Modeling of Landfill Gas From Pretreated MSW

RESEARCH ARTICLE

Modeling and simulation of landll gas production frompretreated MSW landll simulator

Rasool Bux MAHAR ()1, Abdul Razaque SAHITO2, Dongbei YUE3, Kamranullah KHAN4

1 Institute of Environmental Engineering & Management, Mehran University of Engineering & Technology Jamshoro, Sindh 76062, Pakistan2 Department of Mechanical Engineering, Mehran University of Engineering & Technology Jamshoro, Sindh 76062, Pakistan

3 School of Environment, Tsinghua University, Beijing 100084, China4 Pakistan Institute of Engineering and Applied Sciences, P.O. Nilore, Islamabad 76132, Pakistan

Higher Education Press and Springer-Verlag Berlin Heidelberg 2014

Abstract The cumulative landll gas (LFG) productionand its rate were simulated for pretreated municipal solidwaste (MSW) landll using four models namely rst orderexponential model, modied Gompertz model, singlecomponent combined growth and decay model andGaussian function. Considering the behavior of thepretreated MSW landll, a new multi component modelwas based on biochemical processes that occurring inlandlled pretreated MSW. The model was developed onthe basis of single component combined growth and decaymodel using an anaerobic landll simulator reactor whichtreats the pretreated MSW. It includes three components ofthe degradation i.e. quickly degradable, moderatelydegradable and slowly degradable. Moreover, the devel-oped model was statistically analyzed for its goodness oft. The results show that the multi components LFGproduction model is more suitable in comparison to thesimulated models and can efciently be used as a modelingtool for pretreated MSW landlls. The proposed model islikely to give assistance in sizing of LFG collectionsystem, generates speedy results at lower cost, improvescost-benet analysis and decreases LFG project risk. It alsoindicates the stabilization of the landll and helps themanagers in the reuse of the landll space. The proposedmodel is limited to aerobically pretreated MSW landlland also requires the values of delay times in LFGproductions from moderately and slowly degradablefractions of pretreated MSW.

Keywords combine growth and decay model, pretreatedmunicipal solid waste (MSW), multi component landll gas(LFG) model

1 Introduction

Landlling is one of the most commonly adoptedtechnologies for municipal solid waste (MSW) disposalas an alternative to waste burning and composting. In mostof the western countries over the past few years, MSWlandlling has been signicantly developed. These land-lls decrease the environmental consequences caused bythe unsanitary landlls and open dumps [1]. MSW hasdifferent organic and inorganic fractions. Leachate andbiogas are two pollutants which are emitting from MSWlandlls [2]. The production of landll gas (LFG) andleachate are still unavoidable disadvantages, though thelandll technology has been improved from open dumpsites to engineered sanitary landll. These environmentalconsequences of these sanitary landll sites can beaddressed by using different tools of landll managementlike mathematical modeling. Mathematical modeling ishelpful in estimation of quality and quantity of LFG andleachate. Anaerobic degradation of the organic materialresults in formation of LFG, which is the combination ofcarbon dioxide (CO2), methane (CH4) and small quantitiesof other gases. Modeling of the LFG is an exercise ofestimation of gas production its retrieval and its retrievalefciency [3].If LFG emissions are beyond control, it is signicant

hazard to the environment [4]. CH4 and CO2 are the twogases which are primarily accountable for global warming[5]. Though, CH4 concentration in the atmosphere is lowerthan that of CO2, but it is 2125 times more powerfulgreenhouse gas than CO2 [6]. Biogenic processes are themain sources of the release of the CH4 and CO2 [7]. Thesubstantial source of production of CH4 is the MSWlandlls, which accounts about 12% to 18% of yearlyworldwide anthropogenic CH4 productions [8]. Numerousattempts have been made to make the sanitary landlls

Received March 23, 2013; accepted February 24, 2014

E-mail: [email protected]

