthe steam generator are presented The collectedenergy is transferred to the steam generator to gener-ate 38 kghr saturated steam at 1138˚C The solarintensity (I) and calculated absorbed energy by thereceiver is shown in Fig 5(b) The difference is notice-able at mid-day time

Fig 6(a) shows comparison between the VDS sim-ulation and experimental results of oil temperaturerise through the solar collector along the day timeThe oil temperature difference increases as the solar

Fig 4 NF system

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6899

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

intensity increase while the maximum difference inmid-day reach to 25˚C Fig 6(b) shows a comparisonbetween the simulation and experimental results of

the oil temperature drop in the steam generator unitThe maximum heat transfer occurs at the mid-dayand the maximum temperature drop is 14˚C It is

Table 1Typical NF pretreatment experimental results compared with VDS results

VDS simulation Experimental diff VDS simulation Experimental diff

Feed pressure bar 8 ndash 10 ndash

Feed flow rate tonhr 1325 ndash 1375 ndash

Feed salinity ppm 2000 ndash 2000 ndash

Permeate flow rate tonhr 067 06 10 085 076 11

Permeate salinity ppm 648 600 7 761 650 15

Brine flow rate tonhr 065 0725 12 052 06 15

Brine salinity ppm 3395 3158 7 4027 3760 7

Recovery ratio 51 45 12 62 55 11

Salt rejection 676 70 4 62 68 10

Fig 5 (a) VDS simulation results of CSP system at 400 PM June 2012 and (b) Measured solar intensity and absorbedenergy by the receiver

Fig 6 Oil temperature variation through CSP trough amp steam generator

6900 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

noticed that at the same day time the temperaturedrop in steam generator unit is less than the tempera-ture rise in the solar collector This means that part ofgained energy in the collector is absorbed in the steamgenerator and the remaining is maintained with theoutlet oil stream from steam generator and come backthe collector This explains the increase in oil tempera-ture at the concentrated solar collector inlet throughthe day time

The CSP system average efficiency (gCSP) is calcu-lated as the average useful gained poweraveragesolar input power

gCSP frac14 mCp

oilethTooil TioilTHORN

I ACSP

eth1THORN

Fig 7 shows the simulation and experimentalresults of the collector efficiency through the day timeThe collector efficiency decreases along the day timedue to the increase in the average oil temperaturewhich increases the energy loss to the ambient Theexperimental collector efficiency shows relatively lowvalue than that of the simulation collector efficiencydue to (1) inaccurate tracking system that could notfollow the sun movements accurately and (2) the inef-ficient concentrated tubes location in the CSP focusand possible convection loss

Fig 8 shows comparison between the simulatedand the measured generated steam temperature Thewater inlet and steam exit valve remain closed whilethe oil valves are open to allow energy transfer fromoil to heat the enclosed water in the boiler The waterfeed and steam valves are opened when the watertemperature reaches 77˚C The generated steam tem-perature increases as the solar intensity increase andthe maximum temperature reached is 110˚C at mid-day time

The steam valve is opened at 100 PM at steamflow rate of 43 kghr The generated steam is directedto MSF desalination unit as a heating source The con-densate steam in the brine heater of MSF is feed backto the steam generator The amount of generatedsteam flow rate decreased linearly as shown in Fig 9this due to the decreasing in the solar collector effi-ciency The measured generated steam flow rateshows lower values than the simulation results due tothe thermal losses encountered through insufficientinsulation of steam generator and throughout the con-nection pipe between CSP outlet and steam generatorAs shown in Fig 9 the operation of the steam genera-tion extends up 11 PM this is due to heat storage inthe CSP system

Fig 10 shows the simulation and experimental val-ues of TBT variation through day time As the CSPsteam condenses in the brine heater of MSF the TBTraise up due to gained energy of the latent heatUnder the ambient and operating condition of June2012 the mid-day TBT reaches up to 100˚C while theCSP steam condenses at 106˚C that is 6˚C tempera-ture differences is maintained

Fig 7 Solar collector efficiency

Fig 8 Generated steam temperature

Fig 9 Generated steam flow rate

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6901

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

33 New MSF with De-aerator and Brine Mix (DBM)

The permeate water of NF system is directed tothe deaeration and brine mix tower where the feed issprayed for oxygen removal The deaerated water ismixed with part of the brine blow down then itpumped to the MSF condensers The brine mix feedabsorbs latent heat energy in condnesers before passthough brine heater where the brine reaches its TBTThe brine is then directed to the first flash chamberwhere flashing process occurs and vapor releases Thereleased vapor condenses to form a product waterThe flashing process occurs in the successive stagesuntil reach the last stage where the unflashed brineexit as brine blown down The condensate of all stagesis collected and directed to the water product tankThe brine level is adjusted above the interconnectingpipes (interstage gates) to guarantee the flash cham-bers sealing

Under the same feed saline water flow rate (NFpermeate) of 370 kghr and feed temperature of 27˚Cand controlling the brine mix ratio at 20ndash70 of theMSF brine blow down the distillate water is mea-sured and recorded at different TBTs as shown inFig 11 as compared with design calculated valuesThe pressure of saline water before the first chamberis controlled and fixed value of 15 bar absolute (abovesaturation conditions) by partially closing the valveAlso the orifices among chambers are controlled bypartially close the valve between each two successivechambers The in tube water velocity is controlled at2ms

