global methane emissions from pit latrines - princeton · global methane emissions from pit...

Global Methane Emissions from Pit LatrinesMatthew C Reiddaggerperp Kaiyu Guansect Fabian Wagner∥ and Denise L MauzeralldaggerDagger

daggerDepartment of Civil and Environmental Engineering and DaggerWoodrow Wilson School of Public and International Affairs PrincetonUniversity Princeton New Jersey 08544 United StatessectDepartment of Environmental Earth System Science Stanford University Stanford California 94305 United States∥International Institute for Applied Systems Analysis Laxenburg Austria

S Supporting Information

ABSTRACT Pit latrines are an important form of decentral-ized wastewater management providing hygienic and low-costsanitation for approximately one-quarter of the globalpopulation Latrines are also major sources of the greenhousegas methane (CH4) from the anaerobic decomposition oforganic matter in pits In this study we develop a spatiallyexplicit approach to account for local hydrological control overthe anaerobic condition of latrines and use this analysis toderive a set of country-specific emissions factors and toestimate global pit latrine CH4 emissions Between 2000 and2015 we project global emissions to fall from 52 to 38 Tg yminus1or from sim2 to sim1 of global anthropogenic CH4 emissionsdue largely to urbanization in China Two and a half billionpeople still lack improved sanitation services however and progress toward universal access to improved sanitation will likelydrive future growth in pit latrine emissions We discuss modeling results in the context of sustainable water sanitation andhygiene development and consider appropriate technologies to ensure hygienic sanitation while limiting CH4 emissions Weshow that low-CH4 on-site alternatives like composting toilets may be price competitive with other CH4 mitigation measures inorganic waste sectors with marginal abatement costs ranging from 57 to 944 $ton carbon dioxide equivalents (CO2e) in Africaand 46 to 97 $ton CO2e in Asia

INTRODUCTION

Methane (CH4) is produced in wastewater streams by theanaerobic decomposition of organic matter There is activeinterest in reducing anthropogenic CH4 emissions because ofits role as a greenhouse gas (GHG) and precursor totropospheric ozone1 and identifying low-cost mitigationstrategies is an international priority23 While most analysesof CH4 emissions and mitigation opportunities from waste-water have focused on centralized treatment plants45 it hasbecome increasingly clear that on-site wastewater treatmenttechnologies like septic systems and pit latrines are alsoimportant though poorly quantified sources of CH4

6minus8 Pitlatrines alone which are concentrated in rural areas ofdeveloping and middle income countries have been estimatedto emit sim14 Tg CH4 yminus16 or gt4 of global anthropogenicemissions7 The mitigation measures for wastewater CH4 thatare discussed in the literature like upgrading from primary tosecondarytertiary treatment9 are not applicable to on-sitesystems so there is a need to revisit CH4 emissions fromdecentralized wastewater sources and reassess the appropriateactions for emissions reductionsPit latrines are utilized by 177 billion people10 and are low

cost require little maintenance and use little to no water Theyare an essential component of public health campaigns to

provide adequate sanitation for the 25 billion people whocurrently lack improved sanitation services11 which contributesto the more than 800000 deaths annually from poor watersanitation and hygiene12 As the United Nations (UN)sustainable development goals near finalization with a proposedtarget of universal access to adequate sanitation by 203013 theinternational development community is poised to makedecisions that will significantly impact global wastewatermanagement and concomitant CH4 emissions for a gen-eration Given the cross-cutting importance of environmentalsustainability to the post-2015 development agenda it is vitalfor policymakers to have a comprehensive understanding ofGHG emissions from on-site sanitation systems and to beaware of appropriate mitigation technologies and their costsThe implications of sanitation development on global CH4

emissions and how impacts will vary geographically have notbeen discussed in detail in previous research Pit latrine CH4

emissions are controlled in part by local hydrology sincemethanogenesis is contingent on anaerobic conditions that

Received April 3 2014Revised June 23 2014Accepted July 7 2014Published July 7 2014

Article

pubsacsorgest

copy 2014 American Chemical Society 8727 dxdoiorg101021es501549h | Environ Sci Technol 2014 48 8727minus8734

occur when organic waste in pits is submerged beneath thewater table14 Spatial variation in groundwater level and itseffect on biogeochemical decomposition pathways in pitlatrines are not captured by the coarse calculations underlyingpit latrine emissions in the current generation of GHGinventories In order to characterize geographic variations inhydrological control over pit latrine emissions we develop thefirst spatially explicit approach to estimating pit latrine CH4emissions and use the analysis to create a discussion of thelinkages between pit latrines sanitation development andglobal CH4 emissionsOur approach uses spatial analyses of population urban-

ization15 and groundwater level16 in 21 developing and middleincome nations to develop a set of emissions factors (EFs)14

that reflect local hydrology These spatial data sets arecombined with country-level health and sanitation surveys todetermine rural and urban populations using pit latrines and toquantify the resulting CH4 emissions We use this analyticalframework to characterize variations in latrine CH4 emissionswithin and between countries to predict changes in globalemissions between 2000 and 2015 and to evaluate the costs ofmitigation measures like composting toilets that can providelow-CH4 on-site alternatives to pit latrines

