The carbon footprint of water management policy options

Page 1

Energy Policy 42 (2012) 201–212

Contents lists available at SciVerse ScienceDirect

Energy Policy

0301-42

doi:10.1

n Corr

E-m

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

The carbon footprint of water management policy options

Eleeja Shrestha, Sajjad Ahmad n, Walter Johnson, Jacimaria R. Batista

Department of Civil and Environmental Engineering, University of Nevada Las Vegas (UNLV), 4505 Maryland Parkway, Las Vegas, NV 89154-4015, USA

a r t i c l e i n f o

Article history:

Received 24 January 2011

Accepted 24 November 2011Available online 23 December 2011

Keywords:

Water conveyance

Carbon footprint

System dynamics

15/$ - see front matter & 2011 Elsevier Ltd. A

016/j.enpol.2011.11.074

esponding author. Tel.: þ1 702 895 5456; fax

ail address: [email protected] (S. Ahma

a b s t r a c t

The growing concerns of global warming and climate change have forced water providers to scrutinize

the energy for water production and the greenhouse gas (GHG) emissions associated with it. A system

dynamics model is developed to estimate the energy requirements to move water from the water

source to the distribution laterals of the Las Vegas Valley and to analyze the carbon footprint associated

with it. The results show that at present nearly 0.85 million megawatt hours per year (MWh/y) energy

is required for conveyance of water in distribution laterals of the Valley from Lake Mead resulting in

approximately 0.53 million metric tons of CO2 emissions per year. Considering the current mix of fuel

source, the energy and CO2 emissions will increase to 1.34 million MWh/y and 0.84 million metric tons

per year, respectively, by the year 2035. Various scenarios including change in population growth rate,

water conservation, increase in water reuse, change in the Lake level, change in fuel sources, change in

emission rates, and combination of multiple scenarios are analyzed to study their impact on energy

requirements and associated CO2 emissions.

& 2011 Elsevier Ltd. All rights reserved.

1. Introduction

With the growth in both population and economic develop-ment, the demand for water has been increasing (Morrison et al.,2009; Vedwan et al., 2008). Climate variability and changepresents additional water management challenge by impactinghydrologic events such as floods (Forsee and Ahmad, 2011,Mosquera-Machado and Ahmad, 2007; Ahmad and Simonovic,2001, 2005) and droughts (Ahmad et al., 2010; Stephen et al.,2010; Puri et al., 2011). In most developing countries, the qualityof existing freshwater sources is declining due to increasing waterpollution as untreated wastewater is directly disposed intonatural water sources (Eltawil et al., 2009; Von Uexkull, 2004).In addition, over-exploitation of groundwater is affecting theavailability of enough freshwater (Eltawil et al., 2009). In orderto ensure the availability of water for current as well as futureneeds, efficient, and sustainable water production strategies mustbe introduced. Sustainable water production, which involvessatisfying the current needs while ensuring the availability ofwater to meet the future needs (Darwish et al., 2008), also impliesminimizing the use of such resources as energy in the productionof water.

Water and energy are inextricably linked, and both are equallyimportant for economic and population growth (Lampe et al.,2009; Rio Carrillo and Frei, 2009). Water production – which

ll rights reserved.

: þ1 702 895 3936.

d).

involves extraction, treatment, transmission, distribution, use,and disposal of water – requires energy. Reduction in energyuse is a major goal for sustainable development of water supplysystems (Vieira and Ramos, 2009). In order to maintain a safe andreliable water supply, environmental impacts of water productiondue to greenhouse gas emissions should be minimal (Darwishet al., 2008; Strutt et al., 2008).

As a result of the growth in population and economic devel-opment, cities expand, requiring the transport of water from longdistances. Bringing water from these long-distance sourcesrequires massive water infrastructures and extensive use ofenergy. A vast amount of energy is consumed to extract, process,and deliver clean water (Morrison et al., 2009). In fact, electricityused for the purpose of water transport, compared to that neededfor water treatment and distribution, is the major source ofgreenhouse gases as well as the corresponding carbon footprint,and thereby contributes to global warming and climate change(Stokes and Horvath, 2009). The related energy consumptiondepends on the quantity of water and on the topography of thedistribution network (Bakhshi and Demonsabert, 2009; Pelli andHitz, 2000;Reiling et al., 2009). That is, the spatial distribution ofwater users from water sources is major energy use determinant(Pelli and Hitz, 2000).

