lable at ScienceDirect

Journal of Environmental Management 143 (2014) 34e43

Contents lists avai

Journal of Environmental Management

journal homepage: www.elsevier .com/locate/ jenvman

Biosolid stockpiles are a significant point source for greenhouse gasemissions

Ramaprasad Majumder a, Stephen J. Livesley b, David Gregory c, Stefan K. Arndt a,*aDepartment of Forest and Ecosystem Science, The University of Melbourne, 500, Yarra Boulevard, Richmond, VIC 3121, AustraliabDepartment of Resource Management and Geography, The University of Melbourne, 500, Yarra Boulevard, Richmond, VIC 3121, Australiac Technology and Marine Research, Melbourne Water, 990 Latrobe Street, Docklands, VIC 3008, Australia

a r t i c l e i n f o

Article history:Received 22 January 2014Received in revised form11 April 2014Accepted 20 April 2014Available online

Keywords:BiosolidsWaste managementGreenhouse gasesSewage sludgeWater treatmentLabile carbon

* Corresponding author. Tel.: þ61 3 90356819.E-mail address: sarndt@unimelb.edu.au (S.K. Arnd

http://dx.doi.org/10.1016/j.jenvman.2014.04.0160301-4797/� 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

The wastewater treatment process generates large amounts of sewage sludge that are dried and thenoften stored in biosolid stockpiles in treatment plants. Because the biosolids are rich in decomposableorganic matter they could be a significant source for greenhouse gas (GHG) emissions, yet there are nodirect measurements of GHG from stockpiles. We therefore measured the direct emissions of methane(CH4), nitrous oxide (N2O) and carbon dioxide (CO2) on a monthly basis from three different age classesof biosolid stockpiles at the Western Treatment Plant (WTP), Melbourne, Australia, from December 2009to November 2011 using manual static chambers. All biosolid stockpiles were a significant point sourcefor CH4 and N2O emissions. The youngest biosolids (<1 year old) had the greatest CH4 and N2O emissionsof 60.2 kg of CO2-e per Mg of biosolid per year. Stockpiles that were between 1 and 3 years old emittedless overall GHG (w29 kg CO2-e Mg�1 yr�1) and the oldest stockpiles emitted the least GHG (w10 kgCO2-e Mg�1 yr�1). Methane emissions were negligible in all stockpiles but the relative contribution ofN2O and CO2 changed with stockpile age. The youngest stockpile emitted two thirds of the GHG emissionas N2O, while the 1e3 year old stockpile emitted an equal amount of N2O and CO2 and in the oldeststockpile CO2 emissions dominated. We did not detect any seasonal variability of GHG emissions and didnot observe a correlation between GHG flux and environmental variables such as biosolid temperature,moisture content or nitrate and ammonium concentration. We also modeled CH4 emissions based on afirst order decay model and the model based estimated annual CH4 emissions were higher as comparedto the direct field based estimated annual CH4 emissions. Our results indicate that labile organic materialin stockpiles is decomposed over time and that nitrogen decomposition processes lead to significant N2Oemissions. Carbon decomposition favors CO2 over CH4 production probably because of aerobic stockpileconditions or CH4 oxidation in the outer stockpile layers. Although the GHG emission rate decreased withbiosolid age, managers of biosolid stockpiles should assess alternate storage or uses for biosolids to avoidnutrient losses and GHG emissions.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Greenhouse gas (GHG) concentrations of methane (CH4), nitrousoxide (N2O) and carbon dioxide (CO2) have increased in the at-mosphere due to direct and indirect human activities (Dalal et al.,2003). The waste and wastewater industry sectors contributeapproximately 3% to the global anthropogenic emissions of GHG(IPCC, 2007), which has led to increased public awareness and theneed to better estimate or measure GHG emissions from the sector

t).

(Ishigaki et al., 2005). The wastewater sector emits CH4, CO2 andN2O directly through nutrient transformations during the waste-water treatment process and degradation of organic matter(Boldrin et al., 2009; Czepiel et al., 1996) and indirectly throughfossil fuel consumption for pumping, transport, operations andprocessing (Monteith et al., 2005). In all wastewater treatmentplants large quantities of sewage sludge are produced (Brown et al.,2008). The dried sludge is called biosolids and still contains sig-nificant quantities of organic matter, macro-nutrients and traceelements (Bright and Healey, 2003). The quantity of biosolid pro-duced in Australia is around 360,000 dry Mg per year (Pritchardet al., 2010) and in the state of Victoria it is approximately 68,250dry Mg (AWA, 2013). These biosolids are often stored in large

R. Majumder et al. / Journal of Environmental Management 143 (2014) 34e43 35

stockpiles for many years within the wastewater treatment site,before they are moved off-site and used as a resource.

Currently, GHG emissions from waste and wastewater treat-ment plants are often based on simple emission factors or mathe-matical models developed from kinetic relationships of mass andenergy balances (Yerushalmi et al., 2013). One of the main un-certainties in GHG accounting is associated with the method ofcalculation, which is frequently not based on, or validated by, directGHG flux measurements. For example, waste industries use mini-mumdata to predict the GHG emissions based on population, wastegeneration rate, solid waste composition and methods of solidwaste disposal (El-Fadel and Massoud, 2000). Another uncertaintyconfronting GHG accounting in thewastewater industry is access tolong term monitoring of GHG emissions and under a wide range ofclimatic and management conditions (Cowie et al., 2012; Bogneret al., 2008; Czepiel et al., 1993). Hence, to reduce these un-certainties it is important to obtain reliable long-term data to betterunderstand the magnitude of direct GHG emissions and the vari-ation over seasonal or inter-annual time frames. There are manystudies investigating GHG emissions from landfill waste (Bastianet al., 2013; Stanisavljevic et al., 2012; Goldsmith et al., 2012),stockpile composting (Chan et al., 2011; Ahn et al., 2011; Boldrinet al., 2009) and stockpile manure management (Petersen et al.,2013; Wood et al., 2012). However, GHG emissions from stock-piled biosolids have not yet been measured in detail.

The overall objective of this study was therefore to investigatethe direct GHG emissions from various ages of biosolids stockpilesat a wastewater treatment plant. The specific objectives were to:

(1) measure the flux of CH4, CO2 and N2O from different ages ofbiosolid stockpiles across different seasons and investigateGHG concentrations (CH4, CO2) at two different biosolidstockpile depths,

(2) investigate the relationship between GHG emissions frombiosolid stockpiles and environmental variables,

(3) calculate the cumulative magnitude of annual CH4, CO2 andN2O fluxes from different aged biosolid stockpiles andcompare this against modeled emissions.