Front. Environ. Sci. Eng.DOI 10.1007/s11783-014-0685-6

stable; one of them was the pretreatment of volatile solids(VS) from mixed un-shredded MSW through natural airconvection before landlling [9,10]. Employing aerobicpretreatment (natural air convection) can signicantlydecrease the lag phase time of the LFG production by74.1% to 97.0%. This aerobic treatment eventuallyconverts VS present in the MSW into CO2 rather thaninto unrestrained LFG and extremely polluted leachate[11].The path of conversion VS to LFG is somewhat assumed

comparable to that of for anaerobic digestion process. TheVS is transformed to biogas through a sequence ofinterconnected bacteriological breakdowns. The anaerobicdigestion process broadly includes three stages that arehydrolysis, acetogenesis, and methanogenesis [12]. TheLFG production model approaches the landll as batchreactor [2]. The kinetics of the anaerobic digestion isassumed as rst order with respect to biodegradablematerial. Moreover, the accumulation of biogas in thelandll is considered as negligible. As the solid waste isplaced in the landll site, it will breakdown throughchemical and bacterial activities. Primarily, the solid wasteexperiences the process of hydrolysis in which proteins,carbohydrates and fats are decreased to soluble com-pounds. In the subsequent stage the hydrolyzed material isconverted into the organic acids, which are again break-down into acetic acid. Methanogenesis is the last stage,where acetic acid is converted to CH4 and CO2 while,hydrogen (H2) and CO2 react to produce CH4 [13].LFG model is a tool to project LFG production over the

incubation time from a certain mass of waste that has to belandlled. The model gives assistance in sizing of LFGcollection system, forecasting of LFG production poten-tials and their efcient utilization. Moreover, in response toregulatory necessities beneath the Clean Air Act (CAA),LFG production model helps in installation of LFGcollection and treatment mechanisms. The other advan-tages of the LFG models include lower cost and speedyresults of the LFG production and other related parameters,improved cost-benet analysis and decreased LFG projectrisk. The LFG production model also indicates thestabilization of the landll and helps to the managers inthe reuse of the landll space. In past, many models weredeveloped to demonstrate LFG production from sanitarylandlls including modied Gompertz model which isbased on bacteriological development curve, the modelbased on bio-kinetic characteristics [14,15] and the modelbased on environmental features [16]. A landll is verycompound mixed surrounding environment and face upseveral modeling challenges. Landlled MSW undergoesdegradation through complex biological and chemicalprocesses. At the start of landlling, VS follows aerobicdegradation, but by the time as the quantity of the wasteincreases layer by layer it follows anaerobic degradation.Thus, landlls bear a resemblance to that of an anaerobicdigester. All the models employed for prediction of the

LFG throughout the world are based on rst order decay(FOD) model, in which depletion of VS in the landlledwaste with respect to the time is taken into account [17].Moreover, all FODs have linear relation with ultimateproduction of methane per unit mass of landlled wasteand have an exponential relation with rate of degradationand time of incubation. The simplest FOD model isrepresented in Eq. 1 [18].

G m L0 e kt t1, (1)where G is LFG production in m3$year1; m is mass oflandlled waste in tons; L0 is LFG yield potential inm3$ton1 of landlled waste; t is time after waste placementin year; t1 is lag phase time and k is rst order rate constantin year-1. Another FOD model is TNO model asrepresented in Eq. 2 [19].

G 1:87 m C0 k e kt, (2)where G is LFG production in m3$year1; 1.87 is theconversion factor having dimension as m3$kg1 (biode-gradation of the one kg of organic carbon that is landlledproduces 1.87 Nm3 of LFG); is the formation factor(certain fraction landlled waste that is converted intoLFG); m is mass of landlled waste in ton; C0 is theamount of degradable organic carbon in kg$ton1 oflandlled waste; t is time after waste placement in year andk is rst order rate constant in year1. Both the modelsstated above are based on one component degradation ofMSW in the landll and lack the multi componentsapproach, which introduces gross oversimplication.Moreover, both the models have only one exponentiallydecreasing LFG production curve.Afvalzorg model is a three component model for

estimation of LFG production and it seems to be similarto TNO model as stated in Eq. 2, with some modicationand is employed for the mixed solid waste having eightdifferent waste types. This model also considers threefractions of waste i.e. rapid, moderate and slow degradableas given in Eq. 3 [19].

G 1:87 mX3

i1 C0,i ki e kit, (3)Where i represent the degradation fraction of the landlledwaste and have values from 1 to 3 for rapid, moderate andslow waste fractions. In this model the LFG production hasthree curves, which decreases exponentially and does nothave any rising limb, it is not suitable for LFG productionrates.All the LFG models reported in Eqs.(1-3) are for

untreated MSW landlls, while a little literature isavailable on the LFG models for pretreated MSW landlls.The objective of this study was to simulate the cumulativeLFG production and LFG production rate for pretreatedMSW landll by means of four models i.e. rst orderexponential model, modied Gompertz model, single

2 Front. Environ. Sci. Eng.

component combined growth and decay model andGaussian function and to propose a new mathematicalmodel for the prediction of methane production rate. Theresults of proposed model were compared with the abovemodels and two components model available in literature.For this purpose, anaerobic landll simulator reactor(ALSR) was developed with controlled temperature withleachate recycling mechanism. ALSR was fed withpretreated MSW and LFG production was measured. TheMSW was pretreated aerobically by natural convection ofair as described in earlier study [20]. Pretreatment of theMSW decreases some quantity of degradable material andthus remaining residue was landlled.