METHODS

Spatial CH4 Emissions Model Pit latrine CH4 emissionswere estimated by integrating EFs from the 2006 IPCCGuidelines for National Greenhouse Gas Inventories14 withcountry-level latrine utilization data and a high-resolutiongeospatial analysis of population urbanization and water tabledepth in 21 countries in Asia Africa and Latin AmericaCountries were selected to represent a range of geophysicalsettings as well as to capture most global pit latrine usersEmissions were calculated for grid cells at 30 arc-s (sim1 km atthe equator) grid spacing and summed for each country

sum= middot middot middotP TCH emissions EF BODi

i i i i4(1)

P and T are the total population and the fraction of thepopulation using pit latrines respectively in grid cell i T was afunction of the urban or rural classification of the cell BOD isthe country- or region-specific per capita production ofbiochemical oxygen demand (BOD) [kg BOD personminus1 yminus1]taken from the IPCC guidelines (Table 1) EF is the CH4emissions factor [kg CH4 kg

minus1 BOD]

Pit Latrine Prevalence Latrine utilization ratios (T) inurban and rural areas were determined from country-levelhealth and sanitation surveys compiled by the WHOUNICEFJoint Monitoring Program (JMP) for Water Supply andSanitation17 We aggregated data from 1998minus2003 todetermine a mean and standard error of latrine prevalence ca2000 (Table S1 Supporting Information) We followed theclassification scheme from Graham and Polizzotto10 andconsidered the following categories to be pit latrines toiletsthat flush to pits ventilated improved pit (VIP) latrines pitlatrines with concrete slabs traditional latrines and pit latrineswithout slabopen pit Hanging latrines and bucket latrineswere not included since the ultimate disposal location may notbe a pit We do not separate urban populations into ldquourban highincomerdquo and ldquourban low incomerdquo groups as is done in theIPCC guidelines because JMP surveys only include data on anurbanrural basis

Selection of Emissions Factors EFs were determined from

= middotBEF MCFo (2)

Bo is the maximum CH4 producing capacity for domesticwastewater and the default value of 06 kg CH4 kg

minus1 BOD wasused due to the lack of country-specific values14 MCF is themethane correction factor and depends on the type oftreatment andor discharge method ranging from 0 for afully aerobic process to 1 for a fully anaerobic process MCFswere taken from the IPCC methodology and are listed in Table1 The MCF selected for each grid cell depended on the latrinedepth relative to the groundwater level since organic mattersubmerged beneath the water table will decompose anaerobi-cally while dry decomposition will be largely aerobic TheseMCFs will yield conservative emissions estimates that mayunderestimate the true CH4 emissions since they assume thatdecomposition in pits in dry climates will be mostly aerobic Inreality pits with many users those with flush water use orthose in impermeable soils will retain water and remain at leastpartially anaerobic even in dry climates The simplifiedgrouping of pit latrines into ldquowetbelow the water tablerdquo andldquodryabove the water tablerdquo was necessary because detailed dataon latrine water use number of users and water retention wasnot available from JMP country files or other data sourcesA global groundwater level model16 was used to determine

pit depths relative to the groundwater level The modelrepresents a mean annual water table depth and does notaccount for seasonal fluctuations or long-term variations ingroundwater level We assume a globally uniform pit latrinedepth of 25 plusmn 05 m which spans the 2minus3 m depth rangetypical of pit latrines1819 It is good practice for the bottom oflatrines to be above the water table to avoid groundwatercontamination though this recommendation is often dis-regarded in practice102021 P and EF in rural grid cells wereused to determine population-weighted average rural EFs foreach country

=sum middot

sum=

=

P

PEF

EFin

i i

in

i

1

1 (3)

Population and Urbanization The model calculation in2000 used a global gridded population data set for the year2000 and an urban extents mask ca 1995 in 30 arc-sresolution15 2015 population projections were only availablein a coarser 25 arc-min resolution22 so a linear interpolationwas used to generate a 30 arc-s mesh for the 2015 population

Table 1 Emissions Model Parameters14

methane correction factors (MCF) for decentralized treatment systems

disposalpathway description MCF range

pit latrine water table below latrine 3minus5 people 01 005minus015pit latrine water table higher than latrine 07 07minus10septic system half of BOD settles in anaerobic tank 05 05

biochemical oxygen demand (BOD)

regioncountry BOD [kg personminus1 yminus1] range

Africa 1351 1278minus1643Asia Latin America 1460 1278minus1643Turkey 1387 986minus1825India 1241 986minus1497Brazil 1825 1643minus2008

Environmental Science amp Technology Article

dxdoiorg101021es501549h | Environ Sci Technol 2014 48 8727minus87348728

figures which were then adjusted to match 2015 UNpopulation figures23 by multiplying each grid cell by acountry-specific scaling factor An urban extents mask for2015 was not available so the urbanrural classification wasperformed by manually tuning the population threshold forurban areas so that the urbanrural population ratio matchedUN figures Using the water table and urbanrural determi-nation grid cells were grouped into one of four types (1)urban-shallow water table (2) urban-deep water table (3)rural-shallow water table or (4) rural-deep water table Eachgrouping yielded a unique combination of EF and T for eachgrid cell which were used along with P and country-specificBOD to calculate the pit latrine emissions for each cellfollowing eq 12015 Projection A linear regression analysis of country-level