2. Material and methods

2.1. Site description and experiment set up

The Western Treatment Plant (WTP) is located 35 km to thesouth-west Melbourne and services wastewater from a populationof about 1.6 million people from the western and northern suburbsof Melbourne (38�105200S, 144�34’8200E). Climate in this area istemperate with warm dry summer and cool winters withmaximum rainfall occurring during spring. The long-term averagerainfall is 542 mm (Stickland et al., 2013) and mean evaporationis generally highest between December and February(Parameswaran, 1999). The WTP was established more than 100years ago and currently processes 52% of Melbourne sewage. Thesewage is transported from domestic (70%) and industrial (30%)sources to the WTP via pipes and pumping stations. The plant re-ceives an average 500 ML of wastewater per day and has amaximum capacity of 2000 ML per day. Wastewater in the WTP isdistributed to lagoons for processing in two stages. The first stage isanaerobic treatment in an initial deeper lagoon, which is covered toallow capture of CH4 gas produced for biogas energy. The secondstage are aerobic lagoons to provide an active process for nitrogenremoval. The treatment processes produces sludge that settles tothe floor of the lagoons that therefore require removal. The sludgefrom the base of the aerobic and anaerobic settling ponds is

collected using a floating pump and then transported to large openair drying pans to produce biosolids. The length of air-drying de-pends upon the potential evaporation and rainfall conditions, butgenerally ranges between 10 and 20 weeks. The dried biosolidproduct is then stored onsite in a stockpile format. The compositionof each stockpile is likely to vary quite significantly depending onthe relative contribution of sludge that has been collected fromaerobic or anaerobic settling ponds and the time it has taken to drythe sludge in the drying pans. The anaerobic settling ponds had notbeen dredged for a long period of time and continuous dredgingstarted in 2009. Hence, biosolids from anaerobic ponds are ofvarying age and consistency, depending where they were collected.

The total annual biosolid production in WTP is aroundw24,000dry Mg (Stickland et al., 2013). There are currently 15 stockpilesonsite in the WTP. Three stockpiles of different age ranges wereselected to assess the magnitude of GHG flux emissions over a two-year period and to better understand the biosolid characteristicsand environmental conditions that may influence emissions. Weselected three different age ranges to assess if GHG emissionschange during the ageing of biosolids. Themain sources of biosolidsin the oldest stockpile (>3 years old, yo) were from the sludge ofvarious drying pans collected before 1995. The 1e3 yo biosolids and<1 yo biosolids were sourced from the ongoing sludge drying pancollections. The dimensions of the three different biosolid stock-piles were: >3 yo (130e150 m long, 40e50 m wide and 8e10 mhigh), 1e3 yo (130e150 m long, 50e60 mwide and 10e12 m high)and <1 yo (100e120 m long, 30e40 m wide and 6e8 m high).

2.2. Measuring GHG flux from biosolid stockpiles

The closed static chamber technique (Hutchinson and Mosier,1981) was used to quantify the GHG flux rates of CH4, CO2 andN2O from the top surface of each stockpile. Gas samples werecollected every month from December 2009 to November 2011from the >3 yo and 1e3 yo stockpiles, and from March 2010 toNovember 2011 for the <1 yo stockpile. Eight chambers wereplaced in a row about 10e12 m apart on the top surface of eachstockpile. Chambers were made from non-transparent PVC pipe(diameter 25 cm, height 24.5 cm, volume 12.0 L, basal area0.045m2) with a twist-lid incorporating a butyl-rubber septum anda rubber O-ring to form a gas tight seal. Chambers were inserted toa depth of 3e4 cm at least 15e30 min before closing the lid. Onceclosed, 20 mL headspace gas samples were taken at 0, 4, 8 and12 min after closure and stored in 12 mL pre-evacuated Exetainer�(Labco, UK) gas vials. Gas samples were collected between 10:30am and 14.30 pm. After collecting gas samples, the chamber lidswere removed to measure the internal chamber height at fourpoints to calculate the internal headspace volume of each chamber.Gas samples stored in Exetainers� were transferred to the labo-ratory and analysed for CH4, CO2 and N2O concentration throughgas chromatography (Shimadzu GC17A with N2 carrier gas).Methane and CO2 concentrations were determined using a flameionization detector (FID), and N2O concentration determined usingan electron capture detector (ECD). Gas concentrations (%) of CH4,CO2 and O2 were also measured at two different depths (0.5 m and1 m depth) from August 2010 to November 2011 from the >3 yoand 1e3 yo biosolids stockpiles using GA 2000 landfill gas analyser.

2.3. Measuring biosolid properties

The temperature of biosolid (BT) was measured at a depth of10 cm using a short temperature probe (ColeeParmer, USA) for theentire measurement period and at a depth of 50 cm using a longerDigi-Sense� Type K thermocouple thermometer from August 2010onwards. Biosolid samples were collected from 0 to 10 cm using a

R. Majumder et al. / Journal of Environmental Management 143 (2014) 34e4336

stainless steel ring (10 cm diameter, 10.5 cm height) at eachchamber location to measure the moisture content (MC), pH,electrical conductivity (EC) and to analyze NO3 - and NH4 + con-centrations. Biosolid MC was determined gravimetrically afterdrying subsamples at 105 �C for 48 h. A fresh subsample wasanalysed for pH and EC using a pH/EC meter (TPS, lab CHEM) in a1:5 biosolid:water solution. Biosolid samples (8 g) were extractedwith 1 M KCl (1:4, biosolid:KCl) and shaken for 1 h on a flat-bedshaker, then filtered (Whatman 42) and frozen prior to analysis ofNO3 - and NH4 + on a Technicon� auto-analyser. Bulk density wasmeasured by the oven dried biosolid (g) divided by the total volumeof soil (cm3) (Gifford and Roderick, 2003). The samples werecollected every month on the same day as GHG flux measurement.Biosolid samples collected at the beginning and end of the studyperiod were air dried and measured for total volatile solid content(TVS, %) and carbon content (C, %). TVS was determined throughloss on ignition in a muffle furnace at 550 �C. Total organic carbonwas measured after acid treatment to remove inorganic carbonatesfollowed by LECO CHN� combustion analysis (Yeomans andBremner, 1991).

2.4. GHG flux calculation

The fluxes of CH4, CO2 and N2O from biosolid stockpiles werecalculated using a curvilinear function because headspace gasconcentration gradually decreased with length of chamber closure(Matson and Harriss, 1995):

fo ¼ V*ðC1 � C0Þ2.h

A*t*1�2*C1 � C2 � C0

�i

*ln½ðC1 � C0Þ=ðC2 � C1Þ� (1)

where, fo is the flux at time 0, V is the chamber headspace volume(L), A is the biosolid surface area (m2), C0, C1, and C2 are the chamberheadspace gas concentrations ppm(v) at time 0, 4, 8 and 12,respectively, and t1 is the interval between gas sampling points(Minato et al., 2013). The resulting units of fo are mL tracegas m�2 min�1.