Nearly 3–4% of the total U.S. electricity use is for moving andtreating water and wastewater (EPRI, 2002; Reiling et al., 2009;USDOE, 2006; USEPA, 2009a). Costs associated with energy orelectricity use account for nearly 80% of municipal water proces-sing and distribution costs (EPRI, 2002). On average, 85% of thiselectricity is used for pumping water in the distribution system,

Page 2

E. Shrestha et al. / Energy Policy 42 (2012) 201–212202

9% for pumping raw water to the treatment plant and 6% for thetreatment processes (Reiling et al., 2009). The reduction in energyuse has dual benefits, reduction in the cost of water productionand reduction in emissions of greenhouse gases (GHGs).

The use of energy contributes to carbon footprint, which is ameasure of the total amount of greenhouse gases, expressed ascarbon dioxide equivalents (CO2e), that directly and indirectlyresult from an activity or are accumulated over the life stages of aproduct (Strutt et al., 2008; Wiedmann and Minx, 2008). Theprincipal greenhouse gases entering the atmosphere due tohuman activities, and also contributing most to the carbonfootprint, are carbon dioxide (CO2); methane (CH4); nitrous oxide(N2O); and fluorinated gases, such as hydrofluorocarbons, per-fluorocarbons, and sulfur hexafluoride (Strutt et al., 2008; USEPA,2010). Each of these gases has a different potential to trap theheat in the atmosphere, the least being CO2. However, CO2 isproduced in such large quantity that all greenhouse gases areconverted into CO2 equivalent (CO2e) in order to ease thecalculation of the total footprint of all gases. For a 100-year timehorizon, the global warming potential for anthropogenic GHGs, ascompared to CO2, is 21 for CH4 and 310 for N2O; for fluorinatedgases, it varies from 140 to 23,900 (Forster et al., 2007; USEPA,2009b). Depending on the source of energy for electricity gen-eration, the size of carbon footprint varies. For example, fossilfuels have the highest carbon footprint, whereas such renewabletechnologies as geothermal, hydroelectric, solar, and wind havethe lowest. The carbon footprint related to water in the U.S.accounts for 5% of all U.S. carbon emissions (Griffiths-Sattenspieland Wilson, 2009). The emissions due to water use are likely toincrease in the future due to growing water demand, limited andremote locations of the freshwater sources, and stringent andenergy intensive water treatment regulations and technologies(Griffiths-Sattenspiel and Wilson, 2009).

One goal of developing a sustainable water production systemis to reduce the carbon footprint. However, as the populationgrows and economic conditions improve in developing countries,treated water becomes more affordable. Therefore, the carbonfootprint of water is likely to increase unless water managementpolicies are implemented that support sustainability. The carbonfootprint of water could be decreased by addressing both thewater supply side, for instance, by means of water conveyanceand treatment technologies. The demand side could also beaddressed through water reuse and conservation, as an example.

In this research, a dynamic model is developed that canevaluate different options for water systems in terms of energyuse and the associated carbon footprints. Such a model can informpolicymakers when deciding on which policies will reduce thecarbon footprint of water systems.

The water system of the Las Vegas Valley (LVV) in Nevada, U.S.is used to demonstrate the applicability of the proposed dynamicmodel and its value in helping policymakers in making informedchoices. The impacts of water conservation policies, water reuse,and energy source type on the carbon footprint of water transportin the LVV is investigated. However, the approach used in thisstudy as well as the policies tested, have broader application topotable water systems throughout the world.

Specifically, the energy use and CO2 emissions associated withvarious water management policy scenarios were compared:

(i)

A status quo scenario, which provides a baseline for compar-ison of different policy options; status quo relates to thecarbon footprints of the current water system.

(ii)

A population growth scenario, in which the water carbonfootprint is evaluated for increased population growth.

(iii)

A water conservation scenario, where the impact of variousconservation measures on the carbon footprint is evaluated.

(iv)

A water reuse scenario to compare the impact of variouslevels of reuse on the carbon footprint.

(v)

A water source depth scenario, in which the effect on thecarbon footprint of decreasing lake levels due to drought isevaluated.

(vi)

A combination of scenarios, where the impact on the carbonfootprint is examined for various policies that are applied atthe same time.