survey data from 1998 through the latest available datatypically 2010 or 2011 was used to predict latrine usage in2015 (see Figure S1 for representative data) Unambiguouspositive or negative trends were sometimes not clear In orderto filter the data for statistically significant changes in latrineutilization the projection was only used if the slope of theregression was significantly different than 0 at a 90 confidencelevel That is the projections were only used when there was90 confidence that a trend existed in latrine utilization andwhen a positive or negative trend could not be extracted fromthe noise the 2015 latrine utilization was assumed to be themean of the two most recent utilization ratios The standarderror determined for latrine usage ratios in 2000 was applied tothe 2015 projections to account for potential deviations fromthe linear trend which may occur as rural areas approach latrinesaturation Table S1 (Supporting Information) compares thelatrine utilization ratios determined here to estimates in otherrecent studies1024 The same groundwater level model was usedfor 2000 and 2015Model Uncertainties Uncertainty intervals of CH4

emissions were determined by propagating uncertainties inthe model parameters T MCF BOD and latrine depththrough the emissions model Uncertainties in MCF and BOD(Table 1) are based on expert judgment and represent therange of values for domestic wastewater Uncertainties in Trepresent plusmn2 se of latrine usage estimates from JMP countryreports and uncertainties in latrine depth represent a depthrange of 2minus3 m No specific confidence levels for theuncertainty ranges are reported since our input datauncertainties are partially based on expert judgment Anadditional sensitivity analysis into the influence of pit depthon CH4 emissions (Figure S2 Supporting Information) showedthat broadening the range of latrine depth from 1 to 5 mcontributed an additional 10minus20 uncertainty beyond thatdescribed by the uncertainty range of 2minus3 mMarginal Abatement Costs Marginal abatement costs

(MACs) were calculated as the additional cost of a CH4mitigation technology beyond the cost of a simple pit latrineper ton of carbon dioxide equivalents (CO2e) averted

=minus

middot middot middotSMAC

TACH TACH

BOD EF GWPabate pit latrine

(4)

TACH is the total annual cost per household The subscriptabate refers to the abatement technology and pit latrine to asimple pit latrine Regional BOD values are included in Table 1and Table 2 lists the mean regional household size (S) andrepresentative regional EF We use the global warming

potential for CH4 of 21 from the IPCC Second AssessmentReport25 Composting toilets26 and household biogasdigesters27 were considered as abatement technologies andthis approach assumes zero CH4 emissions from eithertreatment system TACH was computed using

= ++ minus

middot +⎡⎣⎢

⎤⎦⎥

r rr

C CTACH(1 )

(1 ) 1

L

L CAP OampM(5)

The capital recovery factor18 is in brackets CCAP is the initialcapital cost of the investment and COampM is the annualoperating and maintenance (OampM) cost per household Theinterest rate (r) was 4 the investment lifetime (L) was 20years and S was used to scale from per capita to householdcosts Capital cost estimates were collected from technicaldocuments and case studies of on-site sanitation developmentin Africa and Asia1928minus32 and annual OampM costs wereassumed to be 5 of capital costs for simple pit latrines and10 of capital costs for more maintenance-intensive compost-ing toilets and biogas33

RESULTS AND DISCUSSIONDeterminants of Pit Latrine CH4 Emissions A schematic

illustrating the inputs to the geospatial model and the estimatedCH4 emissions for the case of China is shown in Figure 1Spatial distributions of population urbanization and watertable depth are combined with pit latrine utilization ratios toevaluate the population using either ldquowetrdquo or ldquodryrdquo pit latrinesThese values are then used with parameters for per-capita BODproduction and ldquowetrdquo and ldquodryrdquo pit latrine emissions factors tocreate spatial emissions maps (also see Figure 2 for maps ofEast Africa and Vietnam)Population-weighted average rural emissions factors (EF)

were developed for each country to evaluate regional variationsin pit latrine CH4 emissions (Figure 3) We focus on rural gridcells because those areas are most likely to be served by pitlatrines Countries with greater population density in areas withshallow water tables have higher EF since a greater share ofhuman excreta is disposed of in wet anaerobic latrines Thisanalysis revealed geographic trends in the colocation ofpopulation and shallow water tables with countries in EastSoutheast and South Asia characterized by high EF becausepopulated areas largely overlap with shallow water tables (egChina in Figure 1 and Vietnam in Figure 2 also see maps of

Table 2 Costs of Traditional and Low CH4 on-SiteSanitation Systems

Africa Asia

capital costs per capita (US$)simple pit latrinea 39 26ventilated improved pit (VIP) latrinea 57 50septic systema 115 104composting toiletb 39minus213 37minus58household biogasc 94minus330 70

regional parametersmean household size (S)d 45 43

EFe 012 023

marginal abatement cost (MAC)f (US$ton CO2e)composting toilet 57minus944 46minus97household biogas 338minus1541 127

aReference 19 bReferences 29minus31 cReferences 28 and 32 dReference42 eRepresentative regional EF (Figure 3) fCalculated from eq 4



Bangladesh and India in the Supporting Information) Thepercentage of the rural population living in areas with watertables shallower than 25 m a standard pit latrine depth inChina India and Bangladesh is 41 48 and 88respectively In contrast many Latin American and Africancountries have lower EF because populated areas are more arid(see maps of East Africa in Figure 2 as well as maps of BrazilEthiopia and South Africa in the Supporting Information) Just18 of Kenyarsquos rural population lives in areas with water tablesshallower than 25 m and in South Africa the ratio is only 4Variability within regions is also significant with Nigeria andCameroon characterized by greater EFs than those in East andSouthern AfricaTrends in Emissions from 2000 to 2015 Pit latrines are

