This content has been downloaded from IOPscience. Please scroll down to see the full text.

Download details:

IP Address: 141.211.4.224

This content was downloaded on 03/11/2015 at 22:21

Please note that terms and conditions apply.

The energy and emissions footprint of water supply for Southern California

View the table of contents for this issue, or go to the journal homepage for more

2015 Environ. Res. Lett. 10 114002

(http://iopscience.iop.org/1748-9326/10/11/114002)

Home Search Collections Journals About Contact us My IOPscience

Environ. Res. Lett. 10 (2015) 114002 doi:10.1088/1748-9326/10/11/114002

LETTER

The energy and emissions footprint of water supply for SouthernCalifornia

A J Fang1, Joshua PNewell2 and Joshua JCousins2

1 Humphrey School of Public Affairs, University ofMinnesota, 301 19th Ave S.,Minneapolis,MN55455,USA2 School of Natural Resources and Environment, University ofMichigan, 440Church Street, AnnArbor,MI 48109,USA

E-mail: jpnewell@umich.edu

Keywords:water supply, carbon footprint, California, drought, water–energy nexus, life cycle assessment, energy

AbstractDue to climate change and ongoing drought, California andmuch of the AmericanWest face criticalwater supply challenges. California’s water supply infrastructure sprawls for thousands ofmiles, fromtheColoradoRiver to the SacramentoDelta. Bringingwater to growing urban centers in SouthernCalifornia is especially energy intensive, pushing local utilities to balancewater security with factorssuch as the cost and carbon footprint of the various supply sources. To enhancewater security, citiesare expanding efforts to increase local water supply. But do these local sources have a smaller carbonfootprint than imported sources? To answer this question and others related to the urbanwater–energy nexus, this study uses spatially explicit life cycle assessment to estimate the energy andemissions intensity of water supply for twoutilities in SouthernCalifornia: Los Angeles Department ofWater and Power, which serves Los Angeles, and the Inland EmpireUtility Agency, which serves theSanBernardino region. This study differs fromprevious research in two significant ways: (1) emissionsfactors are based not on regional averages but on the specific electric utility and generation sourcessupplying energy throughout transport, treatment, and distribution phases of thewater supply chain;(2)upstream (non-combustion) emissions associatedwith the energy sources are included. Thisapproach reveals that in case of water supply to Los Angeles, local recycledwater has a higher carbonfootprint thanwater imported from theColoradoRiver. In addition, by excluding upstreamemissions, the carbon footprint of water supply is potentially underestimated by up to 30%. Theseresults havewide-ranging implications for how carbon footprints are traditionally calculated at localand regional levels. Reducing the emissions intensity of local water supply hinges on transitioning theenergy used to treat and distributewater away from fossil fuel, sources such as coal.

1. Introduction

California’s history is permeated with concerns andconflicts over water supply and water quality. Climatechange and ongoing drought have only exacerbatedthreats to securing future supply, both for the state’svital agriculture sector as well as its growing urbancenters. California’s water supply infrastructuresprawls for thousands of miles across the AmericanWest, with sources ranging from theColorado River tothe Sacramento Delta (Cousins and Newell 2015).Bringing this water to its users is highly energyintensive, with roughly 8% of the state’s electricityconsumption devoted to the sourcing, conveyance,

and treatment of water (Klein et al 2005, CPUC 2010).California’s Renewable Portfolio Standard (RPS) andCap-and-Trade program are also influencing the wayelectricity is produced in the state by requiringreductions in greenhouse gas (GHG) emissions. Facedwith the dual pressure to reduce GHG emissions whilemaintaining a reliable supply, water utilities acrossCalifornia aremeasuring the carbon footprint of wateror the so-calledwater–energy nexus (LADWP2010).

Water supplied to Southern California is of parti-cular interest to water agencies. It is roughly 50 timesmore energy intensive than water supplied to North-ern California (Klein et al 2005) and maintainingdiverse and reliable water supply sources in a

OPEN ACCESS

RECEIVED

22 July 2015

REVISED

25 September 2015

ACCEPTED FOR PUBLICATION

28 September 2015

PUBLISHED

27October 2015

Content from this workmay be used under theterms of theCreativeCommonsAttribution 3.0licence.

Any further distribution ofthis workmustmaintainattribution to theauthor(s) and the title ofthework, journal citationandDOI.

© 2015 IOPPublishing Ltd

continual challenge. Throughout Southern California,cities are focusing on expanding local water supplysources (e.g. recycled water, groundwater recharge,desalination, stormwater capture). But the energy andemissions intensity of increasing these local supplysources remain uncertain. How, for example, does thecarbon footprint of local sources compare to impor-ted ones?