This flux (fo) was then converted to mmol CH4, CO2 andN2O m�2 h�1 by accounting for temperature, pressure and volumebased on the ideal gas law by using Equation 2

Fmmol ¼FmL � PR� T

(2)

where Fmmol is the flux in mmol CH4, CO2 and N2O m�2 h�1, FmL is theflux in mL CH4, CO2 and N2O m�2 h�1, P is the atmospheric pressurein kPa at the site according to altitude, R is 8.3144 (the ideal gasconstant in L kPa mol�1 K�1), and T is air temperature in K(273 þ �C). Fluxes in mmol CH4, CO2 and N2O m�2 h�1 were thenconverted to mg CH4eC, CO2eC and N2OeN m�2 h�1 based onmolecular and elemental mass of each gas.

2.5. Data analysis and presentation

Preliminary flux measurements at different times of the dayindicated only small diurnal variation of flux for all GHG. Hence, wemultiplied the mean hourly flux (mg kg�1 hr�1) of every measure-ment day by 24 to estimate the daily flux of CH4, CO2 and N2O. Toestimate the monthly flux the daily flux of CH4, CO2 and N2O, themean daily flux (mg kg�1 d�1) was multiplied by the days of eachmonth. The monthly flux was converted to kg CO2-e m2 month�1

for CH4 and N2O flux using the global warming potential of 25 forCH4 and 298 for N2O (IPCC, 2007). Annual emissions (CO2-e) werecalculated by total GHG emissions from the total measured months

divided by total measured months and multiplied by 12. We esti-mated CH4 emissions from stockpiles based on first order decay(FOD) model following the National Greenhouse and EnergyReporting system (NGER) guidelines, where biosolid stockpile CH4emission was estimated similar to landfill:

Ej ¼hCH*

4 � g�Qcap þ Qflared þ Qtr

�i� ð1� OFÞ (3)

where Ej ¼ emissions of CH4 released by the landfill during the year[T CO2-e], OF ¼ oxidation factor (0.1) for near surface methane inthe landfill, CH�

4 ¼ estimated quantity of CH4 in landfill gas gener-ated by the landfill during the year and measured in Mg CO2-e.

CH*4 ¼ CH4gen ¼ DCaðtÞ þ ðDCosðtÞÞF � 1:336� 21 (4)

where, CH4gen ¼ methane generated at the landfill, DCa(t) ¼ quantity of decomposable degradable organic carbon in thenewly deposited waste arriving at the landfill lost through decay(decomposition) (in Mg), (DCos(t) ¼ quantity of decomposabledegradable organic carbon from the opening stock of carbon at thelandfill lost through decay (decomposition) (in Mg), F ¼ 0.5, frac-tion of CH4, by volume, generated in landfill gas, 1.333¼ conversionfactor from a mass of C to mass of CH4. The details about the FODmodel and the NGER guidelines are available online (NGER, 2013).

A simple linear regression analysis was used to evaluate thesignificance of the relationship between CH4, CO2 and N2O flux andMC and BT. Annual GHG emissions (kg Mg�1) were calculated bysumming the total GHG emissions per tonne of biosolid from thetotal measured months, dividing it the total months and multi-plying it by 12. For this, total surface area (top, sides and ramps) ofeach stockpile was measured and total biosolid mass (oven driedbasis) was also calculated based on total volume of biosolid, bulkdensity and MC. GHG emissions (kg Mg�1) of biosolid weremeasured total area based GHG emissions of each stockpile dividedby the total biosolid mass (oven dried basis) of each stockpile.

3. Results

3.1. GHG emissions from different aged biosolid stockpiles

We observed different GHG emissions from the different agedbiosolid stockpiles and also different contributions from the GHG.We observed the greatest GHG emissions from the youngeststockpiles and emission decreased with age. Methane emissionswere low in all stockpiles and N2O and CO2 dominated emissions,with N2O having the greatest contribution in the youngest stockpileand CO2 the greatest contribution in the oldest stockpile.

In the >3 yo biosolid stockpile, the magnitude of CO2 flux wasgreater (between 0.8 and 4.5 kg CO2-e m�2 month�1) as comparedto CH4 (<0.03 kg CO2-e m�2 month�1) and N2O (0.15e3.2 kg CO2-e m�2 month�1) (Fig. 1). Methane flux was consistently very low(around 0.1 kg CO2-e m�2 month�1) and only increased to 0.3 kgCO2-e m�2 month�1 between September and November 2010 asMC increased from 20% to 40% due to high rainfall. CO2 flux wasgreatest in the first three measurement months and then stabilizedat an average flux of around 2 kg CO2-e m�2 month�1. Althoughthere was some monthly variation, we did not observe anyapparent seasonal flux differences. N2O fluxes were generallyaround 1 kg CO2-e m�2 month�1 but increased to 2e4 kg CO2-e m�2 month�1 in the period between November 2010 and May2011.

The biosolid temperature showed typical seasonality withgreater temperature observed in the summer months as comparedto the winter months and temperature at 50 cm closely matched

2009-2010 2010-2011

CH

4xulf

(kg

CO

2-e

m-2

mon

th-1

)

0.0

0.5

1.0

1.5

N2

xulfO

OC

gk(2-

e m

-2m

onth

-1)

02468

1012

CO

2xulf O

Cg k(

2

m-2

mon

th-1

)

02468

sdilosoiB

eru tarepmet

(°C

)

010203040 at 10cm depth

at 50cm depth

sdilosoiB

tnetnocerutsio

m(%

)

0

20

40

60 upto 10 cm depth

Dec-09Jan-10

Feb-10Mar-10

Apr-10May-1

0Jun-10

Jul-10Aug-10

Sep-10Oct-10

Nov-10

Dec-10Jan-11

Feb-11Mar-11

Apr-11May-1

1Jun-11

Jul-11Aug-11

Sep-11Oct-11

Nov-11

llafniarylhtn o

M)

mm (

0

50

100

150

Mea

n m

onth

ly ai

r tem

p (°

C)

0

10

20

30Rainfall Air temp

Fig. 1. Monthly fluxes of CH4, N2O and CO2 from the >3 years old biosolid stockpile (SE, n ¼ 8) in the period of 2009e2011 at the Western Treatment Plant, Werribee, Australia.

R. Majumder et al. / Journal of Environmental Management 143 (2014) 34e43 37

temperature at 10 cm. The biosolid MC was generally quite highand, with the exception of a few months at the beginning of themeasurement period, we observed biosolid MC of between 40 and60% from April 2010 onwards. In the >3 yo stockpile, TVS contentdecreased from 36 to 35% and the C content from 20 to 19%,respectively during the study period (Table 1). The pH increasedfrom approximately 4.5e5.3 and EC decreased from 1.9 to0.32 mS cm�1.

