wsaa occasional paper no.24 - direct methane nad nitrous oxide ... a… · and nitrous oxide (n2o)...

www.wsaa.asn.au

Melbourne Office Level 8, 469 Latrobe Street Melbourne VIC 3000

PO Box 13172 Law Courts Post Office Melbourne VIC 8010

Phone: (03) 9606 0678 Fax: (03) 9606 0376

Sydney Office Suite 1, Level 30 9 Castlereagh Street Sydney NSW 2001

GPO Box 915 Sydney NSW 2001

Phone: (02) 9221 5966 Fax: (02) 9221 5977

Occasional Paper No. 24

October 2009

Direct Methane and Nitrous Oxide emissions from full-scale wastewater treatment systems

ii WSAA Occasional Paper No.24 - Direct Methane and Nitrous Oxide emissions from full-scale wastewater treatment systems

Overview of WSAA

This Research Project and Occasional Paper were prepared for the Water Services Association of Australia by:

Authors:Jeff Foley PhD Student Advanced Water Management Centre University of Queensland. Emails: [email protected] [email protected]

and

Professor Paul Lant Advanced Water Management Centre University of Queensland. Email: [email protected]

DisclaimerThis Occasional Paper is issued by the Water Services Association of Australia Ltd. on the understanding that:1. The Water Services Association of Australia Ltd.

and individual contributors are not responsible for the results of any action taken on the basis of information in this Occasional Paper, nor for any errors or omissions.

2. The Water Services Association of Australia Ltd. and individual contributors disclaim all and any liability to any person in respect of anything, and the consequences of anything, done or omit-ted to be done by a person in reliance upon the whole or any part of this Occasional Paper.

CopyrightThis document is copyrighted. Apart from any use as permitted under the Copyright Act 1968, no part of this document may be reproduced or transmitted in any form or by any means, electronically or mechanical, for any purpose, without the express written permission of the Water Services Association of Australia Ltd.

© Water Services Association of Australia Ltd, 2009 ALL RIGHTS RESERVED

ISBN 1 920760 47 4

The Water Services Association (WSAA) is the peak body of the Australian urban water industry.

The Association’s 33 members and 29 associate members provide water and sewerage services to approximately 15 million Australians and to many of our largest industrial and commercial enterprises.

WSAA was formed in 1995 to provide a forum for debate on issues important to the urban water industry and to be a focal point for communicating the industry’s views.

WSAA encourages the exchange of information and cooperation between its members so that the industry has a culture of continuous improvement and is always receptive to new ideas.

The functions of WSAA are:• bethevoiceoftheurbanindustryatthenational

and international level and represent the industry in the development of national water policy,

• facilitatetheexchangeofinformationandcom-munication within the industry,

• undertakeresearchofnationalimportancetotheAustralian urban water industry and coordinate

• keynationalresearchfortheindustry,• developbenchmarkingandimprovementactivi-

ties to facilitate the development and improved productivity of the industry,

• developnationalcodesofpracticeforwaterandsewerage systems,

• assessnewproductsrelatingtowater,sewerageand trade waste systems on behalf of the water industry,

• jointlyoverseetheSmartApprovedWatermarkScheme for products and services involved in conserving water use

• coordinateannualmetricbenchmarkingoftheindustry and publish the National Performance Framework with the Federal and State Govern-ments.

1WSAA Occasional Paper No.24 - Direct Methane and Nitrous Oxide emissions from full-scale wastewater treatment systems

ContentsForeword .................................................................................................................................. 2Executive summary .................................................................................................................. 3

Conclusions and recommendations ...................................................................................................5

1. Introduction ......................................................................................................................... 61.1 Background .................................................................................................................................61.2 Scope of study ............................................................................................................................71.3 Report structure ..........................................................................................................................91.4 Project acknowledgements ......................................................................................................10

2. Methane and Nitrous Oxide mass transfer coefficients ................................................... 122.1 Introduction ...............................................................................................................................122.2 Materials and methods .............................................................................................................132.3 Results ......................................................................................................................................152.4 Results validation and uncertainty assessment ........................................................................182.5 Conclusions...............................................................................................................................19

3. Nitrous Oxide generation in full-scale BNR wastewater treatment systems .................... 203.1 Introduction ...............................................................................................................................203.2 Materials and methods .............................................................................................................213.3 Results ......................................................................................................................................293.4 Discussion.................................................................................................................................313.5 Conclusions...............................................................................................................................35

4. Dissolved Methane generation in full-scale rising main sewerage systems ..................... 374.1 Introduction ...............................................................................................................................374.2 Materials and methods .............................................................................................................384.3 Results and discussion .............................................................................................................404.4 Model development ..................................................................................................................424.5 Conclusions...............................................................................................................................45

5. Methane generation in full-scale wastewater treatment systems .................................... 465.1 Introduction ...............................................................................................................................465.2 Materials and methods .............................................................................................................485.3 Results ......................................................................................................................................525.4 Discussion.................................................................................................................................545.5 Conclusions...............................................................................................................................56

6. Implications for the National Greenhouse and Energy Reporting System ........................ 587. Conclusions and recommendations .................................................................................. 60Appendix A - References ........................................................................................................ 61Appendix B - Glossary ............................................................................................................ 65


2 WSAA Occasional Paper No.24 - Direct Methane and Nitrous Oxide emissions from full-scale wastewater treatment systems

Foreword

Authors’ NoteChapter 3 of this Paper, “Nitrous oxide generation in full-scale BNR wastewater treatment systems” has also been submitted for publication in the journal, Water Research. As of 24 September 2009, the Editor of Water Research requested moderate revisions to this paper for publication.

The impacts of climate change are already being felt hard in Australia and of course water re-sources are the first casualty of reduced rainfall and runoff. The urban water industry in Australia is undertaking a huge adaptation effort to ensure water security into the future.

Chapter 4 of this Paper, “Dissolved methane generation in full-scale rising main sewerage systems” has also been submitted for publication in the journal, Water Science & Technology. As of 23 September 2009, the Editor of Water Science & Technology has advised that this paper is accepted for publication.

Most members of the community know that these new supply options such as desalination are more energy intensive. The urban water industry has committed to purchasing and powering these new sources of water with renewable energy and it is estimated that our industry will be one of the largest purchasers of renewable energy in Australia in the next 2-3 years. Less well known though is that the water industry also directly contributes greenhouse gases (Scope 1 emissions) to the atmosphere through wastewater collection, treatment and discharge. Whilst this is still a very small component of Australia’s overall greenhouse gas inventory at around 0.5%, the water industry is committed to reducing its contribution to climate change through minimising the release of these Scope 1 emissions. The major culprit is nitrous oxide, at 300 times the warming potential as carbon dioxide, its presence is ubiquitous in urban wastewater treatment and discharge due to treatment processes that attempt to convert nitrogen in its many forms to nitrogen gas. There are other Scope 1 emissions: methane is well known greenhouse gas and not surprisingly is also present in wastewater systems. However methane collection has improved for on-site generation of energy its contribution to the industry profile is much less than that of nitrous oxide. In 2007 WSAA commissioned the University of Queensland to undertake a literature survey to determine the current state for the science in Scope 1 direct emissions and it found there were particular knowledge gaps in understanding the generation

of these emissions during wastewater collection and treatment. At the same time other international research agencies and urban water utilities were beginning their own research into this issues and quickly an international collaboration was arranged which remains ongoing. WSAA subsequently commissioned this piece of work with the University of Queensland to help its members get a much better grasp of the total greenhouse gas profile of wastewater operations. The outcomes in this paper will be used in important policy discussions and to help wastewater treatment operators find ways in which greenhouse gas generation could be minimised in collection, treatment and discharge of treated sewage. There remain some important areas for research, including the contribution of Scope 1 emissions from treated wastewater discharged to environmental receiving waters and WSAA will continue to work with its members to improve its understanding of how the industry can also contribute to efforts to mitigate climate change. I would like to thank the WSAA members who contributed to this important study and in particular Jeff Foley from the University of Queensland who has completed a very high quality research project.

Adam LovellManager, Science and Sustainability, WSAA


Executive summary

IntroductionAwareness and concern about greenhouse gas (GHG) emissions has grown in recent times. Members of the Water Services Association of Australia (WSAA) are keenly aware of these concerns and the environmental impacts that may be caused by their activities. In late 2007, WSAA commissioned The University of Queensland (UQ) to undertake a comprehensive review of the current regulatory guidance and scientific literature on GHG emissions, with relation to the water and wastewater industries (Foley and Lant, 2008). This review highlighted that current international practice for estimating GHG emissions focuses mainly on emissions associated with energy use. Quantifying direct emissions of methane (CH4) and nitrous oxide (N2O) from wastewater systems was an area of uncertainty, with less developed and less reliable estimation methodologies. In particular, three key knowledge gaps were identified:1. The magnitude of dissolved CH4 generation,

and subsequent unaccounted loss, in low-strength anaerobic systems (e.g. rising mains, gravity sewers, lagoons) is potentially significant, yet the influencing factors are largely unknown;

2. There is a high level of uncertainty in the magnitude and variability of N2O emissions from biological nutrient removal (BNR) processes, under different physical configurations and process conditions; and

3. Nitrous oxide emissions resulting from effluent discharges to specific riverine, estuarine and oceanic environments are poorly quantified in the scientific literature, with inconsistent advice provided in the regulatory guidelines.

The GHG emissions profile of the wastewater industry was shown to be highly sensitive to these uncertainties in estimating direct CH4 and N2O emissions. In the past, these concerns of uncertainty and sensitivity have been largely overlooked by the industry. However, rapid changes to the Federal regulatory landscape, with the introduction of the National Greenhouse and Energy Reporting System (NGERS) and possibly an emissions trading scheme, combined with voluntary organisational commitments to “carbon neutrality”, mean that this level of uncertainty now represents an unacceptable business risk.

Consequently in May 2008, WSAA commissioned UQ to undertake further field-based research to improve the level of certainty in the estimation methodologies for direct CH4 and N2O emissions from wastewater systems. The four elements of this research program were:1. Determination of CH4 and N2O mass transfer

coefficients in wastewater systems, under different lab-scale stripping conditions;

2. Determination of N2O generation in full-scale BNR wastewater treatment systems of varying physical configuration and under different process conditions. This fieldwork program was conducted at seven full-scale BNR wastewater treatment plants in Australia;.

3. Determination of dissolved CH4 generation in a full-scale rising main sewerage system, and development of a simple model to relate CH4 generation to pipeline geometry and hydraulic retention time (HRT); and

4. Determination of CH4 generation in full-scale wastewater treatment systems of varying physical configuration and under different process conditions. This fieldwork program was conducted at four full-scale wastewater treatment plants in Australia.

This study does not address the uncertainty in emissions from wastewater discharges to different receiving environments. Outlined below are short summaries of the findings from the four research elements. Also discussed are the implications of these findings for Federal reporting obligations under NGERS. In light of on-going consultation between WSAA and the Australian Department of Climate Change (DCC), it is expected that outcomes from this research project and others being conducted internationally, could be considered for revisions to the NGERS Technical Guidelines in 2010.


Methane and Nitrous Oxide Mass Transfer CoefficientsMethane and nitrous oxide have very low solubilities in water, under atmospheric conditions. For both compounds, there exists a strong thermodynamic driving force for mass transfer from a wastewater bioreactor to the atmosphere. The purpose of this research element was to better characterise the rate of that liquid-to-gas mass transfer in wastewater bioreactors, under different aeration/mixing conditions. Laboratory-scale stripping experiments were conducted in both clean water and mixed liquor, under different aeration sparge rates. The results of these trials established a means of estimating CH4 and N2O mass transfer coefficients, based on superficial gas velocity. For N2O, these results were then validated and corrected for vessel depth, using gas stripping data from three full-scale aerated reactors. The N2O results were further tested in a sensitivity analysis against two published mass transfer correlation techniques for aerated reactors. The results of this sensitivity analysis showed that the mass transfer coefficients determined in this study fit well within the band of likely values predicted by the correlation techniques.

Nitrous Oxide Generation in Full-Scale BNR Wastewater Treatment SystemsInternational guidance for estimating emissions of N2O from BNR wastewater systems is presently inadequate. This research element adopted a rigorous mass balance approach to provide comprehensive N2O generation results from seven full-scale BNR wastewater treatment plants (WWTP). Nitrous oxide generation was shown to be always positive, yet highly variable across the seven plants. The calculated range of N2O generation was 0.006 – 0.253 kgN2O-N per kgN denitrified (average: 0.035 ± 0.027). Higher N2O generation was shown to generally correspond with higher nitrite concentrations, but with many competing and parallel nitrogen transformation reactions occurring, it was very difficult to clearly identify the predominant mechanism of N2O production. The WWTPs designed for low effluent TN (i.e. < 10 mgN.L-1) had lower and less variable N2O generation factors than plants that only achieved partial denitrification.

Dissolved Methane Generation in Full-Scale Rising Main Sewerage SystemsAt present, the potential generation of CH4 in wastewater collection systems is ignored under international GHG accounting protocols, despite recent reports of substantial dissolved CH4 formation in sewers. This suggests that the current national GHG inventories for wastewater systems are likely to be underestimated for some situations. This research element presents a new catalogue of field data on CH4 formation in rising main sewerage systems and proposes an empirically-fitted, theoretical model to predict fugitive CH4 emissions, based upon the independent variables of pipeline geometry (i.e. surface area to volume ratio, A/V) and HRT. Systems with longer HRT and/or larger A/V ratios are shown to have higher dissolved CH4 concentrations. This simple predictive model provides a means for water authorities to estimate the CH4 emissions from other pressurised sewerage systems of similar characteristics.

Methane Generation in Full-Scale Wastewater Treatment SystemsGuidance from the Intergovernmental Panel on Climate Change (IPCC) on CH4 emissions from anaerobic and facultative wastewater systems is presently based on very limited data. This research element aimed to determine the CH4 generation and emission rates in four full-scale treatment plants, using a mass balance approach that specifically accounted for both gaseous and dissolved CH4. In the lagoon-based plants, CH4 generation was shown to occur even in shallow and mechanically aerated zones, which suggests that methanogenic activity still occurs in the deeper water column and/or sediments of these lagoons. The results of this study also confirmed the IPCC’s general range of CH4 generation values for covered anaerobic lagoons, high-rate anaerobic reactors and mechanically aerated lagoons. Results from a primary sedimentation tank suggest that it may act as a sink for dissolved CH4 in the raw wastewater. It should be noted however, that this research element was not comprehensive and not as insightful as the N2O research element, since no comparison of results between plants of similar process configuration was possible.

Executive summary (continued)


Conclusions and Recommendations

This study has provided the first set of comprehensive full-scale data on direct CH4 and N2O emissions from a range of wastewater systems in Australia.

Executive summary (continued)

In combination with the earlier WSAA and University of Queensland literature review, and in light of other international work in this field, the results from this study could be used to improve the level of certainty in the estimation methodologies published by the DCC in the NGERS Technical Guidelines. With regard to future work in this field, the following lines of inquiry are recommended for further investigation:1. A greater understanding of the fundamental

kinetics of competing nitrite and nitrous oxide transformation mechanisms in full-scale BNR applications is required. The dynamic influent and process conditions of an operating treatment plant demand comprehensive on-line monitoring in all bioreactor compartments for a thorough characterisation.

2. Methane generation in gravity sewers has not been quantitatively addressed in the scientific literature. This warrants further investigation because the results of this study suggest that CH4 is ubiquitous in wastewater environments.

3. The biological and physicochemical interactions between lagoon sediments, the water column and the air-water interface in open systems need to be better characterised. The possibility of methanotrophic activity in the aerobic surface layers of lagoons and quiescent tanks warrants closer investigation, as these could be substantial sinks for dissolved CH4 in raw wastewater.

4. The mass balance approach adopted in this study is recommended as a robust framework for the recommended lines of future investigation.

5. The emissions factors of nitrous oxide, and to a lesser extent methane, from discharge of treated sewage to environmental waters such as rivers, estuaries and seas/oceans remain inaccurate and poorly defined. This area of research will require cross disciplinary expertise for such a complex environment.

The authors thank and acknowledge WSAA and the many collaborators on this research project.


1. Introduction

1

1. Introduction

1.1 BackgroundAwareness and concern about greenhouse gas (GHG) emissions has grown significantly in recent

times. Members of the Water Services Association of Australia (WSAA) are keenly aware of these concerns and the environmental impacts that may be caused by their activities. In late 2007, WSAA commissioned The University of Queensland (UQ) to undertake a comprehensive review of the current regulatory guidance and scientific literature on greenhouse gas emissions, with relation to the water and wastewater industries (Foley and Lant, 2008).

This review highlighted that current international practice for estimating GHG emissions focuses mainly on emissions associated with energy use. Quantifying direct emissions of methane (CH4) and nitrous oxide (N2O) from wastewater systems was an area of uncertainty for the industry, with less developed and less reliable estimation methodologies. However, even by current estimations, direct CH4 and N2O emissions from Australian wastewater systems were shown to be significant. The contribution of the “Water, sewerage and drainage” economic sector to the Australian national GHG inventory was comparable to many other prominent industry sectors (e.g. air transport, railway transport, machinery and equipment) (Foley and Lant, 2008).

The WSAA-UQ literature review examined the scientific literature available for all the common wastewater system processes in Australia and assessed their likelihood of forming methane and/or nitrous oxide (Figure 1.1). From this assessment, WSAA-UQ highlighted three key knowledge gaps for direct GHG emissions from wastewater systems:

1. The magnitude of dissolved CH4 generation, and subsequent unaccounted loss, in low-strength anaerobic systems (e.g. rising mains and gravity sewers, lagoons) is potentially significant, yet the influencing factors are largely unknown;

2. There is a high level of uncertainty in the magnitude and variability of N2O emissions from biological nutrient removal (BNR) processes, under different physical configurations and process conditions; and

3. Nitrous oxide emissions resulting from effluent discharges to specific riverine, estuarine and oceanic environments are poorly quantified in the scientific literature, with inconsistent advice provided in the regulatory guidelines.

The GHG emissions profile of the wastewater industry was shown to be highly sensitive to the

uncertainty in direct CH4 and N2O emissions, as defined by the above three knowledge gaps. In the past, these concerns of uncertainty and sensitivity have been largely overlooked by the industry. However, in the past two years there have been rapid changes to the Federal regulatory landscape for GHG emissions. The National Greenhouse and Energy Reporting System (NGERS), enacted in 2007, was designed to provide a single, streamlined reporting point for all energy and greenhouse reporting obligations. Under this legislation, WSAA expects that most of its members will, for the first time, be required to report their GHG emissions, at both a corporation and facility level. The NGERS will also underpin the Federal Government’s proposed emissions trading scheme, to be known as the Carbon Pollution Reduction Scheme (CPRS). Under this legislation, WSAA expects that some of its members will have a direct financial liability for the GHG emissions from larger wastewater handling facilities.

In light of these new regulatory drivers, and voluntary organisational commitments to “carbon neutrality”, the current level of uncertainty in the water industry’s direct GHG emissions profile now represents an unacceptable business risk.


1. Introduction (continued)

2

Figure 1.1. Wastewater system processes and likelihood of methane and nitrous oxide production.

1.2 Scope of Study The aim of this study was to improve the level of certainty in the estimation methodologies for

direct CH4 and N2O emissions from wastewater systems, as calculated in the NGERS TechnicalGuidelines (DCC, 2008b). Specifically, this study addressed the first two knowledge gaps identified in the earlier WSAA Literature Review, through a four-part research program based largely on full-scale wastewater systems. The four elements of the study were:

1. Methane and nitrous oxide mass transfer coefficients in wastewater systems, under different lab-scale stripping conditions;

2. Nitrous oxide generation in full-scale BNR wastewater treatment systems; 3. Dissolved methane generation in full-scale rising main sewerage systems; and 4. Methane generation in full-scale wastewater treatment systems;

This study does not address the uncertainty in emissions from wastewater discharges to different

receiving environments.

1.2.1 Methane and Nitrous Oxide Mass Transfer Coefficients The purpose of this research element was to determine the mass transfer rates of methane and

nitrous oxide from liquid to gas phase, under different mixing and aeration conditions. Experiments were undertaken in a lab-scale stripping column, in both clean water and mixed liquor.



3

These lab-scale results were then calibrated for nitrous oxide using data collected from some of the BNR WWTPs, and published correlation techniques. The mass transfer coefficients determined in this research element were subsequently applied in the mass balance calculations for nitrous oxide at the full-scale treatment plants.

1.2.2 Nitrous Oxide Generation in Full-Scale BNR Wastewater Treatment Systems The purpose of this research element was to determine the nitrous oxide generation and emission

rates in full-scale treatment plants, of varying physical configuration and under different process conditions. This fieldwork program was conducted at seven full-scale BNR wastewater treatment plants in Australia:

Gibson Island WWTP, Brisbane City Council (1 – 4 July, 2008): o Anaerobic contact tank, extended aeration oxidation ditch process 1 (effluent total

nitrogen (TN) < 3 mg/L); Wetalla WWTP, Toowoomba Regional Council (8 – 11 July, 2008):

o Compartmentalised “Johannesburg” process 2 (effluent TN < 3 mg/L); Woodman Point WWTP, Western Australian Water Corporation (22 – 25 July, 2008):

o Primary sedimentation and sequencing batch reactor (SBR) process 3 (effluent TN < 15 mg/L);

Subiaco WWTP, Western Australian Water Corporation (29 July – 1 August, 2008); and Glenelg WWTP, SA Water (12 – 15 August, 2008);

o Primary sedimentation and Modified Ludzack-Ettinger (MLE) process 4 (effluent TN < 12 mg/L);

25W Western Treatment Plant, Melbourne Water (2 – 12 September, 2008): o Lagoon treatment system (anaerobic lagoon, aerated lagoon, maturation lagoons)

and MLE process (effluent TN < 13 mg/L); St Mary’s WWTP, Sydney Water (7 – 10 October, 2008):

o Compartmentalised A2/O™ process 5 (effluent TN < 3 mg/L). Up to four sampling rounds were conducted at each plant. For each sampling round, COD, TN

and N2O mass balances were constructed. This mass balance approach allowed for a rigorous determination of N2O generation and emission within each zone of the plant.