2 Methodology

2.1 Composition of MSW

The sample of MSW was picked from Beijings Beishen-shu landll site. Its nature was mixed and comprising 60%of VS. The collected sample was segregated manually andits composition was determined as shown in Fig. 1.

2.2 Characteristics of pretreated MSW

After ve months of aerobic treatment by naturalconvection in aerobic pretreatment simulator (APS), ninepretreated MSW samples were taken from three randompoints of the APS. From each point, three samples weretaken from top, mid and bottom of the height of the APS.The nine pretreated MSW samples were then dried in air

atmospheric temperature and then chopped to have a sizeof less than 20 mm in diameter and were made uniform interms of composition, which made sure that the sampletaken for the analysis and to be landlled was representa-tive.The moisture content and VS in pretreated MSW sample

were determined according to standard method [21]. Thepercentages of elemental hydrogen, carbon, and nitrogenwere determined by ash combustion method in elementalanalyzer (EAI, USA). The gravimetric method wasemployed to determine the percentage lignocellulosematerial [22]. Bulk density was measured by llingpretreated MSW sample in cylinder and weighing cylinderbefore and after sample lling. Bulk density was calculatedby dividing mass of pretreated MSW sample to volume itoccupied. The pH of pretreated MSW was measuredthrough a hydrogen ion sensitive electrode by preparing itsslurry with the distilled water. SCOD was measured forltered sample according to standard method [21]. Thecharacteristics of pretreated MSW are given in Table 1.

2.3 Anaerobic landll simulation reactor

The laboratory scale anaerobic landll simulation reactor(ALSR) was established with an aim to study the behaviorof pretreated MSW in the landll. The sectional view ofALSR assembly is shown in Fig. 2. It comprises of twoshells. The inner shell was air tight having capacity ofabout 25 liters (0.3 m diameter and 0.35 m height). Thebottom of the inner shell was lled with a 25 mm layer ofgravel followed by a membrane. The leachate wascollected through leachate collection hopper. The top ofthe inner shell was provided with water divertingmechanism containing 25% NaCl solution for volumetricmeasurement of LFG and the rain water sump along withmembrane for rain water simulation. The temperature ofthe inner shell was controlled by the water; lled in theouter shell at 401C by using heating element, which islinked with temperature controller as shown in Fig. 2. Byquartering about 10 kg sample, the ALSR was lled withpretreated MSW and of the identical volume of water wasadded followed by properly mixing. It was operated formore than one year and the LFG production during thisperiod was measured. The tap water was also added to theALSR as a rainfall simulation. The tap water was added atthe rate of 3.5 mm fortnightly.

2.4 LFG production simulation

Cumulative LFG production was simulated using two

Fig. 1 Composition of untreated MSW of Beishenshu landllBeijing, China (% TS)

Table 1 Characteristics of pretreated MSW by natural convection of air

parameters moisturecontent/ %

VS/ (%TS)

carbon/ (%TS)

hydrogen/ (%TS)

nitrogen/(%TS)

lignocellulose/(%TS)

bulk density/ (kg$m3)

pH SCOD/ (mg$L1)

pretreated waste 211 26.390.54 150.62 1.610.2 0.760.21 180.7 5203 5.70.1 177315

Rasool Bux MAHAR et al. Modeling and simulation of landll gas production from pretreated MSW landll simulator 3

models i.e. rst order exponential model and modiedGompertz model. Because of the bacteriological involve-ment in the anaerobic digestion process, the rst orderexponential model has been commonly employed tosimulate the anaerobic biodegradation in the landlls[23,24]. First order exponential model for cumulative LFGproduction from pretreated MSW is presented in Eq. 4.