The rate of both the design and the measured dis-tilled water increase as a result of TBT increases asshown in Fig 11 The amount of distillate is lowerthan expected this may be due to the partial loss offlashed vapor through vacuum system and the

irreversibility of flashing process occur within the ori-fices and weirs

Fig 12 shows the design and experimental GOR ofMSF variation with TBT This is defined as the ratiobetween distillate flow rate to the heating steam con-sumption GOR frac14 _m distillate

_m steam The average value of theunit design GOR is 17 which is almost twice the con-ventional MSF GOR The average measured GOR is15 as shown in Fig 12 The small difference betweenthe measured and designed value of GOR is due tothe lower distillate productivity under fixed amountof heating steam flow rate

The MSF specific power consumption (SPC) whichis defined as the ratio between the pumping powerconsumption (kW) and the rate of water distillate

(m3hr) SPC frac14 PumpingPower_m distillate Fig 13 shows that the

SPC decreases as the TBT increase mainly due to theincreases in the water productivity The experimentalSPC is calculated based on the measured distillateflow rate and the rated power consumption The

Fig 10 TBT variation Fig 11 MSF distillate productivity with TBT variation

Fig 12 The GOR of MSF system variation with TBT

6902 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

exprimental SPC is higher however than the designvalue is mainly due to lower experimental distillatefor the same saline water feed and may be due topressure drop in piping and valves that were not con-sidered properly in the design stage The SPC of testpilot unit is relatively higher than that of commercialvalue of large-scale MSF desalination plant due to thevery small test pilot unit productivity

In the next phase of the project it is expected torun MSF system at TBT= 130˚C using heating steam at140˚C from CSP system or auxiliary boiler for highTBT MSF Completing insulation for all piping con-nected CSP and MSF and oil tank need to be finalizedproper operation of tracking system of CSP needed tobe fix also completing installation of the storage tankand filling with molten salt Complete installation ofback up boiler of 10 kghr capacity in case of insuffi-cient steam from CSP at high TBT requirements Alsomatching and synchronize among PV and wind tur-bine electricity in 3 phases instead of 2 phase as perrequired for high-pressure pump of NF system andfeed pump of MSF desalination unit

4 Conclusion

bull A pilot test unit of CSP-NF-MSF-DBM water desali-nation of 10m3day capacity is designed andinstalled at Wadi AL-Natron Egypt

bull The VDS [13ndash16] is upgraded to consider the math-ematical model of CSP and steam generator andused as a process design tool for MSF-DBM NFsteam generator and CSP units

bull Comparison between the numerical simulation andthe experimental performance results of the pilotunitrsquo subsystems are relatively satisfactory

bull The newly developed NF-MSF-DBM configurationis tested under TBT= 100˚C and the GOR is esti-

mated as 15 which almost twice of traditional MSFunder the same operating conditions

bull There is ongoing adjustment to operate the plant atTBT= 130˚C to improve the overall plant perfor-mance with expected GOR=18

bull The newly high-performance NF-MSF-DBM desali-nation unit significantly reduces the unitrsquosinput thermal energy to suits integration with (therelatively expensive) RE as a desalination plantdriver

Acknowledgement

The authors acknowledge the financial support ofresearch development and innovation (RDI) programproject RE-NF-MSF ( C2-S1-148)

Nomenclature

ACSP mdash area of solar collector m2

I mdash solar intensity Wm2

m mdash mass flow rate kgs

MSF mdash multistage flash

MED mdash multieffect distillation

RO mdash reverse osmosis

NF mdash nano filtration

CSP mdash concentrated solar power

TBT mdash top brine temperature ˚C

GOR mdash gain output ratio

SPC mdash specific power consumption kWhm3

Cp mdash specific heat capacity kJkg k

TT mdash transmitter transducer

FT mdash flow meter transducer

PT mdash pressure transducer

References

[1] A Fried Water Industry Segment Report World Trade Cen-ter San Diego 2011 wwwwtcsdorg

[2] JJ Sadhwani JM Veza Desalination and energy consump-tion in Canary Islands Desalination 221 (2008) 143ndash150

[3] The European Academies Science Advisory Council (EASAC)report 2011 wwweasaceu

[4] G Fiorenza V Sharma G Braccio Technoeconomic evalua-tion of a solar powered water desalination plant Energy Con-ver Manage 44 (2003) 2217ndash2240

[5] Amr Mahmoud J Robert Abdelnasser Mabrouk A ImteyazA Nafey JS Choi JK Park S Nied J Detering New anti-sca-lant performance evaluation for MSF technology DesalinWater Treat 1944-39941944-3986 2012 httpdxdoiorg101080194439942012704701

[6] O Hamed A Hassan K Al-Shail K Bamardouf S Al-Sulami Ali Hamza M Farooque A Al-Rubaian HigherTBT more product less scaling Desalin Water Reuse Vol133 IDA NovDec 2003

[7] A Hassan Fully Integrated NF-Thermal Seawater Desalina-tion Process and Equipment US Patents No 20060157410A1 July 20 (2006)