2. Research approach

2.1. System dynamics modeling

A dynamic simulation model using system dynamics (SD) wasdeveloped to facilitate the computation of energy use as well asthe carbon footprint of water conveyance through major lateralsin the Las Vegas Valley. For this purpose, the SD software Stellas

(www.hps-inc.com) was used.System dynamics is a method to understand the behavior of

complex systems over time (Sterman, 2000). It involves theformation of simulation models of complete systems over time;the variable components are linked with each other throughfeedback loops (Spang, 2007). Simulation models play an impor-tant role in understanding complex problems addressed in waterresources management.

A review of system dynamics applications for water manage-ment is provided by Winz et al. (2009). Examples of systemdynamics simulation models used to address water resourcesmanagement problems include: a water consumption model tounderstand the system behavior due to water saving, wastewaterreuse, and water transfer (Zhang et al., 2009); a simulation modelfor municipal water conservation policy analysis (Ahmad andPrashar, 2010; Qaiser et al., 2011); a decision-support model forcommunity-based water planning (Tidwell et al., 2004) as well asfor investigating water trading/leasing and transfer schemes(Gastelum et al., 2010); a water balance model for irrigationmanagement (Khan et al., 2009) and flood management (Ahmadand Simonovic, 2006; Simonovic and Ahmad, 2005); and areservoir operation model (Ahmad and Simonovic, 2000). Othernotable examples of system dynamics models for water manage-ment and policy analysis include: a spatial system dynamicsmodel, developed by integrating system dynamics and a geo-graphic information system (Ahmad and Simonovic, 2004); amodel for water resources policy analysis (Simonovic andFahmy, 1999); a simulation model to gage public awareness ofthe importance of water conservation (Stave, 2003); a dynamicmodel to evaluate salinity load and the impacts of water reuse inenergy consumption and salinity control (Venkatesan et al.,2011a,b); and a simulation model to compute energy use andthe associated carbon footprint for water supply alternatives(Shrestha et al., 2011).

2.2. Model water system

The major water source for the LVV is water from the ColoradoRiver located in southern Nevada (Fig. 1), passing through LakeMead, which is located 32.2 km away from the LVV. Almost 90% ofthe LVV’s water needs are met by Colorado River water (SNWA,2009a). The remaining 10% comes from local groundwatersources (SNWA, 2010a). To move water from Lake Mead to theLVV requires nearly a lift of 365.8 m (m); this requires a greatamount of energy for pumping and has an associated carbonfootprint that is large.

Nevada has the consumptive water use right of 0.4 km3

(300,000 acre-ft) of Colorado River water per year (LVVWAC,

Page 3

E. Shrestha et al. / Energy Policy 42 (2012) 201–212 203

2009). The Southern Nevada Water Authority (SNWA), whichmanages the water supply and distribution to local water agen-cies in the LVV, operates two intake systems that lift ColoradoRiver water from Lake Mead to either of its two water treatmentplants, the Alfred Merritt Smith Water Treatment Facility(AMSWTF) and the River Mountains Water Treatment Facility(RMWTF).

A schematic diagram of water conveyance in the LVV is shownin Fig. 2. Two major intake pumping stations and two boosterpumping stations deliver water to the water treatment plants.

Fig. 2. Schematic of water conveyance system in the La

Fig. 1. Location of the Las Vegas Valley in Southern Nevada, USA.

The AMSWTF is designed to treat 26.3 m3/s (600 mgd) andRMWTF can treat up to 13.1 m3/s (300 mgd) (SNWA, 2010b).RMWTF is designed in such a way that it can expand to 26.3 m3/s(600 mgd) to meet future water needs (SNWA, 2010b). Thetreated water from AMSWTF is transmitted to the LVV througha tunnel of diameter 3 m and five major laterals, namely, BoulderCity lateral (0.9 m diameter), East Valley lateral (2 m diameter),North Las Vegas lateral (1.8 m diameter), Pittman lateral (2.6 mdiameter) and the Henderson lateral (0.9 m diameter). Thetreated water from AMSWTF is also pumped to RMWTF throughthe Foothills Pumping Station when required. Similarly, treatedwater from RMWTF is distributed to the South Valley (2.7 mdiameter) and R-8 (0.8 m diameter) laterals after passing througha 3.7 m diameter tunnel. In addition, untreated water fromupstream of RMWTF is pumped to a golf course in Boulder Citythrough Boulder City Raw Water Pumping Station.