1 For an “oxidation ditch” process description, refer to Tchobanoglous et al (2003), p.794.

2 For a “Johannesburg” process description, refer to Tchobanoglous et al (2003), p.812.

3 For a “sequencing batch reactor” process description, refer to Tchobanoglous et al (2003), p.792.

4 For a “Modified Ludzack-Ettinger” process description, refer to Tchobanoglous et al (2003), p.791.

5 For an “A2/O” process description, refer to Tchobanoglous et al (2003), p.810.



4

1.2.3 Dissolved Methane Generation in Full-Scale Rising Main Sewerage Systems The purpose of this research element was to develop a simple model that related methane

formation in rising mains to the independent variables of pipeline geometry (i.e. surface area to volume ratio) and hydraulic retention time. This simple model could then be used by WSAA members to estimate the direct emissions of methane from pressurised rising main sewer systems.

This fieldwork was undertaken on an operating rising main on the Gold Coast, Australia. Previously reported rising main field data and complex physico-chemical modelling simulations were also incorporated into this assessment.

1.2.4 Methane Generation in Full-Scale Wastewater Treatment Systems The purpose of this research element was to determine the methane generation and emission rates

in full-scale treatment plants, of varying physical configuration and under different process conditions. This fieldwork program was conducted at four full-scale wastewater treatment plants in Australia:

Bird-in-Hand WWTP, SA Water (19 – 22 August, 2008): o Uncovered facultative, aerated and maturation lagoons;

25W Western Treatment Plant, Melbourne Water (2 – 12 September, 2008): o Anaerobic lagoon, aerated lagoon, maturation lagoons and sludge drying pans;

North Head WWTP, Sydney Water (14 – 17 October, 2008): o High-rate primary sedimentation and ocean outfall;

Yatala Brewery Treatment Plant, Fosters (18 – 21 November, 2008); o Upflow Anaerobic Sludge Blanket (UASB) reactor 6, and J-Cell induced air

flotation. Three to four sampling rounds were conducted at each plant. For each sampling round, a

chemical oxygen demand (COD) and CH4 mass balance was constructed. This mass balance approach allowed for determination of methane generation and emission within each zone of the plant. Due to the small number of plants surveyed, and the practical difficulties associated with properly charactering lagoon-based treatment systems, it should be noted that the outcomes of this research element are less insightful than those achieved for nitrous oxide emissions from BNR WWTPs.

1.3 Report Structure The following four chapters of this report are structured around the results of these four research

elements. Chapter 6 discusses the implications of these results for Federal reporting obligations. The Australian Department of Climate Change (DCC) publishes the National Greenhouse and Energy Reporting (Measurement) Determination and associated Technical Guidelines for liable organisations under the scheme to calculate their annual GHG emissions. The DCC and WSAA have been engaged since 2008 in negotiating improvements to the Technical Guidelines. In light of consultation with the DCC, it is expected that outcomes from this research project and others being conducted internationally, could be considered for revisions to the Technical Guidelines in 2010. These amendments would then be applied to the 2010-11 accounting period.

The final chapter presents the conclusions and recommendations flowing from this study. 6 For an “UASB” process description, refer to Tchobanoglous et al (2003), p.1005.



5

1.4 Project Acknowledgements This work was funded by the Water Services Association of Australia and The University of

Queensland. In-kind support was also provided by United Water International (Adelaide), and the participating water authorities.

The members of the project team for this study were:

Principal authors: Jeff Foley (PhD student, UQ) Professor Paul Lant (UQ)

Fieldwork assistants: Bill Fuller (Engineer, United Water International) Peter Muller (Undergraduate engineer, GHD) Con Foley (volunteer) Lyndon Clark and Debbie Chua (Water Corporation) Ian Johnson (Gold Coast Water)

Analytical and modelling support:

Dr Ursula Werner (UQ) Shihu Hu (PhD student, UQ) Dr David de Haas (UQ, GHD) Dr Keshab Sharma (UQ) Dr Albert Guisasola (Universitat Autònoma de Barcelona) Dr Beatrice Keller-Lehmann and Jianguang Li (UQ) Peng Su and Associate Professor Wei Zhang (Flinders University)

Technical review and stewardship:

Professor Zhiguo Yuan (UQ) Professor Jurg Keller (UQ) Dr David de Haas (UQ, GHD) Dr Paul Rasmussen (United Water International) Hamish Reid (South East Water) Michael Greg and Daniel Starrenburg (Brisbane City Council) Greg Allen (Sydney Water) Keith Cadee (Water Corporation) Adam Lovell (Science and Sustainability Manager, WSAA)



6

The authors also thank the plant owners and operators for their cooperation:

Brisbane City Council Keith Barr, Vijay Jayaraman, Kenny Liew and Grahame Simpson

Fosters Brewery Charlie Foxall

Gold Coast Water Shaun Corrie, Ian Johnson and Kelly O’Halloran

Melbourne Water Kim Daire, Peter Gall, Trevor Gulovsen, Suelin Haynes and Erik Ligtermoet

SA Water / United Water Tim Kelly, Bridget Kingham, Ian MacKenzie and Dr Paul Rasmussen

Sydney Water Heri Bustamante, Steven Lin, Mark Ognibene, Dammika Vitanage, Wanxin Wang and Wasantha Wicks

Toowoomba Regional Council

Alan Kleinschmidt and John Paulger

Western Australian Water Corporation

Claire Camplin, Lyndon Clark, Margaret Domurad, Kay Hyde, Pardeep Kumar, Gabrielle O’Dwyer and Leah Rheinberger


2. Methane and Nitrous Oxide mass transfer coefficients

7

2. Methane and Nitrous Oxide Mass Transfer Coefficients

2.1 IntroductionMass transfer is defined as the movement of molecules from a region of high concentration to a

region of low concentration. In the context of methane (CH4) and nitrous oxide (N2O) in wastewater systems, these compounds are generated in relatively high concentrations in the liquid phase of a bioreactor. By contrast, their ambient concentrations in the atmosphere are very low (CH4 1,774 ppbv, N2O 319 ppbv). Therefore, there should exist a strong driving force for CH4 and N2O to be emitted to the atmosphere from wastewater treatment bioreactors. The controlling mechanisms for this mass transfer process are:

1. Thermodynamic equilibrium – at steady-state, every compound achieves a balance between its concentration in the gas phase (defined as its “partial pressure”), and its concentration in the liquid phase (defined as its “solubility”).

2. Mass transfer kinetics – whilst the ultimate steady-state concentration of the compound is defined by thermodynamics, the rate at which the system achieves that equilibrium can vary greatly. This rate is defined by the volumetric mass transfer coefficient (kLa), which describes:

Gas diffusivity – i.e. how well the gas diffuses through the liquid (a fundamental thermodynamic property); and Interfacial surface area available for mass transfer between the phases. The interfacial surface area is dependent on elements such as the reactor geometry, bubble size, degree of turbulence and mixing.

Thermodynamic equilibrium between gas phase and liquid phase concentrations is described by

Henry’s Law at low pressures (Pauss et al., 1990; Tchobanoglous et al., 2003). Shown in Figure2.1 are the Henry’s Law solubilities of CH4 and N2O in water, under different temperatures and partial pressures, at 1 atm. The solubilities of both CH4 and N2O in water under atmospheric conditions are near zero (i.e. in the order of µg/L). This confirms the earlier assumption that there exists a strong equilibrium driving force for CH4 and N2O to be emitted to the atmosphere from liquid phase wastewater processes. However, when these compounds are being continually evolved in a bioreactor, it is the mass transfer kinetics that control how quickly that system moves towards its equilibrium point.

In this project, it is intended to construct a number of liquid phase CH4 and N2O mass balances around the various wastewater treatment plants and their individual bioreactors (refer to Sections 3 and 5). These mass balances account for the generation and movement of dissolved CH4 and N2O into and out of the bioreactors. One of the key routes for these dissolved gases to leave the bioreactor is via mass transfer to the atmosphere. Therefore, it is critical to have an understanding of the liquid-to-gas mass transfer kinetics of CH4 and N2O, under the different mixing and aeration conditions experienced in wastewater treatment plants.

The purpose of this study element was to conduct a number of laboratory-scale stripping experiments on both CH4 and N2O, in clean water and mixed liquor, under a wide range of aeration/mixing conditions. The results of these laboratory-scale studies are to be validated, where possible, with field-based stripping experiments. The combined results can then to be used to predict the CH4 and N2O mass transfer coefficient, for any size reactor and any range of aeration/mixing conditions.


2. Methane and Nitrous Oxide mass transfer coefficients (continued)

8

0

500

1,000

1,500

2,000

2,500

3,000

0 10 20 30 40Temperature (oC)

N2O

Sol

ubili

ty (m

g.L-1

)

100% Pure N2O Atmosphere

80% N2O Atmosphere

60% N2O Atmosphere

40% N2O Atmosphere

20% N2O Atmosphere

Atmospheric Conditions (319 ppbv)

0

10

20

30

40

0 20 40 60 80 100Temperature (oC)

CH

4 Sol

ubili

ty (m

g.L-1

)

100% Pure CH4 Atmosphere

90% CH4 Atmosphere

80% CH4 Atmosphere (typical for anaerobic treatment)

60% CH4 Atmosphere (typical for anaerobic digester)

15% CH4 Atmosphere (Upper Explosive Limit in Air)

5% CH4 Atmosphere (Lower Explosive Limit in Air)

Atmospheric Conditions (1,774 ppbv)

Figure 2.1. Nitrous oxide and methane solubility in water (1 atm) at various temperatures and partial pressures (Perry et al., 1997, Weiss and Price, 1980).

2.2 Materials and Methods

2.2.1 Laboratory Gas Stripping Experiments The stripping experiments were conducted in a 0.05 m diameter, 0.815 m deep glass column, with

a gas sparging bar at its base. Experiments were conducted at 20°C, in both clean tap water and mixed liquor collected from Gibson Island WWTP, Brisbane (mixed liquor suspended solids 4.45 g.L-1). Measurements of dissolved CH4, N2O and oxygen (O2) were made at three depth locations in the liquid column using an on-line Membrane Inlet Mass Spectrometer (MIMS, Hiden Analytical, Warrington, United Kingdom).

In each N2O stripping experiment, the liquid column was initially saturated by bubbling through 0.51% N2O gas (in helium), which equated to an equilibrium dissolved N2O concentration of 6.45 mg.L-1 at 20°C (Weiss and Price, 1980), and dissolved oxygen equal to zero. The sparge gas was then switched to compressed air, and the dissolved concentration of O2 and N2O monitored until oxygen reached saturation (9.07 mg.L-1 at 20°C, corrected for chlorinity) (APHA, 1995) and N2O approached zero. This procedure was repeated for clean water and mixed liquor at three different sparge flowrates (500 mL.min-1, 200 mL.min-1 and 100 mL.min-1).

Similarly for each CH4 stripping experiment, the liquid column was initially saturated by bubbling through 90% CH4 gas (in 5% N2 and 5% CO2), which equated to an equilibrium dissolved CH4 concentration of 21.3 mg.L-1 at 20°C (Perry et al., 1997), and dissolved oxygen equal to zero. The sparge gas was then switched to compressed air, and the dissolved concentration of O2 and CH4 monitored until oxygen reached saturation and CH4 approached zero. This procedure was repeated for clean water and mixed liquor at three different sparge flowrates (500 mL.min-1, 200 mL.min-1 and 100 mL.min-1) and also under non-aerated quiescent conditions. In the mixed liquor stripping experiments for both N2O and CH4, the clean water saturation concentrations were assumed to be reduced by a 95% β correction factor to account for differences in solubility caused by constituents in the mixed liquor such as salts, particulates and surface-active substances (Tchobanoglous et al., 2003). Hence for N2O, the saturation concentration in mixed liquor reduced from 6.45 to 6.13 mg.L-1, and similarly for CH4, the saturation concentration reduced from 21.3 to 20.2 mg.L-1 (refer to Figures 2.3 and 2.4).



9

The dissolved concentration profiles of O2, CH4 and N2O from each stripping experiment were analysed using a non-linear parameter estimation routine, implemented in MS Excel (Stenstrom et al., 1997), to calculate the mass transfer coefficients for each gas. This approach was based upon the industry standard Measurement of Oxygen Transfer in Clean Water (Stenstrom, 2007).

2.2.2 Field Nitrous Oxide Stripping Measurements Six off-gas samples were collected from the aerated bioreactors at Glenelg WWTP, Adelaide

(reactor depth = 4.5 m); two samples from Woodman Point WWTP, Perth (reactor depth = 3.9 m); and one sample from Western Treatment Plant, Melbourne (reactor depth = 4.7 m). A floating gas hood (Flinders University, Adelaide) was deployed on the surface of the aerated bioreactor (Figure2.2), with off-gas samples collected in sealed 1 L Tedlar® gas sample bags (Sigma-Aldrich #24633). Prior to sample collection, the floating hood head space was purged by completely immersing it in the bioreactor, with the vacuum valve open (item 10 in Figure 2.2). The valve was then closed and the hood allowed to float back to the surface, creating a headspace under slight vacuum.

The collected off-gas samples were analysed for CH4, CO2 and N2O using a gas chromatograph (HaiXin Chromatograph Instrument Co., Shanghai), fitted with TCD, FID and ECD detectors. The off-gas flux at the time of sampling was retrieved from the plant supervisory control and data acquisition (SCADA) system, as the aeration flowrate to the bioreactor.

Dissolved N2O concentration measurements were made using a Clark-type microsensor (N2O25 with 70µm outside tip diameter, Unisense A/S, Aarhus, Denmark), logged via a picoammeter to a laptop. Further details of this instrument are provided in Section 3.2.2. Mixed liquor temperature was measured using a portable water quality meter (TPS 90FLMV).

For each off-gas sample, the field mass transfer coefficient was then calculated as per Equation 2.1:

[ ] [ ]( )*

22

,

2

2

SRR

ONGONA

FL ONONV

MWTR

pQ

ak−×

××

×

= (2.1)

where kLaF = volumetric mass transfer coefficient, calculated from field data (d-1)

QA = aeration flowrate supplied to reactor at the time of sampling (m3.d-1) pN2O,G

= partial pressure of N2O in off-gas from reactor (atm) MWN2O = molecular weight of nitrous oxide (44 kg.kmol-1) R = universal gas constant (0.08206 m3.atm.kmol-1.K-1) T = reactor liquid temperature at time of sampling (K)

VR = volume of the reactor (ML) [N2O-N]R = concentration of N2O-N in the reactor (mg.L-1 or kg.ML-1), measured by

microsensor [N2O-N]s

* = saturation concentration of N2O-N in water at atmospheric conditions

= 2.57 × 10-4 kg.ML-1 at 20°C (Weiss and Price, 1980)



10

Figure 2.2. Floating gas hood (Flinders University, Adelaide) – plan view and side view.

2.3 ResultsThe dissolved O2, CH4 and N2O concentration profiles from the laboratory stripping experiments

are shown in Figures 2.3 and 2.4. Also shown in Figures 2.5A and 2.5B are the results of the CH4 and N2O mass transfer coefficient calculations, determined by the non-linear parameter estimation routine. As expected, the mass transfer coefficients decreased with the superficial gas velocity, vg (m3.m-2.s-1) in a power law relationship. This result is similar to other empirical mass transfer modelling approaches for aerated systems, such as bubble columns (Envirosim, 2007; Garcia-Ochoa and Gomez, 2009). However, the volumetric mass transfer coefficient is also proportional to several other variables, including reactor geometry (particularly aerator immersion depth), aeration bubble size and diffuser layout (Gillot et al., 2005). Hence the need to calibrate the lab-scale bubble column results to full-scale reactors that are significantly deeper.



11

A0

2

4

6

8

10

15 25 35 45 55Time (min)

Dis

solv

ed N

2O (m

g.L-1

)

0

2

4

6

8

10

Dis

solv

ed O

2 (m

g.L-1

)

Dissolved N2O (probe 1)

Dissolved N2O (probe 2)

Dissolved O2 (probe 1)Dissolved O2 (probe 2)

B 0

2

4

6

8

10

20 30 40 50Time (min)

Dis

solv

ed N

2O (m

g.L-1

)

0

2

4

6

8

10

Dis

solv

ed O

2 (m

g.L-1

)

C0

2

4

6

8

10

0 20 40 60Time (min)

Dis

solv

ed N

2O (m

g.L-1

)

0

2

4

6

8

10

Dis

solv

ed O

2 (m

g.L-1

)

D 0

2

4

6

8

10

0 20 40 60 80 100Time (min)

Dis

solv

ed N

2O (m

g.L-1

)

0

2

4

6

8

10

Dis

solv

ed O

2 (m

g.L-1

)

E0

2

4

6

8

10

10 30 50 70 90 110Time (min)

Dis

solv

ed N

2O (m

g.L-1

)

0

2

4

6

8

10

Dis

solv

ed O

2 (m

g.L-1

)

F 0

2

4

6

8

10

0 20 40 60 80 100 120Duration (min)

Dis

solv

ed N

2O (m

g.L-1

)

0

2

4

6

8

10

Dis

solv

ed O

2 (m

g.L-1

)

Figure 2.3. Dissolved N2O and O2 profiles from lab-scale stripping experiments: A) Clean water, 500 mL.min-1 sparge flowrate; B) Mixed liquor, 500 mL.min-1 sparge flowrate; C) Clean water, 200 mL.min-1 sparge flowrate; D) Mixed liquor, 200 mL.min-1 sparge flowrate; E) Clean water 100 mL.min-1 sparge flowrate; and F) Mixed liquor, 100 mL.min-1 sparge flowrate.

The results from the field N2O mass transfer coefficients are shown in Figure 2.5B. These are

clearly lower than the mass transfer coefficients measured for the shallow lab-scale column. The N2O kLa power law estimation from the lab-scale mixed liquor experiments can thus be empirically corrected to account for the increased depth of full-scale WWTP reactors, using a sum-of-least-squares fitting algorithm in MS Excel, as per Equation 2.2:

( ) 86.049.0

, 500,342 g

L

RONFL v

DDak ××

⎭⎬⎫

⎩⎨⎧

=−

(2.2)

where DR = depth of the field reactor (m) DL = depth of the lab stripping column (0.815 m) vg = superficial gas velocity of the field reactor (m3.m-2.s-1)



12

A 0

10

20

30

5 15 25 35Time (min)

Dis

solv

ed C

H4 (

mg.

L-1)

0

4

8

12

Dis

solv

ed O

2 (m

g.L-1

)

Dissolved CH4 (probe 1)Dissolved CH4 (probe 2)Dissolved CH4 (probe 3)Dissolved O2 (probe 1)Dissolved O2 (probe 2)Dissolved O2 (probe 3)

B 0

10

20

30

5 15 25 35Time (min)

Dis

solv

ed C

H4 (

mg.

L-1)

0

4

8

12

Dis

solv

ed C

H4 (

mg.

L-1)

C 0

10

20

30

10 20 30 40 50 60Time (min)

Dis

solv

ed C

H4 (

mg.

L-1)

0

4

8

12

Dis

solv

ed O

2 (m

g.L-1

)

D 0

10

20

30

0 10 20 30 40 50 60Time (min)

Dis

solv

ed C

H4 (

mg.

L-1)

0

4

8

12

Dis

solv

ed O

2 (m

g.L-1

)

E 0

10

20

30

0 10 20 30 40 50 60 70 80Time (min)

Dis

solv

ed C

H4 (

mg.

L-1)

0

4

8

12

Dis

solv

ed O

2 (m

g.L-1

)

F 0

10

20

30

0 10 20 30 40 50 60 70 80 90Time (min)

Dis

solv

ed C

H4 (

mg.

L-1)

0

4

8

12

Dis

solv

ed O

2 (m

g.L-1

)

Figure 2.4. Dissolved CH4 and O2 profiles from lab-scale stripping experiments: A) Clean water, 500 mL.min-1 sparge flowrate; B) Mixed liquor, 500 mL.min-1 sparge flowrate; C) Clean water, 200 mL.min-1 sparge flowrate; D) Mixed liquor, 200 mL.min-1 sparge flowrate; E) Clean water 100 mL.min-1 sparge flowrate; and F) Mixed liquor, 100 mL.min-1 sparge flowrate.

Equation 2.2 can be used to calculate the N2O volumetric mass transfer coefficient for any

positively aerated reactor, provided the depth and prevailing specific aeration flowrate are known. No field data were available to undertake a similar validation of the lab-scale methane mass

transfer coefficients.



13

A

y = 121,856x + 0R2 = 1

0

200

400

600

800

0.000 0.001 0.002 0.003 0.004Superficial Gas Velocity (m3.m-2.s-1)

Met

hane

kLa

at 2

0o C (d

-1)

B

y = 34,524.94x0.86

R2 = 0.98

0

100

200

300

400

500


N2O

kLa

at 2

0o C (d

-1)

Lab Values in Clean Water

Lab Values in Mixed Liquor

Field Values

Depth Correction

C

0

50

100

150

200

250

0.000 0.001 0.002 0.003Superficial Gas Velocity (m3.m-2.s-1)

N2O

kLa

in M

ixed

Liq

uor a

t 20o C

(d-1

)

K&S Alpha 0.4, Db 2mmK&S Alpha 0.7, Db 2mmK&S Alpha 0.4, Db 4mmK&S Alpha 0.7, Db 4mmDudley Alpha 0.4, Db 2mmDudley Alpha 0.7, Db 2mmDudley Alpha 0.4, Db 4mmDudley Alpha 0.7, Db 4mmThis Study

Figure 2.5. A) Methane volumetric mass transfer coefficients (kLa) from clean water and mixed liquor stripping experiment in a lab-scale column; B) Nitrous oxide volumetric mass transfer coefficients (kLa) from clean water and mixed liquor stripping experiment in a lab-scale column, and from field measurements; C) Comparison of modelled kLa values (Equation 2.2) against literature correlations for a range of bubble diameters, DB and αF factors. The shaded area indicates the 95% confidence interval adopted for the N2O kLa values.