Fig 13 SPC of MSF with TBT variation

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6903

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

intensity increase while the maximum difference inmid-day reach to 25˚C Fig 6(b) shows a comparisonbetween the simulation and experimental results of

the oil temperature drop in the steam generator unitThe maximum heat transfer occurs at the mid-dayand the maximum temperature drop is 14˚C It is

Table 1Typical NF pretreatment experimental results compared with VDS results

VDS simulation Experimental diff VDS simulation Experimental diff

Feed pressure bar 8 ndash 10 ndash

Feed flow rate tonhr 1325 ndash 1375 ndash

Feed salinity ppm 2000 ndash 2000 ndash

Permeate flow rate tonhr 067 06 10 085 076 11

Permeate salinity ppm 648 600 7 761 650 15

Brine flow rate tonhr 065 0725 12 052 06 15

Brine salinity ppm 3395 3158 7 4027 3760 7

Recovery ratio 51 45 12 62 55 11

Salt rejection 676 70 4 62 68 10

Fig 5 (a) VDS simulation results of CSP system at 400 PM June 2012 and (b) Measured solar intensity and absorbedenergy by the receiver

Fig 6 Oil temperature variation through CSP trough amp steam generator

6900 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

noticed that at the same day time the temperaturedrop in steam generator unit is less than the tempera-ture rise in the solar collector This means that part ofgained energy in the collector is absorbed in the steamgenerator and the remaining is maintained with theoutlet oil stream from steam generator and come backthe collector This explains the increase in oil tempera-ture at the concentrated solar collector inlet throughthe day time

The CSP system average efficiency (gCSP) is calcu-lated as the average useful gained poweraveragesolar input power

gCSP frac14 mCp

oilethTooil TioilTHORN

I ACSP

eth1THORN

Fig 7 shows the simulation and experimentalresults of the collector efficiency through the day timeThe collector efficiency decreases along the day timedue to the increase in the average oil temperaturewhich increases the energy loss to the ambient Theexperimental collector efficiency shows relatively lowvalue than that of the simulation collector efficiencydue to (1) inaccurate tracking system that could notfollow the sun movements accurately and (2) the inef-ficient concentrated tubes location in the CSP focusand possible convection loss

Fig 8 shows comparison between the simulatedand the measured generated steam temperature Thewater inlet and steam exit valve remain closed whilethe oil valves are open to allow energy transfer fromoil to heat the enclosed water in the boiler The waterfeed and steam valves are opened when the watertemperature reaches 77˚C The generated steam tem-perature increases as the solar intensity increase andthe maximum temperature reached is 110˚C at mid-day time

The steam valve is opened at 100 PM at steamflow rate of 43 kghr The generated steam is directedto MSF desalination unit as a heating source The con-densate steam in the brine heater of MSF is feed backto the steam generator The amount of generatedsteam flow rate decreased linearly as shown in Fig 9this due to the decreasing in the solar collector effi-ciency The measured generated steam flow rateshows lower values than the simulation results due tothe thermal losses encountered through insufficientinsulation of steam generator and throughout the con-nection pipe between CSP outlet and steam generatorAs shown in Fig 9 the operation of the steam genera-tion extends up 11 PM this is due to heat storage inthe CSP system

Fig 10 shows the simulation and experimental val-ues of TBT variation through day time As the CSPsteam condenses in the brine heater of MSF the TBTraise up due to gained energy of the latent heatUnder the ambient and operating condition of June2012 the mid-day TBT reaches up to 100˚C while theCSP steam condenses at 106˚C that is 6˚C tempera-ture differences is maintained

Fig 7 Solar collector efficiency

Fig 8 Generated steam temperature

Fig 9 Generated steam flow rate

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6901

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

33 New MSF with De-aerator and Brine Mix (DBM)

The permeate water of NF system is directed tothe deaeration and brine mix tower where the feed issprayed for oxygen removal The deaerated water ismixed with part of the brine blow down then itpumped to the MSF condensers The brine mix feedabsorbs latent heat energy in condnesers before passthough brine heater where the brine reaches its TBTThe brine is then directed to the first flash chamberwhere flashing process occurs and vapor releases Thereleased vapor condenses to form a product waterThe flashing process occurs in the successive stagesuntil reach the last stage where the unflashed brineexit as brine blown down The condensate of all stagesis collected and directed to the water product tankThe brine level is adjusted above the interconnectingpipes (interstage gates) to guarantee the flash cham-bers sealing

Under the same feed saline water flow rate (NFpermeate) of 370 kghr and feed temperature of 27˚Cand controlling the brine mix ratio at 20ndash70 of theMSF brine blow down the distillate water is mea-sured and recorded at different TBTs as shown inFig 11 as compared with design calculated valuesThe pressure of saline water before the first chamberis controlled and fixed value of 15 bar absolute (abovesaturation conditions) by partially closing the valveAlso the orifices among chambers are controlled bypartially close the valve between each two successivechambers The in tube water velocity is controlled at2ms

The rate of both the design and the measured dis-tilled water increase as a result of TBT increases asshown in Fig 11 The amount of distillate is lowerthan expected this may be due to the partial loss offlashed vapor through vacuum system and the

irreversibility of flashing process occur within the ori-fices and weirs

Fig 12 shows the design and experimental GOR ofMSF variation with TBT This is defined as the ratiobetween distillate flow rate to the heating steam con-sumption GOR frac14 _m distillate