The objective of this study, therefore, is to quantifyand compare the energy and emissions footprint ofwater supply sources for Southern California. Specifi-cally, we focus on two urban utilities: (1) The LosAngeles Department of Water and Power (LADWP),which supplies fourmillion residents in the City of LosAngeles; and (2) The Inland Empire Utilities Agency(IEUA), which serves approximately 850 000 residentsin southwest San BernardinoCounty.

LADWP relies heavily on imported water providedby the Metropolitan Water District (MWD), whichadministers supplies from the State Water Project(SWP) and the Colorado River Aqueduct (CRA)(figure 1). LADWP’s ‘local’ supply sources include theLos Angeles Aqueduct (LAA), groundwater, and recy-cled water. IEUA also imports water from MWD but

relies more heavily on local sources, namely recycledwater, surface water, desalted groundwater, anduntreated groundwater (figure 1). In addition, theenergy grid mix of these utilities is significantly differ-ent. LADWP, for example, relies much more heavilyon coal than does IEUA. It is for this reason thatgroundwater and recycled water have similar energyintensities between the two utilities, but IEUA has alower emissions intensity. This is a result of IEUA self-generating more electricity and procuring moreenergy from a utility, Southern California Edison(SCE), with a cleaner generation portfolio thanLADWP.

The traditional approach to quantifying the energyand emissions footprint of products and processesuses life cycle assessment (LCA), environmentalinput–output analysis, or some combination of thetwo (Lifset and Graedel 2002). LCA approaches typi-cally use life cycle inventories based on activity dataand emissions factors that utilize global, regional, ornational averages. This practice tends to render LCAsaspatial as areal differentiation is minimized for thepurposes of expedience and simplicity (Curran 2006,Newell and Vos 2011, Cousins and Newell 2015).

Figure 1. Study Area. Geographic scope ofwater sources supplying Los Angeles Department ofWater and Power and Inland EmpireUtilities Agency Service Areas.

2

Environ. Res. Lett. 10 (2015) 114002 A J Fang et al

Typically, studies of California water utilities utilizeeGRID—an Environmental Protection Agency data-base that provides generalized emissions factors forelectric power plants generating in the United States.These factors are calculated by averaging energy andemissions profiles of plants across the CAMX sub-region, which encompasses California, and parts ofNevada, Arizona, and Mexico. eGRID emissions fac-tors were used, for example, in studies the LADWPand IEUA conducted (or commissioned) of the energyand emissions footprints for portions of their waterdistribution systems (IEUA 2009, LADWP2010).

Previous studies (Filion et al 2004, Stokes andHor-vath 2011, Sanders and Webber 2012, Mo et al 2014)have relied on such regionalized average (e.g. eGRID)emission factors to calculate carbon footprints ofwater supplies in theUS.While the argument for usingthese regional emission factors is that water utilties aresimply purchasing electrons from the grid, water util-ties in California increasingly rely on localized watersupply sources and are self-generating more elec-tricity; making site-specific emission factors moreconsequential. Additionally, given the inter-connectedness of water supplies and hydroelectricpower generation in California, it is important to cap-ture the nuance of this relationship in terms of whosupplies power to this massive infrastructure system.As utilities transition towards localized sources ofwater, they will need to understand both where theiremissions and electrons are coming from.

In addition to omitting this important spatial var-iation in terms of energy grid mix, the eGRIDapproach (hereafter ‘statewide average approach’)often only includes combustion emissions occurringat the power plant, excluding upstream emissionsassociated with the production of energy, such as coal,hydropower, solar, and wind. As a consequence manyrenewable resources become ‘carbon neutral’ and thecarbon footprint of other sources is reduced. Forexample, water supply infrastructures, such as theLAA, that are gravity fed and use hydropower are con-sidered carbon neutral. But upstream emissions can besignificant for infrastructure sectors, such as transpor-tation (Chester andHorvath 2009).

To capture both local variation and upstreamemissions, we developed a spatially explicit approachthat combines LCA with a Geographic InformationSystem (GIS) to quantify the respective carbon foot-prints of the two utilities: LADWP and IEUA.We thencompared our approach with the traditional statewideaverage (e.g. eGRID) approach. The results reveal thatdownscaling to the utility-scale to develop spatiallyexplicit emissions factors can lower the emissionsintensity by 20% or increase it as much as 40%depending on the water source. Again depending onthe water source, including upstream emissions canincrease the carbon intensity by up to 30%. This hasimportant policy implications in terms of developingstrategies for reducing carbon intensity based uponthe differences in the utilities’ energy procurementpractices and for fostering low carbon transitions. As

Figure 2.Utility water source profiles. Energy intensity (kwh AF−1) andwater volume (AF) associatedwith various sources supplyingLADWP (2010) and IEUA (Wilkinson 2000, IEUA 2010).Water purchased fromMetropolitanWaterDistrict of SouthernCalifornia(MWD) by IEUA is not disaggregated by source. Authors assigned transport energy intensity of StateWater Project East and treatmentenergy intensity of theWeymouth treatment plant based upon pipeline infrastructure location.Note: 1 acre-foot (AF)=1.233 cubicdecameters (dam3).