a globally significant CH4 source with aggregate emissions in2000 of 48 Tg yminus1 (33minus94) from the 21 country sampleEmissions are projected to fall to 34 Tg yminus1 (23minus66) for 2015however as a net result of different regional trends In Chinafor example the decrease from 25 to 11 Tg yminus1 (Figure 4) isthe result of sharply declining latrine utilization ratios (TableS1 Supporting Information) This trend is driven by rapidurbanization and is reinforced by the modernization of Chinarsquosurban wastewater infrastructure Modest emissions decreases of10minus30 are projected elsewhere in Asia as a result of similar

social and infrastructure changes In contrast the large relativeincreases from African countries are caused by robustpopulation growth and a sustained and in some cases growingreliance on pit latrines in both rural and urban areas Stronggrowth in emissions in Ethiopia for example is due to majorsanitation improvements in rural areas leading to a projected 3-fold increase in the fraction of the rural population utilizing pitlatrines between 2000 and 2015 (Table S1 SupportingInformation) The magnitude of increases in Africa and LatinAmerica is small relative to the scale of declines in Chinahowever so overall global emissions are expected to fall despitegrowth in some regions China is expected to remain the largestemitter of pit latrine CH4 in 2015 (11 Tg yminus1) followed byBangladesh (061 Tg yminus1) and India (032 Tg yminus1) (Figure 4)

Wastewater CH4 in Emissions Inventories Table S2(Supporting Information) lists CH4 emissions estimated by thepresent study alongside previous estimates of emissions frompit latrines6 decentralized wastewater treatment24 and allwastewater sources734 The total wastewater sector emissionsare included for reference only while the studies of pit latrineor decentralized wastewater sources provide a suitablecomparison to the results of the present study Discrepanciesbetween inventories are due to uncertainties in either the pitlatrine utilization ratios or the appropriate EF The large

Figure 1 Schematic of geospatial emissions model for China in 2000 Population urbanization and water table are gridded at 30 arc-s resolutionlatrine utilization ratios are on an urbanrural basis at the country level and BOD and EF parameters are at country or regional levels In the watertable map red indicates shallow water tables and blue represents deep water tables



difference in estimated emissions from India for example (33Tg yminus1 by USEPA6 vs 044 Tg yminus1 by the present study in2000) is due to an assumption of latrine utilization ratios inrural India of 47 by the earlier study while recent surveyshave shown the actual number is closer to 101017 Thisdiscrepancy has led to a sharp downward revision of emissionsestimates for India There is better agreement between thepresent study and the GAINS integrated assessment model324

for China and India though GAINS uses lower latrineutilization ratios in southeast Asia relative to figures developedin the present study and those reported by Graham andPolizzotto10 (Table S1 Supporting Information) Thisdiscrepancy leads to emissions estimates for BangladeshVietnam and other southeast Asian nations that aresignificantly lower than the estimates reported hereThe default EF employed by GAINS agrees relatively well

with the EFs derived here for Asian countries though it is

Figure 2 Regional variation in the colocation of population and shallow water tables with impacts on pit latrine emissions in East Africa (KenyaTanzania and Uganda) and Vietnam

Figure 3 Population-weighted average emissions factors (EF) in ruralareas Error bars were determined using uncertainties in pit latrinedepth and in MCF for wet and dry pit latrines The default EF used byref 6 of 072 is outside the scale of this figure

Figure 4 Projected percentage change in latrine CH4 emissions from2000minus2015 with estimated 2015 emissions (in Tg yminus1) next to eachbar



greater than our EF estimates for Brazil Bolivia and the Africannations (Figure 3) The default EF of 072 used by ref 6 issignificantly greater than our EF estimates for all countrieswhich explains why their emissions estimates are generallygreater than the emissions reported in the present studyThe 21 countries in our sample include 151 of the estimated

177 billion global pit latrine users and thus contribute a majorfraction of global pit latrine emissions Many of the 260 millionremaining latrine users are in sub-Saharan Africa so using theregional BOD (Table 1) and representative regional EF of 012kg CH4 kg

minus1 BOD we conservatively estimate an additional 04Tg CH4 y

minus1 from populations not included in our 21 countrysample Total global pit latrine emissions are thus estimated tobe 52 Tg yminus1 (37minus98) for 2000 and 38 Tg yminus1 (27minus70) for2015 The 2000 estimate is intermediate between the 14 TgCH4 yminus1 estimated by USEPA6 and the 33 Tg CH4 yminus1

estimated by GAINS For 2015 there is improved agreementbetween this study and GAINS with estimates of 38 and 36Tg yminus1 respectivelyOur global emissions estimate represents sim2 of global

anthropogenic CH4 emissions in 2000 and sim8 of CH4emissions from waste management7 These fractions decreaseto sim1 of total emissions and sim5 of emissions from waste in2015 as pit latrine emissions are projected to fall while totalanthropogenic and waste-related emissions are expected toincrease Pit latrines contribute 25 of anthropogenic CH4emissions from Bangladesh and 5minus10 from many Africannations though the fraction falls to le1 in countries with largeagricultural sectors like Brazil or energy sectors like Bolivia orKazahkstan (Figure S3 Supporting Information)7

Empirical Pit Latrine Emissions Estimates Pit latrineCH4 emissions measurements are not available in the literaturebut gas production from a South African pit latrine sludge hasbeen measured with laboratory incubations35 Surficial sludgeproduced gas at a higher rate (0059 plusmn 0013 mL gas gminus1 sludgedminus1 at standard temperature and pressure) than samples fromthe bottom of the pit (00025 plusmn 0002 mL gas gminus1 sludge dminus1)The fraction of CH4 CO2 and other trace constituents was notdetermined but biogas is typically sim50 CH4