In the 1e3 yo stockpile, the overall magnitude of GHG emissionflux was greater as compared to the >3 yo stockpile. The greatestemission in this stockpile was CO2 flux (0.3e7.5 kg CO2-e m�2 month�1), closely followed by N2O flux (0.3e7.2 kg CO2-e m�2 month�1), whilst CH4 flux (<1.5 kg CO2-e m�2 month�1) wascomparatively low (Fig. 2). Methane flux was greatest at thebeginning of the measurement period in December 2009 anddeclined gradually until October 2010, without showing apparentseasonality. From October 2010 onwards we observed steady andlow CH4 emissions of <0.1 kg CO2-e m�2 month�1. The averagemonthly CO2 flux emission was generally >3.5 kg CO2-e m�2 month�1 but varied throughout the two year study period

Table 1Total volatile solids (TVS%), Carbon (C) content (%), pH and electrical conductivity (EC) ofend of the study period (SE, n ¼ 8) at the Werribee Wastewater Treatment Plant, Austra

Stockpiles

>3 years old 1e3 y

Parameters Dec 09 Nov 11 Dec 0

TVS% 36.0 � 0.7 34.8 � 0.9 21.0 �C content% 19.8 � 0.4 19.1 � 0.5 11.6 �pH 4.5 � 0.2 5.3 � 0.1 6.1 �EC (mS cm�1) 1.9 � 0.2 0.32 � 0.2 3.7 �

and did not show any obvious seasonal pattern with changes inbiosolid temperature and MC. N2O fluxes were generally around3 kg CO2-e m�2 month�1 but increased to 4e7 kg CO2-e m�2 month�1 between September 2010 and December 2010. Inthis stockpile, the TVS % decreased from 21.0 to 20.3% and C %decreased from 11.6 to 11.1%, respectively during the study period(Table 1). The pH decreased slightly from 6.1 to 5.8 whereas ECdecreased sharply from 5.1 to 1.1 mS cm�1.

A different GHG flux pattern was observed in the <1 yostockpile where N2O emissions (0.8e10.3 kg CO2-e) were greaterthan CO2 emissions (0.5e5.0 kg CO2), but CH4 emissions wereagain negligible (Fig. 3). Methane emissions were consistently low(<0.1 kg CO2-e m�2 month�1) during the entire measurementperiod.

The average monthly CO2 flux was >2 kg CO2-e m�2 month�1

but varied considerably throughout entire study period withoutshowing obvious seasonal patterns. The N2O emissions were alsovariable and we observed the greatest emissions in the springmonths (October/November) in each measurement year, while theother months showed considerably lower fluxes. In this stockpile,

different aged biosolid stockpiles, sampled at 0e10 cm depths at the beginning andlia.

ears old <1 year old

9 Nov 11 Mar 10 Nov 11

1.2 20.3 � 0.6 38.3 � 3.0 34.5 � 1.00.7 11.1 � 0.4 21.0 � 1.0 19.0 � 0.60.1 5.8 � 0.1 6.8 � 0.1 5.6 � 0.10.2 1.1 � 0.1 5.0 � 0.3 1.9 � 0.2

CH

4xulf

(kg

CO

2-e

m-2

mon

th-1

)

0.0

0.5

1.0

1.5

N2

xulfO

OC

gk(2-

e m

-2m

onth

-1)

02468

1012

CO

2 xulf O

Cg k(

2

m-2

mon

th-1

)

02468

sdilosoiB

erutarepmet

(°C

)

010203040 at 10cm depth

at 50cm depth

sdilosoiB

tnetno ceruts io

m(%

)

0

20

40

60 upto 10cm depth

Dec-09Jan-10

Feb-10Mar-10

Apr-10May-1

0Jun-10

Jul-10Aug-10

Sep-10Oct-10

Nov-10

Dec-10Jan-11

Feb-11Mar-11

Apr-11May-1

1Jun-11

Jul-11Aug-11

Sep-11Oct-11

Nov-11

llafniarylhtno

M)

mm(

0

50

100

150

Mea

n m

onth

ly ai

r tem

p (°

C)

0

10

20

30RainfallAir temp

2009-2010 2010-2011

Fig. 2. Monthly fluxes of CH4, N2O and CO2 from the 1e3 year old biosolid stockpile (SE, n ¼ 8) in the period of 2009e2011 at the Western Treatment Plant, Werribee, Australia.

CH

4xulf

(kg

CO

2-e

m-2

mon

th-1

)

0.0

0.5

1.0

1.5

N2

xulfO

(kg

CO

2-e

m-2

mon

th-1

)

02468

1012

CO

2xulf

(kg

CO

2

m-2

mon

th-1

)

02468

sdilosoiB

erutarepmet

(°C

)

010203040

at 10 cm depthat 50cm depth

sdilosoiB

tnetnocerutsio

m(%

)

0

20

40

60upto 10 cm depth

Dec-09Jan-10

Feb-10Mar-10

Apr-10May-1

0Jun-10

Jul-10Aug-10

Sep-10Oct-10

Nov-10

Dec-10Jan-11

Feb-11Mar-11

Apr-11May-1

1Jun-11

Jul-11Aug-11

Sep-11Oct-11

Nov-11

llafniarylhtno

M)

mm(

0

50

100

150

Mea

n m

onth

ly ai

r tem

p (°

C)

0

10

20

30RainfallAir temp

2010-20112009-2010

Fig. 3. Monthly fluxes of CH4, N2O and CO2 from the <1 year old biosolid stockpile (SE, n ¼ 8) in the period of 2009e2011 at the Western Treatment Plant, Werribee, Australia.

R. Majumder et al. / Journal of Environmental Management 143 (2014) 34e4338

>3 years old stockpile

CH

4noitartnecnoc

(%)

-10123 at 0.5m depth

at 1m depth

1-3 years old stockpile

-10123

>3 years old stockpile

CO

2noitartn ecno c

(%)

05

10152025

1-3 years old stockpile

05

10152025

>3 years old stockpile

O2

noitartnecn oc(%

) 05

101520

1-3 years old stockpile

MonthAug-10

Sep-10Oct-10

Nov-10Dec-10

Jan-11Feb-11

Mar-11Apr-11

May-11Jun-11

Jul-11Aug-11

Sep-11Oct-11

Nov-1105

101520

Fig. 4. Monthly concentration (%) of CH4, CO2 and O2 at two different depth from the >3 year and the 1e3 year old biosolid stockpiles (SE, n ¼ 3) in the period of 2009e2011 at theWestern Treatment Plant, Werribee, Australia.

R. Majumder et al. / Journal of Environmental Management 143 (2014) 34e43 39

TVS decreased from 38.3 to 34.5% and C % from 21 to 19% during thestudy period (Table 1). The pH decreased from 6.8 to 5.6 and ECdecreased from 5.0 to 1.9 mS cm�1.