3

Environ. Res. Lett. 10 (2015) 114002 A J Fang et al

utilities increase their reliance on localized water sour-ces, a spatially explicit approach provides improvedmeasures tomanage the carbon intensity of water sup-ply by capturing the differences.

2.Methods

This study uses a spatially explicit LCA approach tomeasure the energy and emissions intensity of thewater supply infrastructure for two Southern Califor-nia utilities (LADWP and IEUA). This approachenables one to assess the importance of localized,regionally-specific emissions factors when quantifyingthe carbon footprinting of utilities and associated gridmixes. The system boundary of the study is limited tothe delivery phases of the water–energy nexus mostlikely to show geographic variation: sourcing andconveying, treatment, and distribution to consumers(figure 1). This builds on work by Wilkinson(2000, 2007) who found that differences across watersources stem primarily from the varying pumping,treatment, and distribution processes required. Addi-tionally, the energy and emissions related to theconstruction and maintenance of water infrastructureare generally overshadowed by the operation of theinfrastructure (Stokes and Horvath 2006, 2009). Theuse and disposal phases, in-home energy usage forheating and cooling water were excluded. Due to dataconstraints, temporal variation was also excluded.Ideally, time-of-day could be used to determineelectricity generation sources using location marginalpricing (Rogers et al 2013). Instead, the average annual

consumption and generation was used to calculateemissions. This approach, utilized by Weber et al(2010), both eliminates seasonal variation and isolatesthe geographic distinctions between water and electri-city sources.

One acre-foot (AF) of water represents the func-tional unit and the emissions burden was measured ingrams of CO2e generated per acre foot (gCO2e/AF).The activity data, such as the volume of water bysource (AF), energy intensity (KWh AF−1), utility gridmix, water pumping, recycling, and treatment plantefficiency came primarily from the utilities’ UrbanWater Management Plans (IEUA 2009,LADWP 2010). Determining emissions factors forboth the utility emissions and the upstream emissionsof the energy sources consisted of three primary steps:

2.1. Assign specific utility for pumping andtransport, treatment, and distribution phasesIn GIS, we mapped the water supply infrastructure ofLADWP and IEUA. Data that were not publiclyavailable were obtained through correspondence withLADWP, MWD, and IEUA. Agency publications(MWD 2009, CDWR 2011) provided pumping plantlocations that were geocoded by cross-referencing theestimated X, Y coordinates in Google Maps. Each ofthese plants was then assigned to a geographically-defined Electricity Utility Service Area (EUSA) asspecified in the California Energy Almanac(CEC 2015). Groundwater and recycled water sourcesremain within the LADWP service area and weredesignated emissions burdens accordingly. IEUA and

Table 1.Electricity andwater source for water treatment plants. Distribution ofwater sources to treatment plants is based upon analysis ofpiping infrastructure locations and plant capacities (LADWP2010). Electricity supplier for each plant is based upon its geographic location.

4

Environ. Res. Lett. 10 (2015) 114002 A J Fang et al

LADWP provided data to match the utilities with theproper treatment plants (table 1). We assume potablewater, after leaving the treatment plant, is distributeduniformly throughout the city, regardless of thesource, requiring consistent energy inputs for all watersources of 196 kWh AF−1 (159 kWh dam−3) fromLADWP and 220 kWh AF−1 (178 kWh dam−3)from IEUA.

2.2.Determine gridmix for each utility andemissions factors for energy sourcesState-mandated power content labels provided thegrid mix for each utility and their corresponding

emissions factors (figure 3). The direct and ‘upstream’

emissions associated with the various energy sourceswere calculated by averaging the results of studiesobtained through a review of the literature (EPA 2009,Blanco et al 2012). To select these studies, three criteriawere used: (1) must be geographically appropriate tothe California region; (2) must consist of cradle-to-grave analysis covering emissions associated withconstruction, on-site assembly, production, transport,waste and disposal; and (3)must reflect similar outputcapacities and facility life expectancies to those facil-ities currently supplying or projected to supply theutilities with power.