_m steam The average value of theunit design GOR is 17 which is almost twice the con-ventional MSF GOR The average measured GOR is15 as shown in Fig 12 The small difference betweenthe measured and designed value of GOR is due tothe lower distillate productivity under fixed amountof heating steam flow rate

The MSF specific power consumption (SPC) whichis defined as the ratio between the pumping powerconsumption (kW) and the rate of water distillate

(m3hr) SPC frac14 PumpingPower_m distillate Fig 13 shows that the

SPC decreases as the TBT increase mainly due to theincreases in the water productivity The experimentalSPC is calculated based on the measured distillateflow rate and the rated power consumption The

Fig 10 TBT variation Fig 11 MSF distillate productivity with TBT variation

Fig 12 The GOR of MSF system variation with TBT

6902 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

exprimental SPC is higher however than the designvalue is mainly due to lower experimental distillatefor the same saline water feed and may be due topressure drop in piping and valves that were not con-sidered properly in the design stage The SPC of testpilot unit is relatively higher than that of commercialvalue of large-scale MSF desalination plant due to thevery small test pilot unit productivity

In the next phase of the project it is expected torun MSF system at TBT= 130˚C using heating steam at140˚C from CSP system or auxiliary boiler for highTBT MSF Completing insulation for all piping con-nected CSP and MSF and oil tank need to be finalizedproper operation of tracking system of CSP needed tobe fix also completing installation of the storage tankand filling with molten salt Complete installation ofback up boiler of 10 kghr capacity in case of insuffi-cient steam from CSP at high TBT requirements Alsomatching and synchronize among PV and wind tur-bine electricity in 3 phases instead of 2 phase as perrequired for high-pressure pump of NF system andfeed pump of MSF desalination unit

4 Conclusion

bull A pilot test unit of CSP-NF-MSF-DBM water desali-nation of 10m3day capacity is designed andinstalled at Wadi AL-Natron Egypt

bull The VDS [13ndash16] is upgraded to consider the math-ematical model of CSP and steam generator andused as a process design tool for MSF-DBM NFsteam generator and CSP units

bull Comparison between the numerical simulation andthe experimental performance results of the pilotunitrsquo subsystems are relatively satisfactory

bull The newly developed NF-MSF-DBM configurationis tested under TBT= 100˚C and the GOR is esti-

mated as 15 which almost twice of traditional MSFunder the same operating conditions

bull There is ongoing adjustment to operate the plant atTBT= 130˚C to improve the overall plant perfor-mance with expected GOR=18

bull The newly high-performance NF-MSF-DBM desali-nation unit significantly reduces the unitrsquosinput thermal energy to suits integration with (therelatively expensive) RE as a desalination plantdriver

Acknowledgement

The authors acknowledge the financial support ofresearch development and innovation (RDI) programproject RE-NF-MSF ( C2-S1-148)

Nomenclature

ACSP mdash area of solar collector m2

I mdash solar intensity Wm2

m mdash mass flow rate kgs

MSF mdash multistage flash

MED mdash multieffect distillation

RO mdash reverse osmosis

NF mdash nano filtration

CSP mdash concentrated solar power

TBT mdash top brine temperature ˚C

GOR mdash gain output ratio

SPC mdash specific power consumption kWhm3

Cp mdash specific heat capacity kJkg k

TT mdash transmitter transducer

FT mdash flow meter transducer

PT mdash pressure transducer

References

[1] A Fried Water Industry Segment Report World Trade Cen-ter San Diego 2011 wwwwtcsdorg

[2] JJ Sadhwani JM Veza Desalination and energy consump-tion in Canary Islands Desalination 221 (2008) 143ndash150

[3] The European Academies Science Advisory Council (EASAC)report 2011 wwweasaceu

[4] G Fiorenza V Sharma G Braccio Technoeconomic evalua-tion of a solar powered water desalination plant Energy Con-ver Manage 44 (2003) 2217ndash2240

[5] Amr Mahmoud J Robert Abdelnasser Mabrouk A ImteyazA Nafey JS Choi JK Park S Nied J Detering New anti-sca-lant performance evaluation for MSF technology DesalinWater Treat 1944-39941944-3986 2012 httpdxdoiorg101080194439942012704701

[6] O Hamed A Hassan K Al-Shail K Bamardouf S Al-Sulami Ali Hamza M Farooque A Al-Rubaian HigherTBT more product less scaling Desalin Water Reuse Vol133 IDA NovDec 2003

[7] A Hassan Fully Integrated NF-Thermal Seawater Desalina-tion Process and Equipment US Patents No 20060157410A1 July 20 (2006)

Fig 13 SPC of MSF with TBT variation

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6903

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

noticed that at the same day time the temperaturedrop in steam generator unit is less than the tempera-ture rise in the solar collector This means that part ofgained energy in the collector is absorbed in the steamgenerator and the remaining is maintained with theoutlet oil stream from steam generator and come backthe collector This explains the increase in oil tempera-ture at the concentrated solar collector inlet throughthe day time