LFG LFGu1 e kt, (4)where, LFG is the cumulative LFG production in m3$ton1

MSW(DM), t is the time on day over the digestion period.LFGu is the LFG production potential in m

3$ton1 MSW(DM) and k is the rst order kinetic constant day1. TheLFGu and kwere estimated analytically through nonlinearregression using least square method. Modied Gompertzmodel was another commonly used model for simulationof cumulative LFG production [25,26]. The modiedGompertz model for cumulative LFG production frompretreated MSW is given in Eq. 5.

LFG LFGu:exp expRm:e

Pl t 1

, (5)

Where, LFG is the cumulative LFG production in m3$ton1

MSW(DM), t is the time in day over the digestion period,LFGu is the LFG production potential in m

3$ton1 MSW(DM), Rm is the LFG production rate in m

3$ton1 MSW(DM)$day1, is the lag-phase time in day, and e is theexponential of 1. The LFGu,Rm and were estimatedanalytically through nonlinear regression using leastsquare method.In addition to the cumulative LFG production, LFG

production rates of pretreated MSW were also simulatedusing combined growth and decay model as dened byZachrarof & Butler [3] and Gaussian function. Thecombined growth and decay model for cumulative LFGproduction from pretreated MSW is given in Eq. 6.

LFGt Ate kt, (6)Where LFG (t) is the LFG production rate in m3$ton1

MSW(DM)$day1 at time t in day, t is the time over thedigestion period, A is the amplitude in m3$ton1 MSW(DM)$day2) and k is reaction rate constant in day1. The Aand k were estimated analytically through nonlinearregression using least square method.

Fig. 2 Sectional view of ALSR


Gaussian model was another model used for simulationof LFG production rates from pretreated MSW [24], whichassumes that the LFG production rate follow the normaldistribution. Gaussian model can be employed to simulateLFG production rates including climbing and descendentmember and is presented in Eq. 7.

LFGt ae 0:5t t0b

2h i, (7)

Where, LFG(t) is the LFG production rate in m3$ton1

MSW(DM)$day1 at time t in day, t is the time over thedigestion period, a is the ultimate LFG production rate inm3$ton1 MSW(DM)$day1 and b is constant in day and t0is the time in day where the peak (maximal) LFGproduction rate occurred. The parameters a, b and t0were estimated analytically through nonlinear regressionusing least square method.

2.5 Development of LFG production model

The mathematical representation of the LFG productionrate follows one of the three approaches. In the rstapproach, the LFG production rate is represented as single/multiple empirical functions of a general kinetic factorwhich mostly appears in the literature [2729]. The secondapproach includes the representation of the LFG produc-tion rate in a complex sum of mathematical functions,which are based on the physical, chemical and biologicalprocesses of biodegradation and variables includingmoisture content, operating temperature, elemental com-position, volatilization, dilution, precipitation, oxidation,evaporation, reduction, adsorption, absorption, ltration,complexion and neutralization. To determine these char-acteristics inside the landll, a signicant amount ofdetailed analysis is required. The complex mathematicalmodel available in literature for landlled waste isHalvadakis model [14], which is rst order model and isbased on the consecutive biological growth. The thirdapproach includes the models that consider LFG produc-tion rate in digits and are called numerical models. Thenumerical models are robust in estimation of production ofLFG, if all the phenomena takes place within thedegradation of landlled waste are known [18]. The mostconcerned variable in landll modeling is time as theanaerobic degradation of the MSW extremely depends ontime. If the combined growth and decay rates are dened asfunctions of time, then there will be dual advantage. Theforemost advantage is related to the digestion of waste as itis more readily understood and on second unrepresentativeuse of biomass concentration is avoided. This will not onlysimplify the structure of the model, but it will also improvemodels functionality for which it is developed. Theproposed model was constructed as multiple empiricalfunctions having multiple kinetic factors and was based onthe combined growth and decay function as dened byZachrarof and Butler [3] in the following form:

Rt Ate kt, (8)Where, R(t), is the rate of reaction at time t in kg$year1; Ais an amplitude term in kg$year2 and k, is rate constant inyear1. The rate of reaction is an illustration of the outcomeseen in bacteriological development and it requires onlytwo values that are an amplitude and rate constant. Asshown in Fig. 3, the curve produced by using Eq. 8increases abruptly, which represents fast growth and thenafter reaches at the highest point it follows an exponentialdecline to zero. This behavior is result of the hydrolysisstage of the anaerobic digestion. Furthermore, Zachrarof &Butler [3] suggested that the algorithm of the model in Eq.8 can be enhanced by incorporating time variant uxes.The MSW is heterogeneous in nature. It comprises ofvarious types of biodegradable components and hasdifferent degradation rates. Thus in this study we proposethe multi-components model as given in Eq. 9.