2.4 Results Validation and Uncertainty Assessment To confirm the validity of this kLa estimation technique and to determine the uncertainty

associated with the calculated values, two independent mass transfer correlation techniques were applied to field data collected from seven WWTPs (refer to Section 5). The first correlation technique (Khudenko and Shpirt, 1986) empirically related the clean water oxygen kLa to vg, reactor geometry and aeration bubble diameter, DB. The N2O kLa can then be estimated in accordance with Higbie’s penetration model (Pauss et al., 1990; Capela et al., 2001):

2

2

22,

,

OF

ONFOLONL D

Dakak ×= (2.4)



14

where DF,N2O = Molecular diffusivity of N2O in water

= 1.84 × 10-9 m2.s-1 at 20°C (Tamimi et al., 1994) DF,O2 = Molecular diffusivity of oxygen in water

= 1.98 × 10-9 m2.s-1 at 20°C (Ferrell and Himmelblau, 2002) The second correlation (Dudley, 1995) empirically related the clean water N2O kLa to vG, DF,N2O,

DB and viscosity. Both correlations were also corrected for temperature, using a standard θ factor of 1.024 (Tchobanoglous et al., 2003).

Both correlations are sensitive to bubble diameter, which was unknown for all WWTPs. Furthermore, the clean water kLa values must be corrected for mixed liquor conditions and diffuser fouling (i.e. αF factor), which was also unknown. Therefore, a range of likely values was evaluated for both parameters: DB = 2 – 4mm (Hasanen et al., 2006) and αF = 0.4 – 0.9 (Tchobanoglous et al., 2003). The results of these sensitivity analyses are shown in Figure 2.5C for the two correlations.

Importantly, it can be seen that the N2O kLa model proposed for this study (Equation 2.2) fits well within the band of likely values, and thus provides an independent validation of the adopted approach. The correlations also provide an uncertainty range for the N2O kLa value. A 95% confidence interval of the values shown in Figure 2.5C is adopted for each kLa determined by Equation 2.2.

Similarly, the N2O and CH4 mass transfer coefficients for quiescent reactor zones (i.e. primary sedimentation tanks, anaerobic zones, anoxic zones, secondary sedimentation tanks) can be estimated using the empirical correlation technique of Van’t Riet (1979). This relates kLa to the volumetric power input (P/V) for mixing. Similar to the approach for aerated zones, a range of likely P/V values was surveyed (2 – 8 W.m-3) and used to construct a 95% confidence interval for each kLa value. However, the kLa values in these quiescent zones are an order of magnitude smaller than those in the aerated zones.

2.5 ConclusionsMethane and nitrous oxide have very low solubilities in water, under atmospheric conditions. For

both compounds, there exists a strong thermodynamic driving force for mass transfer from a wastewater bioreactor to the atmosphere. The purpose of this study was to better characterise the rate of that liquid-to-gas mass transfer in wastewater bioreactors, under different aeration/mixing conditions.

Laboratory-scale gas stripping experiments were conducted in both clean water and mixed liquor, under different aeration sparge rates. The results of these lab trials established a means of estimating CH4 and N2O mass transfer coefficients, based on the superficial gas velocity. For N2O, these results were then validated and corrected for vessel geometry, using gas stripping data from three full-scale WWTPs.

Using this combination of lab-scale and full-scale gas stripping data, estimation methodologies are proposed for determining CH4 and N2O mass transfer coefficients, based on bioreactor depth and superficial gas velocity. The kLa estimation methodology for N2O was then tested in a sensitivity analysis against two published kLa correlation techniques for aerated reactors. The results of this sensitivity analysis showed that the mass transfer coefficients determined in this study fit well within the band of likely values predicted by the correlation techniques. This analysis also provided a 95% confidence interval for the kLa values calculated by Equations 2.2 and 2.3.


3. Nitrous Oxide generation in full-scale BNR wastewater treatment systems

15

3. Nitrous Oxide Generation in Full-Scale BNR Wastewater Treatment Systems

3.1 IntroductionThe United Nations Framework Convention on Climate Change (UNFCCC) is the globally

recognised basis for collective action on the reduction of anthropogenic greenhouse gas (GHG) emissions (UNFCCC, 2007). One of the key obligations for signatory countries under the UNFCCC is the compilation of an annual national GHG inventory, covering four general sectors (energy; industrial processes; agriculture, forestry and other land use; and waste). Emissions of methane (CH4) and nitrous oxide (N2O) from wastewater treatment and discharge are reported under the waste sector (IPCC, 2006b). However, GHG emissions are not usually measured directly, but are rather estimated through the application of models that link emissions to data on observable activities.

The Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC, 1997) estimation methodology for N2O emissions from wastewater handling assumed minimal nitrogen removal occurs during treatment, and hence all influent nitrogen is discharged into rivers and/or estuaries, where it is mineralised, nitrified and denitrified under natural environmental processes. During these transformations, some of the discharged nitrogen will be emitted to the atmosphere as N2O, at a default rate of 0.01 kgN2O-N.kgN-1

discharged (uncertainty range: 0.002 – 0.12) (IPCC, 1997). The 2006 IPCC Guidelines for National Greenhouse Inventories subsequently revised this default emission factor to 0.005 kgN2O-N.kgN-1

discharged (uncertainty range: 0.0005 – 0.25). The assumption that minimal nitrogen removal occurs in wastewater treatment plants (WWTPs)

is incorrect for many countries. Recognising this, the 2006 IPCC Guidelines also updated the N2O estimation methodology to include direct emissions from WWTPs with “controlled nitrification and denitrification steps” (IPCC, 2006a). The proposed default emission factor was 0.0032 kgN2O.person-1.yr-1 (uncertainty range: of 0.002 – 0.008), based on one full-scale study by Czepiel et al (1995) on a basic secondary treatment plant in New Hampshire, USA. Assuming a wastewater nitrogen loading of 16 g.person-1.d-1 for developed countries (i.e. high protein intake) (Tchobanoglous et al., 2003; IPCC, 2006a; DCC, 2008b), this equates to approximately 3.5 × 10-4 kgN2O-N.kgN-1

influent. This 2006 approach is little different from that of the Revised 1996 IPCC Guidelines, in that the application of a basic emissions factor lacks scientific verification. The process description in the New Hampshire study was not sufficient to determine the extent of nitrification-denitrification activity (if any), nor does it seem reasonable to extrapolate the very low result from this one plant for use as an international default N2O emission factor for biological nutrient removal (BNR) WWTPs.

Foley and Lant (2008) recently reviewed 11 published studies on N2O emissions from full-scale and lab-scale wastewater systems with BNR. For low to medium strength municipal systems, a median emission factor of 0.01 kgN2O-N.kgN-1

influent was found (range: 0.0003 – 0.03). Importantly, the wide range of process conditions surveyed in these studies highlighted a number of significant influencing variables, such as carbon substrate availability, dissolved oxygen (DO) concentration and the presence of potentially inhibitory intermediates (i.e. nitrite NO2

- and nitric oxide NO). Limited carbon substrate was shown by several authors to have a strong influence on N2O

emissions, due to incomplete denitrification (Itokawa et al., 1996; Schulthess and Gujer, 1996; Parket al., 2000; Schalk-Otte et al., 2000; Itokawa et al., 2001). Oxygen limitation or low DO concentration was considered to have a number of effects in BNR systems, such as controlling the metabolism of ammonia-oxidising bacteria (AOB) (Poth and Focht, 1985; Kampschreur et al., 2008), and acting as an inhibitor to the heterotrophic N2O reductase (Schulthess and Gujer, 1996; Gejlsbjerg et al., 1998; Kimochi et al., 1998; Schalk-Otte et al., 2000). Many authors reported a positive correlation between bulk NO2

- concentration and N2O emissions (Gejlsbjerg et al., 1998; Park et al., 2000; Beline et al., 2001). This was postulated to be due to the inhibition of the N2O reductase by high concentrations of NO, NO2

-, or its equilibrium partner, HNO2 (Itokawa et al.,


3. Nitrous Oxide generation in full-scale BNR wastewater treatment systems (continued)

16

2001; Zhou et al., 2008). The production of N2O in the presence of NO2- could also be attributed to

denitrification by AOB, under oxygen-limited conditions (Hynes and Knowles, 1984; Bock et al., 1995).

International guidance on N2O emissions from wastewater systems is presently inadequate for the advanced BNR process configurations being used in many developed countries. Furthermore, there is a lack of comprehensive studies on full-scale WWTPs that would allow for better characterisation of the N2O emissions potentially occurring under different physical configurations and process conditions. The purpose of this study was to address this knowledge gap by comprehensively surveying a range of full-scale BNR WWTPs to determine the magnitude and variability of N2O generation under different designs and operating conditions.


3.2.1 Field Sampling Sites For this study, seven full-scale BNR WWTPs were sampled. Their basic features are listed in

Table 3.1. These plants were chosen to provide a range of plant sizes, process configurations, effluent qualities and climatic conditions. The widest possible range of WWTP types (in Australia) was selected with the expectation that differences between plant design and process conditions might help elucidate the factors influencing N2O production.

3.2.2 Samples Collection and Analysis For each WWTP, it was intended to conduct four intensive sampling rounds (2 – 4 h duration

each, morning and afternoon on two consecutive days). Due to a combination of circumstances, this was not possible at all plants. However, 20 of the intended 28 sampling rounds were completed over a five month timeframe in the Australian winter/spring of 2008.

For each sample round, data was collected to enable the construction of chemical oxygen demand (COD), total nitrogen (TN) and N2O-N mass balances over the entire WWTP. The sampling locations and types of data collected are illustrated in Figure 3.1. Field data collection consisted of a combination of 1) wastewater grab samples; 2) measurement of process conditions, namely temperature, pH, dissolved oxygen and oxidation-reduction potential (ORP), using a portable water quality meter (TPS 90FLMV); and 3) dissolved N2O concentration measurements using a Clark-type microsensor (N2O25 with 70µm outside tip diameter, Unisense A/S, Aarhus, Denmark), logged via a picoammeter to a laptop. The electrochemical microsensor was two-point calibrated before and after each field sampling round at ambient temperature, using distilled water (zero point) and a freshly prepared 0.15 mM N2O solution. Laboratory trials with the N2O25 microsensor in 0.27 mM N2O solution showed its maximum measurement error to be ± 0.3%. The response of the electrochemical microsensor is known to be linear in the range of 0 – 1.2 mM (Andersen et al., 2001). At each sampling location, the microsensor was fully immersed in the reactor and allowed to stabilise and log data at 3 s intervals over a 5 – 10 minute period. The signal data from this sampling period at each location was then statistically analysed to calculate its 95% confidence interval (t-dist, α = 0.05). The microsensor was also zero point calibrated with distilled water in the field after every measurement to provide a baseline signal and correct for any drift.



17

Tabl

e 3.

1.

Was

tew

ater

trea

tmen

t pla

nt si

tes.

No.

Abb

revi

ated

Nam

e

Loca

tion

Influ

ent

Flow

rate

(ML.

d-1)

App

rox.

SRT

(d)

Efflu

ent T

otal

N

itrog

en

(mg.

L-1

)

Proc

ess D

escr

iptio

n

(Bol

dite

ms a

re th

e pr

oces

s uni

ts sa

mpl

ed in

this

stud

y)

1 “Ox.

Ditc

h”

Bris

bane

, Q

ueen

slan

d 38

13

3

Inle

t wor

ks, a

naer

obic

con

tact

tank

, 2 ×

ext

ende

d ae

ratio

n ox

idat

ion

ditc

hes(

in p

aral

lel)

with

diff

used

aer

atio

n, se

cond

ary

sedi

men

tatio

n;

mec

hani

cal s

ludg

e th

icke

ning

and

dew

ater

ing.

2 “Joh

anne

sbur

g”

Toow

oom

ba,

Que

ensl

and

10

20

5 In

let w

orks

, 2 ×

ext

ende

d ae

ratio

n Jo

hann

esbu

rg b

iore

acto

rs (i

n pa

ralle

l)w

ith su

bmer

ged

aspi

ratin

g O

KI™

aer

ator

s, se

cond

ary

sedi

men

tatio

n, sl

udge

th

icke

ning

, aer

obic

dig

estio

n, m

echa

nica

l and

sola

r dew

ater

ing.

3 “SB

R”

Perth

, W

este

rn A

ustra

lia

137

16

15

Inle

t wor

ks, p

rim

ary

sedi

men

tatio

n, se

quen

cing

bat

ch r

eact

or (4

co

mpa

rtmen

ts)w

ith d

iffus

ed a

erat

ion

and

bio-

sele

ctor

zon

e, sl

udge

th

icke

ning

, ana

erob

ic d

iges

tion

and

mec

hani

cal d

ewat

erin

g.

4 “MLE

(1)”

Perth

, W

este

rn A

ustra

lia

63

13

12

Inle

t wor

ks, p

rim

ary

sedi

men

tatio

n, 1

1 ×

cove

red

Mod

ified

Lud

zack

Et

tinge

r (M

LE) b

iore

acto

rs (i

n pa

ralle

l)w

ith d

iffus

ed a

erat

ion,

seco

ndar

y se

dim

enta

tion,

slud

ge th

icke

ning

, mec

hani

cal d

ewat

erin

g an

d lim

e st

abili

satio

n.

5 “MLE

(2)”

Ade

laid

e,

Sout

h A

ustra

lia

49

8 11

In

let w

orks

, pri

mar

y se

dim

enta

tion,

2 ×

Inte

grat

ed F

ixed

Film

Act

ivat

ed

Slud

ge b

iore

acto

rs +

1×

MLE

bio

reac

tor

(in p

aral

lel)

with

diff

used

aer

atio

n,

seco

ndar

y se

dim

enta

tion,

slud

ge th

icke

ning

and

ana

erob

ic d

iges

tion.

6 “MLE

(3)”

Mel

bour

ne,

Vic

toria

20

0 15

13

A

naer

obic

lago

on, 1

× M

LE b

iore

acto

r w

ith d

iffus

ed a

erat

ion,

seco

ndar

y se

dim

enta

tion,

mat

urat

ion

lago

ons,

and

slud

ge w

astin

g to

an

aera

ted

facu

ltativ

e la

goon

.

7 “A2 /O

”

Sydn

ey,

New

Sou

th W

ales

25

14

3

Inle

t wor

ks, 2

× p

aral

lel t

rain

s: 1

) Prim

ary

sedi

men

tatio

n, 4

-sta

ge B

arde

npho

bi

orea

ctor

with

diff

used

aer

atio

n, se

cond

ary

sedi

men

tatio

n; 2

) Pre

-fer

men

ter,

four

-sta

ge b

iore

acto

r (s

imila

r to

A2 /O

con

figur

atio

n) w

ith d

iffus

ed a

erat

ion

and

supp

lem

enta

l CO

D d

osin

g by

prim

ary

slud

ge fr

om T

rain

1, s

econ

dary

se

dim

enta

tion,

terti

ary

filtra

tion,

slud

ge th

icke

ning

, aer

obic

dig

estio

n, sl

udge

la

goon

and

mec

hani

cal d

ewat

erin

g.



18

Figure 3.1. Typical data (types and locations) collected for the construction of the COD and total nitrogen mass balances over the entire process at each WWTP.

The TPS 90FLMV water quality meter was calibrated before and after the two-day sampling

exercise at each WWTP. Grab samples for soluble species were immediately filtered using 0.22µm syringe filters, acid-preserved and kept on ice before analysis.

COD was measured by the colorimetric method described in APHA (1995) using commercial Lovibond tubes in a range of 0 to 150 mgCOD.L-1. The ammonium (NH4

+), nitrate (NO3-) and

nitrite (NO2-) concentrations were analysed using a Lachat QuikChem8000 Flow Injection Analyser

(Lachat Instrument, Milwaukee, USA). Total Kjeldahl Nitrogen (TKN), mixed liquor suspended solids (MLSS) and volatile solids (MLVSS) were analysed according to Standard Methods (APHA, 1995).

Physical plant data (e.g. reactor dimensions, plant flowrates, diurnal aeration flowrates) were provided by the WWTP operators for the specific days and times of field sampling. Where possible, the WWTP operators also supplied their own routine process and analytical data that provided a useful cross-check against results from the grab samples collected during the field study.

Off-gas samples were also collected from the aerated reactor zones at each WWTP. A floating gas hood (Flinders University, Adelaide) was deployed on the surface of the aerated bioreactor, with off-gas samples collected in sealed 1 L Tedlar® sample bags (Sigma-Aldrich #24633). Prior to sample collection, the floating hood head space was purged by opening an air relief valve on top of the hood, and then completely immersing it in the bioreactor. The relief valve was then closed and the hood allowed to float back to the surface, creating a headspace under slight vacuum. The collected samples were then analysed for N2O using a gas chromatograph (HaiXin Chromatograph Instrument Co., Shanghai), fitted with TCD, FID and ECD detectors. However, due to difficulties with sample preservation and analysis, off-gas results were only obtained from WWTP No.5 (six samples), WWTP No. 3 (two samples) and WWTP No. 6 (one sample).



19

3.2.3 COD and Total Nitrogen Mass Balances over Entire WWTP Processes At each of the seven WWTP sites, total COD and TN mass balances were constructed across the

entire process. These mass balances drew upon the analytical data collected in the field (i.e. COD, TKN, NO3

-, NO2-, MLSS, MLVSS concentrations), as well as the plant data (i.e. flowrates, reactor

volumes, solids capture efficiencies, biosolids tonnages and composition, biogas production and composition) supplied by the WWTP operators (refer to Figure 3.1). The purpose of this initial mass balance analysis was to: 1) ensure an accurate characterisation of the WWTP operation, such that both COD and TN balances over the WWTP generally achieved greater than 90% closure; and 2) to determine the mass of nitrogen denitrified and emitted to the atmosphere, according to Equation 3.1:

SNEffNInfNAtmDNN MMMM ,,,, −−= (3.1)

where MN,AtmDN = Mass of N denitrified to atmosphere, either as N2 or N2O gas (kg.d-1)

MN,Inf = Mass of nitrogen in influent (kg.d-1) MN,Eff = Mass of nitrogen in effluent (kg.d-1) MN,S = Mass of nitrogen in wasted solids (kg.d-1)

Equation 3.1 assumes that the WWTP is operating at steady-state, with no net accumulation of

nitrogen within the biomass inventory. All of the plants investigated had medium to long solids retention times (SRT) (i.e. 8 – 20 d). Hence, the change in biomass inventory is relatively slow, and the assumption of near steady-state conditions should hold for the two day sampling period at each WWTP.

Given the large variation in physical size and treated load of the seven WWTPs surveyed, Equation 3.1 provided a means of normalising the generation and emissions of nitrous oxide (i.e. as a percentage of the total nitrogen denitrified to atmosphere).

3.2.4 Liquid Phase Nitrous Oxide Mass Balances over Individual WWTP Zones The second phase of mass balance analysis examined bulk liquid phase nitrous oxide across the

individual zones of each WWTP. For five of the seven sites, the WWTP was divided into five reactor zones for this mass balance analysis (i.e. primary sedimentation tank or anaerobic zone; anoxic zone; highly aerated aerobic zone; less aerated aerobic zone; and secondary sedimentation tank). At WWTP No.6, the plant was divided into seven reactor zones (three anoxic zones, three tapered flow aerobic zones, and secondary sedimentation tank). At WWTP No.3 (SBR), the mass balance was divided across each operational phase (i.e. fill/aerate, settle, decant). The general formulation of the mass balance construction is given in Equation 3.2:

RNONRNONOutNONInNONRNON GTrMM

dtdM

,,,,,

2222

2−−−−

− +−∑−∑= (3.2)

where

dtdM RNON ,2 − = change in mass of N2O-N in the reactor zone, over time (kg.d-1)

InNONM ,2 −∑ = sum of i mass flows of N2O-N into the reactor zone (kg.d-1)

OutNONM ,2 −∑ = sum of j mass flows of N2O-N out of the reactor zone (kg.d-1)

RNONTr ,2 − = mass transfer of N2O-N from the reactor liquid to gas phase (kg.d-1)



20

RNONG ,2 − = net generation of N2O-N in the reactor zone (kg.d-1) (i.e. net result of N2O-N production and consumption due to biological reactions in the reactor)

Similar to Equation 3.1, it was assumed that the reactor zones operate at near steady-state

conditions, and are well-mixed. Equation 3.2 can then be expanded and re-formulated to solve for RNONG ,2 − :

[ ] [ ] [ ] [ ]( )*22,2,2,,2 SRLRiIn

iiInR

jjOutRNON NONNONakVNONQNONQG −−−××+−×−−×= ∑∑−

(3.3) where QIn,i , QOut,j = individual flows in and out of the reactor zone (ML.d-1)

[N2O-N]In,i = concentration of N2O-N in the incoming streams (mg.L-1 or kg.ML-1), which is generally equal to the N2O-N concentration in the originating reactor

[N2O-N]R = concentration of N2O-N in the reactor zone (mg.L-1 or kg.ML-1) VR = volume of the reactor zone (ML) kLa = volumetric mass transfer coefficient (d-1) [N2O-N]s

* = saturation concentration of N2O-N in water at atmospheric conditions

= 2.57 × 10-4 kg.ML-1 at 20°C (Weiss and Price, 1980) This calculation was completed for each reactor zone at each WWTP. It was then repeated at each

reactor zone for the “best-case” combination of the lower limit values of the 95% confidence intervals of the measured N2O-N concentrations (refer to section 2.2) and estimated mass transfer coefficients (refer to Figure 3.2). The calculation was then repeated again for the “worst-case” combination of the upper limit values of the 95% confidence intervals. These calculations determined the uncertainty range of net N2O-N generation (i.e. production minus consumption) in each reactor zone. Negative values of RNONG ,2 − indicate N2O-N consumption is greater than N2O-N production in that particular zone.

The net generation and emissions of N2O-N in each reactor zone were then summed to give the net generation, WWTPNONG ,2 − , and emissions, WWTPNONTr ,2 − , of N2O-N for the whole WWTP:

RNONWWTPNON GG ,, 22 −− ∑= (3.4A)

[ ] [ ]( ){ }*22,, 22 SRLRRNONWWTPNON NONNONakVTrTr −−−××∑=∑= −− (3.4B)

The ratio of N2O-N mass transfer emissions to net generation was then calculated for each

WWTP:

%100%,

,

2

2 ×=−

−

WWTPNON

WWTPNONWWTP G

TrTr (3.5)



21

3.2.5 Normalisation of Nitrous Oxide Mass Balance Results

To compare results between different reactor zones, RNONG ,2 − was normalised by dividing by the corresponding total mass of nitrogen denitrified to the atmosphere from the entire WWTP:

AtmDNN

RNONR M

GGF

,

,2 −= (3.6A)

where GFR = N2O-N generation factor for each reactor zone (kgN2O-N.kgN-1

denitrified) MN,AtmDN is calculated according to Equation 3.1.