27 Based on arepresentative fecal production of 400 g personminus1 dminus1 for alargely vegetarian diet18 the biogas production from the freshsurficial sludge corresponds to 110 plusmn 025 kg CH4 person

minus1

yminus1 Using South Africarsquos EF of 0075 (range 0047minus0112) kgCH4 kg

minus1 BOD and the per capita BOD production for Africathe model developed in the present study predicts a range ofemissions from 064minus151 kg CH4 personminus1 yminus1 with a bestestimate of 101 kg CH4 person

minus1 yminus1 The model predictionthus agrees well with the incubation measurementPost-2015 Outlook Future changes in CH4 emissions

from pit latrines will depend on the balance betweenurbanization which decreases reliance on latrines and thusreduces emissions and expanded latrine coverage in rural areasthat have been historically underserved by improved sanitationHigh rates of open defecation (Figure S4 SupportingInformation) indicate countries likely to be the foci of futuresanitation interventions and pit latrines are expected to remainthe dominant form of rural sanitation due to cost and growingwater scarcity in developing regions3637 While the expansionof pit latrine use in these regions is difficult to predict countrieswith growing rural populations (Figure S5 SupportingInformation) along with high rates of open defecation aremost likely to experience growth in pit latrine utilization India

Pakistan and Nigeria each combine limited sanitation coveragerural population growth and high EFs and thus may experiencestrong growth in pit latrine CH4 emissions In contrasturbanization in East and Southeast Asia will drive declines inrural population that may decrease latrine emissions dependingon the sanitation services available in cities

Uncertainty Analysis The uncertainty intervals reportedin this study reflect uncertainties in pit latrine depth and themodel parameters BOD MCF and T Other sources of errormay influence pit latrine CH4 emissions in some settings butwere difficult to quantify within the model framework Raisedpit latrines are sometimes built in locations with shallow watertables (eg Bangladesh) or where rocky soils prevent theexcavation of deep pits18 Elevated latrines store excreta aboveground level or in shallow pits leading to a smaller fraction ofpits below the water table and consequently lower emissionsrelative to estimates generated with our approach The staticgroundwater level16 used in our analysis is another potentialerror source particularly in monsoon climates where watertables fluctuate seasonally by several meters The net effect onCH4 dynamics is unclear since more aerobic conditions inpremonsoon periods could be partially or fully balanced byflooded anaerobic conditions during the monsoon A thirdimportant uncertainty is the distribution of household andpublic pit latrines as latrines with many users are associatedwith more anaerobic conditions and a higher EF14

Mitigation Opportunities and Costs CH4 emissionsfrom on-site sanitation can be reduced by using aerobictreatment or by capturing anaerobically produced CH4 before itis released to the atmosphere Aerobic decomposition can bemost simply achieved by digging shallow pits that remain abovethe water table which is also preferable for limiting ground-water pollution10 This approach will only reduce rather thaneliminate CH4 emissions since even ldquodryrdquo latrines are partiallyanaerobic due to the commingling of liquid and solid wasteMore fully aerobic disposal can be achieved through the use ofwell-maintained composting toilets also known as ecologicalsanitation (ldquoecosanrdquo) methods Composting toilets separateliquid and solid waste and with proper maintenance the solidsdecompose aerobically to a nutrient-rich compost within a fewmonths18 Composting toilets have traditionally been promotedfor their low water use avoided groundwater contaminationrelative to pit latrines and the opportunity for nutrientrecycling26

Small-scale biogas digesters in which human excreta arecombined with manure and the generated CH4 burned as anenergy source are another potential mitigation option27 Whiletheir advantages as renewable and clean-burning energy sourcesare well-recognized38 their climate benefits are equivocal dueto the potential for significant leakage from poorly maintainedsystems which may negate emissions reductions due to fuelsubstitution or reduced latrine emissions39 Adoption of biogasmay also be limited by the need for a reliable supply of manureas feedstock27 and possible failure in cold climates40

We estimate MACs for reducing emissions with thesealternative on-site technologies in Africa and Asia (Table 2 andeq 4-5) since these are the regions likely to see the greatestgrowth in latrine emissions MACs for composting toilets rangefrom 46 to 97 $ton CO2e in Asia and 47 to 944 $ton CO2e inAfrica The upper limit of MACs is lower in Asia due to thecombination of lower sanitation costs and higher emissions perlatrine These costs lie above the 80th percentile on the globalcurve of potential CH4 mitigation costs though they are



competitive with some other measures in the waste manage-ment sector like source separation of municipal food waste(weighted average of 134 $ton CO2e) or upgradingwastewater treatment plants to anaerobic treatment with biogasrecovery (average 193 $ton CO2e)

3 Benefits of CH4emissions reductions are estimated to range from 33 to 238$ton CO2e

2 so in some settings composting toilets may beeffective from a cost-benefit perspective without including theancillary benefits of fertilizer and avoided groundwatercontamination The high MACs associated with householdbiogas digesters are a further drawback to their utility asmitigation measuresComposting toilets have traditionally been promoted for

reasons unrelated to CH4 mitigation so the recognition of anew benefit in CH4 emissions reductions will only add to theirexisting advantages and may attract financing based on GHGmitigation opportunities Before recommending specific miti-gation actions however it is critical that both CH4 and nitrousoxide (N2O) emissions from on-site sanitation systems becharacterized with greater certainty N2O is another potentGHG and is emitted from aerobic processes in municipaltreatment plants41 though emissions from on-site systems arenot known Direct measurements of CH4 and N2O from pitlatrines and composting toilets are not present in the literaturebut are needed to validate and improve emissions factors andinventories We additionally caution that the adoption ofcomposting toilets may be limited in some areas due to culturalsensitivities to the handling of human excreta18