Figure 3.Electricity generation gridmix for selected electric utilities andwater sources. (a)Generation sources of electric utilities (LosAngeles Department ofWater and Power, City of Riverside, SouthernCalifornia Edison) are determined using PowerContent Labels,which are produced annually by retail utilities inCalifornia. The eGridCAMXgridmix (EPA2009) is representative of theCalifornia–Mexico Power Area (CAMX) sub-region of theWestern Electric CoordinatingCouncil. (b)Profile of electricity generation sources isbased upon utility service agreements with the ColoradoRiver Aqueduct (MWD2006), StateWater Project (Cooke et al 2012).Generation profile is based upon the annual electricity (kwh) used forwater pumping/conveyance by theCRA and SWP. The IEUAprofile was generated based upon power purchased fromSCE andmonthly self- generated electricity data provided by IEUA.

5

Environ. Res. Lett. 10 (2015) 114002 A J Fang et al

Including upstream emissions is in contrast toeGRID, which only includes combustion emissions(i.e. those occurring at the plant). Emissions factors foreGRID (CAMX) were calculated as the weighted aver-age of the power plants in the database based onannual electric output (MWh). Including upstreamemissions results in higher emissions footprints for theenergy sources, with the exception of biomass andgeothermal, which are likely skewed due to a lownum-ber of generators reported for the CAMX region. Forrenewable sources, the effect is more nuanced becauseeGRID assumes these sources have zero emissions (noemissions from combustion). The primary differenceis the additional+240 gCO2e/kWh attributed to largehydropower generation sources (table 2).

AMonteCarlo Simulationmeasured the sensitivityof the carbon footprint of each water source to therange of upstream emissions factors shown in table 2.Assuming a normal distribution over the range ofvalues for each generation source (e.g. coal, natural gas,solar, wind, hydro), 1000 random trials were run todetermine the minimum and maximum range of thecarbon footprints. All local sources of water varybetween ±1%, while SWP varies between ±5%. CRAvaries the most at ±10%. The larger variation in SWPand CRA water is due to a greater reliance on hydro-power, which has the largest range of estimated emis-sions. Overall, the sensitivity analysis reveals thatalthough there is uncertainty in the upstream emis-sions factors of each power plant, the carbon footprintof the water sources will not be significantly affected.Thus, the increase in emissions for eachwater source infigure 6 cannot be explained solely by the uncertainty inthe upstream electricity emissions factors (table 2) andusing an average, rather than plant-specific emissionsfactorwill not disproportionately skew the results.

2.3. Calculate the energy and emissions footprint forthe three life-cycle phasesThe activity data and the emissions factors weremultiplied for all three phases. Emissions profiles foreach pumping and treatment plant were generated

based upon the distribution of net electricity con-sumption that could be attributed to each EUSA. Inaddition to adjusting the emissions factors of theelectricity, also adjusted was the distribution ofelectricity generation sources attributed to waterconveyance. This change directly altered SWP andCRA emissions; previously the eGrid CAMX emis-sions factor had been used to approximate theiremissions intensity. Power generation associated withthe CRA and SWP conveyance systems required acalculation of self-generated electricity. Long-termpower agreements with Pacific Gas & Electric (PG&E)and SCE support the electricity demanded by thepumping plants. This allows the operators of the SWPandCRA to sell their hydropower during peak demandtimes and buy electricity from the grid at cheaper off-peak rates.

In figure 3(b), the portfolio of electricity genera-tion sources is aggregated for the CRA and SWP. Theuse of hydropower in the conveyance systems forpumping water was added to purchased electricityfrom the EUSA in order to determine an overall emis-sions factor for each conveyance system. For example,the SWP generates approximately 38%(2189 GWh yr−1) of its electricity from large hydro-power projects (Cooke et al 2012), while CRA receivesapproximately 80% (1517 GWh yr−1) from damsalong the Colorado River (MWD 2006). Using annualaverage electricity purchase data for each pumpingplant (Cooke et al 2012), we estimated the electricitypurchased by the SWP from PG&E (3183 GWh yr−1)and SCE (349 GWh yr−1). Based upon the proportionof self-generated and purchased electricity by theSWP, a weighted average of electricity generationsources (figure 3(b)) is calculated using 2010 PowerContent Labels of SCE (SCE 2010) and PG&E(PG&E 2010).

3. Results

The conveyance of water constitutes the largestcomponent of LADWP’s and IEUA’s water supply

Table 2.Upstream emissions of electricity generation sources. Estimated eGRID emissions are based upon aweighted average by net genera-tion (MWh) of the plants located in theCAMX subregion. Upstream emissions ranges are based upon a literature review (Blanco et al 2012);themean is used to calculate the difference between upstream and eGRID reported emissions factors. All values are in units of gCO2e/kWh.

6

Environ. Res. Lett. 10 (2015) 114002 A J Fang et al

carbon footprint, followed by treatment and distribu-tion (figure 4). The result is due to their reliance onimported water from the MWD, which is pumpedover great distances to Southern California. Forsupplies typically transported over short distances,such as recycled water and desalted groundwater, thetreatment phase comprises the largest portion of thefootprint. In order to ensure sufficient quality for non-potable use, recycled water treatment for the twoutilities includes, aeration, microfiltration, anddisinfection.