LFGt Xni1

Xm 1j0

Aj1ti tje kj1ti tj, (9)

where LFG (t) is the LFG production at time t, in m3$ton1

MSW(DM)$day1; A is the amplitude in m3$ton1MSW(DM)$day2; k is the LFG reaction rate constant in day1; nis total number of days; m is number of biodegradablecomponents of heterogeneous pretreated MSW. Moreover,i j and tj is the delayed time, which is dened as periodbetween the beginning times to the end of biodegradablecomponents.

Analyzing the LFG production data, obtained fromALSR, it was observed that the pretreated MSW may havethree types of the biodegradable components i.e. m = 3;and are considered as quickly degradable, moderatelydegradable and slowly degradable components. Consider-ing three degradable components Eq. 9 was expanded as

Fig. 3 Combined growth and decay function (A = 1.7 kg$day2

and k = 0.07 year1)


Eqs.(10 12):

LFGt A1te k1t,at j 0 and t

ied Gompertz model showed better R2 of 0.995 than rstorder exponential model of 0.972 as shown in Fig. 5 (a andb). A similar trend was observed in the values of RMSEthat the modied Gompertz model evidenced better t andhad lower RMSE of only 1.477 m3$ton1 MSW(DM) incomparison to the RMSE of 10.52 m3$ton1 MSW(DM)for rst order exponential model.

Considering LFG production rate simulation, Gaussianfunction showed better R2 of 0.954 than single componentcombined growth and decay model of 0.912 as depicted inFig. 6 (a and b). A similar fashion was observed in thevalues of RMSE that the Gaussian function demonstratedbetter t and had lower RMSE of only 0.185m3$ton1MSW(DM)$day1 in comparison to the RMSE of0.256 m3$ton1MSW(DM)$day1 for single componentcombined growth and decay model.The single component combined growth and decay

model and Gaussian function were tted to experimentaldata and observed that Gaussian function is better t thanformer one. This is because of considering only one type ofmaterial in the present study (pretreated MSW). On thecontrary, by considering the multi components of degrada-tion of pretreated MSW (quickly degradable, moderatelydegradable and slowly degradable), it was found that thenewly developed multi components combined growth anddecay model had become more accurate than the Gaussianfunction.The multi components combined growth and decay LFG

production model along with experimental data wasplotted as shown in Fig. 7. The delayed time of LFGproduction was taken from the experimental data trend ofLFG production. For moderately and slowly degradablematters, delay time t1 and t2 were 15 and 50 daysrespectively. As the quickly and moderately degradablecomponents were used up till LFG production rate reached

Fig. 5 Cumulative LFG production (a) rst order exponentialmodel and (b) modied Gompertz model

Fig. 6 LFG production rate (a) combined growth and decaymodel and (b) Gaussian function


to peak then exponential decay followed to back zero. Incomparison to single component combined growth anddecay model and Gaussian function, the multi componentLFG production model demonstrates better t with R2

value of 0.969 and RSME of only 0.151 m3$ton1MSW(DM)$day1. All these three models were comparedstatistically and it was observed that the multi-componentsmodel has better correlation with the experimental data andhas 40% and 20% less RMSE as compared to the singlecomponent model and Gaussian function respectively.In recent past, Gioannis et al. [23] has developed a two

stage model, which is based on the rst-order exponentialmodel. The model was capable to predict LFG productionrate for aerobically mechanical biological treated waste(MBTW) and the development of the LFG production ratewas estimated for the three different ratios of waste. As pera two stage model, as the time passes the LFG productionrate rises and is comparative to the volume of LFG that hasbeen previously formed. Throughout the second stage, theLFG production rate is comparative to the remainingquantity of biodegradable material and it declines as thetime passes. The values of the R2 were in the range of 0.81to 0.90 for the rst and second stages respectively. Incomparison to two stages model, based on rst-order

exponential model developed by Gioannis et al. [23], multicomponent model proposed in this study has higher valueof R2 (0.969), thus later predicts more accurate LFGproduction rates over earlier.As the LFG production rates for each type of the

degradable fraction of MSW is different, so the LFGproduction rate constants will also be different. The LFGproduction rate constants and amplitudes for singlecomponent combined growth and decay model and formulti component LFG production model parameters wereestimated by using experimental data through MATLABprogram and are given in the Table 2. The proposed modelhas some limitations that it is applicable to aerobically(natural convection of air) pretreated MSW landll and italso requires the values of delay times in LFG productionsfrom moderately and slowly degradable fractions ofpretreated MSW.