The CSP system average efficiency (gCSP) is calcu-lated as the average useful gained poweraveragesolar input power

gCSP frac14 mCp

oilethTooil TioilTHORN

I ACSP

eth1THORN

Fig 7 shows the simulation and experimentalresults of the collector efficiency through the day timeThe collector efficiency decreases along the day timedue to the increase in the average oil temperaturewhich increases the energy loss to the ambient Theexperimental collector efficiency shows relatively lowvalue than that of the simulation collector efficiencydue to (1) inaccurate tracking system that could notfollow the sun movements accurately and (2) the inef-ficient concentrated tubes location in the CSP focusand possible convection loss

Fig 8 shows comparison between the simulatedand the measured generated steam temperature Thewater inlet and steam exit valve remain closed whilethe oil valves are open to allow energy transfer fromoil to heat the enclosed water in the boiler The waterfeed and steam valves are opened when the watertemperature reaches 77˚C The generated steam tem-perature increases as the solar intensity increase andthe maximum temperature reached is 110˚C at mid-day time

The steam valve is opened at 100 PM at steamflow rate of 43 kghr The generated steam is directedto MSF desalination unit as a heating source The con-densate steam in the brine heater of MSF is feed backto the steam generator The amount of generatedsteam flow rate decreased linearly as shown in Fig 9this due to the decreasing in the solar collector effi-ciency The measured generated steam flow rateshows lower values than the simulation results due tothe thermal losses encountered through insufficientinsulation of steam generator and throughout the con-nection pipe between CSP outlet and steam generatorAs shown in Fig 9 the operation of the steam genera-tion extends up 11 PM this is due to heat storage inthe CSP system

Fig 10 shows the simulation and experimental val-ues of TBT variation through day time As the CSPsteam condenses in the brine heater of MSF the TBTraise up due to gained energy of the latent heatUnder the ambient and operating condition of June2012 the mid-day TBT reaches up to 100˚C while theCSP steam condenses at 106˚C that is 6˚C tempera-ture differences is maintained

Fig 7 Solar collector efficiency

Fig 8 Generated steam temperature

Fig 9 Generated steam flow rate

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6901

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

33 New MSF with De-aerator and Brine Mix (DBM)

The permeate water of NF system is directed tothe deaeration and brine mix tower where the feed issprayed for oxygen removal The deaerated water ismixed with part of the brine blow down then itpumped to the MSF condensers The brine mix feedabsorbs latent heat energy in condnesers before passthough brine heater where the brine reaches its TBTThe brine is then directed to the first flash chamberwhere flashing process occurs and vapor releases Thereleased vapor condenses to form a product waterThe flashing process occurs in the successive stagesuntil reach the last stage where the unflashed brineexit as brine blown down The condensate of all stagesis collected and directed to the water product tankThe brine level is adjusted above the interconnectingpipes (interstage gates) to guarantee the flash cham-bers sealing

Under the same feed saline water flow rate (NFpermeate) of 370 kghr and feed temperature of 27˚Cand controlling the brine mix ratio at 20ndash70 of theMSF brine blow down the distillate water is mea-sured and recorded at different TBTs as shown inFig 11 as compared with design calculated valuesThe pressure of saline water before the first chamberis controlled and fixed value of 15 bar absolute (abovesaturation conditions) by partially closing the valveAlso the orifices among chambers are controlled bypartially close the valve between each two successivechambers The in tube water velocity is controlled at2ms

The rate of both the design and the measured dis-tilled water increase as a result of TBT increases asshown in Fig 11 The amount of distillate is lowerthan expected this may be due to the partial loss offlashed vapor through vacuum system and the

irreversibility of flashing process occur within the ori-fices and weirs

Fig 12 shows the design and experimental GOR ofMSF variation with TBT This is defined as the ratiobetween distillate flow rate to the heating steam con-sumption GOR frac14 _m distillate

_m steam The average value of theunit design GOR is 17 which is almost twice the con-ventional MSF GOR The average measured GOR is15 as shown in Fig 12 The small difference betweenthe measured and designed value of GOR is due tothe lower distillate productivity under fixed amountof heating steam flow rate

The MSF specific power consumption (SPC) whichis defined as the ratio between the pumping powerconsumption (kW) and the rate of water distillate

(m3hr) SPC frac14 PumpingPower_m distillate Fig 13 shows that the

SPC decreases as the TBT increase mainly due to theincreases in the water productivity The experimentalSPC is calculated based on the measured distillateflow rate and the rated power consumption The

Fig 10 TBT variation Fig 11 MSF distillate productivity with TBT variation

Fig 12 The GOR of MSF system variation with TBT

6902 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

exprimental SPC is higher however than the designvalue is mainly due to lower experimental distillatefor the same saline water feed and may be due topressure drop in piping and valves that were not con-sidered properly in the design stage The SPC of testpilot unit is relatively higher than that of commercialvalue of large-scale MSF desalination plant due to thevery small test pilot unit productivity