At present, there are more than two dozen pumping stations tofacilitate the conveyance of the treated water. The associatedenergy requirements and the corresponding carbon footprint formoving water are likely to increase in future due to increasedwater demand because of population growth. In addition, theincreased pumping head due to declining water levels (static lift)in Lake Mead and increased friction head (dynamic head). Thewater supplied in the Valley is either used indoors or outdoors.The water used outdoor for landscape or in golf courses irrigation,due to the arid environment, is lost to the atmosphere throughevaporation and evapotranspiration, and contributes to shallowsubsurface soil moisture, or flows to the Las Vegas Wash as urbanrunoff (Stave, 2003). The indoor used water is sent to one of thethree wastewater treatment plants. The treated effluent from thewastewater treatment plants is returned back to Lake Meadthrough the Las Vegas Wash.

s Vegas Valley (adapted from Shrestha et al., 2011).

Page 4

E. Shrestha et al. / Energy Policy 42 (2012) 201–212204

According to Clark County Sewage and Wastewater AdvisoryCommittee (SWAC) reports, 43% of the water supplied is currentlyused indoors, while 57% is used outdoors and is generally forlandscape purposes. The indoor used water is treated in threewastewater treatment plants. Almost 90% of the treated effluentis discharged back into Lake Mead through the Las Vegas Washwhile the remaining is used for landscape irrigation and coolingtower make-up water. Depending on the amount of treatedwastewater discharge, Nevada can actually withdraw additionalwater from the Colorado River. This additional amount is knownas return flow credits. The Las Vegas Wash flows are comprised ofnot only treated wastewater effluent, but also urban runoff,intercepted shallow groundwater, and stormwater. Nevada actu-ally receives return flow credits only for the Colorado River waterreturned back to the Lake Mead (LVWCAMP, 1999).

2.3. Data sources

The data to construct the model were obtained from varioussources:

�

Population data were obtained from Center for Business andEconomic Research (CBER, 2009) and Clark County Depart-ment of Comprehensive Planning, Demographics (www.accessclarkcounty.com). This includes only permanent populationof the Valley and does not include tourist population. Thepermanent population in the year 2003 was nearly 1.6 million,which gradually increased to around 1.9 million in the year2009 and is projected to reach approximately 3.2 million bythe year 2035. The historical annual population growth ratehas averaged 3.4% per year between 2003 and 2009. Theaverage annual forecasted population growth rate is estimatedto be 1.6%. The future population growth rate used in themodel is in accordance with the CBER forecasted growth rate. � Per capita water demand data were obtained from Southern

Nevada Water Authority (SNWA, 2009a,b). The per capitawater demand in the LVV has decreased from 1,113 l percapita per day (l pcd) (294 gallons per capita per day (gpcd)) in2003 to 908 lpcd (240 gpcd) in 2009, and it is expected todecrease to 753 lpcd (199 gpcd) by the year 2035.

� Lake level was obtained from U.S. Bureau of Reclamation

(2010). The future lake level is assumed to be constant at335 m (1099 ft) above mean sea level (amsl), which is theaverage lake level for the year 2009.

� Indoor and outdoor water use rate was obtained from Clark

County Sewage and Wastewater Advisory Committee (SWAC,2009). Approximately 57% of the water pumped into the Valleyis used outdoors for landscape irrigation and is lost throughinfiltration and evapotranspiration. The remaining 43% of thewater is used indoors and ends up as wastewater.

� Reuse rate of treated wastewater was obtained from SWAC

(2009) and Clark County, Nevada (CCN, 2000). On average, 10%of the treated effluent from wastewater treatment is reused.However, the reuse of treated effluent has increased from25 MCM (18 mgd) in 2003 to nearly 30 MCM (22 mgd) in 2008and is expected to reach 77 MCM (56 mgd) by 2020.

� Urban runoff and intercepted shallow groundwater was obtained

from Las Vegas Wash Comprehensive Adaptive ManagementPlant (LVWCAMP, 1999). The urban runoff and interceptedshallow groundwater was assumed to be 30 MCM (25,000) afythroughout the study.

� Source of energy for electricity was obtained from U.S. Energy

Information Administration (USEIA, 2009). The state of Nevada’senergy mix for the year 2007 was used for future computations.