The future path of pit latrine CH4 emissions depends on thespread of latrines into previously underserved areas in SouthAsia and sub-Saharan Africa and on the policies promotingspecific sanitation technologies Recognizing both the globalimportance of pit latrine emissions and the availability ofappropriate on-site mitigation measures highlights potentialsynergies between water and sanitation development and GHGmitigation efforts Opportunities for abating CH4 emissionscould mobilize financial resources to promote compostingtoilets which are broadly preferable from water quality andsustainable development perspectives as well due to avoidedgroundwater contamination10 and the opportunity for nutrientrecycling26 This analysis demonstrates that the problem of pitlatrine CH4 emissions can be reframed as an opportunity toincentivize progress up the sanitation ladder to compostingtoilets or more advanced systems yielding cobenefits for bothGHG mitigation and water and sanitation development

ASSOCIATED CONTENT

S Supporting InformationTables with complete sociodemographic data and emissionsmodeling results additional figures and population water tableand CH4 emissions maps for selected countries This material isavailable free of charge via the Internet at httppubsacsorg

AUTHOR INFORMATION

Corresponding AuthorE-mail matthewcharlesreidgmailcom

Present AddressperpEnvironmental Microbiology Laboratory Ecole PolytechniqueFederale de Lausanne Lausanne CH 1015 Switzerland

NotesThe authors declare no competing financial interest

ACKNOWLEDGMENTSS Brink H Godfrey S Mignotte and M Sengupta contributedto early discussions on this topic We thank three anonymousreviewers for their constructive comments MCR acknowl-edges support from the Princeton Environmental Institute andthe Woodrow Wilson School of Public and InternationalAffairs

REFERENCES(1) West J Fiore A Horowitz L Mauzerall D Global HealthBenefits of Mitigating Ozone Pollution with Methane EmissionControls Proc Natl Acad Sci U S A 2006 103 3988minus3993(2) Shindell D et al Simultaneously Mitigating Near-Term ClimateChange and Improving Human Health and Food Security Science2012 335 183minus190(3) Hoglund-Isaksson L Global Anthropogenic Methane Emissions2005minus2030 Technical Mitigation Potentials and Costs Atmos ChemPhys 2012 12 9079minus9096(4) Czepiel P Crill P Harriss R Methane Emissions fromMunicipal Wastewater Treatment Processes Environ Sci Technol1993 27 2472minus2477(5) Daelman M van Voorthuizen E van Dongen U Volcke Evan Loosdrecht M Methane emission during municipal wastewatertreatment Water Res 2012 46 3657minus3670(6) Quantif ication of Methane Emissions and Discussion of NitrousOxide and Ammonia Emissions f rom Septic Tanks Latrines andStagnant Open Sewers in the World United States EnvironmentalProtection Agency Research Triangle Park NC 1999(7) Global Anthropogenic Non-CO2 Greenhouse Gas Emissions 1990 -2030 United States Environmental Protection Agency (USEPA)Washington DC 2012(8) Diaz-Valbuena L R Leverenz H L Cappa C DTchobanoglous G Horwath W R Darby J L Methane CarbonDioxide and Nitrous Oxide Emissions from Septic Tank SystemsEnviron Sci Technol 2011 45 2741minus2747(9) Near-term Climate Protection and Clean Air Benef its Actions forControlling Short-Lived Climate Forcers United Nations EnvironmentProgram (UNEP) Nairobi Kenya 2011(10) Graham J Polizzotto M Pit Latrines and Their Impacts onGroundwater Quality A Systematic Review Environ Health Perspect2013 121 521minus530(11) Progress on Sanitation and Drinking-Water - 2013 Update WorldHealth Organization (WHO) and United Nations Childrenrsquos Fund(UNICEF) 2013(12) Pruss-Ustun A et al Burden of disease from inadequate watersanitation and hygiene in low- and middle-income settings aretrospective analysis of data from 145 countries Tropical Medicineand International Health 2014 in press DOI 101111tmi12329(13) Introduction and Proposed Goals and Targets on SustainableDevelopment for the Post 2015 Development Agenda Open WorkingGroup on Sustainable Development Goals 2014 https u s t a i n a b l e d e v e l o pmen t u n o r g c o n t e n t d o c umen t s 4044140602workingdocumentpdf(14) Doorn M Towprayoon S Vieira S Irving W Palmer CPipatti R Wang C 2006 IPCC Guidelines for National GreenhouseGas Inventories Chapter 6 Wastewater Treatment and Discharge IPCC2006(15) Center for International Earth Science Information Network(CIESIN) Global Rural-Urban Mapping Project Version 1(GRUMPv1) httpsedacciesincolumbiaedudataset(16) Fan Y Li H Miguez-Macho G Global Patterns ofGroundwater Table Depth Science 2013 339 940minus944(17) WHOUNICEF Joint Monitoring Programme (JMP) for WaterSupply and Sanitation Country Files httpwwwwssinfoorgdocuments-linksdocumentstx_displaycontroller[type]=country_files(18) Franceys R Pickford J Reed R A Guide to the Development ofOn-Site Sanitation World Health Organization Geneva 1992