In the next phase of the project it is expected torun MSF system at TBT= 130˚C using heating steam at140˚C from CSP system or auxiliary boiler for highTBT MSF Completing insulation for all piping con-nected CSP and MSF and oil tank need to be finalizedproper operation of tracking system of CSP needed tobe fix also completing installation of the storage tankand filling with molten salt Complete installation ofback up boiler of 10 kghr capacity in case of insuffi-cient steam from CSP at high TBT requirements Alsomatching and synchronize among PV and wind tur-bine electricity in 3 phases instead of 2 phase as perrequired for high-pressure pump of NF system andfeed pump of MSF desalination unit

4 Conclusion

bull A pilot test unit of CSP-NF-MSF-DBM water desali-nation of 10m3day capacity is designed andinstalled at Wadi AL-Natron Egypt

bull The VDS [13ndash16] is upgraded to consider the math-ematical model of CSP and steam generator andused as a process design tool for MSF-DBM NFsteam generator and CSP units

bull Comparison between the numerical simulation andthe experimental performance results of the pilotunitrsquo subsystems are relatively satisfactory

bull The newly developed NF-MSF-DBM configurationis tested under TBT= 100˚C and the GOR is esti-

mated as 15 which almost twice of traditional MSFunder the same operating conditions

bull There is ongoing adjustment to operate the plant atTBT= 130˚C to improve the overall plant perfor-mance with expected GOR=18

bull The newly high-performance NF-MSF-DBM desali-nation unit significantly reduces the unitrsquosinput thermal energy to suits integration with (therelatively expensive) RE as a desalination plantdriver

Acknowledgement

The authors acknowledge the financial support ofresearch development and innovation (RDI) programproject RE-NF-MSF ( C2-S1-148)

Nomenclature

ACSP mdash area of solar collector m2

I mdash solar intensity Wm2

m mdash mass flow rate kgs

MSF mdash multistage flash

MED mdash multieffect distillation

RO mdash reverse osmosis

NF mdash nano filtration

CSP mdash concentrated solar power

TBT mdash top brine temperature ˚C

GOR mdash gain output ratio

SPC mdash specific power consumption kWhm3

Cp mdash specific heat capacity kJkg k

TT mdash transmitter transducer

FT mdash flow meter transducer

PT mdash pressure transducer

References

[1] A Fried Water Industry Segment Report World Trade Cen-ter San Diego 2011 wwwwtcsdorg

[2] JJ Sadhwani JM Veza Desalination and energy consump-tion in Canary Islands Desalination 221 (2008) 143ndash150

[3] The European Academies Science Advisory Council (EASAC)report 2011 wwweasaceu

[4] G Fiorenza V Sharma G Braccio Technoeconomic evalua-tion of a solar powered water desalination plant Energy Con-ver Manage 44 (2003) 2217ndash2240

[5] Amr Mahmoud J Robert Abdelnasser Mabrouk A ImteyazA Nafey JS Choi JK Park S Nied J Detering New anti-sca-lant performance evaluation for MSF technology DesalinWater Treat 1944-39941944-3986 2012 httpdxdoiorg101080194439942012704701

[6] O Hamed A Hassan K Al-Shail K Bamardouf S Al-Sulami Ali Hamza M Farooque A Al-Rubaian HigherTBT more product less scaling Desalin Water Reuse Vol133 IDA NovDec 2003

[7] A Hassan Fully Integrated NF-Thermal Seawater Desalina-tion Process and Equipment US Patents No 20060157410A1 July 20 (2006)

Fig 13 SPC of MSF with TBT variation

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6903

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

33 New MSF with De-aerator and Brine Mix (DBM)

The permeate water of NF system is directed tothe deaeration and brine mix tower where the feed issprayed for oxygen removal The deaerated water ismixed with part of the brine blow down then itpumped to the MSF condensers The brine mix feedabsorbs latent heat energy in condnesers before passthough brine heater where the brine reaches its TBTThe brine is then directed to the first flash chamberwhere flashing process occurs and vapor releases Thereleased vapor condenses to form a product waterThe flashing process occurs in the successive stagesuntil reach the last stage where the unflashed brineexit as brine blown down The condensate of all stagesis collected and directed to the water product tankThe brine level is adjusted above the interconnectingpipes (interstage gates) to guarantee the flash cham-bers sealing

Under the same feed saline water flow rate (NFpermeate) of 370 kghr and feed temperature of 27˚Cand controlling the brine mix ratio at 20ndash70 of theMSF brine blow down the distillate water is mea-sured and recorded at different TBTs as shown inFig 11 as compared with design calculated valuesThe pressure of saline water before the first chamberis controlled and fixed value of 15 bar absolute (abovesaturation conditions) by partially closing the valveAlso the orifices among chambers are controlled bypartially close the valve between each two successivechambers The in tube water velocity is controlled at2ms

The rate of both the design and the measured dis-tilled water increase as a result of TBT increases asshown in Fig 11 The amount of distillate is lowerthan expected this may be due to the partial loss offlashed vapor through vacuum system and the

irreversibility of flashing process occur within the ori-fices and weirs

Fig 12 shows the design and experimental GOR ofMSF variation with TBT This is defined as the ratiobetween distillate flow rate to the heating steam con-sumption GOR frac14 _m distillate

_m steam The average value of theunit design GOR is 17 which is almost twice the con-ventional MSF GOR The average measured GOR is15 as shown in Fig 12 The small difference betweenthe measured and designed value of GOR is due tothe lower distillate productivity under fixed amountof heating steam flow rate