3.2. CH4, CO2 and O2 concentration (%) at two different depths ofthe 1e3 year old and the >3 year old biosolid stockpile

The average CH4 concentration (0.62%) at 1.0 m depth in the >3yo stockpile was greater than at 0.5 m (0.06%) (Fig. 4) but thisdifference was driven by four months in which the CH4 concen-tration at depth was substantially greater. In contrast, there was nodifference between CH4 concentrations (%) at 0.5 and 1.0 m in the1e3 yo stockpile and the CH4 concentrations generally were verylow. CO2 concentrations at 1.0 m depth of both the stockpiles weregreater than at 0.5 m depths. The average CO2 concentrations weregreater at both 0.5 m (13%) and 1.0 m depth (20%) in 1e3 yostockpile as compared to 0.5 m (5%) and 1.0 m (15%) in the >3 yostockpile. In contrast, the average O2 concentrations were greater atboth 0.5 m (17%) and 1.0 m (9%) in the >3 yo stockpile as comparedto 0.5 m (10%) and 1.0 m (4%) in the 1e3 yo stockpile.

3.3. Relationships between biosolid moisture content, temperatureand GHG flux

Generally, we observed no, or a very weak, relationship betweenMC and GHG fluxes of the three biosolid stockpiles. However, therewas a significant (p � 0.05) weak correlation between MC and CH4flux of 1e3 yo biosolid stockpile (BSP) (Table 2).

There was no significant relationship between MC and CH4 fluxfor the <1 yo stockpile and >3 yo stockpile. A very weak relation-ship was apparent between MC and CO2 flux from the >3 yostockpile, but there was no correlation in the <1 year and 1e3 yobiosolid stockpiles. We also observed only very weak relationshipsbetween BT and GHG flux in all of the studied stockpiles. However,there was a significant relationship (p � 0.05) between BT at 10 cmand CO2 and N2O flux in the >3 yo and 1e3 yo stockpiles. Nitrateconcentration was significantly correlated with N2O flux in the twoolder stockpiles and could explain around 30% of the variation, butnot in the youngest stockpile. Ammonium concentration was notcorrelated with N2O flux in any stockpile.

3.4. Biosolid NO3 - -N and NH4 + -N contents

The biosolids contained significant amounts of mineral nitrogenwith NO3 - being the dominant form (Fig. 5). The average NO3 - -Nconcentration in the <1 yo stockpile was with 457 mg N kg�1

Table 2The relationship between greenhouse gas flux ofmethane (CH4), nitrous oxide (N2O)and carbon dioxide (CO2), biosolid moisture content (MC) and temperature (BT) at10 cm depth and nitrate (NO3 - ) and ammonium (NH4 + ) concentration in threedifferent aged biosolid stockpiles (BSP) at the Western Treatment Plant, Werribee,Australia.

Stockpiles GHG flux

CH4 N2O CO2

R2 p value R2 p value R2 p value

MC at 10 cm>3 years old BSP 0.01 ns 0.01 ns 0.19 <0.051e3 years old BSP 0.23 <0.05 0.01 ns 0.08 ns<1 year old BSP 0.01 ns 0.06 ns 0.13 ns

BT at 10 cm>3 years old BSP 0.03 ns 0.20 <0.05 0.36 <0.051e3 years old BSP 0.07 ns 0.24 <0.05 0.22 <0.05<1 year old BSP 0.01 ns 0.07 ns 0.01 ns

NO3 - -N content>3 years old BSP 0.004 ns 0.30 <0.05 0.14 ns1e3 years old BSP 0.64 <0.05 0.34 <0.05 0.04 ns<1 year old BSP 0.007 ns 0.008 ns 0.21 <0.05

NH4 + -N content>3 years old BSP 0.02 ns 0.01 ns 0.48 <0.051e3 years old BSP 0.60 <0.05 0.02 ns 0.27 <0.05<1 year old BSP 0.35 <0.05 0.01 ns 0.02 ns

R. Majumder et al. / Journal of Environmental Management 143 (2014) 34e4340

almost two-fold greater than that in the 1e3 yo stockpile (245mg Nkg�1), and more than three-fold greater than that in the >3 yostockpile (119 mg N kg�1).

There was no apparent seasonal variation of NO3 - concentra-tion in either of the stockpiles. The NH4 + concentration was anorder of magnitude lower compared to NO3 - , but we also observedthe greatest average NH4 + -N concentration in the <1 yo stockpile(40 mg N kg�1) which was five-fold greater than in the 1e3 yostockpile (8 mg N kg�1) and more than six-fold greater than in the>3 yo stockpile (6 mg N kg�1). However, this difference wasgenerated by very high NH4 + concentrations in the<1 yo stockpilebetween Aug and October 2010, for the remainder of the time theNH4 + concentration was in a similar range in all stockpiles. Wealso did not observe any apparent seasonal variation of NH4 +concentration in either stockpile.

NO

3-tnetnoc

N-g k

Ng

m(-1

)

0200400600800

Dec-09Jan-10

Feb-10Mar-10

Apr-10May-10

Jun-10Jul-10

Aug-10Sep-10

Oct-10Nov

NH

4+tnetnoc

N-gk

Ng

m(-1

)

050

100150200250

Fig. 5. Monthly NO3 - and NH4 + concentration (mg N kg�1) from the >3 year old, 1e3 yeaWestern Treatment Plant, Werribee, Australia.

3.5. Estimated annual GHG emissions from direct field basedmeasures and modeling

Based on our monthly field measurements we calculated theannual GHG emissions from the different biosolid stockpiles. Theoverall annual GHG emissionwas greatest in the youngest stockpile(90.3 kg CO2-e Mg�1 yr�1), followed by 1e3 yo (59.3 kg CO2-e Mg�1 yr�1), whilst the oldest stockpile had the lowest annualemissions, less than a third of those of the youngest stockpile (27.5CO2-e Mg�1 yr�1) (Table 3).

In the youngest stockpile two thirds of the GHG emissions werecaused by N2O emissions, whilst the 1e3 yo stockpile had equalcontributions from N2O and CO2 and the oldest stockpile hadalmost double the CO2 emission compared to N2O. Methaneemissions were very low in all stockpile ages and contributed lessthan 3% of the total GHG emissions of the stockpiles.

But multiplying the average annual GHG emission by the totalbiosolid stockpile mass for each age class we were able to deter-mine a total annual GHG emission of 44,049 Mg of CO2-e from allthe stored biosolid stockpiles at the Werribee treatment plant(Table 3). Although the younger age (<1 yo) stockpile has thegreatest emissions, there is a greater mass of stockpiles that are >3yo, such that overall GHG emission was greatest from >3 yostockpiles (91%), followed by 1e3 years (5%) and<1 yo (4%) biosolidstockpile. The overall the contribution of stockpile CH4 emissionswas negligible (<2%) as compared to the proportional contributionof N2O (37%) and CO2 emissions (61%) to total GHG emissions.

The first order decay (FOD) model estimated approximately ten-fold greater annual CH4 emissions than those estimated from directfield-based measures of this study (Fig. 6).