For LADWP, the most energy and emissionsintensive water comes from the SWP (West and EastBranch), followed by local recycled water (figure 4).This is an interesting finding as SWP-West and recy-cled water have carbon footprints of 0.88 gCO2e/AF(0.71 gCO2e/dam

3) and 0.87 gCO2e/AF(0.71 gCO2e/dam

3) respectively, countering theassumption that local sources necessarily have a lowercarbon footprint. LADWP’s emissions intensity forboth recycledwater and groundwater is approximatelytwice that of IEUA’s, due to the latter’s greater use ofelectricity from hydropower, solar and biomass sour-ces (figure 2). The third most emissions intensivesource for LADWP is water from the CRA. Althoughmore energy intensive than recycled water, the CRAmainly uses electricity generated from hydropower,which reduces its emissions footprint. The least inten-sive water source, both in terms of energy and emis-sions, is water from the LAA, which is gravity fed andtherefore requires no net energy to transport thewater.

For IEUA, water from the SWP and the CRA is themost energy and emissions intensive (figure 2). In con-trast to LADWP, local sources have a smaller footprintbecause of the cleaner grid mix. However, in the caseof IEUA the desalted water has a higher footprint thanrecycled water (figure 4). This is because the treatmentphase is highly energy intensive due to the reverseosmosis and ion exchange used to remove nitrates andtotal dissolved solids. Surface water requires a

relatively small amount of energy for transport andtreatment, with emissions largely in the distributionphase.

3.1. Spatial-upstream versus statewide average (e.geGRID) approachThe fact that the carbon footprint of LADWP’s localrecycled water is higher than that sourced from theColorado River was made evident by the spatial-upstream approach used in this study. With respect tobroadly comparing the spatial-upstream approachwith the statewide average approach, the formerresulted in higher emissions profiles for the watersupplies of the two utilities (figure 5). Specifically, thespatial-upstream approach resulted in an increase inthe weighted average of the emissions intensity(gCO2e/AF) for both LADWP (+6%) and IEUA(+7%). Increases in emissions vary by water source.Emissions intensity for the CRA actually decreased by7%. Differences in grid mixes can further be seen inthe distinctions between IEUA’s and LADWPsgroundwater and recycled water emission footprints,because embodied energy for the water sources issimilar. However, IEUA’s groundwater and recycledwater increased by 18% and 2% respectively, whileLADWP’s increased by 65%and 79%.

The overall higher carbon footprints of both uti-lities are primarily due to the inclusion of upstreamemissions. Figure 6 illustrates the impact of upstreamemissions. For example, including upstream emis-sions in the statewide approach resulted in a 28% and30% weighted average increase for LADWP and IEUArespectively. Sources dominated by transport (MWD,SWP, CRA) have greater absolute differences on agCO2e/AF basis. The MWD carbon intensity increa-ses by 29% and the SWP East and SWPWest increaseby 28%, but the smallest increase is the LAA at 9%.

In terms of comparing the spatially explicitapproach with the statewide approach without includ-ing upstream emissions, for both utilities the overall

Figure 4.Comparative emissions burden for LAWDPand IEUAwater sources. Greenhouse gas emissions (gCO2e per acre feet) arelisted for eachwater sourcewith the shading indicatingwhich operational phase of water supply the emissions are attributable to.

7

Environ. Res. Lett. 10 (2015) 114002 A J Fang et al

Figure 5.Comparing emissions of eGRID (Statewide average) versus spatially explicit approach. For eachwater source, emissionsintensity (gCO2e/AF) is compared using a spatialmethodwith upstream emissions included and the conventional approach, using aregional emissions factor from eGRID’s CAMX subregion.

Figure 6.Relative impact of spatial and upstream emission factors. (a)Emissions intensity for eachwater source using eGRIDgeneration profiles and spatially explicit generation profiles, with the percentage indicating the percent difference using ‘eGRIDwithno upstream’ as a base case. (b)Emissions intensity for eachwater source using eGRID generation profiles and varying the inclusion ofupstream emissions factors. The percentage shown indicates the percentage difference using ‘eGRIDwith no upstream’ as the basecase.

8

Environ. Res. Lett. 10 (2015) 114002 A J Fang et al

emissions footprint is significantly lower (figure 6).LADWP’s average emissions intensity decreases by38% due to its heavy reliance on the SWP and CRA,where over 50% of its water supply is sourced. Emis-sions intensity of the SWP water falls by 40% and theCRA emissions falls by 70% as the eGRID CAMXemissions profile effectively underrepresents the pro-portion of electricity generated by hydropower used totransport the water. However, this is not the case forlocal sources of water (LAA, groundwater, recycledwater) which shows increases in emissions intensitydue to reliance on LADWP’s relatively dirty electricitygeneration mix. This has important implications asthe City of Los Angeles looks to expand efforts to gen-erate more supply through recycled water, stormwatercapture, and groundwater replenishment. All IEUAsources of water show a decrease in emissions intensityusing a spatialized emissions factor, resulting in a 29%decrease in theweighted average (figure 6).