4 Conclusions

On the subject of cumulative LFG production simulation,modied Gompertz model depicted better t in comparisonto the rst order exponential model. On the other hand,regarding the LFG production rate, the newly developedmulti-components mathematical model for LFG produc-tion, from anaerobic landll simulator reactor, based onbiochemical processes is more accurate in comparison tosingle component combined growth and decay model andGaussian function. All these three models were comparedstatistically and it was observed that multi-componentsmodel has better t with experimental data and has 40%less RMSE as compare to the single component model.Correspondingly, multi-components model has about 20%less RMSE as compare to the Gaussian function. Thus, it isconcluded that the multi component LFG productionmodel provides better t in comparison to the simulatedmodels and was well described by a three components.Additionally, un-shredded aerobically pretreated MSWincludes three components of the degradation of thepretreated MSW i.e. quickly degradable, moderatelydegradable and slowly degradable. The proposed modelcan efciently be used as modeling tool for pretreatedMSW landlls.

Fig. 7 Multi component combined growth and decay LFGproduction model versus experimental data

Table 2 Model parameters and statistical analysis of LFG production models of pretreated MSW

combined growthand decay model Eq.

amplitude/(m3$ton1MSW(DM)$day2) rate constant/(day1)R2

RMSE /(m3$ton1MSW(DM)$day1)

A1 A2 A3 k1 k2 k3

single component (9) 0.2066 0.0296 0.912 0.255

multi components (10) 0.3414 0.0699 0.969 0.151

(11) 0.1400 0.1625 0.0016 0.0026

(12) 0.1035 0.0538 0.1223 0.0150 0.0150 0.0165


Acknowledgements The authors wish to acknowledge Mehran Universityof Engineering & Technology Jamshoro, Sindh, Pakistan and HigherEducation commission Pakistan for their nancial support and TsinghuaUniversity, Beijing, China technical support to carry out this research work.

References

1. Gourc J P, Staub M J, Conte M. Decoupling MSW settlement into

mechanical and biochemical processes modelling and validation

on large-scale setups. Waste Management (New York N.Y.), 2010,

30(89): 15561568

2. Manna L, Zanetti M C, Genon G. Modeling biogas production at

landll site. Resources, Conservation and Recycling, 1999, 26(1):

114

3. Zacharof A I, Butler A P. Stochastic modelling of landll processes

incorporating waste heterogeneity and data uncertainty. Waste

Management (New York N.Y.), 2004a, 24(3): 241250

4. Pohland F G. Landll bioreactors: fundamentals and practice. In

International Trends in Water Environment Management, Japan

Society on Water Environment, Tokyo, Japan, 1996, 95110

5. Xiaoli C, Ziyang L, Shimaoka T, Nakayama H, Ying Z, Xiaoyan C.

Characteristics of environmental factors and their effects on CH4and CO2 emissions from a closed landll: an ecological case study

of Shanghai. Waste Management (New York), 2010, 30(3):

446451

6. Talyan V, Dahiya R P, Anand S, Sreekrishnan T R. Quantication of

methane emission from municipal solid waste disposal in Delhi.

Resources, Conservation and Recycling, 2007, 50(3): 240259

7. Bingemer H G, Crutzen P J. The production of methane from solid

wastes. Journal of Geophysical Research: Atmospheres, 1987, 92

(D2): 21812187

8. Bogner J, Pipatti R, Hashimoto S, Diaz C, Mareckova K, Diaz L.

Mitigation of global greenhouse gas emissions from waste:

conclusions and strategies from the Intergovernmental Panel on

Climate Change (IPCC) 4th Assessment Report. Working Group III

(Mitigation). Waste Manage Research, 2008, 26(1): 1132

9. Mahar R B, Liu J, Yue D, Nie Y. Biodegradation of organic matters

from mixed unshredded municipal solid waste through air

convection before landlling. Journal of Air Waste Manage

Association, 2007, 57(1): 3946

10. Mahar R B, Liu J, Li H, Nie Y. Bio-pretreatment of municipal solid

waste prior to landlling and its kinetics. Biodegradation, 2009, 20

(3): 319330

11. Zhang Y, Yue D, Nie Y. Greenhouse gas emissions from two-stage

landlling of municipal solid waste. Atmospheric Environment,

2012, 55: 139143

12. Gerardi MH. TheMicrobiology of Anaerobic Digesters. JohnWiley

& Sons, Inc. ISBN 0-471-20693-8, 2003

13. Zacharof A I, Butler A P. Stochastic modelling of landll leachate

and biogas production incorporating waste heterogeneity. Model

formulation and uncertainty analysis. Waste Management (New

York N.Y.), 2004b, 24(5): 453462

14. El-Fadel M, Findikakis A N, Leckie J O. A numerical model for

methane production on managed sanitary landlls. Waste Manage-

ment and Research, 1989, 7(1): 3142

15. Findikakis A N, Papelis C, Halvadakis C P, Leckie J O. Modeling

gas production in managed sanitary landlls. Waste Management

and Research, 1988, 6(2): 115123

16. Gurijala K R, Sa P, Robinson J A. Statistical modeling of methane

production from landll samples. Applied and Environmental

Microbiology, 1997, 63(10): 37973803

17. Ozakaya B, Demir A, Bilgili M. Neural network prediction model

for the methane fraction in biogas from eld-scale landll

bioreactors. Environmental Modelling and Software, 2007, 22(6):

815822

18. Kamalan H, Sabour M, Shariatmad N. A review on available landll

gas models. Journal of Environmental Science and Technology,

2011, 4(2): 7992

19. Scharff H, Jacobs J. Applying guidance for methane emission

estimation for landlls. Waste Management (New York N.Y.), 2006,

26(4): 417429

20. Mahar R B, Liu J, Li H, Nie Y. Landlling of pretreated municipal

solid waste by natural convection of air and its effects. Journal of

Environmental Science and Health, Part A: Toxic/Hazardous

Substances and Environmental Engineering, 2007, 42(3): 351359

21. APHA. Standard methods for the examination of water and

wastewater. 20th ed. Washington, D C: American Public Health

Association, 1998

22. Method for Determination of Crude Fiber in Feedstuff (GB/T6434).

Standard Method of Peoples Republic of China. Beijing, China,

1994

23. Gioannis G D, Muntoni A, Cappai G, Milia S. Landll gas

generation after mechanical biological treatment of municipal solid

waste. Estimation of gas generation rate constants. Waste Manage-

ment (New York, N.Y.), 2009, 29(3): 10261034

24. Lo H M, Kurniawan T A, Sillanp M E, Pai T Y, Chiang C F,

ChaoK P. Modeling biogas production from organic fraction of

MSW co-digested with MSWI ashes in anaerobic bioreactors.

Bioresource Technology, 2010, 101(16): 63296335

25. Mali Sandip T, Khare Kanchan C, Biradar Ashok H. Enhancement

of methane production and bio-stabilisation of municipal solid waste

in anaerobic bioreactor landll. Bioresource Technology, 2012, 110:

1017

26. Zhu B, Gikas P, Zhang R, Lord J, Jenkins B, Li X. Characteristics

and biogas production potential of municipal solid wastes pretreated

with a rotary drum reactor. Bioresource Technology, 2009, 100(3):

11221129

27. Findikakis A N, Leckie J O. Numerical simulation of gas ow in

sanitary landlls. Journal of the Environmental Engineering

Division, 1979, 105(5): 927945

28. Gardner N, Probert S D. Forecasting Landll-Gas Yields. England:

Science Publishers Ltd, 1993, 131163

29. Hartz K E, Ham R K. Gas generation rates of landll samples.

Conservation and Recycling, 1982, 5(23): 133147

30. Heyer K U, Stegmann R. Leachate management: leachate genera-

tion, collection, treatment and costs. 2001, http://www.ifashamburg.

de/pdf/leachate.pdf


Outline placeholderbmkcit1bmkcit2bmkcit3bmkcit4bmkcit5bmkcit6bmkcit7bmkcit8bmkcit9bmkcit10bmkcit11bmkcit12bmkcit13bmkcit14bmkcit15bmkcit16bmkcit17bmkcit18bmkcit19bmkcit20bmkcit21bmkcit22bmkcit23bmkcit24bmkcit25bmkcit26bmkcit27bmkcit28bmkcit29bmkcit30

Modeling of Landfill Gas From Pretreated MSW

Documents

Transcript of Modeling of Landfill Gas From Pretreated MSW