The MSF specific power consumption (SPC) whichis defined as the ratio between the pumping powerconsumption (kW) and the rate of water distillate

(m3hr) SPC frac14 PumpingPower_m distillate Fig 13 shows that the

SPC decreases as the TBT increase mainly due to theincreases in the water productivity The experimentalSPC is calculated based on the measured distillateflow rate and the rated power consumption The

Fig 10 TBT variation Fig 11 MSF distillate productivity with TBT variation

Fig 12 The GOR of MSF system variation with TBT

6902 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

exprimental SPC is higher however than the designvalue is mainly due to lower experimental distillatefor the same saline water feed and may be due topressure drop in piping and valves that were not con-sidered properly in the design stage The SPC of testpilot unit is relatively higher than that of commercialvalue of large-scale MSF desalination plant due to thevery small test pilot unit productivity

In the next phase of the project it is expected torun MSF system at TBT= 130˚C using heating steam at140˚C from CSP system or auxiliary boiler for highTBT MSF Completing insulation for all piping con-nected CSP and MSF and oil tank need to be finalizedproper operation of tracking system of CSP needed tobe fix also completing installation of the storage tankand filling with molten salt Complete installation ofback up boiler of 10 kghr capacity in case of insuffi-cient steam from CSP at high TBT requirements Alsomatching and synchronize among PV and wind tur-bine electricity in 3 phases instead of 2 phase as perrequired for high-pressure pump of NF system andfeed pump of MSF desalination unit

4 Conclusion

bull A pilot test unit of CSP-NF-MSF-DBM water desali-nation of 10m3day capacity is designed andinstalled at Wadi AL-Natron Egypt

bull The VDS [13ndash16] is upgraded to consider the math-ematical model of CSP and steam generator andused as a process design tool for MSF-DBM NFsteam generator and CSP units

bull Comparison between the numerical simulation andthe experimental performance results of the pilotunitrsquo subsystems are relatively satisfactory

bull The newly developed NF-MSF-DBM configurationis tested under TBT= 100˚C and the GOR is esti-

mated as 15 which almost twice of traditional MSFunder the same operating conditions

bull There is ongoing adjustment to operate the plant atTBT= 130˚C to improve the overall plant perfor-mance with expected GOR=18

bull The newly high-performance NF-MSF-DBM desali-nation unit significantly reduces the unitrsquosinput thermal energy to suits integration with (therelatively expensive) RE as a desalination plantdriver

Acknowledgement

The authors acknowledge the financial support ofresearch development and innovation (RDI) programproject RE-NF-MSF ( C2-S1-148)

Nomenclature

ACSP mdash area of solar collector m2

I mdash solar intensity Wm2

m mdash mass flow rate kgs

MSF mdash multistage flash

MED mdash multieffect distillation

RO mdash reverse osmosis

NF mdash nano filtration

CSP mdash concentrated solar power

TBT mdash top brine temperature ˚C

GOR mdash gain output ratio

SPC mdash specific power consumption kWhm3

Cp mdash specific heat capacity kJkg k

TT mdash transmitter transducer

FT mdash flow meter transducer

PT mdash pressure transducer

References

[1] A Fried Water Industry Segment Report World Trade Cen-ter San Diego 2011 wwwwtcsdorg

[2] JJ Sadhwani JM Veza Desalination and energy consump-tion in Canary Islands Desalination 221 (2008) 143ndash150

[3] The European Academies Science Advisory Council (EASAC)report 2011 wwweasaceu

[4] G Fiorenza V Sharma G Braccio Technoeconomic evalua-tion of a solar powered water desalination plant Energy Con-ver Manage 44 (2003) 2217ndash2240

[5] Amr Mahmoud J Robert Abdelnasser Mabrouk A ImteyazA Nafey JS Choi JK Park S Nied J Detering New anti-sca-lant performance evaluation for MSF technology DesalinWater Treat 1944-39941944-3986 2012 httpdxdoiorg101080194439942012704701

[6] O Hamed A Hassan K Al-Shail K Bamardouf S Al-Sulami Ali Hamza M Farooque A Al-Rubaian HigherTBT more product less scaling Desalin Water Reuse Vol133 IDA NovDec 2003

[7] A Hassan Fully Integrated NF-Thermal Seawater Desalina-tion Process and Equipment US Patents No 20060157410A1 July 20 (2006)

Fig 13 SPC of MSF with TBT variation

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6903

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

exprimental SPC is higher however than the designvalue is mainly due to lower experimental distillatefor the same saline water feed and may be due topressure drop in piping and valves that were not con-sidered properly in the design stage The SPC of testpilot unit is relatively higher than that of commercialvalue of large-scale MSF desalination plant due to thevery small test pilot unit productivity

In the next phase of the project it is expected torun MSF system at TBT= 130˚C using heating steam at140˚C from CSP system or auxiliary boiler for highTBT MSF Completing insulation for all piping con-nected CSP and MSF and oil tank need to be finalizedproper operation of tracking system of CSP needed tobe fix also completing installation of the storage tankand filling with molten salt Complete installation ofback up boiler of 10 kghr capacity in case of insuffi-cient steam from CSP at high TBT requirements Alsomatching and synchronize among PV and wind tur-bine electricity in 3 phases instead of 2 phase as perrequired for high-pressure pump of NF system andfeed pump of MSF desalination unit