Similarly, the mass transfer emissions to atmosphere from each reactor zone, RNONTr ,2 − , were normalised by dividing by MN,AtmDN :

AtmDNN

RNONR M

TrEF

,

,2 −= (3.6B)

where EFR = N2O-N emissions factor for each reactor zone (kgN2O-N.kgN-1

denitrified)

To compare results across sites, WWTPNONG ,2 − for each WWTP was normalised by dividing by its corresponding total mass of nitrogen denitrified to the atmosphere:

AtmDNN

WWTPNONWWTP M

GGF

,

,2 −= (3.7)

where GFWWTP = N2O-N generation factor for entire WWTP (kgN2O-N.kgN-1

denitrified).

3.2.6 Nitrous Oxide Volumetric Mass Transfer Coefficients To determine the N2O mass transfer coefficient (kLa) for each WWTP reactor zone (refer to

Equation 3.3), a series of lab-scale stripping experiments was conducted. These experiments were conducted in both clean tap water and mixed liquor (from WWTP No.1, MLSS 4.45 g.L-1) in a 0.05 m diameter, 0.815 m deep glass column, with a gas sparging bar at its base. Measurements of dissolved nitrous oxide and dissolved oxygen were made at two locations in the liquid column using an on-line Membrane Inlet Mass Spectrometer (Hiden Analytical, Warrington, United Kingdom).

In each stripping experiment, the liquid column was initially saturated by bubbling through 0.51% N2O gas (in helium), which equated to an equilibrium dissolved N2O concentration of 6.45 mg.L-1 at 20°C (Weiss and Price, 1980), and dissolved oxygen equal to zero. The sparge gas was then switched to compressed air, and the dissolved concentration of O2 and N2O monitored until oxygen reached saturation (9.07 mg.L-1 at 20°C, corrected for chlorinity) (APHA, 1995) and N2O approached zero. This procedure was repeated for clean water and mixed liquor at three different sparge flowrates (500 mL.min-1, 200 mL.min-1 and 100 mL.min-1). In the mixed liquor stripping experiments, the clean water saturation concentrations were reduced by a 95% β correction factor to



22

account for differences in solubility caused by constituents in the mixed liquor such as salts, particulates and surface-active substances (Tchobanoglous et al., 2003).

The nitrous oxide mass transfer coefficient for each experiment was then calculated using a non-linear parameter estimation routine (Stenstrom et al., 1997). The results of these calculations are shown in Figure 3.2A. As expected, the mass transfer coefficient decreased with superficial gas velocity, vg (m3.m-2.s-1), in a power law relationship. This result is similar to other empirical mass transfer modelling approaches for aerated systems, such as bubble columns (Envirosim, 2007; Garcia-Ochoa and Gomez, 2009).

A B

y = 34,524.94x0.86

R2 = 0.98

0

100

200

300

400

500


N2O

kLa

at 2

0o C (d

-1)

Lab Values in Clean Water

Lab Values in Mixed Liquor

Field Values

Depth Correction

0

50

100

150

200

250

0.000 0.001 0.002 0.003Superficial Gas Velocity (m3.m-2.s-1)

N2O

kLa

in M

ixed

Liq

uor a

t 20o C

(d-1

)

K&S Alpha 0.4, Db 2mmK&S Alpha 0.7, Db 2mmK&S Alpha 0.4, Db 4mmK&S Alpha 0.7, Db 4mmDudley Alpha 0.4, Db 2mmDudley Alpha 0.7, Db 2mmDudley Alpha 0.4, Db 4mmDudley Alpha 0.7, Db 4mmThis Study

Figure 3.2. A) Nitrous oxide volumetric mass transfer coefficients (kLa) from clean water and mixed liquor stripping experiment in a lab-scale column, and from field measurements; B) Comparison of modelled kLa values for all WWTPs (Equation 3.9) against literature correlations for a range of bubble diameters, DB and αF factors. The shaded area indicates the 95% confidence interval adopted for the kLa values in Equation 3.3.

However, the volumetric mass transfer coefficient is also related to several other variables, including reactor geometry (particularly aerator immersion depth), aeration bubble size, diffuser layout and liquid viscosity (Gillot et al., 2005). To better correlate the lab-scale bubble column results to full-scale reactors that are significantly deeper (i.e. 3.68 – 5.95 m), six off-gas samples were collected from the aerated zones of WWTP No.5 (depth = 4.5 m), two off-gas samples from WWTP No. 3 (depth = 3.9 m) and one sample from WWTP No. 6 (depth = 4.7 m) (refer to Section3.2.2). For each sample, the mass transfer coefficient was then calculated as per Equation 3.8:

[ ] [ ]( )*

22

,

2

2

SRR

ONGONA

FL ONONV

MWTR

pQ

ak−×

××

×

= (3.8)

where kLaF = volumetric mass transfer coefficient, calculated from field data (d-1)

QA = aeration flowrate supplied to reactor at the time of sampling (m3.d-1) pN2O,G

= partial pressure of N2O in off-gas from reactor (atm) MWN2O = molecular weight of nitrous oxide (44 kg.kmol-1) R = universal gas constant (0.08206 m3.atm.kmol-1.K-1)



23

T = reactor liquid temperature at time of sampling (K) The results for these field kLa estimates are also shown in Figure 3.2A, and are lower than the

mass transfer coefficients measured for the shallower lab-scale column. Therefore, the power law estimation based on vg only, from the lab-scale mixed liquor experiments (refer to Figure 3.2A)

was modified by the addition of a correction factor κ

⎭⎬⎫

⎩⎨⎧

L

R

DD

, to account for the increased depth of

full-scale WWTP reactors. The value of κ was empirically determined using a sum-of-least-squares fitting algorithm in MS Excel. The resulting depth-corrected kLa correlation is shown in Equation 3.9:

( ) 86.049.0

* 500,34 gL

RFL v

DDak ××

⎭⎬⎫

⎩⎨⎧

=−

(3.9)

where DR = depth of the field reactor (m) DL = depth of the lab stripping column (0.815 m) vg = superficial gas velocity of the field reactor (m3.m-2.s-1) Using Equation 3.9, the nitrous oxide volumetric mass transfer coefficient was calculated for all

aerated reactor zones in the seven WWTPs. To determine the uncertainty associated with these field kLa values, two independent mass transfer correlation techniques were applied to the field data. The first correlation technique (Khudenko and Shpirt, 1986) empirically related the clean water oxygen mass transfer coefficient (

2OL ak ) to vg, reactor geometry and aeration bubble diameter, DB. The N2O mass transfer coefficient ( ONL ak

2) was then estimated in accordance with Higbie’s penetration

model (Pauss et al., 1990; Capela et al., 2001):

2

2

22,

,

OF

ONFOLONL D

Dakak ×= (3.10)

where DF,N2O = Molecular diffusivity of N2O in water

= 1.84 × 10-9 m2.s-1 at 20°C (Tamimi et al., 1994) DF,O2 = Molecular diffusivity of oxygen in water

= 1.98 × 10-9 m2.s-1 at 20°C (Ferrell and Himmelblau, 2002) The second correlation (Dudley, 1995) empirically related the clean water N2O kLa to vG, DF,N2O,

DB and viscosity. Both correlations were also corrected for temperature, using a standard θ factor of 1.024 (Tchobanoglous et al., 2003):

2020

−×= TCLCTL oo akak θ (3.11)

Both correlations are sensitive to bubble diameter, which was unknown for all WWTPs.

Furthermore, the clean water kLa values must also be corrected for mixed liquor conditions and diffuser fouling (i.e. αF factor), which was also unknown. Therefore, a range of likely values was evaluated for both parameters: DB = 2 – 4mm (Hasanen et al., 2006) and αF = 0.4 – 0.9



24

(Tchobanoglous et al., 2003). The results of these sensitivity analyses are shown in Figure 3.2B for the two correlations. It can be seen that the kLa model proposed for this study (Equation 3.9) fits well within the band of likely values and thus provides an independent validation of the adopted approach. The correlations also provide an uncertainty range for the kLa value used in Equation3.3. Using the full range of sensitivity analysis values calculated from the two correlations, a 95% confidence interval (t-dist, α = 0.05) of kLa values was calculated (i.e. shaded area in Figure 3.2B). This confidence interval was then applied to each kLa determined by Equation 3.9. From Figure3.2B, it can be seen that the values from Equation 3.9 sit in the lower quartile of the confidence interval generated from the literature correlations.

The mass transfer coefficients for quiescent reactor zones (i.e. primary sedimentation tanks, anaerobic zones, anoxic zones, secondary sedimentation tanks) were estimated using the empirical correlation technique of Van’t Riet (1979). This relates kLa to the volumetric power input (P/V) for mixing. Similar to the approach for aerated zones, a range of likely P/V values was surveyed (2 – 8 W.m-3) and used to construct a 95% confidence interval for each field reactor zone. However, the kLa values in these quiescent zones (indicatively 3 – 4 d-1) were an order of magnitude smaller than those in the aerated zones. Therefore, the results are not especially sensitive to these values.

3.3 ResultsShown in Figure 3.3 are the dissolved N2O concentrations (and associated 95% confidence

intervals) measured by the microsensor in each reactor zone, at each WWTP. A wide range of values is apparent across the reactor zones, sampling rounds and WWTPs, from near zero to greater than 1 mg.L-1. Shown in Figure 3.4 are the normalised net N2O-N generation (GFR) and mass transfer emission profiles (EFR), per reactor zone and sampling round, at each of the continuous flow WWTPs.

1 - Ox. Ditch

0.00

0.02

0.04

0.06

0.08

Rd1 Rd2 Rd3 Rd4

Dis

solv

ed N

2O C

once

ntra

tion

(mg.

L-1)

AnAxHI AeLO AeSST

2 - Johannesburg

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

Rd1 Rd2 Rd3

Dis

solv

ed N

2O C

once

ntra

tion

(mg.

L-1) An

AxHI AeLO AeSST

3 - SBR

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Rd1 Rd2 Rd3

Dis

solv

ed N

2O C

once

ntra

tion

(mg.

L-1)

PSTBioselectorTurbulent BioselectorAeratedDecant

4 - MLE (1)

0.0

0.1

0.2

0.3

0.4

Rd1

Dis

solv

ed N

2O C

once

ntra

tion

(mg.

L-1) PST

AxHI AeLO AeSST

5 - MLE (2)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Rd1 Rd2 Rd3 Rd4

Dis

solv

ed N

2O C

once

ntra

tion

(mg.

L-1)

PSTAxHI AeLO AeSST

6 - MLE (3)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Rd1 Rd2 Rd3

Dis

solv

ed N

2O C

once

ntra

tion

(mg.

L-1) Ax1

Ax2Ax3Ae1Ae2Ae3SST

7 - A2/O

0.00

0.01

0.02

0.03

0.04

Rd1 Rd2

Dis

solv

ed N

2O C

once

ntra

tion

(mg.

L-1) An

AxHI AeLO AeSST

Figure 3.3. Dissolved N2O concentrations measured at each WWTP. Error bars indicate the 95% confidence interval, based on the microsensor sampling data.



25

1 - Ox. Ditch

-0.02

-0.01

0.00

0.01

0.02

0.03

An Ax HI Ae LO Ae SSTs

N2O

-N G

F R a

nd E

F R (k

gN2O

-N.k

gND

N-1

)

2 - Johannesburg

-0.02

-0.01

0.00

0.01

0.02

0.03

0.04


N2O

-N G

F R a

nd E

F R (k

gN2O

-N.k

gND

N-1

)

4 - MLE(1)

-0.04

-0.02

0.00

0.02

0.04

PSTs Ax HI Ae LO Ae SSTs

N2O

-N G

F R a

nd E

F R (k

gN2O

-N.k

gND

N-1

)

5 - MLE(2)

-0.04

-0.02

0.00

0.02

0.04

0.06

0.08

0.10

PSTs Ax HI Ae LO Ae SSTs

N2O

-N G

F R a

nd E

F R (k

gN2O

-N.k

gND

N-1

)

6 - MLE(3)

-0.02

-0.01

0.00

0.01

0.02

0.03

0.04

Ax1 Ax2 Ax3 Ae1 Ae2 Ae3 SSTs

N2O

-N G

F R a

nd E

F R (k

gN2O

-N.k

gND

N-1

)

-0.20

-0.10

0.00

0.10

0.20

0.30

0.40

7 - A2/O

-0.02

-0.01

0.00

0.01

0.02

0.03

0.04


N2O

-N G

F R a

nd E

F R (k

gN2O

-N.k

gND

N-1

)

Rd1 EF Rd2 EF Rd3 EF Rd4 EFRd1 GF Rd2 GF Rd3 GF Rd4 GF

Figure 3.4. Net N2O-N generation (GFR) and mass transfer emissions (EFR) profiles, per reactor zone, in each sampling round at the six continuous flow WWTPs. “An” – anaerobic zone, “Ax” – anoxic zone, “HI Ae” – highly aerated aerobic zone, “LO Ae” – less aerated aerobic zone, “PSTs” – primary

Secondary axis



26

sedimentation tanks, “SSTs” – secondary sedimentation tanks. Results from Round 1 at WWTP No. 6 are plotted on the secondary axis – note the change in scale. Error bars indicate the combined uncertainty from the 95% confidence intervals of both kLa and measured dissolved N2O concentrations.

The calculated total WWTP generation factors (GFWWTP) for the 20 sampling rounds are reported in Table 3.2 and show a wide range of results across two orders of magnitude. The minimum generation factor was 0.006, and the maximum 0.253 kgN2O-N.kgN-1

denitrified. The average of the 20 samples was 0.035 ± 0.027 (t-dist, α = 0.05). However, close inspection of these results shows the average to be skewed upwards by four results greater than 0.03 kgN2O-N per kgNdenitrified. Excluding these results from the statistical sample, the average of the remaining 16 results is 0.013 ± 0.003 kgN2O-N.kgN-1

denitrified. Table 3.2 also reports the percentage of generated N2O-N that is lost via mass transfer to the

atmosphere. Generated N2O-N can also be lost as dissolved N2O-N in the WWTP effluent and waste solids. However, it is clear from the results in Table 3.2, that these losses are generally minor (i.e. < 5%). Whilst N2O is a highly soluble gas in water, it also has a high mass transfer coefficient (similar to oxygen), so it is not unexpected that the majority of N2O generated in the bioreactors is quickly stripped to the atmosphere.

Shown in Figure 3.5 are the N2O-N generation factor results plotted against bulk bioreactor NO2-

-N concentration, and two process design parameters, namely effluent total nitrogen and a-recycle rate (i.e. recycle rate from aerobic zone to anoxic zone). Figure 3.5A suggests that the high N2O-N generation factors generally correspond with high bulk NO2-N concentrations in the bioreactor. Figures 3.5B and 3.5C together show particular design parameters that potentially influence the stability of WWTPs in relation to N2O generation.

3.4 Discussion

3.4.1 N2O Generation in Full-scale WWTPs This study has provided the first set of comprehensive N2O-N generation results for a range of

full-scale BNR WWTPs. The mass balance approach adopted for individual reactor zones allowed for more detailed profiling of N2O-N generation and mass transfer emissions than has previously been reported (cf. Czepiel et al., 1995; Kimochi et al., 1998; Peu et al., 2006; Tallec et al., 2006). In particular, the N2O-N generation/emission profiles in Figure 3.4 highlight several key points:

1. Mass transfer emissions of N2O-N to atmosphere occur predominantly in the aerated zones, due to significantly larger mass transfer coefficients;

2. The net generation of N2O-N in the various anaerobic zones, PSTs and SSTs is generally positive, but small in comparison to the net generation in the anoxic and aerobic zones;

3. The net generation of N2O-N in the various anoxic zones is almost always positive; and 4. The net generation of N2O-N in the various highly aerated (HI Ae) and less aerated (LO

Ae) aerobic zones varies considerably between large positive values (indicating N2O-N production > N2O-N consumption) and large negative values (indicating N2O-N production < N2O-N consumption). This suggests that the aerobic zones tend to “swing” in terms of their net N2O-N production potential, likely depending on changes in environmental conditions.



27

Tabl

e 3.

2.

Nitr

ous o

xide

gen

erat

ion

fact

or (G

F WW

TP) a

nd %

mas

s tra

nsfe

r em

issi

ons (

%Tr

WW

TP) f

or e

ach

WW

TP a

nd sa

mpl

e ro

und.

N2O

-NG

FW

WTP

(kg

N2O

-N.k

gN-1

deni

trifi

ed)

Not

e 2

N2O

-N M

ass T

rans

fer

Em

issi

ons (

%Tr

WW

TP)

WW

TP

No.

Rou

nd 1

R

ound

2

Rou

nd 3

R

ound

4

Rou

nd 1

R

ound

2

Rou

nd 3

R

ound

4

1 –

Ox.

Ditc

h

0.00

8

(0.0

05 –

0.0

16)

0.00

6

(0.0

03 –

0.0

12)

0.01

3

(0.0

08 –

0.0

23)

0.00

6

(0.0

03 –

0.0

13)

98.2

– 9

9.0%

98

.5 –

99.

3%

95.3

– 9

8.2%

96

.3 –

98.

7%

2 –

Joha

nnes

burg

0.01

4

(0.0

08 –

0.0

25)

0.02

1

(0.0

11 –

0.0

39)

0.01

1

(0.0

04 –

0.0

25)

96.8

– 9

8.8%

99

.0 –

99.

6%

88.4

– 9

7.7%

3 –

SBR

0.01

0

(0.0

09 –

0.0

21)

0.01

7

(0.0

15 –

0.0

34)

0.07

11

(0.0

64 –

0.1

46)

94

.6 –

97.

4%

94.6

– 9

7.5%

93

.7 –

97.

2%

4 –

MLE

(1)

0.02

7

(0.0

18 –

0.0

45)

67

.7 –

86.

3%

5 –

MLE

(2)

0.02

2

(0.0

14 –

0.0

34)

0.07

81

(0.0

56 –

0.1

34)

0.00

7

(0.0

04 –

0.0

10)

0.09

61

(0.0

68 –

0.1

66)

88.3

– 9

5.3%

94

.1 –

97.

6%

71.4

– 8

7.4%

94

.5 –

97.

8%

6 –

MLE

(3)

0.25

31

(0.1

77 –

0.4

51)

0.01

0

(0.0

05 –

0.0

21)

0.00

6

(0.0

03 –

0.0

11)

99

.6 –

99.

8%

99.7

– 9

9.9%

99

.9%

7 –

A2 O

0.01

8

(0.0

14 –

0.0

36)

0.01

0

(0.0

07 –

0.0

19)

95

.1 –

98.

0%

94.0

– 9

7.6%

Not

es:

1. B

old

resu

lts a

re g

reat

er th

an 0

.03

kgN

2O-N

.kgN

-1de

nitri

fied.

2. F

igur

es in

bra

cket

s re

pres

ent t

he u

ncer

tain

ty r

ange

for

GF W

WTP

res

ults

, bas

ed o

n th

e co

mbi

natio

n of

95%

con

fiden

ce in

terv

als

for

both

kLa

val

ues a

nd m

easu

red

diss

olve

d N

2O c

once

ntra

tions

.



28

A)

0.00

0.10

0.20

0.30

0.40

0.0 0.2 0.4 0.6 0.8Bioreactor NO2-N (mgN.L-1)

N2O

-N G

F WW

TP (k

gN2O

-N.k

gND

N-1

)Ox. DitchJo'BurgMLE(1)MLE(2)A2/OMLE(3)SBR

B)

0.00

0.10

0.20

0.30

0.40

0 5 10 15 20

Effluent Total Nitrogen (mgN.L-1)

N2O

-N G

F WW

TP (k

gN2O

-N.k

gND

N-1

)

C)

0.00

0.10

0.20

0.30

0.40

0 10 20 30 40 50 60

a-recycle rate (xQ)

N2O

-N G

F WW

TP (k

gN2O

-N.k

gND

N-1

)

Figure 3.5. Net N2O-N generation factor, GFWWTP, in each sampling round, plotted against A) bulk bioreactor nitrite-N concentration, B) effluent total nitrogen, and C) a-recycle rate (aerobic → anoxic) as a multiple of the average influent flowrate, Q.

The predominance of mass transfer emissions from the aerated zones is in agreement with other

full-scale studies of aerated and quiescent bioreactors (Czepiel et al., 1995). The profiles in Figure3.4 show that whilst little atmospheric emission occurs from the quiescent anoxic zones, they actually are responsible for a substantial portion of the N2O generated in the bioreactor. Examination of the net generation profile, rather than the emission profile, is therefore more useful in the context of understanding (and possibly controlling) the underlying mechanisms of N2O formation.

The measurement of positive dissolved N2O concentrations in all reactor zones across all seven WWTPs (refer to Figure 3.3) suggests that N2O is pervasive in most BNR plants. Furthermore, the calculation of positive N2O-N generation factors for all WWTPs in Table 3.2 confirms that these facilities are sources of anthropogenic GHG emissions for UNFCCC accounting purposes. It should also be noted that the generation factors for all sample rounds were positive, even when accounting for the uncertainty of a best-case combination of a low mass transfer coefficient and low dissolved N2O concentration.



29

The variation in generation factor results in Table 3.2 highlights two key points. Firstly, WWTP Nos. 1 and 2 had low generation factors, with low variability. This was true also for the limited sampling rounds at WWTP No.7. Secondly, WWTP Nos. 3, 5 and 6 generally had higher generation factors, but also exhibited significantly higher variability over a short time frame. For example at WWTP No.5, the generation factor varied from 0.022 kgN2O-N.kgN-1

denitrified in the morning sampling round, to 0.078 in the afternoon sampling, decreasing to 0.007 the following morning, and then increased again to 0.096 that afternoon. The results also highlight that some WWTPs might operate steadily with relatively low N2O generation, and then suffer some process perturbation that leads to a temporary spike in N2O formation. This was witnessed most noticeably at WWTP No.6, which had a peak N2O GF of 0.253 kgN2O-N.kgN-1

denitrified in the first sampling round, but relatively low GFs in the remaining sampling rounds. Together, these two points indicate that both process design and variability in operating process conditions are likely to influence the magnitude and variability of N2O-N generation factors.