� Average emission rates for coal (1022.9 g CO2e/kWh), oil

(779.6 g CO2e/kWh), natural gas (605.9 g CO2e/kWh), sloar/PV

(70.8 g CO2e/kWh), hydroelectric (25.4 g CO2e/kWh), andgeothermal (66.7 g CO2e/kWh) were obtained from the litera-ture (Shrestha et al., 2011).

2.4. Model components

The SD model used in this study developed estimates for theenergy requirement and consequent carbon footprint of watersupply and conveyance in the LVV. The model is comprised ofthree major sectors: water demand sector; water supply andenergy use sector; and carbon footprint sector. These sectors aredirectly or indirectly connected and influence the behavior of oneanother.

2.4.1. Water demand sector

The water demand sector computes the total water demandand demand fulfilled by Colorado River water based on thepopulation and per capita water demand for a simulation periodranging from 2003 to 2035. However, the model allows forvariation of the future population growth rate. The water demandto be fulfilled by Colorado River water is computed by subtractingthe groundwater resource and wastewater reuse.

2.4.2. Water supply and energy use sector

Water supply and energy use sector is the main sector of thesystem that incorporates all the major pumping stations andcomputes the energy requirements. Water flow in the systemshown in Fig. 2 is captured in this sector along with water use inthe Valley, wastewater collection, water reuse and discharge oftreated effluent back into the Lake Mead. The pumping powerrequirement is calculated within the SD model using the equa-tion:

P¼gQH

Zð1Þ

where P is the power, g is the specific weight of water, Q is theflow rate in, H is the total dynamic head, and Z is the overallpump efficiency. A pump efficiency of 93% and a motor efficiencyof 80% were used. These values reflect the average efficiency forthe water pump systems in the LVV.

The total dynamic head includes the static head and head lossdue to friction while the other minor losses are ignored. The headloss due to friction is calculated using the equation:

hL ¼f LV2

2gDð2Þ

where f is a coefficient of friction, L is the length of pipe, V is thevelocity, g is the acceleration due to gravity, and D is the insidepipe diameter.

The coefficient of friction is calculated using an empiricalequation developed by Swamee and Jain (Jones et al., 2008):

f ¼0:25

½log10ðððe=DÞ=3:7Þþð5:74=R0:9ÞÞ�2

ð3Þ

where e is the absolute roughness and R is Reynolds number.The pumping energy is calculated assuming the pumps are

operated 90% of the time. The annual pumping energy is calcu-lated for each pumping station in kilowatt hour per year. Theenergy calculation is only for moving water from the source to thedistribution laterals. It does not include the energy requirementsfor water moving in the potable water distribution system, or theenergy requirements in the wastewater collection and treatmentsystems.

Page 5

E. Shrestha et al. / Energy Policy 42 (2012) 201–212 205

2.4.3. Carbon footprint sector

The carbon footprint sector calculates the associated carbonfootprint of moving water in the system based on the energysource used in pumping water. The electric power sources for thestate of Nevada until 2006 were coal, natural gas, petroleum,hydroelectric power, and geothermal (USEIA, 2009). In 2007,solar/PV provided 0.13% of the state’s electric power supply(Table 1). The 2007 energy source distribution and 2025 targetof 25% renewable energy sources in electricity generation, set byPublic Utilities Commission of Nevada (PUCN) (2009), were usedto compute the future carbon emissions. The total carbon foot-print is calculated using the CO2 emission rates. The emissionrates vary depending upon the electricity generating plant effi-ciency, its technological options and carbon/heat content of thefuel when electricity generation is due to direct combustion offuel (Evans et al., 2009; Weisser, 2007).

3. Results

The SD model is developed to analyze energy and conse-quently the associated carbon footprint that is required to movewater in the conveyance system of the LVV. Before any policy isanalyzed, the model should be validated against the observeddata so that a sense of credibility and confidence is established,and historical behavior is realistically replicated (Sterman, 2000).A seven-year period from 2003 to 2009 is used as the verificationperiod in the model and a 26 year period from 2010 to 2035 isused as a planning horizon with a yearly time step. The model isable to accurately replicate the historical population trend

Table 12003 and 2007 electricity source distribution for

the state of Nevada (USEIA, 2009).