4. Discussion

4.1. GHG emissions from biosolid stockpiles of different age

Our data clearly show that age of the stockpiles is important forthe total amount of GHG emissions (CO2-e) and that the contribu-tion of the different GHG species was not equal in the different agedbiosolid stockpiles. The youngest stockpile had the greatest GHGemissions and these decreased with stockpile age. We alsoobserved that N2O had the greatest contribution to total GHGemissions in the youngest stockpile and this decreased in favor ofCO2 as the stockpiles aged. Methane only played a minor role in thetotal GHG emissions.

The average monthly CH4 emissions were less than 1% of thetotal GHG emissions in the youngest and oldest stockpiles and wereonly slightly greater in 1e3 yo biosolid stockpile (3%). The mostimportant factor determining CH4 production, if organic substrates

>3 yo1-3 yo<1 yo

Month

-10Dec-10

Jan-11Feb-11

Mar-11Apr-11

May-11Jun-11

Jul-11Aug-11

Sep-11Oct-11

Nov-11

r old and <1 year old biosolid stockpiles (SE, n ¼ 8) in the period of 2009e2011 at the

Table 3Estimated annual GHG emissions (kg CO2-e dry Mg�1 yr�1) and total GHG emissions from three different aged biosolid stockpiles (BSP) at the Western Treatment Plant,Werribee, Australia.

Stockpile Annual GHG emissions (kg CO2-e dry Mg�1 yr�1) Biosolid mass (dry Mg) Annual total GHG (kg CO2-e yr�1)

CH4 N2O CO2 CH4 þ N2O CH4 N2O CO2 CH4 þ N2O

>3 yo BSP

0.3

9.8 17.4 10.1 1,455,900 437 14,268 25,333 14,705

1e3 yo BSP 2.0 26.8 30.5 28.8 38,400 77 1029 1171 1106<1 yo BSP 0.2 60.0 30.1 60.2 19,200 4 1152 578 1156Total 2.5 96.6 78.0 e 1,513,500 518 16,449 27,082 16,967

Year1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

CH

4O

C(xulf

2-e)

ennotyrd

gk(-1

)dilosoibfo

0

5

10

15

20FOD model Direct measures

Fig. 6. Comparison of annual CH4 emissions (kg CO2-e dry tonne�1) from three different ages of biosolid stockpile emission with model based CH4 in the period of 2009e2011 at theWestern Treatment Plant Werribee, Australia.

R. Majumder et al. / Journal of Environmental Management 143 (2014) 34e43 41

are available, is the intensity of the anaerobic conditions within thebiosolid stockpile (Peigne and Girardin, 2004). Amlinger et al.(2008) noted in a composting study that O2 availability was amajor factor for controlling CH4 formation dynamics. We observedthat O2 concentration in the deeper 1 m biosolid layer were alwayslower compared to the 0.5m layer, indicating that wemay have hadanaerobic microsites that could have led to CH4 production bybacterial decomposition of organic matter. Concurrently, we alwaysobserved greater CH4 concentration in the deeper biosolid layer,confirming that there was CH4 production in these layers. The factthat we did not observe greater CH4 fluxes are likely related toconsumption processes of CH4 in the biosolid stockpile. Methano-trophic bacteria utilize CH4 as an energy source and convert it intoCO2 in well-aerated soil layers. Given the high O2 concentration inthe upper biosolid layers it is likely that the methanotrophs in thesurface layers of the biosolid stockpile may consume some, or eventhe majority, of CH4 that passes through converting it into CO2.Similar trends have been observed from landfill studies, where theaerobic microbial community on the surface of the pile oxidizedCH4 before it was released into the environment (Bogner et al.,1997). A review study by Lou and Nair (2009) observed that themajority of CH4 from landfills is oxidized and released as CO2 fromthe surface of the pile which reduces the overall CH4 emissions.

The bulk of the annual carbon emissions from stockpiles wasCO2, confirming that most of the decomposable organic materialwas oxygenized and that anaerobic conditions in the stockpilesseem to be an exception rather than the rule. In static compostpiles, emissions of CH4 and CO2 peaked a few days after construc-tion and gradually decreased and diminished over time (Ahn et al.,2011). In our study, the newer biosolid stockpile had greater CH4

and CO2 emissions as compared to the older stockpiles. The annualstockpile CO2 emissions were similar in the two younger stockpilesbut much lower in the >3 yo. This indicates that the amount ofdecomposable material decreases over time as the stockpile ages.Labile fractions of organic carbon are decomposed and released tothe atmosphere as CO2 but over time the labile fractions probably

decrease and as a consequence we observe lower CO2 emissions asthe stockpile ages.

A similar trend was observed for N2O emissions. N2O emissionswere greatest in the youngest stockpile followed by 1e3 yo andwere least in the >3 yo biosolid stockpile. This coincided withsimilar trends of NH4 + and NO3 - concentrations in biosolids,where NO3 - concentrations were an order of magnitude greaterthan NH4 + concentrations. This suggests high rates of nitrificationin all stockpiles and that a large fraction of N2O is produced vianitrification. Given that the NO3 - pool is relatively stable it is alsolikely that N2O is at least partially also produced by denitrificationprocesses that may occur deeper in the stockpile where anaerobicconditions may occur (see above). Hence, similar to the CO2 emis-sions, it is also likely that labile organic nitrogen forms decrease asthe stockpiles age and that as a consequence we measure lowerN2O emissions from older stockpiles. Research on solid animalmanure stockpiles also indicates that stockpile structure and con-ditions are critical in influencing N2O emissions. Solid manurestockpiles provide aerobic and anaerobic conditions within closeproximity, there is a tendency for N2O emissions to increase withincreasing density of composting manure heaps and N2O emissionscan be substantial (Chadwick et al., 2011; Hansen et al., 2006;Sommer et al., 2000). However, animal manure stockpiles areusually much smaller andmanure inputs have a greater consistencycompared to biosolid stockpiles.

4.2. Relationship between environmental parameters and GHG fluxof the studied stockpiles

The emissions of GHG are often well correlated with substratemoisture and temperature because these environmental variablesgreatly influence the activity of microbes and thus ultimately theflux of GHG (Livesley et al., 2011). However, biosolid stockpilespresent a well-mixed substrate that is rich in organic material anddecomposition processes are likely to occur throughout the stock-pile. However, the environmental conditions at the surface areunlikely to represent the conditions inside the stockpile, where

R. Majumder et al. / Journal of Environmental Management 143 (2014) 34e4342

most of the biosolids are located. The moisture content and tem-perature of the biosolids at the surface undergo quite drastic sea-sonal changes whilst the conditions inside the stockpile arebuffered and change more conservatively, i.e. the temperature ofbiosolids at 50 cm was more stable as compared to at 10 cm in thetwo older stockpiles, and was in general also greater. Hence,biosolid stockpiles are likely to have their own climate that can berather independent of the external conditions and this is onereason why we did not observe strong correlations between GHGfluxes and moisture and temperature of the surface layers. Thesame probably applies to the NO3 - and NH4 + concentrations,although fluctuations in NO3 - concentrations could explain a smallpercentage of the variation of N2O fluxes in the older stockpiles,indicating that nitrification processes are responsible for a largeproportion of the flux. Hence, it will be difficult to model andpredict GHG emissions from environmental conditionsmeasured atthe surface of biosolid stockpiles. Biosolid properties, such as %Cand %TVS were also poor predictors for GHG emissions. It is sur-prising that the %C and also %TVS did not show a more lineardecline with biosolid age and that the 1e3 yo stockpile had thelowest %C and %TVS values. However, the biosolids that werecollected at that time from the anaerobic ponds could have beenrather old sludge that had more time to decompose as the anaer-obic ponds had not been de-sludged for decades.