(19) The World Bank Sanitation Hygiene and Water ResourceGuide Infrastructure - Costing Sanitation Technologies 2013 httpwaterworldbankorgshw-resource-guideinfrastructurecosting-sanitation-technologies(20) Dzwairo B Hoko Z Love D Guzha E Assessment of theimpacts of pit latrines on groundwater quality in rural areas A casestudy from Marondera district Zimbabwe Phys Chem Earth 2006 31779minus788(21) Lin J Aoll J Niclass Y Velazco M I Wunsche L Pika JStarkenmann C Qualitative and Quantitative Analysis of VolatileConstituents from Latrines Environ Sci Technol 2013 47 7876minus7882(22) Center for International Earth Science Information Network(CIESIN) Gridded Population of the World Version 3 (GPWv3)Population Density Grid Future Estimates httpsedacciesincolumbiaedudatasetgpw-v3-population-density-future-estimates(23) United Nations Department of Economic and Social AffairsWorld Urbanization Prospects The 2009 Revision PopulationDatabase 2013 httpesaunorgwup2009unupindexasppanel=1(24) Amann M et al Cost-effective control of air quality andgreenhouse gases in Europe Modeling and policy applicationsEnviron Modell Softw 2011 26 1489minus1501(25) Houghton J et al Climate Change 1995 The Science of ClimateChange Contribution of Working Group I to the Second AssessmentReport of the Intergovernmental Panel on Climate Change CambridgeUniversity Press 1995(26) Calvert P Morgan P Rosemarin A Sawyer R Xiao J InEcological Sanitation Winblad U Simpson-Hebert M Eds Stock-holm Environment Institute 2004(27) Bond T Templeton M History and future of domestic biogasplants in the developing world Energy for Sustainable Development2011 15 347minus354(28) Chen Y Yang G Sweeney S Feng Y Household biogas usein rural China A study of opportunities and constraints Renew SustEnerg Rev 2010 14 454minus549(29) Study for Financial and Economic Analysis of Ecological Sanitationin sub-Saharan Africa The World Bank Water and Sanitation Program- Africa Nairobi Kenya 2009(30) Lamichhane K On-site sanitation a viable alternative tomodern wastewater treatment plants Water Sci Technol 2007 55443minus440(31) Technical Handbook - Construction of ecological sanitationlatrine Water Aid in Nepal 2011 httpwwwwateraidorgsimmediaPublicationsconstruction-ecological-sanitation-latrine-technical-handbookpdf(32) Mwirigi J Makenzi P Ochola W Socio-economic constraintsto adoption and sustainability of biogas technology by farmers inNakuru Districts Kenya Energy for Sustainable Development 2009 13106minus115(33) Hutton G Haller L Evaluation of the Costs and Benefits ofWater and Sanitation Improvements at the Global Level World HealthOrganization Geneva 2004(34) European Commission Joint Research Centre NetherlandsEnvironmental Assessment Emission Database for Global Atmos-pheric Research (EDGAR) Release Version 42 2010 httpedgarjrceceuropaeu(35) Couderc A A-L Foxon K Buckley C A Nwaneri C FBakare B F Gounden T Battimelli A The effect of moisturecontent and alkalinity on the anaerobic biodegradation of pit latrinesludge Water Sci Technol 2008 58 1461minus1466(36) Black M Fawcett B The Last Taboo Opening the Door on theGlobal Sanitation Crisis Earthscan London 2008(37) Narain S Sanitation for all Nature 2012 14 185(38) Anenberg S Balakrishnan K Jetter J Masera O Mehta SMoss J Ramanathan V Cleaner Cooking Solutions to AchieveHealth Climate and Economic Cobenefits Environ Sci Technol 201347 3944minus3952(39) Bruun S Jensen L S Vu V T K Sommer S Small-scalehousehold biogas digesters An option for global warming mitigation

or a potential climate bomb Renew Sust Energ Rev 2014 33 736minus741(40) Terradas-Ill G Pham C H Triolo J M Mart-Herrero JSommer S G Thermic Model to Predict Biogas Production inUnheated Fixed-Dome Digesters Buried in the Ground Environ SciTechnol 2014 48 3253minus3262(41) Ahn J Kim S Park H Rahm B Pagilla K Chandran KN2O Emissions from Activated Sludge Processes 2008minus2009 Resultsof a National Monitoring Survey in the United States Environ SciTechnol 2010 44 4505minus4511(42) United Nations Center for Human Settlements (UN - Habitat)Cities in a Globalizing World Global Report on Human Settlements2001 Earthscan 2001



occur when organic waste in pits is submerged beneath thewater table14 Spatial variation in groundwater level and itseffect on biogeochemical decomposition pathways in pitlatrines are not captured by the coarse calculations underlyingpit latrine emissions in the current generation of GHGinventories In order to characterize geographic variations inhydrological control over pit latrine emissions we develop thefirst spatially explicit approach to estimating pit latrine CH4emissions and use the analysis to create a discussion of thelinkages between pit latrines sanitation development andglobal CH4 emissionsOur approach uses spatial analyses of population urban-

ization15 and groundwater level16 in 21 developing and middleincome nations to develop a set of emissions factors (EFs)14

that reflect local hydrology These spatial data sets arecombined with country-level health and sanitation surveys todetermine rural and urban populations using pit latrines and toquantify the resulting CH4 emissions We use this analyticalframework to characterize variations in latrine CH4 emissionswithin and between countries to predict changes in globalemissions between 2000 and 2015 and to evaluate the costs ofmitigation measures like composting toilets that can providelow-CH4 on-site alternatives to pit latrines