Source Percent of total electric powersector consumption in

2003 2007

Coal 52.67 25.95

Natural gas 35.26 58.59

Oil 0.06 0.03

Hydro 5.35 6.57

Geothermal 6.66 8.73

Solar/PV – 0.13

0.2

0.4

0.6

0.8

1.0

1.2

1.4

2003

Ener

gy(m

illio

n M

Wh/

y)

YearTotal LM to WTP

WTP to DS

2013 2023 2033

Fig. 3. Energy for moving water from Lake Mead (LM) to water treatment plant (WTP),

corresponding CO2 emissions. (a) Energy and (b) Co2 emissions.

obtained from the Clark County Department of Comprehensiveplanning demographics.

In a similar way, the model simulation for water demand ofthe LVV was comparable to the historic water demand of theValley. The model was also tested for extreme conditions.Extreme condition checks if the behavior of the model is appro-priate when the extreme values are provided as an input(Sterman, 2000). Some of the extreme condition tests includedzero population, no change in population and zero Lake level.In all these tests, the model behavior was as anticipated.

3.1. Status quo scenario

For the status quo scenario, it is assumed that the populationincreases as predicted by the CBER. In addition, the per capitademand is assumed to remain constant at 908 lpcd (240 gpcd) asfor the 2009 and onwards. Also, of the total water supplied, 43% isassumed for indoor use, while the remaining is assumed foroutdoor use. The reuse of treated effluent from wastewatertreatment plants is assumed to remain constant at nearly 30million cubic meters (MCM) (22 mgd) (2008 value; the latestavailable) throughout the period from 2009 and onwards. Theremaining treated effluent is returned back to Lake Mead throughthe Las Vegas wash. The supply of water is assumed to beunlimited. The Lake level does not fluctuate. There is no variationin the state’s fuel source for electricity. The same assumptions areused for other scenarios as well unless otherwise mentioned.Some of these assumptions are later challenged by means ofsensitivity analysis.

For status quo scenario, Fig. 3 shows the total energy andassociated carbon footprint for moving water from the source tothe conveyance system in the LVV. It also shows in the disag-gregate form in terms of moving water from the source to watertreatment plants and then from the water treatment plants to theconveyance system of the Valley. The total energy consumption inthe year 2009 is nearly 0.85 million MWh enough to light nearly77,000 homes on average for a year in the United States, a statisticthat is based on an average annual electricity consumption of11,040 kWh for a US residential home in 2008 (USEIA, 2010).

In order to lift water from Lake Mead to the water treatmentplants, approximately 35% of the total energy use is required.However, there are only four pumping stations for this purposewith the pump horsepower varying from 1000 HP to 4000 HP. As

0.0

0.2

0.4

0.6

0.8

1.0

2003

CO

2 em

issi

ons

(mill

ion

met

ric to

ns/y

)

YearTotal LM to WTP

WTP to DS

2013 2023 2033

from WTP to distribution system (DS), and total energy for the whole system, and

Page 6

E. Shrestha et al. / Energy Policy 42 (2012) 201–212206

compared to more than 2 dozen pumping stations (pump horse-power range from 60 HP to 3500 HP) in the distribution system, a35% of the total energy consumption for only four pumpingstations is substantial.

There is a gradual rise in energy consumption due to anincreasing demand for water; this trend is predicted to continue.Energy consumption is directly proportional to the water demand.The CO2 emissions are based on the state’s electricity mix and theemission rates for each energy source. The CO2 emissions graduallyincreased with each year until 2005 when there was a sudden dropto 0.09 million metric tons of CO2, a nearly 15.5% decrease—eventhough the energy consumption during that period increased by1.3%. This is due to the fact that in the year 2005, the coalconsumption rate was decreased by nearly 45% and in turn theconsumption rate of natural gas was increased approximately bythe same amount. Because coal has higher CO2 emission potentialas compared to natural gas, there was a decrease in the total CO2

emission by nearly 0.09 million metric tons.The emission of greenhouse gases depend on the carbon

content of the fuel (for fuels such as black and brown coal);electricity generation technologies (such as steam turbine, opencycle gas turbine, and combined cycle gas turbine), the thermalefficiency of fuel, and plant capacity factor (IPCC, 2000; Lenzen,2008). Greenhouse gas emissions can also vary depending onlocation, therefore, the use of the average emission rate based ondifferent literature review may not be realistic. In this research, toaccount for the uncertainty associated with the average emissionrate, a thousand iterations of a Monte-Carlo simulation (IPCC,2000) was conducted, each time with an uncertain emissionfactor chosen randomly by the model within the distribution ofuncertainty that was initially specified in order to calculate thetotal CO2 emissions for water distribution. A uniform distributionwas chosen for the purpose since there was no useful informationavailable on the distribution of emission factors (Winiwarter andRypdal, 2001).