4.3. Estimated annual GHG emissions from the studied stockpiles

Our direct measurements of GHG emissions revealed that theyoungest biosolid stockpile had the greatest emissions in kg CO2-eper Mg dry biosolids and that emissions decreased as stockpilesaged. However, because older stockpiles account for over 96% of thebiosolid material that is stored at the WTP this picture is reversedwhen total annual emissions are concerned. We calculated a totalannual GHG emission from all the stored stockpiles of 44,049 MgCO2-e and the older stockpiles accounted for 91% of these emis-sions, whilst the younger stockpiles contributed less (1e3years ¼ 5%; and <1 year ¼ 4%). However, the Australian NationalGreenhouse and Energy Reporting (NGER) guidelines only includeCH4 and N2O emissions from wastewater handling whilst CO2emissions are not considered because the process based CO2emissions are of biogenic origin and should not be included innational total emissions. Consequently, the total non-CO2 emissionsfrom biosolid stockpiles were 16,967 Mg CO2-e from all stockpiles.Based on the NGER guidelines the annual direct GHG emissions(scope 1) from the WTP in 2012/13 were about 91,350 Mg (LiChoong, Melbourne Water, personal communication). Theseemissions included CH4 and N2O emissions from wastewaterhandling, natural gas used for the WTP office, liquefied petroleumgas (LPG) for flare ignition and fugitive flare, but they did notinclude emissions from stockpiles. Hence, the GHG emissions frombiosolid stockpiles are a substantial additional GHG emission pointsource (an additional 20%) that needs to be considered in the future.Our study demonstrated that the prediction of the magnitude ofGHG emissions stockpile is not straightforward. Often GHG emis-sions can be predicted using process-based models that are drivenby easy to measure environmental and edaphic variables. The lackof relationship between the environmental variables and GHG fluxfrom stockpiles indicates that this is not possible for biosolidstockpiles. The first order decay model for CH4, which is success-fully used for landfill waste, over-predicted the annual CH4 emis-sions by orders of magnitude in our study. It is possible that theenvironmental conditions in biosolid stockpiles are less anaerobiccompared to landfill or that biosolids have greater activity ofmethanotrophic bacteria in surface layers, which leads to lower CH4emissions. Future research should consider improving model

predictions for CH4 and N2O emissions from biosolids, given thatthey are such a significant source of GHG.

Another issue for managers at wastewater treatment plants isstockpile management. At present biosolids are stored in largequantities in treatment plants before they are used and in manycases they are stored indefinitely. Given the significant GHGemissions from these biosolid stockpiles we also need futureresearch that can investigate management options to reduce theseemissions. It is possible that the stockpiling of biosolids contributesthe magnitude of emissions as stockpiles create and control theirown climate with at times very high temperatures inside thestockpile e we observed spontaneous ignition of biosolids insidesome young stockpiles at WTP in Werribee, indicating the hugetemperatures these stockpiles can generate. Hence, different stor-age options like shallower stockpiles or different shapes couldpossibly lead to different emissions. A better option may be the usebiosolids for other purposes so that they will not have to bestockpiled at all. Depending on their characteristics biosolids can beused as a soil amendment in forestry and agriculture, for producingenergy via combustion or pyrolysis or as structural fill such as fillfor road embankments. Research into new markets and technolo-gies that have the potential to enhance the value of biosolids for aparticular market is warranted.

5. Conclusions

Our study evaluated the GHG emission strength of differentlyaged biosolid stockpiles at a wastewater treatment plant by directGHG measurements and confirmed that biosolid stockpiles are asignificant GHG source. Very young biosolid stockpiles emit largeamounts of GHG, mainly in the form of N2O and the GHG emissionssource strength decreases as biosolids age and labile organic carbonand nitrogen sources decline. Methane emissions were negligiblefrom biosolid stockpiles, mainly because of aerobic conditions inthe stockpile or activity of methanotrophic bacteria in the outerlayers of the stockpile. Our study therefore confirms that biosolidstockpiles contribute significantly to the GHG emission profile ofwastewater treatment plants and that biosolid stockpile manage-ment should be a priority to avoid substantial long-term emissionsin the future.

Acknowledgements

The authors would like thank Dr Scott W Laidlaw and KevinGillett for assistance with the field based measurements. The studywas supported by funding from the Australian Research Councilgrants LE0882936, LP0883573 and DP120101735 and the Mel-bourne Water Corporation (LP0883873).

References

Ahn, H.K., Mulbry, W., White, J.W., Kondrad, S.L., 2011. Pile mixing increasesgreenhouse gas emissions during composting of dairy manure. Bioresour.Technol. 102, 2904e2909.

Amlinger, F., Peyr, S., Cuhls, C., 2008. Green house gas emissions from compostingand mechanical biological treatment. Waste Manage. Res. 26, 47e60.

AWA, 2013. Biosolids Production and End Use in Australia [Online]. Available:http://www.biosolids.com.au/bs-australia.php (accessed 5.11.13.).

Bastian, L., Yano, J., Hirai, Y., Sakai, S., 2013. Greenhouse gas emissions from biogenicwaste treatment: options and uncertainty. J. Mater. Cycles Waste Manage 15,49e60.

Bogner, J., Meadows, M., Czepiel, P., 1997. Fluxes of methane between landfillsand the atmosphere: natural and engineered controls. Soil. Use Manage 13,268e277.

Bogner, J., Pipatti, R., Hashimoto, S., Diaz, C., Mareckova, K., Diaz, L., Kjeldsen, P.,Monni, S., Faaij, A., Gao, Q.X., Zhang, T., Ahmed, M.A., Sutamihardja, R.T.M.,Gregory, R., 2008. Mitigation of global greenhouse gas emissions from waste:conclusions and strategies from the Intergovernmental Panel on Climate

R. Majumder et al. / Journal of Environmental Management 143 (2014) 34e43 43

Change (IPCC) Fourth Assessment Report. Working Group III (Mitigation).Waste Manage. Res. 26, 11e32.

Boldrin, A., Andersen, J.K., Moller, J., Christensen, T.H., Favoino, E., 2009. Compostingand compost utilization: accounting of greenhouse gases and global warmingcontributions. Waste Manage. Res. 27, 800e812.