3.4.2 Influence of Process Conditions on N2O Generation Figure 3.5A suggests that high N2O-N generation factors generally correspond with higher bulk

NO2-N concentrations in the bioreactor. Although the number of data points is limited, there appears to be a threshold value at approximately 0.3 – 0.5 mg.L-1 NO2

--N at which the generation factor jumps to be >> 0.03 kgN2O-N.kgN-1

denitrified. All four results > 0.03 kgN2O-N.kgN-1denitrified

occurred under conditions where bulk NO2-N concentrations exceeded 0.4 mg.L-1. Under aerobic conditions, nitrite is known to accumulate when the rate of ammonia oxidation to

nitrite by the ammonia oxidising bacteria (AOB) exceeds the rate of nitrite oxidation to nitrate by the nitrite oxidising bacteria (NOB). In full-scale WWTPs, the raw wastewater characteristics and process operating conditions are very dynamic and it is likely that the competing rates of AOB and NOB oxidation are constantly fluctuating. Sudden process perturbations such as a shift in bioreactor pH, an increase in raw wastewater TKN loading or a drop in DO concentration (due to increased loading or aeration capacity limitations) could therefore lead to transient spikes in reactor nitrite concentrations (Butler et al., 2005; Park et al., 2007; Sinha and Annachhatre, 2007).

A second possible consequence of a change in process conditions, especially increased nitrogen load or aeration capacity limitations, is that the local DO concentration will decrease. Under oxygen-limited conditions, it is known that AOB have the capability to switch from oxidising ammonium using oxygen as the electron acceptor, to oxidising ammonium with the concomitant reduction of nitrite to N2O (Hynes and Knowles, 1984; Poth and Focht, 1985; Bock et al., 1995; Shiskowski and Mavinic, 2006; Kampschreur et al., 2008). It is not entirely clear whether all AOB have the ability to reduce nitrite completely to N2 or only to N2O (Wrage et al., 2001; Shiskowski and Mavinic, 2006). In their 15N labeled denitrification experiments on Nitrosomonas europaea, Poth and Focht (1985) found N2O was the only reduction product, and that nitrite was its primary source. Similarly, Chain et al (2003) reported that the N. europaea genome possesses the denitrification enzymes for NO2

- to N2O, but not the enzymes for NO3- or N2O. Nonetheless, this

capacity of some AOB to switch from classical ammonia oxidation with oxygen, to oxidation with nitrite producing N2O, could have contributed to the wide variation in net N2O-N generation seen in the aerobic zones (refer Figure 3.4).

A third possible consequence of changed process conditions and nitrite accumulation in the aerobic zone is increased recycle of nitrite back to the anoxic zone. It is postulated that nitrite is the limiting substrate for AOB oxidation in the anoxic zone, since it is normally found at low concentrations and certainly much lower than the concentration of influent ammonium. Therefore, a short-term increase in the nitrite concentration recycled from the aerobic zone could lead to increased N2O generation by AOB in the anoxic zone. This is supported by the net N2O-N generation profiles of Figure 3.4. All sample rounds with high overall generation factors had high N2O-N generation in the anoxic zone.



30

A further possible consequence of higher NO2- concentrations in the anoxic zone is inhibition of

the classical heterotrophic denitrification pathway (Wrage et al., 2001; Zhou et al., 2008). In this pathway, the final step in the reduction sequence from NO3

- to N2 is left incomplete due to inhibition of the N2O reductase. It is postulated that inhibition of heterotrophic denitrification constitutes a second possible mechanism for anoxic N2O formation that could proceed in parallel with AOB oxidation with nitrite. This postulate is also in accordance with the observed peaks in N2O-N formation in the anoxic zone, under conditions of high bulk NO2

--N concentration.

3.4.3 Influence of Process Design on N2O Generation Figure 3.5B and 3.5C together show particular design parameters that potentially influence the

stability of WWTPs in relation to N2O generation. Both WWTP Nos. 1 and 2 have very high a-recycle rates and correspondingly low effluent TN concentrations. The WWTPs that had more highly variable N2O-N generation factors tended to have lower a–recycle rates and correspondingly higher effluent TN concentrations. From Figure 3.5C, there appears to be a threshold at approximately TN 10 mg.L-1, below which the N2O-N generation factor is relatively low and more stable.

The results of this study suggest that the process characteristics of WWTPs designed for near-complete denitrification lead to lower and more stable N2O generation than plants with partial denitrification. The lower N2O generation is probably an unintended consequence of the features usually employed in WWTPs designed to achieve low effluent TN concentrations. Such design features include influent flow (and load) balancing, high a-recycle rate, large bioreactor volume, long SRT and sometimes external carbon dosing. The very high recycle rates in such plants (e.g. approximately 100 times influent flowrate in Carrousel™-type oxidation ditches) also tend to substantially dilute the concentrations of all the intermediates of nitrification-denitrification, including nitrite and nitric oxide, thereby reducing their inhibitory effect (Casey et al., 1999b, a). Larger reactors and/or influent load balancing, together with sufficient aeration capacity and a rapidly responding DO control system that meets peak oxygen demand would also reduce the risk of the transient spikes in reactor nitrite (even at relatively low concentrations) that are postulated to provide a precursor for N2O formation.

In light of the possible N2O formation mechanisms, it is suggested that plants designed for low effluent TN concentrations and that approach “ideal” well-mixed conditions (i.e. high recycle rates) are expected to have relatively low N2O generation factors. This is true even for such plants that might rely upon a significant degree of simultaneous nitrification-denitrification and operate at relatively low reactor DO concentrations (e.g. oxidation ditches). On the other hand, plants that do not have a high degree of denitrification and approach more “plug flow” mixing conditions (i.e. low recycle rates) are more likely to be more susceptible to process perturbations, nitrite accumulation and subsequently high N2O generation factors. This is also likely to be true for sequencing batch reactors, which operate timed process sequences that are similar to plug flow continuous reactors.

3.5 ConclusionsInternational guidance on N2O emissions from wastewater systems is presently inadequate for the

advanced BNR process configurations being used in many developed countries. Furthermore, there is a lack of comprehensive studies on full-scale WWTPs to better characterise the extent of likely N2O emissions. This study has adopted a rigorous mass balance approach to provide the first set of comprehensive N2O-N generation results for a range of full-scale BNR WWTP configurations and process conditions.

Nitrous oxide concentration and net N2O generation was shown to be positive in all 20 sample rounds at the seven WWTPs surveyed. The generation factors calculated across the plants varied considerably in the range 0.006 – 0.253 kgN2O-N.kgN-1

denitrified (average: 0.035 ± 0.027), with the majority of N2O formed being subsequently emitted to the atmosphere (i.e. > 95%).



31

High N2O-N generation factors were shown to generally correspond with elevated bulk NO2--N

concentrations in the bioreactor. However, the mechanisms of nitrite and nitrous oxide formation in full-scale WWTPs are complex. There are many competing and parallel reactions of production and consumption for both compounds. In particular, nitrite is simultaneously a product, a substrate and an inhibitor, which can be formed and utilised under both aerobic and anoxic conditions, by several different types of microorganisms. Therefore, it is very difficult in full-scale operating WWTPs to clearly identify the predominant mechanism of N2O production.

However, the results presented in this study suggest that WWTPs designed for low effluent TN (i.e. < 10 mgN.L-1) generally have lower and less variable N2O generation factors than plants that only achieve partial denitrification. This is likely to be a fortunate consequence of the more generous process design in such “advanced” WWTPs (e.g. influent load balancing, high a-recycle rate, large bioreactor volume).

Future work in this field should concentrate upon gaining a greater understanding of the fundamental kinetics of the competing nitrite and nitrous oxide transformation mechanisms in full-scale BNR applications. The dynamic influent and process conditions of an operating WWTP demand comprehensive on-line monitoring in all bioreactor compartments for a thorough characterisation. The mass balance approach adopted in this study is recommended as a robust framework for the subsequent data analysis.


4. Dissolved Methane generation in full-scale rising main sewerage systems

32

4. Dissolved Methane Generation in Full-Scale Rising Main Sewerage Systems

4.1 IntroductionThe United Nations Framework Convention on Climate Change is the globally recognised basis

for collective action on the reduction of anthropogenic greenhouse gas emissions (UNFCCC, 2007). One of the key obligations for signatory countries under the UNFCCC is the compilation of an annual national greenhouse gas (GHG) inventory, covering four general sectors (energy; industrial processes; agriculture, forestry and other land use; and waste). Emissions of methane and nitrous oxide from wastewater treatment and discharge are reported under the waste sector (IPCC, 2006b). However, GHG emissions are not usually measured directly, but are rather estimated through the application of models that link emissions to data on observable activities.

Presently, the inventory guidelines provided by the Intergovernmental Panel on Climate Change (IPCC) state that “wastewater in closed underground sewers is not believed to be a significant source of methane”, and hence no emission estimation methodology is provided (IPCC, 2006a). However, the plentiful supply of readily biodegradable carbon and the presence of anaerobic biofilms, particularly in fully surcharged rising mains, suggests that there is potential for the formation of methane in the raw sewage, and this is presently unaccounted in national GHG inventories (Foley et al., 2008).

Rising main sewers are anaerobic environments where both sulphate reduction and methanogenesis occur simultaneously using the organic material in wastewater as electron donors. However, most of the previous research on sewer modelling focused only on sulphide formation. Pomeroy (1959) and Thistlethwayte (1972) proposed empirical equations to predict sulphide formation based on wastewater organic strength, pipe characteristics and flow conditions. More recently, the WATS (Wastewater Aerobic/Anaerobic Transformations in Sewers) model was developed to provide a more detailed description of the carbon and sulphur transformations in sewers (Hvitved-Jacobsen et al., 2000; Yongsiri et al., 2003; Abdul-Talib et al., 2005; Nielsen et al., 2005a; Nielsen et al., 2005b; Nielsen et al., 2006). Successful attempts have also been completed to unify the traditional transport and conversion models for sewers with the Activated Sludge Model No.3 (ASM3) (Henze et al., 2000; Huisman et al., 2003). Most recently, Sharma et al (2008) developed a comprehensive model to describe the dynamics of sulphide production in sewer systems. However, none of these models considered methane formation in sewers. In their investigations of a pilot plant pressure sewer, Tanaka and Hvitved-Jacobsen (2002) did consider methane formation, but assumed it to be negligible.

Recently, Guisasola et al. (2008) showed a positive correlation between dissolved methane concentration and hydraulic retention time (HRT) in fully-surcharged, pressurised rising mains on the sub-tropical Gold Coast, Australia. Dissolved methane concentrations were shown to increase along the length of the sewer, which is reflective of increasing HRT, the plug-flow nature of fluid transport in fully surcharged pipelines and the methanogenic biofilm activity on the inner pipe surface. These authors showed a loss of 60 – 70 mg.L-1 soluble chemical oxygen demand (COD) in laboratory-scale sewer reactors, of which 72% was demonstrated to be utilised by methanogens. These direct measurements of dissolved methane in full-scale rising mains (and corroborated by laboratory-scale reactors) confirmed the substantial generation potential of methane in anaerobic rising mains.

The purpose of this study was to develop a theoretically robust, yet easily accessible model that related methane formation in rising mains to the independent variables of pipeline geometry (i.e. surface area to volume ratio) and hydraulic retention time. This functional model is proposed for water authorities to estimate the direct emissions of methane from similar pressurised rising main sewer systems, and hence address an acknowledged uncertainty in the IPCC GHG inventory guidelines for wastewater systems.


4. Dissolved Methane generation in full-scale rising main sewerage systems (continued)

33


4.2.1 Field Sampling Site Field samples were obtained from along a single rising main (CO16) of the Gold Coast area

(Queensland, Australia), as shown in Figure 4.1. Samples were collected at four locations: 1) at the CO16 pump station wet well; 2) a pressurised sampling point 500 m downstream; 3) a pressurised sampling point 1,100 m downstream; and 4) a pressurised sampling point 1,900 m downstream of the CO16 pump station. At 1,350 m (i.e. intersection of Amity Rd and Foxwell Rd), a second rising main (CO19) discharges into the larger CO16 rising main. Therefore, the samples collected at the 1,900 m sample point include the sewage from both rising mains. Samples from the first three locations were collected hourly between 5:00 a.m. and 8:00 a.m. to cover a wide range of HRTs, as the morning peak of raw wastewater entered the pump station’s collection wet well. Samples at the fourth location (1,900 m) were collected at 5:00 a.m. and 8:00 a.m. only.

Figure 4.1. Aerial photograph of CO16 and CO19 rising mains. Samples were collected from the CO16 pump station wet well (0 m), CO16 sample point 1 (at 500 m), CO16 sample point 2 (at 1,100 m) and CO16 sample point 3 (at 1,900 m, downstream CO19 rising main injection point). CO19 discharges into CO16 at 1,350 m.

The CO16 rising main had an internal pipe diameter of 300 mm (surface area to volume ratio, A/V = 13.3 m2.m-3), a total daily pumped flow of 707.4 m3, with 33 pump start events (typically 4 – 6 minutes in duration). The volume of wastewater pumped into the pipe during each pump run was calculated from the physical dimensions of the CO16 wet well and measured stop/start water levels.

The 1,700 m long CO19 rising main had an internal pipe diameter of 250 mm (A/V = 16.0 m2.m-

3), a total daily pumped flow of 92.6 m3 and 34 pump start events (typically 2 – 3 minutes in duration). The volume of wastewater pumped into the pipe during each pump run was calculated from the physical dimensions of the CO19 wet well and measured stop/start water levels.

N



34

The HRT of a wastewater sample was defined as the time spent by that particular “slug” in the rising main up until the time of sampling. This was calculated from the operational data of the single pump stations feeding the CO16 and CO19 rising mains respectively (i.e. pump start/stop times recorded at 15 s intervals by the on-line supervisory control and data acquisition (SCADA) system), the pipeline geometry, and assuming ideal plug flow conditions. The HRT for the first three sample points ranged from 0.7 to 8.7 h (refer Table 4.1), depending on the time of day, incoming flowrate of raw sewage to the pump station wet well and the associated frequency of pumping events.

It was not possible to determine the HRT of the samples taken at the 1,900 m location, based on the pump stations’ run data alone, since the intersection of the two rising mains nullified the assumption of ideal plug flow conditions. Accurate determination of the instantaneous HRT in this connected rising main system would require a tracer study, or more detailed network modelling than was possible under this study.

4.2.2 Sample Collection and Analysis Dissolved methane was sampled and analysed using the methodology described by Alberto et al.

(2000) and Guisasola et al. (2008). Sewage (containing dissolved methane) was sampled into freshly vacuumed BD Vacutainer® (BD Bioscience #367895) tubes using a hypodermic needle and 5 mL plastic syringe, attached directly to the pressurised rising main via a flexible hose (refer Figure 4.2). This procedure avoided any contact of the wastewater with atmosphere and possible oxygen interference. The Vacutainer tubes were weighed before and after sampling to determine the sample volume collected and mixed overnight in a shaker to allow equilibration of gas and liquid phases. Most of the methane (~ 97 % at 25 ºC) was transferred to the gas phase in this process (Alberto et al., 2000). The methane concentration in the gas phase of the tubes was measured using a Shimadzu GC-9A Gas Chromatograph equipped with a flame ionization detector (FID). The concentration of methane in the initial liquid phase was then calculated using mass balance and Henry’s law. The measurement error of this procedure is estimated to be ± 4.8% (t-disb., α = 0.05).

Samples were also collected for COD, soluble COD, and volatile fatty acids (VFAs). Where necessary, samples were immediately filtered using 0.22µm syringe filters, acid-preserved and kept on ice before analysis. Temperature, pH, redox and salinity were also recorded for all field samples.

VFAs were measured by gas chromatography using a Perkin Elmer Autosystem equipped with a polar capillary column DB-FFAP and a FID. COD was measured using the colorimetric method described in APHA (1995) using commercial Lovibond tubes in a range of 0 to 150 mgCOD.L-1.

Figure 4.2. Collection of dissolved methane sample directly from the rising main into an airtight syringe.



35

4.3 Results and Discussion The field measurements collected from rising main CO16 are shown in Table 4.1 and Figure 4.3.

By comparison, the saturation concentration for a 100% methane gas phase at 1 atm and 25°C is 21.5 mgCH4 L-1 (Perry et al., 1997). Also shown in Table 4.1 are field measurements previously reported for another Gold Coast rising main, UC09 (Guisasola et al., 2008). Figure 4.3 highlights the increase in dissolved methane concentration along the length of the rising main, with a concomitant decrease in pH. This is also reflected in the accumulation of VFAs and decrease in redox potential (refer Table 4.1), and indicates the onset of increasingly anaerobic conditions along the length of the rising main. This corresponds well with the theory and results presented in Guisasola et al (2008), which suggested that anaerobic methanogenesis increases with HRT.

Table 4.1. Field data collected from the CO16 and UC09 rising mains (Gold Coast, Australia). Data from UC09 rising main (150 mm diameter) previously reported by Guisasola et al (2008).

Location Time Temp.

(°C)

pH Redox(mV)

HRT(h)

VFAs(kg.m-3)

Dissolved CH4

(kg.m-3)

CO16 P/Station 05:05 22.9 7.00 -120 0.0 0.0444 0.00125

CO16 SP1 – 500m 05:23 22.9 6.83 -210 5.4 0.0570 0.00567

CO16 SP2 – 1,100m 05:35 22.5 6.69 -217 7.3 0.0549 0.00658

CO16 SP3 – 1,900m 05:49 24.2 6.70 -252 0.0566 0.00884

CO16 P/Station 06:06 23.1 6.94 -187 0.0 0.0389 0.00172

CO16 SP1 – 500m 06:18 24.1 6.77 -236 0.8 0.0431 0.00458

CO16 SP2 – 1,100m 06:26 24.1 6.66 -244 7.3 0.0571 0.00567

CO16 P/Station 07:03 22.9 7.17 -205 0.0 0.0487 0.00192

CO16 SP1 – 500m 07:16 23.8 6.81 -220 0.7 0.0576 0.00276

CO16 SP2 – 1,100m 07:25 24.2 6.66 -252 1.6 0.0513 0.00499

CO16 P/Station 08:04 22.9 7.36 -137 0.0 0.0333 0.00100

CO16 SP1 - 500m 08:16 24.3 7.37 -235 0.9 0.0545 0.00280

CO16 SP2 – 1,100m 08:23 23.0 6.77 -243 1.5 0.0524 0.00336

CO16 SP3 – 1,900m 08:35 23.4 6.65 -249 0.0464 0.00929

UC09 – 828m 14:06 27.6 7.13 - 3.8 - 0.00563

UC09 – 828m 14:30 27.4 7.06 - 4.0 - 0.00609

UC09 – 828m 15:01 27.7 7.09 - 4.2 - 0.00600

UC09 – 828m 15:30 27.5 7.31 - 4.4 - 0.00589

UC09 – 828m 16:00 27.1 7.25 - 4.6 - 0.00547

UC09 – 828m 16:30 26.4 7.17 - 4.6 - 0.00528



36

Location Time Temp.

(°C)

pH Redox(mV)

HRT(h)

VFAs(kg.m-3)

Dissolved CH4

(kg.m-3)

UC09 – 828m 17:00 26.5 7.15 - 4.6 - 0.00441

UC09 – 828m 17:30 26 7.10 - 3.9 - 0.00452

UC09 – 828m 18:00 25.7 7.15 - 3.1 - 0.00456

0.000

0.004

0.008

0.012

0.016

0.020

CO16 P/Station SP1 - 500m SP2 - 1100m SP3 - 1900m

Dis

solv

ed C

H4 C

once

ntra

tion

(kg.

m-3

)

5.4

5.8

6.2

6.6

7.0

7.4

pH

5:00AM

6:00AM

7:00AM

8:00AM

Figure 4.3. Dissolved methane concentrations (filled points) and pH (hollow points) measured at the four sampling points along the length of the CO16 rising main.

The results also illustrate the persistence of dissolved methane in fresh raw sewage, even when

the system is open to the atmosphere. The measurements taken at the CO16 pump station wet well all recorded concentrations of between 0.0010 – 0.0019 kgCH4.m-3 (i.e. 1.0 – 1.9 mgCH4.L-1), regardless of the time of sampling. Such results have not been previously reported and indicate that even fresh domestic sewage, such as that received directly into the wet well, has some small amount of methanogenic activity. Hence methane exchange was occurring continuously between the liquid phase and the gas phase. The sample collected at 5:00 a.m. was taken from sewage that had been accumulating in the wet well for up to 90 minutes (i.e. previous pumping event at 3:35 a.m.). Whereas the samples collected at 6:00 a.m., 7:00 a.m. and 8:00 a.m. were taken from sewage that had been accumulating in the wet well for only up to 11, 4 and 17 minutes respectively. Yet, the difference in dissolved methane concentration between these hourly samples is minimal. This indicates that there was continual methanogenic activity occurring upstream and within the wet well sediments and biofilms, and continual exchange of methane between the liquid and gas phase in the wet well. The dissolved methane concentration of 1.0 – 1.9 mgCH4.L-1 appears to represent a stable equilibrium between the rate of methane generation in the liquid phase and the rate of methane exchange to the gas phase. Quantifying the methane emissions from these open environments (e.g.



37

gravity sewers and collection wet wells) requires further characterisation of the liquid-to-gas phase mass transfer coefficient, and certainly warrants further investigation.

The CO16 rising main is a total of 4,880 m in length, and at its end discharges into the PA9 pump station, where it mixes with another gravity sewer. The results from Figure 4.3 suggest that dissolved methane would continue to accumulate in the rising main along the full length of the rising main. At the point of discharge into the pump station, the wastewater could be subject to a high degree of turbulence as the CO16 wastewater mixes with the joining gravity sewer flows, depending on the structure design and prevailing flow conditions. Under these energetic conditions, it is possible that the much of the accumulated dissolved methane is stripped from the liquid phase wastewater into the sewer headspace, from where it ultimately leaks to the atmosphere. Similar stripping processes are also possible at air relief valves along the rising main, at the inlet works of a treatment plant, or at other points of discharge and turbulence in the sewer system. However, it is also possible that a substantial fraction of the dissolved methane is consumed via aerobic methane oxidation or anaerobic methane-driven denitrification. These possibilities warrant further investigation under full-scale conditions.