4.Discussion

In addition to providing a detailed account of theenergy and emissions footprint of two major waterutilities in Southern California, this study exposes twoimportant considerations when making such esti-mates. First, using spatially explicit emissions factorsrather than regionalized averages provides a moreaccurate accounting of emissions intensity associatedwith sourcing, conveying, and treating water. As suchthe results call into question the validity of relying oneGRID emissions factors (e.g. regionalized averages)to estimate the carbon footprint of localized watersupply, which has been the predominant approach(Filion et al 2004, Stokes and Horvath 2011, SandersandWebber 2012,Mo et al 2014) to calculating carbonfootprints of water supplies in the US. For relativelyindependent grids, where little electricity tradingoccurs; this approach is valid (Weber et al 2010). Butusing a regional emissions factor omits importantvariations in state and site-specific electricity genera-tion, as Marriott and Matthews (2005) and Colett et al(2015) have shown. For water supplies in the WesternUnited States, if the energy or carbon intensity isconcentrated in the treatment stage, typical of loca-lized water sources, then a spatially explicit emissionsfactor is especially important.

Second, the results demonstrate the importance ofincluding upstream emissions for the various elec-tricity generation sources. This study has revealed thatthe carbon footprint of water supply increases by20%–30% (figure 6) depending on the portfolio ofelectricity it relies on. This has important implicationsnot only for local entities like water utilities, but alsofor the calculation of GHG emissions at the state,regional, and federal levels.

Uncertainty in future supply, coupled with envir-onmental mitigation in the Sacramento-San JoaquinDelta, Owens Valley, and Mono Lake basin, is drivingthe development of new sources to maintain a reliablewater supply. To meet demand and account for short-falls caused by drought and climate change, strategiesin Southern California are increasingly focused onexpanding local supply options such as large-scaledesalination, recycled water, stormwater and ground-water projects (Hughes et al 2013). As our study hasdemonstrated however, for Los Angeles in particular,these local supply sources (e.g. recycled water) canhave a higher carbon footprint than imported sources,such as water fromCRA and especially the LAA, whichis gravity fed. As a consequence, this may compoundthe water–energy nexus and produce an ironic situa-tion where those sources advocated to mitigate theeffects of climate changemay actually exacerbate them(Cousins andNewell 2015).

The environmental sustainability of local watersupply (in terms of carbon footprint), therefore, hin-ges on the electricity grid mix used to treat and dis-tribute water. To gain a sense of just how influentialthe local grid mix is consider the example of LADWPand its proposed transition away from coal andtowards a cleaner grid mix. Using LADWP’s fore-casted generation sources for 2020 and 2030(LADWP 2011) and Integrated Resource Plans fromrelevant utilities, we calculated future GHG emissionsto understand how LADWP’s energy transitionimpacts the carbon footprint of its water supply sys-tem (figure 7). In this scenario, LADWP decreases itscoal generation from 40% in 2010 to 28% in 2020 andto 0% in 2030. Meanwhile, the percentage of renew-able generation increases from 18% (2010) to 40%(2030) and the percentage of natural gas increasesfrom 30% (2010) to 50% (2030). Under this scenario,the reduction in carbon intensity of local water sources(LAA, groundwater, recycled) is especially pro-nounced (54%). Imported water, by comparison, isreduced just by 6%, 8%, and 10% for the SWP East,SWPWest, and CRA, respectively. LADWP could fol-low IEUA, which was able to mitigate the emissionsassociated with recycled water and groundwater byself-generatingmore of the electricity needed to powerits local water treatment plants. Of course, in additionto the grid mix, the energy intensity of the technologyhas to be considered as well. This is especially the casewith desalination, which is highly energy intensive andeven with a relatively clean energy grid, would have aconsiderable carbon footprint. Other considerationsalso include the extent to which efficiency improve-ments throughout the various phases of water pump-ing and transport, treatment, and distribution mayyield greater overall emissions reductions as result ofeconomies of scale.