4 Conclusion

bull A pilot test unit of CSP-NF-MSF-DBM water desali-nation of 10m3day capacity is designed andinstalled at Wadi AL-Natron Egypt

bull The VDS [13ndash16] is upgraded to consider the math-ematical model of CSP and steam generator andused as a process design tool for MSF-DBM NFsteam generator and CSP units

bull Comparison between the numerical simulation andthe experimental performance results of the pilotunitrsquo subsystems are relatively satisfactory

bull The newly developed NF-MSF-DBM configurationis tested under TBT= 100˚C and the GOR is esti-

mated as 15 which almost twice of traditional MSFunder the same operating conditions

bull There is ongoing adjustment to operate the plant atTBT= 130˚C to improve the overall plant perfor-mance with expected GOR=18

bull The newly high-performance NF-MSF-DBM desali-nation unit significantly reduces the unitrsquosinput thermal energy to suits integration with (therelatively expensive) RE as a desalination plantdriver

Acknowledgement

The authors acknowledge the financial support ofresearch development and innovation (RDI) programproject RE-NF-MSF ( C2-S1-148)

Nomenclature

ACSP mdash area of solar collector m2

I mdash solar intensity Wm2

m mdash mass flow rate kgs

MSF mdash multistage flash

MED mdash multieffect distillation

RO mdash reverse osmosis

NF mdash nano filtration

CSP mdash concentrated solar power

TBT mdash top brine temperature ˚C

GOR mdash gain output ratio

SPC mdash specific power consumption kWhm3

Cp mdash specific heat capacity kJkg k

TT mdash transmitter transducer

FT mdash flow meter transducer

PT mdash pressure transducer

References

[1] A Fried Water Industry Segment Report World Trade Cen-ter San Diego 2011 wwwwtcsdorg

[2] JJ Sadhwani JM Veza Desalination and energy consump-tion in Canary Islands Desalination 221 (2008) 143ndash150

[3] The European Academies Science Advisory Council (EASAC)report 2011 wwweasaceu

[4] G Fiorenza V Sharma G Braccio Technoeconomic evalua-tion of a solar powered water desalination plant Energy Con-ver Manage 44 (2003) 2217ndash2240

[5] Amr Mahmoud J Robert Abdelnasser Mabrouk A ImteyazA Nafey JS Choi JK Park S Nied J Detering New anti-sca-lant performance evaluation for MSF technology DesalinWater Treat 1944-39941944-3986 2012 httpdxdoiorg101080194439942012704701

[6] O Hamed A Hassan K Al-Shail K Bamardouf S Al-Sulami Ali Hamza M Farooque A Al-Rubaian HigherTBT more product less scaling Desalin Water Reuse Vol133 IDA NovDec 2003

[7] A Hassan Fully Integrated NF-Thermal Seawater Desalina-tion Process and Equipment US Patents No 20060157410A1 July 20 (2006)

Fig 13 SPC of MSF with TBT variation

ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904 6903

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014

[8] M Al Sofi A Hassan G Mustafa A Dalvi N Kithar Nano-filtration as means of achieving higher TBT of gt120C in MSFDesalination 118 (1998) 123ndash129

[9] A Hassan M Al Sofi A Al-Amoudi A Jamaluddin AFarooque A Rowaili A Dalvi N Kithar G Mustafa I Al-Tisan A new approach to membrane and thermal seawaterdesalination processes using nanofilteration membranes (PartI) Desalination 118 (1998) 35ndash51

[10] L Awerbuch Salt Water Desalination Process Using IonSelective Membranes International Patent (PCT) No WO0114256 A1 March 1 2001

[11] L Awerbuch Water Desalination Process Using Ion SelectiveMembranes US Patent No 20050011832 A1 Jan 20 2005

[12] X Wang T Tsuru M Togoh S Nakao S Kimura Evalua-tion of pore structure and electrical properties of nonfiltrationmembranes J Chem Eng Jpn (1995) 186ndash192

[13] Abdel Nasser Mabrouk H Fath Techno-economic analysis ofhybrid high performance MSF desalination plant with NFmembrane Desalin Water Treat 2012 httpdxdoi101080194439942012714893

[14] AS Nafey HES Fath AA Mabrouk A new visual packagefor design and simulation of desalination processes Desalina-tion 194 (2006) 281ndash296

[15] Abdel Nasser Mabrouk AS Nafey Hassan Fath Steam elec-tricity and water evaluation of cogeneration powerndashdesalina-tion plants Desalin Water Treat 22 (2010) 56ndash64

[16] Abdel Nasser Mabrouk Techno-economic Analysis of Seawa-ter Desalination Plants ISBN 978-3-659-17934-1 LAP LAM-BERT Academic Publishing Germany 2012

6904 ANA Mabrouk and HES Fath Desalination and Water Treatment 51 (2013) 6895ndash6904

Dow

nloa

ded

by [

UQ

Lib

rary

] at

00

14 1

4 N

ovem

ber

2014