4.4 Model Development

4.4.1 Theoretical Development As discussed in Section 4.2, the rising main can be assumed to approximate an ideal plug flow

reactor. At each pumping event, a new “slug” of wastewater is pushed into the pipeline and this moves along the pipeline with each subsequent pumping event. For the purposes of this model, ideal conditions were assumed, whereby negligible longitudinal dispersion of the plug occurs over time. As the slug moves along the pipeline, it is exposed to the methanogenic biofilm on the pipe wall. The length of exposure of each wastewater slug to the pipeline biofilm reactor depends upon the HRT of the slug, or the frequency of pumping events, which changes throughout the day. From first principles then, the mass of methane generated in the wastewater slug can be expressed as:

HRTARateM CH ××=4 (4.1)

where MCH4 = mass of dissolved methane (kg) Rate = specific rate of methanogenic activity of the biofilm (kg.m-2.h-1) A = surface area of the biofilm (m2) HRT = hydraulic retention time of the particular slug (h)

Dividing through by the pipe volume to determine the dissolved methane concentration yields:

HRTVARate

VM

C CHCH ××== 4

4 (4.2)

where CCH4 = concentration of dissolved methane (kg.m-3);

V = volume of pipe (m3) A/V = surface area to volume ratio of pipe (m-1)



38

The physical expression of Equation 4.2 is illustrated in Figure 4.4, where dissolved methane concentration (y-axis) is plotted against the lumped parameter [HRT × A/V] (x-axis). By Equation 4.2, the gradient of this curve describes the methanogenic activity rate of the biofilm in units of kg.m-2.h-1. This relationship is shown in Figure 4.4, with the measured data from the CO16 rising main in this study and the previously reported dissolved methane concentrations from the UC09 rising main. The two data points from the 1,900 m sample point on CO16 were excluded because of the difficulty in accurately calculating their HRT (refer to Section 4.2.1).

0.000

0.002

0.004

0.006

0.008

0.010

0.012

0 50 100 150 200A/V x HRT (h.m-1)

Dis

solv

ed C

H4 C

once

ntra

tion

(kg.

m-3

)

CO16 DataUC09 DataLinear RegressionGuisasola Model (2009)

Figure 4.4. Dissolved methane concentration (kg.m-3) vs (A/V × HRT) (h.m-1) for measured CO16 and UC09 rising main data.

4.4.2 Empirical Model Fitting In Figure 4.4, a simple linear regression model has been applied to the CO16 and UC09 field

data for the purposes of extrapolation and application to other rising main systems:

0015.0.1024.5 54 +⎥⎦

⎤⎢⎣⎡ ××= − HRTVACCH (4.3)

where

5.24 × 10-5 kg.m-2.h-1 equals the rate of methanogenic activity of the pipeline biofilm (i.e. 52.4 mg.m-2.h-1), empirically derived using a sum of least squares fitting algorithm in MS Excel 0.0015 kg.m-3 equals the average residual concentration of dissolved methane (i.e. 1.5 mg.L-1) measured in raw wastewater samples, at the CO16 pump station wet well



39

This model formulation provides a robust and easily accessible means for water authorities to estimate the methane emissions resulting from other pressurised rising mains, that have similar characteristics to the systems described in this study (i.e. temperature, organic matter composition). Only two easily determined independent variables are required. Firstly, rising main geometry (i.e. A/V ratio) is generally well-catalogued by water authorities in corporate databases. Secondly, the average hydraulic retention time of a particular rising main can be easily calculated from SCADA system data on flowrates and/or pump run times and nominal pumping capacities. Whilst it has been demonstrated in this study that rising main HRT has a marked diurnal profile, for the purposes of annual GHG inventory reporting, an average HRT is considered to be an acceptable approximation.

4.4.3 Calculation of Methane Emissions from a Rising Main As an example, consider a 300 mm diameter rising main that has an average flow of 7,000 m3.d-1

and HRT of 4.0 h, before discharging into vented headworks at a wastewater treatment plant. Applying Equation 4.3:

Surface area to volume ratio, 12

4

33.133.044

.... −==== m

mDLDLD

VA

π

π

Dissolved methane concentration (kg.m-3)

= 5.24 × 10-5 kg.m-2.h-1 × [13.33 m-1 × 4.0 h] + 0.0015 kg.m-3 = 0.0043 kg.m-3 Assuming all dissolved methane is stripped to atmosphere upon discharge from the rising main

into the vented headworks:

Methane emissions = 7,000 m3.d-1 × 0.0043 kg.m-3 × 100% stripping rate = 30.1 kgCH4.d-1

Multiplying up to a yearly inventory and converting to UNFCCC carbon dioxide equivalents (i.e. 21 kg CO2-e per kg CH4):

Methane emissions = 30.1 kg.d-1 × 365 d.y-1 × 21 kgCO2-e.kgCH4-1 × 10-3 t.kg-1

= 230 tCO2-e.y-1

As a comparison, the total GHG emissions from a typical biological nutrient removal wastewater treatment plant receiving 7,000 m3.d-1 would be in the order of 2,500 – 3,800 tCO2-e per year (De Haas et al., 2009). Therefore in this example, the methane emissions from the inlet rising main would account for an additional 6 – 9% of the annual GHG inventory for the plant. Larger rising main sewerage networks, with longer retention times, could be expected to contribute a much higher percentage to the annual GHG inventory for a water authority.

In the authors’ experience, rising main pipelines are typically in the range of 150 – 600 mm diameter, with nominal design HRTs of less than eight hours. Therefore, it is expected that the proposed model and the range of [A/V × HRT] values shown in Figure 4.4 (i.e. 0 – 200 h.m-1) will cover the majority of rising main configurations in practice.

4.4.4 Comparison with State-of-the-Art Model A recent publication by Guisasola et al. (2009) has extended upon their previous work, and

proposed a comprehensive model that characterises the biological and physicochemical processes in sewers. This extended model accounts for the competitive interactions of sulfate-reducing bacteria (SRB) and methanogenic archaea (MA) in sewers. The model development is based primarily upon lab-scale experimental data, and partly validated using the UC09 field data previously reported (Guisasola et al., 2008).



40

The dissolved methane concentrations from this complex model were overlaid with the CO16 and UC09 field data in Figure 4.4. Note that the “Guisasola Model” has been modified by increasing all values by 0.0015 kg.m-3 (i.e. 1.5 mg.L-1) to account for the influent dissolved methane concentration in the raw wastewater. The data points from the Guisasola Model were developed by modelling a large combination of theoretical A/V and HRT combinations. It is therefore interesting to note that there was very little scatter in the y-direction for these different A/V – HRT combinations. This suggests that the use of the lumped parameter [A/V × HRT] in the simplified model proposed by this study is a valid approximation for GHG inventory purposes.

In the region of [A/V × HRT] = 50 – 100 h.m-1, there is a very good fit between the CO16-UC09 field data, the proposed model and the complex Guisasola Model. However at lower values, the field data clearly suggests that both models underestimate the dissolved methane concentration. This is believed to be caused by competitive interaction between SRBs and MAs for early fermentation products. Further field investigations are required to determine the true kinetics that operate at these very short residence times.

Finally, the biofilm methanogenic activity rate (i.e. slope of the line) calculated for the proposed model was shown to be a good fit with the complex Guisasola Model. This suggests that for the rising main systems investigated in this study, the proposed model has valid predictive power beyond the limited range of HRT conditions tested in the field. This good fit between the measured field data, the comprehensive dynamic biological and physicochemical model and the simplified model proposed in this study, suggests that whilst an empirical approach has been useful in this instance, there is indeed some fundamental validity to the linear regression in Figure 4.4.

4.5 ConclusionsAt present, the potential generation of methane in wastewater collection systems is ignored under

the UNFCCC/IPCC greenhouse gas inventory guidelines, as it is “not believed to be a significant source of methane”. The field data presented in this paper suggest that for the operational rising mains investigated, this is a false assumption.

Dissolved methane concentrations were measured at four different longitudinal locations, and at four different times during a typical morning peak event, and were shown to be dependent upon the pipeline A/V ratio and the instantaneous HRT of the associated wastewater slug. This relationship is consistent with first principles reaction theory and with earlier data and findings by other authors (Guisasola et al., 2008; Guisasola et al., 2009).

Based on this theoretical relationship with two independent variables, an empirically derived model is proposed from the field data for the estimation of dissolved methane concentrations in other rising main sewerage systems, that have similar operational characteristics (i.e. temperature, organic matter composition). The empirically derived model of this study was shown to fit well with the more theoretically grounded biological and physicochemical sewer processes model, proposed by Guisasola et al. (2009).

This study addresses the widely recognised gap in the current international GHG accounting guidelines for methane emissions from pressurised rising main sewerage systems. For the particular systems investigated, it provides a robust, empirically fitted, theoretical model for water authorities to easily estimate the dissolved methane concentrations in their rising main networks, and hence the upper limit of methane emissions to the atmosphere. Future studies should address the equally important potential for methane generation and emission from gravity sewers with an open headspace.


5. Methane generation in full-scale wastewater treatment systems

41

5. Methane Generation in Full-Scale Wastewater Treatment Systems

5.1 IntroductionThe United Nations Framework Convention on Climate Change is the globally recognised basis

for collective action on the reduction of anthropogenic greenhouse gas emissions (UNFCCC, 2007). One of the key obligations for signatory countries under the UNFCCC is the compilation of an annual national greenhouse gas (GHG) inventory, covering four general sectors (energy; industrial processes; agriculture, forestry and other land use; and waste). Emissions of methane and nitrous oxide from wastewater treatment and discharge are reported under the waste sector (IPCC, 2006b). However, GHG emissions are not usually measured directly, but are rather estimated through the application of models that link emissions to data on observable activities.

The Australian Department of Climate Change (DCC, 2008b) and the Intergovernmental Panel on Climate Change (IPCC, 2006a) prescribe the following generalised approach for estimating methane emissions from wastewater treatment systems:

( )4

.4 CHwww RMCFEFCODkgEmissionsCH −××∆= (5.1)

Similarly, for methane emissions from sludge treatment systems:

( )4

.4 CHslslsl RMCFEFCODkgEmissionsCH −××∆= (5.2)

where

∆CODw, ∆CODsl = Mass of chemical oxygen demand (COD) consumed / removed over the wastewater and sludge treatment processes, respectively (kg), determined by a simplified COD mass balance

EF = Maximum methane production / emission factor (0.25 kg CH4 per kg COD)

MCFw, MCFsl = Methane correction factor for the type of process employed for wastewater treatment and sludge treatment, respectively (refer to Table 5.1)

RCH4 = Mass of methane captured for combustion and/or flaring on the plant, or transfer out of the plant (kg)

In essence, this approach is a reconciliation of the estimated mass of methane produced in the

treatment process (i.e. ∆COD × EF × MCF), with the measured mass of methane captured in the associated biogas system (i.e. RCH4 ). Any difference in these figures is assumed to be a loss of methane to the atmosphere. For treatment systems that are uncovered, RCH4 is zero. Notwithstanding the practical imprecision of measuring COD, biogas flowrates and biogas composition, it is clear that the accuracy of this estimation methodology is heavily dependent upon the factors, EF and MCF.

The theoretical maximum yield of methane from influent organic matter can be calculated by considering the simple conversion of a typical sugar (e.g. glucose) to carbon dioxide and methane, under anaerobic conditions (Crites and Tchobanoglous, 1998):

484

1322

1806126 33 CHCOOHC +→

(5.3) CH4 subsequently has an oxygen demand for complete conversion to carbon dioxide and water:


5. Methane generation in full-scale wastewater treatment systems (continued)

42

1082

1322

1922

484 6363 OHCOOCH +→+

(5.4) Therefore, the oxygen demand per kg of glucose is:

180192

6126

2 =OHkgC

kgO (5.5)

And the yield of methane per kg of glucose is:

18048

6126

4 =OHkgC

kgCH (5.6)

Therefore, the yield of methane per kg of oxygen demand is:

25.0

180/192180/48

2

4 ==kgO

kgCH

(5.7) In wastewater terms, oxygen demand is measured by COD. Therefore, the maximum theoretical

production of methane from the removal of COD is 0.25 kgCH4 per kgCOD removed. This factor is incontestable, since it is governed by the chemical stoichiometry of methane.

The methane correction factor however, is only an estimation of the likely conversion efficiency of COD to CH4 in a given type of process. Default values and ranges are published by the IPCC for a limited range of wastewater systems, as shown in Table 5.1. Based on the accompanying notes in the IPCC guidelines, these values are not based on any published studies, but rather “on expert judgment by lead authors”. Therefore, there clearly exists a strong need to address this acknowledged uncertainty in the prevailing GHG estimation methodologies.

Table 5.1. Default Methane Correction Factors for Domestic Wastewater (IPCC, 2006a; after Table 6.3)

Type of Treatment Comments MCF Range CH4 Production

(i.e. EF × MCF)

Centralised aerobic treatment plant

Well-managed Over-loaded

0.0 0.3

0.0 – 0.1 0.2 – 0.4

0.00 kgCH4.kgCOD-1

0.08 kgCH4.kgCOD-1

Anaerobic digester or reactor

Does not include CH4 recovery

0.8 0.8 – 1.0 0.20 kgCH4.kgCOD-1

Shallow anaerobic lagoon

Depth < 2m 0.2 0.0 – 0.3 0.05 kgCH4.kgCOD-1

Deep anaerobic lagoon

Depth > 2m 0.8 0.8 – 1.0 0.20 kgCH4.kgCOD-1

Foley and Lant (2008) recently reviewed the scientific literature on methane emissions from

various wastewater systems. They found that for closed anaerobic systems, such as covered high-rate anaerobic reactors, many international studies reported biogas CH4 production significantly less than the 0.20 kgCH4.kgCOD-1 suggested by the IPCC (Toprak, 1995; Kalogo and Verstraete, 1999; Paing et al., 2000). This discrepancy was mainly due to the failure of most studies to consider the loss of dissolved CH4 in the effluent. Depending on mass transfer kinetics, it is possible that generated CH4 can remain dissolved in the reactor liquid phase at supersaturated concentrations. In their pilot study of an Anaerobic Migrating Bed Reactor (AMBR), Hartley and Lant (2006) reported CH4 supersaturation ratios up to 2.2. These authors also surveyed similar studies from the literature, reporting supersaturation ratios ranging from 1.9 to 6.9, and associated losses of



43

dissolved CH4 in the effluent of up to 85%. The issue of methane supersaturation in anaerobic systems is not well understood and the assumption of thermodynamic equilibrium in COD mass balances appears to have been a common underlying mistake (Pauss et al., 1990; Hartley and Lant, 2006). Where dissolved methane losses were considered, several data sets from anaerobic digesters, covered anaerobic lagoons (CAL) and high-rate anaerobic reactors did support the IPCC’s suggested 0.80 MCF (Uemura and Harada, 2000; operational data from three Australian WWTPs presented in Foley and Lant, 2008).

For other basic facultative treatment systems, such as uncovered anaerobic lagoons, aerated lagoons and maturation lagoons, there was found to be no scientific literature or operational data to support the IPCC’s suggested methane correction factors (Foley and Lant, 2008). These systems are quite common in regional Australian towns and certainly warrant further investigation.

Therefore, the purpose of this study was to determine the methane generation and emission rates in full-scale treatment plants, of varying physical configuration and process conditions, to better inform the MCF guidance provided by the IPCC and DCC. This was undertaken using a rigorous mass balance approach, that specifically accounted for both gaseous and dissolved methane in the various wastewater treatment systems.


5.2.1 Field Sampling Sites For this study, four full-scale WWTPs were sampled. Their basic characteristics are listed in

Table 5.2. These sites were chosen to provide a selection of plant sizes, process configurations, and treatment applications. Three of the WWTPs treated largely domestic wastewater, whilst the fourth WWTP treated only brewery wastewater. This fourth site was selected to provide a comparison between low-strength and medium-strength anaerobic treatment applications. Also highlighted in Table 5.2 are the individual reactor zones sampled at each WWTP under this study.

5.2.2 Sample Collection and Analysis For each WWTP, it was intended to conduct four intensive sampling rounds (2 – 4 h duration,

morning and afternoon on two consecutive days). At the Bird-in-Hand WWTP, only three sampling rounds were completed, due to bad weather. This sampling program was completed over a four month timeframe in the Australian winter/spring of 2008.

For each sample round, sufficient data was collected to construct chemical oxygen demand (COD) and CH4 mass balances over the reactor zones of interest. For each reactor zone, field data was collected from its inlet and outlet, and consisted of a combination of 1) wastewater grab samples for COD, solids and volatile fatty acids (VFA); 2) measurement of process conditions, namely temperature, pH and oxidation-reduction potential (ORP), using a portable water quality meter (TPS 90FLMV); and 3) dissolved CH4 grab samples.

Dissolved methane was sampled and analysed using the methodology described by Alberto et al (2000) and Guisasola et al (2008). Sewage (containing dissolved methane) was sampled into freshly vacuumed BD Vacutainer® (BD Bioscience #367895) tubes using a hypodermic needle and 5 mL plastic syringe. The Vacutainer tubes were weighed before and after sampling to determine the sample volume collected and mixed overnight in a shaker to allow equilibration of gas and liquid phases. Most of the methane (~ 97 % at 25 ºC) was transferred to the gas phase in this process (Alberto et al., 2000). The methane concentration in the gas phase of the tubes was measured using a Shimadzu GC-9A Gas Chromatograph equipped with a flame ionization detector (FID). The concentration of methane in the initial liquid phase was then calculated using mass balance and Henry’s law. The measurement error of this procedure is estimated to be ± 4.8% (t-disb., α = 0.05).



44

Tab

le 5

.2.

Was

tew

ater

trea

tmen

t pla

nt si

tes.

No.

Abb

revi

ated

Nam

e L

ocat

ion

Flow

rate

(M

L.d

-1)

Proc

ess D

escr

iptio

n (B

old

item

s are

the

proc

ess u

nits

sam

pled

in th

is st

udy)

1 “Lag

oons

” B

ird-in

-Han

d W

WTP

, A

dela

ide

Hill

s, So

uth

Aus

tralia

1.

2 C

ombi

natio

n of

dom

estic

was

tew

ater

from

Lob

etha

l and

Woo

dsid

e to

wns

hips

, and

indu

stria

l was

tew

ater

from

the

Lobe

thal

aba

ttoir

– de

liver

ed v

ia ri

sing

mai

n.

2 ×

unco

vere

d an

aero

bic/

facu

ltativ

e la

goon

s (U

AFL

) (in

par

alle

l),ae

rate

d la

goon

(2 ×

mec

hani

cal s

urfa

ce a

erat

ors)

, 7 ×

mat

urat

ion

lago

ons.

2 “CA

L”

25W

Lag

oon

at W

este

rn

Trea

tmen

t Pla

nt (W

TP),

Mel

bour

ne, V

icto

ria

230

Com

bina

tion

of d

omes

tic a

nd in

dust

rial w

aste

wat

er fr

om w

este

rn

subu

rbs o

f Mel

bour

ne, d

eliv

ered

via

ope

n gr

avity

cha

nnel

. 65

0 M

L c

over

ed a

naer

obic

lago

on (C

AL

) with

bio

gas r

ecov

ery

and

ener

gy g

ener

atio

n, 7

13 M

L a

erat

ed la

goon

(50 ×

mec

hani

cal s

urfa

ce

aera

tors

) rec

eivi

ng a

ppro

x. 3

0 M

L.d-1

from

CA

L, a

ctiv

ated

slud

ge p

lant

re

ceiv

ing

appr

ox. 2

00 M

L.d-1

from

CA

L, 9

× m

atur

atio

n la

goon

s.

3 “PST

” N

orth

Hea

d W

WTP

, Sy

dney

, New

Sou

th W

ales

27

5 C

ombi

natio

n of

dom

estic

and

indu

stria

l was

tew

ater

from

Syd

ney

subu

rbs

(wes

t to

Seve

n H

ills,

sout

h to

Ban

ksto

wn

and

north

to K

u-rin

g-ga

i and

C

olla

roy)

, del

iver

ed v

ia g

ravi

ty se

wer

.

Inle

t wor

ks, 4

× p

rim

ary

sedi

men

tatio

n ta

nks (

appr

ox. 0

.5 h

HR

T),

slud

ge th

icke

ning

, ana

erob

ic d

iges

tion,

cen

trifu

ge m

echa

nica

l de

wat

erin

g.

4 “UA

SB”

Fost

ers B

rew

ery,

Y

atal

a, Q

ueen

slan

d 3.

3 B

rew

ery

was

tew

ater

onl

y.

Inle

t wor

ks, p

re-a

cidi

ficat

ion,

4 ×

0.66

ML

Upf

low

Ana

erob

ic S

ludg

e B

lank

et (U

ASB

) rea

ctor

s, 2 ×

J-C

ell I

nduc

ed A

ir F

lota

tion

(IA

F),

Mov

ing

Bed

Bio

Rea

ctor

(MB

BR

), D

isso

lved

Air

Flot

atio

n an

d Fi

ltrat

ion

(DA

FF),

mic

rofil

tratio

n, re

vers

e os

mos

is, s

ludg

e de

wat

erin

g.



45

The water quality meter was calibrated before and after the two-day sampling exercise at each WWTP. Grab samples for soluble species were immediately filtered using 0.22µm syringe filters, acid-preserved and kept on ice before analysis.

VFA were measured by gas chromatography using a Perkin Elmer Autosystem equipped with a polar capillary column DB-FFAP and a FID. COD was measured using the colorimetric method described in APHA (1995) using commercial Lovibond tubes in a range of 0 to 150 mg COD/L. Total suspended solids (TSS) and volatile suspended solids (VSS) were analysed according to Standard Methods (APHA, 1995).

Physical plant data (i.e. reactor dimensions, wastewater flowrates, biogas flowrates and composition) were provided by the WWTP operators for the specific days and times of field sampling. Where possible, the WWTP operators also supplied their own routine process and analytical data that provided a useful cross-check against results from the grab samples collected during the field study.