9

Environ. Res. Lett. 10 (2015) 114002 A J Fang et al

5. Conclusion

Faced with unprecedented drought and climatechange vulnerabilities, water utilities across Californiaand the American Southwest are focused on increasinglocal sources of water supply. But as this study hasshown local sources can have a higher carbon footprintthan imported sources, at least with respect to thetransport, distribution and treatment phases. Theenergy grid mix of the electricity that is used in thesethree phases is crucial with respect to assessing thecarbon footprint of local versus imported watersupply. This was made apparent by the spatiallyexplicit approach utilized to determine the energy andemissions intensity of water supply for two utilitiessupplying Southern California: LADWP and IEUA.Comparing these two utilities revealed that the rela-tively dirty grid mix that LADWP relies on leads tomuch higher carbon footprints for local water sourcesthan for IEUA. As utilities in California work to meetnew state mandated emissions standards the spatiallyexplicit measurement of emissions provides a morerobust method than conventional eGRID emissionsfactors which do not sufficiently take into accountgeographic variability.

This study also clarified how upstream emissionsfor all types of energy sources is significant and needsto be considered as entities from local to federal scales

estimate the GHGs associated with their resource con-sumption. In the case of the two utilities we studied,the carbon footprint was 20%−30% higher whenupstream emissions were included, in addition tocombustion emissions.

Moving forward, utilities such as LADWP have anopportunity to greatly influence the emissions of theirlocal sources of water supply. This was clearly illu-strated in the case of LADWP’s transitioning awayfrom coal-based electricity towards more renewablesources. But this sustainability transition can only becaptured using the spatially explicit/upstreammodel-ing approach that we have introduced in this paper.This approach has implications not only for water andelectric utilities but also for local entities of all kinds asthey consider how best to transition to a lower carbonfuture.

Finally, it would be useful to extend this analysis toconsider how green water (i.e. utilized precipitationand/or soil moisture) and greywater (i.e. wastewaterpreviously used by residential/commercial end users)resources could be further utilized to help meet waterdemand. This study considered all water supplied tobe blue (i.e. potable water), while acknowledging thatrecycled water is currently not utilized for direct pota-ble water reuse. These types of water will becomeincreasingly important for places that are seeking todiversify water sources in order to enhance resilience,such asCalifornia and theAmericanWest.

Acknowledgments

The authors would like to thank Jordan Garfinkle,JohnWillard, and Brandon List for their contributionsto this project, the Haynes Foundation for providingfunding for the research, and the Inland EmpireUtilityAgency (IEUA), Metropolitan Water District (MWD),and Los Angeles Department of Water and Power(LADWP) for data and support.

References

BlancoH,Newell J P, Stott L andAlbertiM2012Water SupplyScarcity in Southern California: AssessingWater District LevelStrategies (LosAngeles, CA: Center for Sustainable Cities,Price School of Public Policy, University of SouthernCalifornia)

CDWR2011California StateWater Project at a Glance (Sacramento,CA:California Department ofWater Resources) (http://water.ca.gov/recreation/brochures/swp.cfm)

CEC2015California Electric Utility Service AreasMap (Sacramento,CA:California EnergyCommission) (http://energy.ca.gov/maps/serviceareas/Electric_Utility_Service_Areas.html)

ChesterMVandHorvathA 2009 Environmental assessment ofpassenger transportation should include infrastructure andsupply chainsEnviron. Res. Lett. 4 024008

Colett J S, Kelly J C andKeoleianGA2015Using nested averageelectricity allocation protocols to characterize electrical gridsin life cycle assessment: a case study ofUS primary aluminumproduction J. Ind. Ecol. at press ( doi:10.1111/jiec.12268)

CookeR, Jercich S, Trombly C andVillalobos P 2012 Managementof the California StateWater Project—Bulletin 132–12

Figure 7.Projected impact of gridmix changes for LAWDP.The projected carbon intensity (gCO2e/AF) of each LADWPwater source based upon electricity generation source projec-tions (LADWP2011) for 2010, 2020, and 2030.

10

Environ. Res. Lett. 10 (2015) 114002 A J Fang et al

pp 185–99 (www.water.ca.gov/swpao/docs/bulletins/bulletin132/Bulletin132-12.pdf)

Cousins J J andNewell J P 2015A political–industrial ecology ofwater supply infrastructure for Los AngelesGeoforum 5838–50

CPUC2010Embedded Energy inWater Studies, Study 1: StatewideandRegionalWater–Energy Relationship GEIConsultantsandNavigant Consulting for the California PublicUtilitiesCommission (31August 2010)—appendixN

CurranMA2006Report on activity of task force 1 in the life cycleinventory programme: data registry—global life cycleinventory data resources Int. J. Life Cycle Assess. 11 284–9

EPA 2009Clean Energy—eGRID (WashingtonDC:USEnvironmental ProtectionAgency) (http://epa.gov/cleanenergy/energy-resources/egrid/index.html)

FilionYR,MacLeanHL andKarney BW2004 Life-cycle energyanalysis of awater distribution system J. Infrastruct. Syst. 10120–30