METHODS

Spatial CH4 Emissions Model Pit latrine CH4 emissionswere estimated by integrating EFs from the 2006 IPCCGuidelines for National Greenhouse Gas Inventories14 withcountry-level latrine utilization data and a high-resolutiongeospatial analysis of population urbanization and water tabledepth in 21 countries in Asia Africa and Latin AmericaCountries were selected to represent a range of geophysicalsettings as well as to capture most global pit latrine usersEmissions were calculated for grid cells at 30 arc-s (sim1 km atthe equator) grid spacing and summed for each country

sum= middot middot middotP TCH emissions EF BODi

i i i i4(1)

P and T are the total population and the fraction of thepopulation using pit latrines respectively in grid cell i T was afunction of the urban or rural classification of the cell BOD isthe country- or region-specific per capita production ofbiochemical oxygen demand (BOD) [kg BOD personminus1 yminus1]taken from the IPCC guidelines (Table 1) EF is the CH4emissions factor [kg CH4 kg

minus1 BOD]

Pit Latrine Prevalence Latrine utilization ratios (T) inurban and rural areas were determined from country-levelhealth and sanitation surveys compiled by the WHOUNICEFJoint Monitoring Program (JMP) for Water Supply andSanitation17 We aggregated data from 1998minus2003 todetermine a mean and standard error of latrine prevalence ca2000 (Table S1 Supporting Information) We followed theclassification scheme from Graham and Polizzotto10 andconsidered the following categories to be pit latrines toiletsthat flush to pits ventilated improved pit (VIP) latrines pitlatrines with concrete slabs traditional latrines and pit latrineswithout slabopen pit Hanging latrines and bucket latrineswere not included since the ultimate disposal location may notbe a pit We do not separate urban populations into ldquourban highincomerdquo and ldquourban low incomerdquo groups as is done in theIPCC guidelines because JMP surveys only include data on anurbanrural basis

Selection of Emissions Factors EFs were determined from

= middotBEF MCFo (2)

Bo is the maximum CH4 producing capacity for domesticwastewater and the default value of 06 kg CH4 kg

minus1 BOD wasused due to the lack of country-specific values14 MCF is themethane correction factor and depends on the type oftreatment andor discharge method ranging from 0 for afully aerobic process to 1 for a fully anaerobic process MCFswere taken from the IPCC methodology and are listed in Table1 The MCF selected for each grid cell depended on the latrinedepth relative to the groundwater level since organic mattersubmerged beneath the water table will decompose anaerobi-cally while dry decomposition will be largely aerobic TheseMCFs will yield conservative emissions estimates that mayunderestimate the true CH4 emissions since they assume thatdecomposition in pits in dry climates will be mostly aerobic Inreality pits with many users those with flush water use orthose in impermeable soils will retain water and remain at leastpartially anaerobic even in dry climates The simplifiedgrouping of pit latrines into ldquowetbelow the water tablerdquo andldquodryabove the water tablerdquo was necessary because detailed dataon latrine water use number of users and water retention wasnot available from JMP country files or other data sourcesA global groundwater level model16 was used to determine

pit depths relative to the groundwater level The modelrepresents a mean annual water table depth and does notaccount for seasonal fluctuations or long-term variations ingroundwater level We assume a globally uniform pit latrinedepth of 25 plusmn 05 m which spans the 2minus3 m depth rangetypical of pit latrines1819 It is good practice for the bottom oflatrines to be above the water table to avoid groundwatercontamination though this recommendation is often dis-regarded in practice102021 P and EF in rural grid cells wereused to determine population-weighted average rural EFs foreach country

=sum middot

sum=

=

P

PEF

EFin

i i

in

i

1

1 (3)

Population and Urbanization The model calculation in2000 used a global gridded population data set for the year2000 and an urban extents mask ca 1995 in 30 arc-sresolution15 2015 population projections were only availablein a coarser 25 arc-min resolution22 so a linear interpolationwas used to generate a 30 arc-s mesh for the 2015 population

Table 1 Emissions Model Parameters14

methane correction factors (MCF) for decentralized treatment systems

disposalpathway description MCF range

pit latrine water table below latrine 3minus5 people 01 005minus015pit latrine water table higher than latrine 07 07minus10septic system half of BOD settles in anaerobic tank 05 05

biochemical oxygen demand (BOD)

regioncountry BOD [kg personminus1 yminus1] range

Africa 1351 1278minus1643Asia Latin America 1460 1278minus1643Turkey 1387 986minus1825India 1241 986minus1497Brazil 1825 1643minus2008







=minus


TACH TACH


(4)



= ++ minus

middot +⎡⎣⎢

⎤⎦⎥

r rr

C CTACH(1 )

(1 ) 1

L

L CAP OampM(5)






Africa Asia



EFe 012 023




























minus1

















ASSOCIATED CONTENT


AUTHOR INFORMATION
















=minus


TACH TACH


(4)



= ++ minus

middot +⎡⎣⎢

⎤⎦⎥

r rr

C CTACH(1 )

(1 ) 1

L

L CAP OampM(5)






Africa Asia



EFe 012 023




























minus1

















ASSOCIATED CONTENT


AUTHOR INFORMATION



































minus1

















ASSOCIATED CONTENT


AUTHOR INFORMATION

















ASSOCIATED CONTENT


AUTHOR INFORMATION












global methane emissions from pit latrines - princeton · global methane emissions from pit...

Documents