5.2.3 COD and Methane Mass Balances over Entire WWTP Processes At each of the WWTP sites, a steady-state total COD mass balance was constructed across

the entire process. This mass balance drew upon the analytical data collected in the field (e.g. COD and dissolved CH4 concentrations), as well as the plant data (i.e. flowrates, reactor volumes, solids capture efficiencies, biogas production and composition) supplied by the WWTP operators. Note that the Standard Methods COD test does not capture dissolved CH4 (Hartley and Lant, 2006), so it was necessary to specifically include dissolved CH4 in the calculations for an accurate mass balance characterisation. Methane has a COD value of 4 kgCOD.kgCH4

-1. The purpose of this initial mass balance analysis was to ensure an accurate characterisation of the

WWTP operation, such that the COD balance over each WWTP generally achieved greater than 90% closure.

5.2.4 Methane Mass Balances over Individual Reactor Zones The second phase of mass balance analysis examined methane across the liquid phase of the

individual reactors at each WWTP. The general formulation of the mass balance construction is given in Equation 5.5:

RCHRCHOutCHInCHRCH GTrMM

dtdM

,,,,,

4444

4 +−∑−∑= (5.5)

where

dtdM RCH ,4 = change in mass of CH4 in the reactor zone, over time (kg.d-1)

InCHM ,4∑ = sum of i mass flows of CH4 into the reactor zone (kg.d-1)

OutCHM ,4∑ = sum of j mass flows of CH4 out of the reactor zone (kg.d-1)

RCHTr ,4 = mass transfer of CH4 from the reactor liquid to gas phase (kg.d-1)

RCHG ,4 = net generation of CH4 in the reactor zone (kg.d-1) (i.e. net result of CH4

production and consumption due to biological reactions in the reactor) It was assumed that for the brief 2 – 4 h window of analysis, the reactor zones operate at

near steady-state conditions, and are well-mixed. Equation 5.5 can then be expanded and re-formulated to solve for GCH4,R:



46

[ ] [ ] [ ] [ ]( )*44,4,4,,4 SRLRiIn

iiInR

jjOutRCH CHCHakVCHQCHQG −××+×−×= ∑∑ (5.6)

where QIn,i , QOut,j = individual flows in and out of the reactor zone (ML.d-1)

[CH4]In,i = concentration of dissolved CH4 in the incoming streams (mg.L-1 or kg.ML-1), which is generally equal to the dissolved CH4 concentration in the originating reactor

[CH4]R = concentration of dissolved CH4 in the reactor zone (mg.L-1 or kg.ML-1)VR = volume of the reactor zone (ML) kLa = volumetric mass transfer coefficient (d-1) (refer to Table 5.3)[CH4]s

* = saturation concentration of CH4 in water at atmospheric conditions

= 4.19 × 10-5 kg.ML-1 at 20°C (Perry et al., 1997)

For the CAL at WWTP No.2 and UASB at WWTP No.4, it was not necessary to calculate the mass transfer, RCHTr ,4

using kLa, since the biogas flowrate and composition were measured directly by the plant operators. For the other reactors, the kLa values used in Equation 5.6 were estimated using the empirical correlation technique of Van’t Riet (1979). This technique relates kLa to volumetric power input (P/V) and superficial gas velocity, vg.(refer to Table 5.3). The lab-scale methane kLa values determined in Section 2.3 were not suitable for these calculations, since those experiments were based on a diffused aeration system, whereas the field situations investigated in this study used mechanical surface aeration.

The mass calculations of Equations 5.5 and 5.6 was completed for the relevant process units at each WWTP, and then repeated for best-case and worst-case combinations of the 95% confidence intervals of the dissolved CH4 concentrations and mass transfer terms. These calculations determined the uncertainty range of CH4 emissions to atmosphere and net CH4 generation (i.e. production minus consumption) in each reactor zone. Negative values of GCH4,R indicate CH4consumption is greater than CH4 production in that particular reactor.

To compare results across sites and to relate back to the IPCC methane estimation model, GCH4,Rfor each reactor was normalised by dividing by the corresponding total mass of COD removed:

RCOD

RCHR M

GGF

,

,4= (5.7)

whereGFR = CH4 generation factor for each reactor zone (kgCH4.kgCOD-1

removed)MCOD,R = Steady-state mass of COD (kg) removed over the reactor zone

The GFR for each process unit captures both the methane lost by mass transfer to the gas phase, and the dissolved methane lost in the effluent. Finally, the GFR was divided by the IPCC maximum methane emission factor, EF (i.e. 0.25 kgCH4.kgCOD-1

removed) to determine the MCF for each reactor type.



47

Table 5.3. Volumetric mass transfer coefficients adopted for mass balance calculations.

WWTP No. Process Unit kLa Range (d-1)

Comments

1 – Lagoons N/A N/A Mass balance could not be completed.

2 – CAL Covered anaerobic lagoon

N/A Biogas flowrate directly measured.

Aerated lagoon 3.8 – 5.5 Based on Van’t Riet (1979) correlation 1,with 1.40 – 1.54 W.m-3 aerator power input. Superficial gas velocity calculated assuming 0.7 – 1.2 kgO2.kWh-1 (Tchobanoglous et al.,2003).

3 – PST Primary sedimentation tank

2.2 – 4.8 Based on Van’t Riet (1979) correlation, assuming minimal mixing power input (i.e. 0.1 W.m-3) and superficial gas velocity9.4 × 10-6 – 5.4 × 10-5 m3.m-2.s-1 (based on PST measurements at Glenelg WWTP, Adelaide – data not shown).

4 - UASB UASB N/A Biogas flowrate directly measured.

Notes: 1. For further details of the Van’t Riet (1979) kLa correlation, refer to Section 2.4.

5.3 ResultsShown in Figure 5.1 are the dissolved CH4 concentrations (and associated measurement errors)

measured in the various process units at each WWTP. At WWTP No.1, the raw wastewater is received via a rising main. The elevated dissolved CH4 concentration measured at the inlet to this plant (3.2 – 7.2 mgCH4.L-1) is consistent with the data reported in Section 4 for rising mains on the Gold Coast. However, the hydraulic retention time, flowrate and diameter of the rising main at WWTP No.2 are unknown. Therefore, it is not possible to compare these results against the methane generation model proposed in Section 4.

At WWTP Nos. 2 and 3, the raw wastewater is received via gravity sewers. At both plants, the measured dissolved CH4 was in the range 0.5 – 1.5 mgCH4.L-1. Again, this is consistent with the raw wastewater dissolved CH4 concentration measured in the pump station wet well of the Gold Coast rising main in Section 4. This supports the hypothesis that measurable methane generation and exchange occurs in the raw wastewater and biofilms of gravity sewers.

The raw data reported in Figure 5.1 form the basis of the methane mass balance calculations described in Section 5.2.4. Unfortunately, the data collected from the uncovered lagoon system at WWTP No.1 were not sufficiently accurate or reliable to be able to achieve a suitable COD mass balance closure. WWTP No.1 is a small, unattended treatment plant that intermittently receives raw wastewater of highly variable strength from two townships and an abattoir. Minimal routine process data are collected, and no plant design information (e.g. lagoon volumes, depths, aerator sizes) was available. Therefore, it was not possible to complete a representative mass balance analysis at this particular plant.



48

1 - Lagoons

0.0

2.0

4.0

6.0

8.0

Rd1 Rd2 Rd3

Dis

solv

ed C

H4 C

once

ntra

tion

(mg.

L-1)

Raw WWAn. LagoonAe. LagoonMat. Lagoon

2 - CAL

0.0

2.0

4.0

6.0

8.0

10.0

12.0

Rd1 Rd2 Rd3 Rd4D

isso

lved

CH

4 Con

cent

ratio

n (m

g.L-1

)

Raw WWCAL (1)CAL (2)Ae. Lagoon (1)Ae. Lagoon (2)ASP Ax. Zone

3 - PST

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

16.0

Rd1 Rd2 Rd3 Rd4

Dis

solv

ed C

H4 C

once

ntra

tion

(mg.

L-1) Raw WW

PST Eff.Raw SludgeCentrateDigested Sludge

4 - UASB

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

Rd1 Rd2 Rd3 Rd4

Dis

solv

ed C

H4 C

once

ntra

tion

(mg.

L-1)

PAT UASB IAF

Figure 5.1. Dissolved methane concentrations (mg.L-1) measured at each WWTP. “Raw WW” – raw wastewater, “An. Lagoon” – anaerobic lagoon, “Ae. Lagoon” – aerated lagoon, “Mat. Lagoon” – maturation lagoon, “CAL” – covered anaerobic lagoon, “ASP Ax. Zone” – anoxic zone of activated sludge plant, “PST Eff.” – effluent from primary sedimentation tanks, “PAT” – outlet from pre-acidification tanks, “UASB” – outlet from upflow anaerobic sludge blanket reactor, “IAF” – outlet from induced air flotation cell.

The outcomes of the mass balances analyses at the other three WWTPs are shown in Figure 5.2.

This figure illustrates the absolute magnitude of methane sources and fates at each WWTP. The error bars indicate the range of likely values based on the underlying uncertainty in the dissolved CH4 measurements (refer to Figure 5.1) and the estimated volumetric mass transfer coefficients (refer to Table 5.3).



49

2 - CAL

0

800

1,600

2,400

3,200

CH4 in Raw

WW

CH4 to A

SP

CH4 to A

e. La

goon

CH4 in B

iogas

Met

hane

(kg.

d-1)

0

8,000

16,000

24,000

32,000

Met

hane

in B

ioga

s (k

g.d-1

)

2 - Aerated Lagoon

0

100

200

300

400

500

600

CH4 from

An.La

g.

CH4 from

ASP W

AS

CH4 in E

ffluen

t

CH4 in O

ff-gas

CH4 Gen

erated

Met

hane

(kg.

d-1)

0

1,000

2,000

3,000

4,000

5,000

6,000

Met

hane

Gen

erat

ed a

nd in

Off-

gas

(kg.

d-1)

3 - PST

0

100

200

300

400

500

CH4 in R

aw W

W

CH4 in O

utfall

CH4 inSlud

ge

CH4 to A

tm. F

rom P

STs

CH4 Lost

/ Con

sumed

Met

hane

(kg.

d-1)

4 - UASB

0

2

4

6

8

10

12

14

CH4 from P

AT

CH4 in U

ASBEff.

CH4 in IA

F Eff.

CH4 Los

t / Con

sumed

CH4 in Biog

as

Met

hane

(kg.

d-1)

0

200

400

600

800

1,000

1,200

1,400

Met

hane

in B

ioga

s (k

g.d-1

)

Figure 5.2. Sources and fates of methane from mass balance analyses. Light shaded columns are measured by the secondary y-axis.

5.4 Discussion

5.4.1 Dissolved Methane Concentration Profiles At WWTP No.1, the concentration of dissolved CH4 in the effluent from the anaerobic/facultative

lagoon was 3.5 ± 0.2 mg.L-1. After passing through the mechanically surface aerated lagoon, this concentration dropped to 0.3 ± 0.2 mg.L-1. This suggests that a substantial amount of CH4 consumption and/or emission to the atmosphere occurred in this lagoon. However, in the downstream maturation lagoon, the dissolved CH4 concentration slightly rose again to 0.8 ± 0.4 mg.L-1. Despite being quite shallow (< 2 m depth), it appears that there was some methanogenic activity occurring in the water column and/or sediments of this lagoon. It was not possible to further characterise the methane transformations occurring at WWTP No.1 because the COD mass balance failed to close with a sufficient degree of accuracy.



50

At WWTP No.2, the concentration of dissolved CH4 in the effluent immediately exiting from the CAL was 9.9 ± 2.3 mg.L-1. Approximately 10 m downstream, the measured concentration dropped to 5.3 ± 1.8 mg.L-1, without the intervention of any mechanical mixing or surface aeration. This suggests that a substantial fraction of dissolved methane in the CAL effluent quickly volatilised under natural mass transfer driving forces. Approximately 500 m further downstream in the mechanically surface aerated lagoon, the measured dissolved CH4 concentration dropped to 0.8 ± 0.4 mg.L-1. At the end of the 1,000 m long lagoon, the dissolved CH4 concentration was 0.7 ± 0.4 mg.L-1. The turbulence and aeration provided by the mechanical surface aerators appears to have provided further energy for stripping the dissolved CH4 to atmosphere. However, the persistence of dissolved CH4, even at the end of a 1,000 m long surface aerated lagoon suggests that some level of methanogenic activity continued to occur in the water column and/or sediments of this lagoon. Anecdotally, this lagoon appeared to have very high solids accumulation, with “sludge islands” breaking through the water surface at some points.

At WWTP No.3, the concentration of dissolved CH4 in the raw wastewater was approximately halved to 0.6 ± 0.1 mg.L-1 in the effluent to the ocean outfall. This could have been the result of stripping to atmosphere at the inlet works and/or the PST air-water surface. It is also possible that methanotrophic consumption of the dissolved methane occurred in the aerobic surface layer of the PSTs. Relatively high concentrations of dissolved methane were measured in the various sludge liquor streams, but at very low flowrates, these did not contribute greatly to the overall mass balance profile.

At WWTP No.4, the concentration of dissolved CH4 in the UASB effluent was 10.5 ± 0.8 mg.L-1. After passing through the IAF, the dissolved CH4 concentration dropped to 3.1 ± 0.3 mg.L-1. Not unexpectedly, the turbulence and aeration provided in the IAF stripped a substantial portion of the dissolved CH4 to atmosphere. The residual CH4 in the IAF effluent is more likely due to the relatively short residence time of the IAF, rather than any significant methanogenic activity within the vessel.

5.4.2 Methane Fate Profiles From Figure 5.2, it can be seen that the majority of methane generated in the CAL at WWTP

No.2 was recovered in the biogas. Only 7 – 11% was lost as dissolved methane in the CAL effluent to the ASP and the downstream aerated lagoon. However, for a large plant such a WWTP No.2, this still represents a potentially large source of GHG emissions (13,400 – 21,600 tCO2-e per year).

Based on the mass balance analysis of the CAL, it is calculated that 0.22 – 0.23 kg CH4 is generated per kg COD removed. In the context of the IPCC GHG accounting guidelines, this represents a MCF value of 0.87 – 0.93 (refer to Table 5.4). This is higher than the recommended default MCF of 0.8 (IPCC, 2006a).

The mass balance analysis of the aerated lagoon at WWTP No.2 shows that there is significant methane generation and emission to atmosphere. Perhaps due to significant sludge accumulation, it appears that this lagoon operates in a partly facultative regime. The estimated MCF for this lagoon is 0.03 – 0.20 (refer to Table 5.4), which is similar to the “shallow anaerobic lagoon” MCF of the IPCC guidelines (IPCC, 2006a). However, there is a significant degree of uncertainty in the results for this lagoon (refer to Figure 5.2) because of the difficulty in estimating a mass transfer coefficient for such a large and heterogeneously aerated volume.



51

Table 5.4. Methane Correction Factors determined from field results.

Type of Treatment Comments MCF Range / Comments Covered anaerobic lagoon

8 m depth 0.90 0.87 – 0.93

Aerated lagoon Approx. 3 m depth Approx. 1.5 W/m3 aeration power input

0.10 0.03 – 0.20

UASB 5.5 m depth Medium-strength brewery wastewater

1.00 0.80 – 1.00

PST 3.2 m depth 0.5 h HRT

0.00 Possible sink, if CH4 in raw wastewater is considered in the mass balance

The mass balance analysis on the PST at WWTP No.3 indicates that there was definitely no

generation of CH4. In fact, it is possible that the PSTs act as a sink for the dissolved CH4 in the raw wastewater. Exposure of the wastewater to a large air-water interface, such as the PSTs, may allow methanotrophic activity to occur. Methanotrophic bacteria typically occur in aerobic surface environments, close to a source of methane, such as in a landfill capping layer or a wetland. Under these micro-aerated conditions, methanotrophic bacteria utilise methane as their sole source of carbon and energy, to produce organic products and ultimately carbon dioxide (Thalasso et al., 1997; Rajapakse and Scutt, 1999; Eisentraeger et al., 2001; Waki et al., 2005). Nozhevniikova et al (2003) cite reports of between 10 – 80% of methane produced in landfills and wetlands being oxidised by methanotrophic bacteria in the aerobic surface layer. In summer periods, this may increase to 100% consumption. It is also possible that the dissolved methane was simply stripped to atmosphere in the turbulence of the inlet works. Unfortunately, there is insufficient data from this study to better characterise the relative fate of dissolved methane in raw wastewater as it passed through the turbulent inlet works and the relatively quiescent PSTs. However, the potential for PSTs to act as methane sinks warrants further detailed investigation.

In the UASB of WWTP No.4, only 0.8 – 1.2% of the generated CH4 was lost in the effluent. Consequently, the MCF for this type of strictly anaerobic, enclosed, high-strength reactor is close to 1.00 (refer to Table 4.4). This is higher than the default 0.8 MCF recommended under the IPCC guidelines (IPCC, 2006a).

5.5 ConclusionsInternational guidance on CH4 emissions from anaerobic and facultative wastewater treatment

systems is presently based on very limited data from full-scale operational facilities. In fact, characterisation of methane emissions from open lagoon-based systems appears to have been almost completely ignored to date. This study determined the methane generation and emission rates in four full-scale treatment plants, including covered and uncovered lagoons, to better inform the GHG accounting guidance provided by the IPCC. This was undertaken using a rigorous mass balance approach, that specifically accounted for both gaseous and dissolved methane. It should be noted however, that this study is not comprehensive. No comparison of results between plants of similar process configuration was undertaken, due to the limited number of disparate sites sampled. Therefore, the outcomes of this work are less insightful than those presented earlier for nitrous oxide emissions from biological nutrient removal WWTPs (refer to Section 3). Nonetheless, some findings of real interest were made.



52

Dissolved CH4 concentrations of 0.5 – 1.5 mg.L-1 were measured in the raw wastewater delivered by gravity sewer to two municipal WWTPs. This result supported the earlier findings from the Gold Coast rising main (refer to Section 4) that methane is ubiquitous in raw wastewater.

At each WWTP surveyed, a distinct and consistent profile of dissolved methane concentration was determined. In the case of the lagoon-based WWTPs, it was clear that methane generation and exchange continued to occur, even in shallow and/or mechanically aerated zones. This suggests that such treatment lagoons operate, at least partly, in a facultative regime with methanogenic activity occurring in the water column and/or sediments. Whilst no supersaturated concentrations of dissolved methane were found in this study, the loss of dissolved methane in low-strength, high volumetric throughput, anaerobic treatment systems, can still be substantial.

The mass balance analyses conducted in this study confirmed the general range of MCF values published by the IPCC for covered anaerobic lagoons and high-rate anaerobic reactors. Higher default MCF values of 0.90 and 1.00 are recommended for CALs and UASB reactors, respectively. The mechanically aerated lagoon surveyed in this study also fell within the expected range of MCF values. This lagoon appeared to suffer from very high solids accumulation, which may not make it a representative indication of this particular treatment process. The results from the PST surveyed in this study suggested that it may act as a methane sink for the dissolved methane in the raw wastewater.

In this study, the field work and analysis required to characterise CH4 generation from simple, open systems (e.g. lagoons) was more difficult than similar work undertaken for the characterisation of nitrous oxide emissions from biological nutrient removal WWTPs. Lagoon treatment systems, by their nature, are not well-mixed, can have substantial solids accumulation and are heavily influenced at the surface by prevailing wind conditions. Therefore, they are very difficult to properly characterise and do not yield easily to the mass balance approach adopted in this study. Future work in this field should concentrate upon gaining a better understanding of open treatment systems, such as facultative lagoons. The biological interactions between the lagoon sediments, water column and air-water interface need to be better characterised from a CH4 production/consumption perspective. The possibility of methanotrophic activity in the aerobic surface layers of lagoons and quiescent tanks also warrants closer investigation, as these could be substantial sinks for dissolved methane in raw wastewater. The dynamic influent and process conditions of an operating WWTP probably demand comprehensive on-line monitoring for a thorough characterisation. The mass balance approach adopted in this study is recommended as a robust framework for the subsequent data analysis.


6. Implications for the National Greenhouse and Energy Reporting System

53

6. Implications for The National Greenhouse and Energy Reporting System The National Greenhouse and Energy Reporting Act (NGER Act) 2007 was passed by the

Australian Government on 29 September 2007. It was designed to provide a single, streamlined system for the mandatory reporting of corporate greenhouse gas (GHG) emissions, energy production and energy consumption. The subordinate items of legislation supporting the Act are:

The National Greenhouse and Energy Reporting Regulations (NGER Regulations) 2008; The National Greenhouse and Energy Reporting (Measurement) Determination 2008; and External Audit Legislative Instrument.

In addition to this legislative framework, the Department of Climate Change (DCC) has published

the National Greenhouse and Energy Reporting Guidelines (DCC, 2008c), which have been developed to help corporations understand their obligations under the NGER Act and the NGERRegulations. Incorporated into the NGER Guidelines are the GHG and energy reporting thresholds for the scheme (Figure 6.1). Any corporation or facility which exceeds these thresholds is required to report to the DCC on an annual basis. Based on this framework, WSAA expects that most of its members will, for the first time, be required to report their GHG emissions and energy consumption/production, at both corporation and facility levels.

Figure 6.1. NGERS reporting thresholds for facilities and corporations (DCC, 2008c).

To assist stakeholders with this calculation and reporting task, the DCC has also published the National Greenhouse and Energy Reporting (Measurement) Technical Guidelines (DCC, 2008b). This document is intended to assist stakeholders understand and apply the NGER (Measurement) Determination 2008. The Technical Guidelines outline the calculation methods and criteria for determining GHG emissions, energy production and consumption at a facility level. The methods are generally based on those used for the National Greenhouse Accounts (DCC, 2008a).


6. Implications for the National Greenhouse and Energy Reporting System (continued)

54

Since the enactment of NGERS in late 2007, WSAA has been regularly consulting with the DCC’s Inventory Team to inform and influence the development and application of the proposed GHG calculation methodologies for the “wastewater handling (domestic and commercial)” sector. In May 2009, the DCC invited comments on draft amendments to the NGER (Measurement) Determination 2008. In the associated Department commentary, the following was noted:

“New developments in estimation technologies and methods both in Australia and internationally mean that it is likely that certain aspects of the Determination will be amended again next year. In consultation with the relevant industries, future amendments to the Determination will focus, among others, on review and refinement to the methods for coal mining, carbon capture and storage and wastewater treatment.”