Hughes S, Pincetl S and BooneC2013Triple exposure: regulatory,climatic, and political drivers of watermanagement changesin the city of Los AngelesCities 32 51–9

IEUA2009AnnualWaterUse Report for IEUA Service Area (ChinoHills, CA: Inland EmpireUtilities Agency) (http://ieua.org/wp-content/uploads/2014/11/fy09_10_annualwaterusereport.pdf)

IEUA2010UrbanWaterManagement Plan (ChinoHills, CA: InlandEmpireUtilities Agency) (http://ieua.org/urban-water-management-plan-2010-june-01-2011/)

KleinG, KrebsM,Hall V,O’Brien T andBlevins B B 2005California’sWater—Energy Relationship (http://energy.ca.gov/2005publications/CEC-700-2005-011/CEC-700-2005-011-SF.PDF)

LADWP2010UrbanWaterManagement Plan (Los Angeles: LosAngeles Department ofWater and Power) (http://water.ca.gov/urbanwatermanagement/2010uwmps/Los%20Angeles%20Department%20of%20Water%20and%20Power/LADWP%20UWMP_2010_LowRes.pdf)

Lifset R andGraedel T E 2002 Industrial ecology: goals anddefinitionsAHandbook of Industrial Ecology edRUAyres andLWAyres (Northampton,MA: Edward Elgar) pp 3–15

Marriott J andMatthewsHS 2005 Environmental effects ofinterstate power trading on electricity consumptionmixesEnviron. Sci. Technol. 39 8584–90

MoW,WangR andZimmerman J B 2014 Energy–water nexusanalysis of enhancedwater supply scenarios: a regionalcomparison of Tampa bay, Florida, and SanDiego, CaliforniaEnviron. Sci. Technol. 48 5883–91

MWD2006Power Integrated Resource Plan forMetropolitan’sColorado River Aqueduct Operations—2006 Revised (LosAngeles, CA:MetropolitanWaterDistrict of SouthernCalifornia) (http://ww2.wapa.gov/sites/western/es/irp/Documents/MetWaterDistIRP.pdf)

MWD2009EnergyManagement and Reliability Study (LosAngeles,CA:MetropolitanWaterDistrict of SouthernCalifornia)(http://collab.waterrf.org/Workshops/FLwater/WaterEnergy%20Issues%20Library/MWD%20Energy%20Mgmt%20and%20Reliability%20Study.pdf)

Newell J P andVosRO2011 ‘Papering’ over space and place:product carbon footprintmodeling in the global paperindustryAnn. Assoc. Am.Geogr. 101 730–41

LADWP2011Power Integrated Resources Plan (LosAngeles, CA: LosAngeles Department ofWater and Power) (http://swrcb.ca.gov/water_issues/programs/ocean/cwa316/docs/energy_comp/irp_2012_1.pdf)

PG&E2010 PowerContent Label (http://energy.ca.gov/sb1305/labels/2010_labels/PGE_PCL.pdf)

RogersMM,WangY,WangC,McElmurry S P andMiller C J 2013Evaluation of a rapid LMP-based approach for calculatingmarginal unit emissionsAppl. Energy 111 812–20

Sanders KT andWebberME 2012Evaluating the energy consumedforwater use in theUnited States Environ. Res. Lett. 7034034

SCE 2010 PowerContent Label (https://sce.com/wps/wcm/

connect/861a514f-a20b-4548-9526-66ce424b3baa/2010_SCE_PowerContentLabel.pdf?MOD=AJPERES)

Stokes J andHorvath A 2006 Life cycle energy assessment ofalternative water supply systems Int. J. Life Cycle Assess. 11335–43

Stokes J R andHorvath A 2009 Energy and air emission effects ofwater supplyEnviron. Sci. Technol. 43 2680–7

Stokes J andHorvath A 2011 Life-cycle assessment of urbanwaterprovision: tool and case study inCalifornia J. Infrastruct. Syst.17 15–24

WeberC L, Jaramillo P,Marriott J and Samaras C 2010 Life cycleassessment and grid electricity: what dowe know andwhatcanwe know? Environ. Sci. Technol. 44 1895–901

WilkinsonR 2000Methodology for Analysis of the Energy Intensity ofCalifornia’sWater Systems, and anAssessment ofMultiplePotential Benefits through IntegratedWater–Energy EfficiencyMeasures California Institute for Energy Efficiency

WilkinsonR 2007Analysis of the Energy Intensity ofWater SuppliesforWest BasinMunicipalWater District (Carson, CA:WestBasinMunicipalWaterDistrict) (http://westbasin.org/files/general-pdfs/Energy–UCSB-energy-study.pdf)

11

Environ. Res. Lett. 10 (2015) 114002 A J Fang et al