Bright, D.A., Healey, N., 2003. Contaminant risks from biosolids land application:contemporary organic contaminant levels in digested sewage sludge from fivetreatment plants in Greater Vancouver, British Columbia. Environ. Pollut. 126,39e49.

Brown, S., Kruger, C., Subler, S., 2008. Greenhouse gas balance for composting op-erations. J. Environ. Qual. 37, 1396e1410.

Chan, Y.C., Sinha, R.K., Wang, W.J., 2011. Emission of greenhouse gases from homeaerobic composting, anaerobic digestion and vermicomposting of householdwastes in Brisbane (Australia). Waste Manage. Res. 29, 540e548.

Chadwick, D., Sommer, S., Thorman, R., Fangueiro, D., Cardenas, L., Amon, B.,Misselbrook, T., 2011. Manure management: Implications for greenhouse gasemissions. Anim. Feed Sci. Tech. 166-167, 514e531.

Cowie, A., Eckard, R., Eady, S., 2012. Greenhouse gas accounting for inventory,emissions trading and life cycle assessment in the land-based sector: a review.Crop Pasture Sci. 63, 284e296.

Czepiel, P.M., Crill, P.M., Harriss, R.C., 1993. Methane emissions from municipalwaste-water treatment processes. Environ. Sci. Technol. 27, 2472e2477.

Czepiel, P.M., Mosher, B., Crill, P.M., Harriss, R.C., 1996. Quantifying the effect ofoxidation on landfill methane emissions. J. Geophy. Res. Atmos. 101, 16721e16729.

Dalal, R.C., Wang, W.J., Robertson, G.P., Parton, W.J., 2003. Nitrous oxide emissionfrom Australian agricultural lands and mitigation options: a review. Aust. J. SoilRes. 41, 165e195.

El-Fadel, M., Massoud, M., 2000. Emissions from landfills: a methodologycomparative assessment. Environ. Technol. 21, 965e978.

Gifford, R.M., Roderick, M.L., 2003. Soil carbon stocks and bulk density: spatial orcumulative mass coordinates as a basis of expression? Glob. Chang. Biol. 9,1507e1514.

Goldsmith, C.D., Chanton, J., Abichou, T., Swan, N., Green, R., Hater, G., 2012.Methane emissions from 20 landfills across the United States using verticalradial plume mapping. J. Air Waste Manage. Assoc. 62, 183e197.

Hansen, M.N., Henriksen, K., Sommer, S.G., 2006. Observations of productionand emission of greenhouse gases and ammonia during storage of solidsseparated from pig slurry: effects of covering. Atmos. Environ. 40, 4172e4181.

Hutchinson, G.L., Mosier, A.R., 1981. Improved soil cover method for field mea-surement of nitrous-oxide fluxes. Soil Sci. Soc. Am. J. 45, 311e316.

IPCC, 2007. Climate Change 2007: the physical Science Basis. Contibution ofWorking Group I to the Fourth Assessment Report sof the IntergoverenmentalPannel on Climate Change. Cambridge University Press, Cambridge, UK.

Ishigaki, T., Yamada, M., Nagamori, M., Ono, Y., Inoue, Y., 2005. Estimation ofmethane emission fromwhole waste landfill site using correlation between fluxand ground temperature. Environ. Geol. 48, 845e853.

Livesley, S.J., Grover, S., Hutley, L.B., Jamali, H., Butterbach-Bahl, K., Fest, B.,Beringer, J., Arndt, S.K., 2011. Seasonal variation and fire effects on CH4, N2O andCO2 exchange in savanna soils of northern Australia. Aric. For. Meteorol. 151,1440e1452.

Lou, X.F., Nair, J., 2009. The impact of landfilling and composting on greenhouse gasemissions - a review. Bioresour. Technol. 100, 3792e3798.

Matson, P.A., Harriss, R.C. (Eds.), 1995. Biogenic Trace Gases: Measuring Emissionsfrom Soil and Water. Blackwell Science Ltd.

Minato, K., Kouda, Y., Yamakawa, M., Hara, S., Tamura, T., Osada, T., 2013. Deter-mination of GHG and ammonia emissions from stored dairy cattle slurry byusing a floating dynamic chamber. Anim. Sci. J. 84, 165e177.

Monteith, H.D., Sahely, H.R., MacLean, H.L., Bagley, D.M., 2005. A rational procedurefor estimation of greenhouse-gas emissions from municipal wastewater treat-ment plants. Water Environ. Res. 77, 390e403.

NGER, 2013. Technical Guidelines for the estimation of Greenhouse gas emissionsby facilities in Australia. Department of Climate change and Energy Efficiency[Online]. Available: http://www.climatechange.gov.au/climate-change/greenhouse-gas-measurement-and-reporting/company-emissions-measurement/technical/national-greenhouse-and-energy-reporting-measurement-technical-guidelines-2013 (accessed 10.11.13.).

Parameswaran, M., 1999. Urban wastewater use in plant biomass production.Resour. Conserv. Recy. 27, 39e56.

Peigne, J., Girardin, P., 2004. Environmental impacts of farm-scale compostingpractices. Water Air Soil Pollut. 153, 45e68.

Petersen, S.O., Dorno, N., Lindholst, S., Feilberg, A., Eriksen, J., 2013. Emissions ofCH4, N2O, NH3 and odorants from pig slurry during winter and summer storage.Nutri. Cycl. Agroecosyst. 95, 103e113.

Pritchard, D.L., Penney, N., McLaughlin, M.J., Rigby, H., Schwarz, K., 2010. Landapplication of sewage sludge (biosolids) in Australia: risks to the environmentand food crops. Water Sci. Technol. 62, 48e57.

Sommer, S.G., Petersen, S.O., Sogaard, H.T., 2000. Greenhouse gas emission formstored livestock slurry. J. Environ. Qual. 29, 744e751.

Stanisavljevic, N., Ubavin, D., Batinic, B., Fellner, J., Vujic, G., 2012. Methane emis-sions from landfills in Serbia and potential mitigation strategies: a case study.Waste Manage. Res. 30, 1095e1103.

Stickland, A.D., Rees, C.A., Mosse, K.P.M., Dixon, D.R., Scales, P.J., 2013. Dry stackingof wastewater treatment sludges. Water Res. 47, 3534e3542.

Wood, J.D., Gordon, R.J., Wagner-Riddle, C., Dunfield, K.E., Madani, A., 2012. Re-lationships between dairy slurry total solids, gas emissions, and surface crusts.J. Environ. Qual. 41, 694e704.

Yeomans, J.C., Bremner, J.M., 1991. Carbon and nitrogen analysis of soils by auto-mated combustion techniques. Commun. Soil Sci. Plant Anal. 22, 843e850.

Yerushalmi, L., Ashrafi, O., Haghighat, F., 2013. Reductions in greenhouse gas (GHG)generation and energy consumption in wastewater treatment plants. Water Sci.Technol. 67, 1159e1164.