The DCC is aware of this current study and the progress of other similar studies being conducted

internationally, under the auspices of the Global Water Research Coalition (GWRC). It is understood by WSAA and UQ that the results of these studies could be considered by the DCC in late 2009 for a further revision to the NGER (Measurement) Determination 2008, and associated Technical Guidelines, in 2010.


7. Conclusions and recommendations

The current international guidelines for estimating greenhouse gas (GHG) emissions focus mainly on emissions associated with energy use. Quantifying direct emissions of methane (CH4) and nitrous oxide (N2O) from wastewater systems is an area of uncertainty, with less developed and less reliable methodologies. Recent changes to the Federal regulatory landscape, with the introduction of the National Greenhouse and Energy Reporting System (NGERS) and possibly an emissions trading scheme, combined with voluntary organisational commitments to “carbon neutrality”, mean that this level of uncertainty now represents an emerging business risk for the water industry.This study has provided the first set of comprehensive full-scale data on direct CH4 and N2O emissions from a range of wastewater systems in Australia. In combination with the earlier WSAA-UQ literature review, and in light of other international work in this field, the results from this study could be used to improve the level of certainty in the estimation methodologies published by the Department of Climate Change (DCC) in the NGERS Technical Guidelines. With regard to future work in this field, the following lines of inquiry are recommended for further investigation:1. A greater understanding of the fundamental kinetics of competing nitrite and nitrous oxide

transformation mechanisms in full-scale BNR applications is required. The dynamic influent and process conditions of an operating treatment plant demand comprehensive on-line monitoring in all bioreactor compartments for a thorough characterisation.

2. Methane generation in gravity sewers has not been quantitatively addressed in the scientific literature. This warrants further investigation because the results of this study suggest that CH4 is ubiquitous in wastewater environments.

3. The biological and physicochemical interactions between lagoon sediments, the water column and the air-water interface in open systems need to be better characterised. The possibility of methanotrophic activity in the aerobic surface layers of lagoons and quiescent tanks warrants closer investigation, as these could be substantial sinks for dissolved CH4 in raw wastewater.

4. The mass balance approach adopted in this study is recommended as a robust framework for the recommended lines of future investigation.

5. The emissions factors of nitrous oxide, and to a lesser extent methane, from discharge of treated sewage to environmental waters such as rivers, estuaries and seas/oceans remain inaccurate and poorly defined. This area of research will require cross disciplinary expertise for such a complex environment.


Appendix A - References

56

APPENIDX A: REFERENCES Abdul-Talib, S., Ujang, Z., Vollertsen, J., and Hvitved-Jacobsen, T., 2005. Model concept for

nitrate and nitrite utilization during anoxic transformation in the bulk water phase of municipal wastewater under sewer conditions. Water Science and Technology 52 (3), 181-189.

Alberto, M., Arah, J. R. M., Neue, H., Wassmann, R., Lantin, R., Aduna, J., and Bonson, K., 2000. A sampling technique for the determination of dissolved methane in soil solution. Chemosphere: Global Change Science 2 (1), 57-63.

Andersen, K., Kjaer, T., and Revsbech, N. P., 2001. An oxygen insensitive microsensor for nitrous oxide. Sensors and Actuators B: Chemical 81 (1), 42-48.

APHA, 1995. Standard Methods for the Examination of Water and Wastewater. American Public Health Association, Washington DC.

Beline, F., Martinez, J., Marol, C., and Guiraud, G., 2001. Application of the 15N technique to determine the contributions of nitrification and denitrification to the flux of nitrous oxide from aerated pig slurry. Water Research 35 (11), 2774-2778.

Bock, E., Schmidt, I., Stuven, R., and Zart, D., 1995. Nitrogen loss caused by denitrifying Nitrosomonas cells using ammonium or hydrogen as electron donors and nitrite as electron acceptor. Archives of Microbiology 163 16-20.

Butler, M. D., Stephenson, T., Stokes, L., and Stuetz, R. M., 2005. Dinitrogen oxide detection for nitrification failure early warning systems. Water Science and Technology 52 (8), 249-256.

Capela, S., Roustan, M., and Heduit, A., 2001. Transfer number in fine bubble diffused aeration systems. Water Science and Technology 43 (11), 145-152.

Casey, T. G., Wentzel, M. C., and Ekama, G. A., 1999a. Filamentous organism bulking in nutrient removal activated sludge systems - Paper 10: Metabolic behaviour of heterotrophic facultative aerobic organisms under aerated/unaerated conditions. Water Sa 25 (4), 425-442.

Casey, T. G., Wentzel, M. C., and Ekama, G. A., 1999b. Filamentous organism bulking in nutrient removal activated sludge systems Paper 9: Review of biochemistry of heterotrophic respiratory metabolism. Water Sa 25 (4), 409-424.

Chain, P., Lamerdin, J., Larimer, F., Regala, W., Lao, V., Land, M., Hauser, L., Hooper, A., Klotz, M., Norton, J., Sayavedra-Soto, L., Arciero, D., Hommes, N., Whittaker, M., and Arp, D., 2003. Complete genome sequence of the ammonia-oxidizing bacterium and obligate chemolithoautotroph Nitrosomonas europaea. Journal of Bacteriology 185 (9), 2759-2773.

Crites, R., and Tchobanoglous, G., 1998. Small and Decentralized Wastewater Management Systems. WCB McGraw-Hill, Boston.

Czepiel, P., Crill, P., and Harriss, R., 1995. Nitrous oxide emissions from municipal wastewater treatment. Environmental Science and Technology 29 (9), 2352-2356.

DCC, 2008a. National Greenhouse Accounts (NGA) Factors. Department of Climate Change, Commonwealth of Australia, Canberra.

DCC, 2008b. National Greenhouse and Energy Reporting (Measurement) Technical Guidelines, 1.1 edition. Department of Climate Change, Commonwealth of Australia, Canberra.

DCC, 2008c. National Greenhouse and Energy Reporting Guidelines. Department of Climate Change, Commonwealth of Australia, Canberra.

De Haas, D., Foley, J., and Lant, P., 2009. Energy and greenhouse footprints of wastewater treatment plants in south-east Queensland. in OzWater '09. Australian Water Association, Melbourne.

Dudley, J., 1995. Mass transfer in bubble columns: A comparison of correlations. Water Research 29 (4), 1129-1138.

Eisentraeger, A., Klag, P., Vansbotter, B., Heymann, E., and Dott, W., 2001. Denitrification of groundwater with methane as sole hydrogen donor. Water Research 35 (9), 2261-2267.

Envirosim, 2007. BioWin Process Simulator, v.3.0, Envirosim Associates Ltd, Flamborough, Ontario.

Ferrell, R. T., and Himmelblau, D. M., 2002. Diffusion coefficients of nitrogen and oxygen in water. Journal of Chemical and Engineering Data 12 (1), 111-115.


Appendix A - References (continued)

57

Foley, J., and Lant, P., 2008. Fugitive Greenhouse Gas Emissions from Wastewater Systems: WSAA Literature Review No.01. Water Services Association of Australia, Melbourne.

Foley, J., Lant, P., and Donlon, P., 2008. Fugitive greenhouse gas emissions from wastewater systems. Water Journal of the Australian Water Association 38 (2), 18-23.

Garcia-Ochoa, F., and Gomez, E., 2009. Bioreactor scale-up and oxygen transfer rate in microbial processes: An overview. Biotechnology Advances 27 (2), 153-176.

Gejlsbjerg, B., Frette, L., and Westermann, P., 1998. Dynamics of N2O production from activated sludge. Water Research 32 (7), 2113-2121.

Gillot, S., Capela-Marsal, S., Roustan, M., and Heduit, A., 2005. Predicting oxygen transfer of fine bubble diffused aeration systems-model issued from dimensional analysis. Water Research 39 (7), 1379-1387.

Guisasola, A., de Haas, D., Keller, J., and Yuan, Z., 2008. Methane Formation in Sewer Systems. Water Research 42 (6-7), 1421-1430.

Guisasola, A., Sharma, K. R., Keller, J., and Yuan, Z., 2009. Development of a model for assessing methane formation in rising main sewers. Water Research 43 (11), 2874-2884.

Hartley, K., and Lant, P., 2006. Eliminating non-renewable CO2 emissions from sewage treatment: An anaerobic migrating bed reactor pilot plant study. Biotechnology and Bioengineering 95(3), 384-398.

Hasanen, A., Orivuori, P., and Aittamaa, J., 2006. Measurements of local bubble size distributions from various flexible membrane diffusers. Chemical Engineering and Processing 45 (4), 291-302.

Henze, M., Gujer, W., Mino, T., and van Loosdrecht, M. C. M., 2000. Activated Sludge Models ASM1, ASM2, ASM2d and ASM 3. International Water Association, London.

Huisman, J. L., Krebs, P., and Gujer, W., 2003. Integral and unified model for the sewer and wastewater treatment plant focusing on transformations. Water Science and Technology 47(12), 65-71.

Hvitved-Jacobsen, T., Vollertsen, J., and Tanaka, N., 2000. An integrated aerobic/anaerobic approach for prediction of sulfide formation in sewers. Water Science and Technology 41(6), 107-116.

Hynes, R. K., and Knowles, R., 1984. Production of nitrous oxide by Nitrosomonas europaea: effects of acetylene, pH and oxygen. Canadian Journal of Microbiology 30 1397-1404.

IPCC, 1997. Module 6 - Waste. in Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories, vol.2 - Workbook. Intergovernmental Panel on Climate Change.

IPCC, 2006a. Ch.6 - Wastewater Treatment and Discharge. in H. S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe, editors. 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Prepared by the National Greenhouse Gas Inventories Programme, vol.5 - Waste. IGES, Japan.

IPCC, 2006b. Vol.1 - General Guidance and Reporting. in H. S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe, editors. 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Prepared by the National Greenhouse Gas Inventories Programme, IGES, Japan.

Itokawa, H., Hanaki, K., and Matsuo, T., 1996. Nitrous oxide emission during nitrification and denitrification in a full-scale night soil treatment plant. Water Science and Technology 34(1-2), 277-284.

Itokawa, H., Hanaki, K., and Matsuo, T., 2001. Nitrous oxide production in high-loading biological nitrogen removal process under low cod/n ratio condition. Water Research 35 (3), 657-664.

Kalogo, Y., and Verstraete, W., 1999. Development of anaerobic sludge bed (ASB) reactor technologies for domestic wastewater treatment: motives and perspectives. World Journal of Microbiology & Biotechnology 15 (5), 523-534.

Kampschreur, M. J., Tan, N. C. G., Kleerebezem, R., Picioreanu, C., Jetten, M. S. M., and Loosdrecht, M. C. M., 2008. Effect of dynamic process conditions on nitrogen oxides emission from a nitrifying culture. Environmental Science & Technology 42 (2), 429-435.



58

Khudenko, B. M., and Shpirt, E., 1986. hydrodynamic parameters of diffused air systems. WaterResearch 20 (7), 905-915.

Kimochi, Y., Inamori, Y., Mizuochi, M., Xu, K.-Q., and Matsumura, M., 1998. Nitrogen removal and N2O emission in a full-scale domestic wastewater treatment plant with intermittent aeration. Journal of Fermentation and Bioengineering 86 (2), 202-206.

Nielsen, A. H., Hvitved-Jacobsen, T., and Vollertsen, J., 2005a. Kinetics and stoichiometry of sulfide oxidation by sewer biofilms. Water Research 39 (17), 4119-4125.

Nielsen, A. H., Vollertsen, J., and Hvitved-Jacobsen, T., 2006. Kinetics and stoichiometry of aerobic sulfide oxidation in wastewater from sewers - Effects of pH and temperature. WaterEnvironment Research 78 (3), 275-283.

Nielsen, A. H., Yongsiri, C., Hvitved-Jacobsen, T., and Vollertsen, J., 2005b. Simulation of sulfide buildup in wastewater and atmosphere of sewer networks. Water Science and Technology 52(3), 201-208.

Nozhevnikova, A., Glagolev, M., Nekrasova, V., Einola, J., Sormunen, K., and Rintala, J., 2003. The analysis of methods for measurement of methane oxidation in landfills. Water Science and Technology 48 (4), 45-52.

Paing, J., Picot, B., Sambuco, J. P., and Rambaud, A., 2000. Sludge accumulation and methanogenic activity in an anaerobic lagoon. Water Science and Technology 42 (10-11), 247-255.

Park, K. Y., Inamori, Y., Mizuochi, M., and Ahn, K. H., 2000. Emission and control of nitrous oxide from a biological wastewater treatment system with intermittent aeration. Journal of Bioscience and Bioengineering 90 (3), 247-252.

Park, S., Bae, W., Chung, J., and Baek, S. C., 2007. Empirical model of the pH dependence of the maximum specific nitrification rate. Process Biochemistry 42 (12), 1671-1676.

Pauss, A., Andre, G., Perrier, M., and Guiot, S. R., 1990. Liquid-to-Gas Mass Transfer in Anaerobic Processes: Inevitable Transfer Limitations of Methane and Hydrogen in the Biomethanation Process. Applied and Environmental Microbiology 56 (6), 1636-1644.

Perry, R. H., Green, D. W., and Maloney, J. O., editors. 1997. Perry's Chemical Engineers' Handbook, 7th edition. McGraw-Hill, New York.

Peu, P., Beline, F., Picard, S., and Heduit, A., 2006. Measurement and quantification of nitrous oxide emissions from municipal activated sludge plants in France. in IWA World Water Congress. International Water Association, Beijing.

Pomeroy, R., 1959. Generation and Control of Sulfide in Filled Pipes. Sewage and Industrial Wastes 31 (9), 1082-1095.

Poth, M., and Focht, D. D., 1985. 15N kinetic analysis of N2O production by Nitrosomonaseuropaea: an examination of nitrifier denitrification. Applied and Environmental Microbiology 49 (5), 1134-1141.

Rajapakse, J. P., and Scutt, J. E., 1999. Denitrification with natural gas and various new growth media. Water Research 33 (18), 3723-3734.

Schalk-Otte, S., Seviour, R. J., Kuenen, J. G., and Jetten, M. S. M., 2000. Nitrous oxide (N2O) production by Alcaligenes faecalis during feast and famine regimes. Water Research 34 (7), 2080-2088.

Schulthess, R. V., and Gujer, W., 1996. Release of nitrous oxide (N2O) from denitrifying activated sludge: Verification and application of a mathematical model. Water Research 30 (3), 521-530.

Sharma, K. R., Yuan, Z. G., de Haas, D., Hamilton, G., Corrie, S., and Keller, J., 2008. Dynamics and dynamic modelling of H(2)Sproduction in sewer systems. Water Research 42 (10-11), 2527-2538.

Shiskowski, D. M., and Mavinic, D. S., 2006. The influence of nitrite and pH (nitrous acid) on aerobic-phase, autotrophic N2O generation in a wastewater treatment bioreactor. Journal of Environmental Engineering and Science 5 273-283.



59

Sinha, B., and Annachhatre, A. P., 2007. Partial nitrification - operational parameters and microorganisms involved. Reviews in Environmental Science and Biotechnology 6 (4), 285-313.

Stenstrom, M. K., 2007. Measurement of Oxygen Transfer in Clean Water, ASCE/EWRI 2-06. American Society of Civil Engineers

Environmental and Water Resources Institute, Reston, VA. Stenstrom, M. K., Lee, H., and Ma, J., 1997. DO Parameter Estimation Program, v.1.8, UCLA,

Los Angeles. Tallec, G., Garnier, J., Billen, G., and Gousailles, M., 2006. Nitrous oxide emissions from

secondary activated sludge in nitrifying conditions of urban wastewater treatment plants: Effect of oxygenation level. Water Research 40 (15), 2972-2980.

Tamimi, A., Rinker, E. B., and Sandall, O. C., 1994. Diffusion Coefficients for Hydrogen Sulfide, Carbon Dioxide and Nitrous Oxide in Water over the Temperature Range 293-368 K. Journal of Chemical and Engineering Data 39 (2), 330-332.

Tanaka, N., and Hvitved-Jacobsen, T., 2002. Anaerobic transformations of wastewater organic matter and sulfide production - investigations in a pilot plant pressure sewer. Water Science and Technology 45 (3), 71-79.

Tchobanoglous, G., Burton, F. L., and Stensel, H. D., 2003. Wastewater Engineering, Treatment and Reuse, 4th edition. McGraw Hill, Boston.

Thalasso, F., Vallecillo, A., Garcia-Encina, P., and Fdz-Polanco, F., 1997. The use of methane as a sole carbon source for wastewater denitrification. Water Research 31 (1), 55-60.

Thistlethwayte, D. K. B., 1972. The control of sulphides in sewerage systems. Butterworth, Sydney. Toprak, H., 1995. Temperature and Organic Loading Dependency of Methane and Carbon-Dioxide

Emission Rates of a Full-Scale Anaerobic Waste Stabilization Pond. Water Research 29 (4), 1111-1119.

Uemura, S., and Harada, H., 2000. Treatment of sewage by a UASB reactor under moderate to low temperature conditions. Bioresource Technology 72 (3), 275-282.

UNFCCC, 2007. Uniting on Climate. United Nations Framework Convention on Climate Change, Bonn.

Van't Riet, K., 1979. Review of Measuring Methods and Results in Nonviscous Gas-Liquid Mass Transfer in Stirred Vessels. Industrial and Engineering Chemistry Process Design and Development 18 (3), 357-364.

Waki, M., Suzuki, K., Osada, T., and Tanaka, Y., 2005. Methane-dependent denitrification by a semi-partitioned reactor supplied separately with methane and oxygen. BioresourceTechnology 96 (8), 921-927.

Weiss, R. F., and Price, B. A., 1980. Nitrous oxide solubility in water and seawater. MarineChemistry 8 (4), 347-359.

Wrage, N., Velthof, G. L., van Beusichem, M. L., and Oenema, O., 2001. Role of nitrifier denitrification in the production of nitrous oxide. Soil Biology and Biochemistry 33 (12-13), 1723-1732.

Yongsiri, C., Hvitved-Jacobsen, T., Vollertsen, J., and Tanaka, N., 2003. Introducing the emission process of hydrogen sulfide to a sewer process model (WATS) (vol 47, pg 85, 2003). WaterScience and Technology 47 (12), 319-320.

Zhou, Y., Pijuan, M., Zeng, R. J., and Yuan, Z., 2008. Free Nitrous Acid Inhibition on Nitrous Oxide Reduction by a Denitrifying-Enhanced Biological Phosphorus Removal Sludge. Environmental Science & Technology 42 (22), 8260-8265.


Appendix B - Glossary

vii

GlossaryA Surface area NH3 / NH4

+ Ammonia / ammonium Ae Aerobic NO2

- / HNO2 Nitrite / Nitrous acid AMBR Anaerobic migrating bed reactor NO2

--N Nitrogen in nitrite An Anaerobic NO3

- Nitrate AOB Ammonia oxidising bacteria NO Nitric oxide A/V Surface area to volume ratio NOB Nitrite oxidising bacteria APHA American Public Health Association NGER National Greenhouse and Energy Reporting ASM3 Activated Sludge Model No.3 O2 Oxygen Ax Anoxic ORP Oxidation-reduction potential BNR Biological nutrient removal p partial pressure CH4 Methane ppbv Parts per billion (volumetric basis) CO2 Carbon dioxide PST Primary sedimentation tank CO2-e Carbon dioxide equivalent P/V Volumetric power input COD Chemical oxygen demand Q Volumetric flowrate CPRS Carbon Pollution Reduction Scheme QA Volumetric aeration flowrate DB Bubble diameter R Universal gas constant

(0.08206 m3.atm.kmol-1.K-1) DCC Department of Climate Change RCH4 Recovered methane from anaerobic facilities DF Molecular diffusivity SBR Sequencing batch reactor D Depth SCADA Supervisory control and data acquisition ECD Electron capture detector SRB Sulfate reducing bacteria EF Emission factor SRT Solids retention time FID Flame ionization detector SST Secondary sedimentation tank G Generation T Temperature GF Generation factor TCD Thermal conductivity detector GHG Greenhouse gas t-disb. Student’s t-distribution GWRC Global Water Research Coalition TKN Total Kjeldahl Nitrogen HRT Hydraulic retention time TN Total nitrogen IPCC Intergovernmental Panel on Climate

Change Tr Mass transfer

kLa Volumetric mass transfer coefficient TSS Total suspended solids M Mass UASB Upflow anaerobic sludge blanket MA Methanogenic archaea UNFCCC United Nations Framework Convention on

Climate Change MCF Methane correction factor UQ University of Queensland MIMS Membrane inlet mass spectrometer UWI United Water International MLE Modified Ludzack-Ettinger V Volume MLSS Mixed liquor suspended solids VFA Volatile fatty acid MLVSS Mixed liquor volatile suspended solids vG Superficial gas velocity MW Molecular weight VSS Volatile suspended solids N2O Nitrous oxide WATS Wastewater Aerobic / Anaerobic

Transformation in Sewers N2O-N Nitrogen in nitrous oxide WSAA Water Services Association of Australia [N2O-N] Concentration of N2O-N WWTP Wastewater treatment plant α Confidence interval (i.e. α = 0.05 describes a 95% confidence interval) αF Mixed liquor / fouling correction factor for clean water kLa β Salinity correction factor for clean water solubility concentrations θ Temperature correction factor (1.024)

www.wsaa.asn.au

Melbourne Office Level 8, 469 Latrobe Street Melbourne VIC 3000

PO Box 13172 Law Courts Post Office Melbourne VIC 8010

Phone: (03) 9606 0678 Fax: (03) 9606 0376

Sydney Office Suite 1, Level 30 9 Castlereagh Street Sydney NSW 2001

GPO Box 915 Sydney NSW 2001

Phone: (02) 9221 5966 Fax: (02) 9221 5977

Occasional Paper No. 24

September 2009


wsaa occasional paper no.24 - direct methane nad nitrous oxide ... a… · and nitrous oxide (n2o)...

Documents