Fig. 4 shows the box plot of the range of total CO2 emissionsassociated with the water production in the LVV due to change inemission factors. The center line in the rectangular box representsthe median of the data set. The upper and lower lines of the

2015201020052003

1.3

1.2

1.1

1.0

0.9

0.8

0.7

0.6

0.5

0.4

Y

CO

2 em

issi

ons

(mill

ion

met

ric to

ns/y

)

Fig. 4. Box plot of tot

rectangular box stand for the third quartile (75th percentile) andfirst quartile (25th percentile), respectively. The lines that extendfrom the rectangular box (whiskers) give the minimum andmaximum value of the data set. In 2035, the CO2 emissions areestimated to vary between 0.73 million metric t/yr (first quartile)to 1.02 million metric t/yr (third quartile).

The total CO2 as shown in Fig. 3b is due to the aggregation ofCO2 due to individual energy sources in accordance with thestate’s electricity mix. Except for oil, the non-renewable energysources are the major contributors of total CO2 emissions. Emis-sions due to oil consumption and other renewable resources arealmost negligible. The use of oil for electricity generation ascompared to other sources is quite small.

Nearly 85% of the current electricity resource mix for the stateof Nevada is composed of such non-renewable resources as coal,oil and natural gas; the remaining 15% comes from renewableresources such as solar, geothermal, and hydroelectric. In order tocompare CO2 emission due to change in resource mix, a modelsimulation was carried out varying the contribution of non-renewable resources in the generation mix from 100% to 0% andcorrespondingly the percent contribution due to renewableresources. The change in CO2 emissions is shown in Fig. 5. Theuse of 100% renewable resources may not be a completelyrealistic scenario from an operational point of view. However,according to a Renewable Portfolio Standard (RPS) established byPublic Utilities Commission of Nevada (PUCN) (2009), the goal foruse of renewable energy sources in electricity generation is set to25% by 2025. Hence, changing the resource mix such that 25% ofthe energy comes from renewable resources decreased the totalCO2 emissions by nearly 10.4% (0.09 million metric t/yr) by 2035.

3.2. Change in estimated population growth rate scenario

If the population grows as predicted by CBER, then by 2035,nearly 1.34 million MWh/y of energy will be required to movewater from source to the distribution system. As a consequence,nearly 0.84 million metric tons of CO2 per year will be released(Fig. 6). If the predicted population growth rate is varied by 70.5%,

2035203020252020ear

al CO2 emissions.

Page 7

E. Shrestha et al. / Energy Policy 42 (2012) 201–212 207

the energy and associated CO2 will vary by 12.8% on average. Thismeans that by year 2035, even a 0.5% change in predicted popula-tion growth rate may lower or augment the energy requirements by

0.7

0.9

1.1

1.3

1.5

1.7

2003

Ener

gy(m

illio

n M

Wh/

y)

YearStatus quo Change by ±0.5%

2013 2023 2033

Fig. 6. Energy and corresponding CO2 emissions when annual population change rate is inc

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

2003

Ener

gy(m

illio

n M

Wh/

y)

YearStatus quo Conservation

2013 2023 2033

Fig. 7. Energy and associated CO2 emissions for indoor and outd

Fig. 5. CO2 emissions in Nevada due to varying non-renewable resource con-

tribution in the total resource mix.

0.17 million MWh/y, which is adequate to light nearly 15,400homes for a year in the U.S. or 0.11 million metric tons of CO2. A0.5% change in the estimated population growth rate will result in achange in population by 0.41 million, as compared to the 3.2 millionstatus quo population in the year 2035.

3.3. Water conservation scenario

The per capita water demand has decreased from 1113 lpcd(294 gpcd) in the year 2003 to 908 lpcd (240 gpcd) in the year2009; the goal is to further decrease it to 753 lpcd (199 gpcd) by theyear 2035. Fig. 7 shows the energy and corresponding CO2 emissionsassuming that the conservation goal of 753 lpcd (199 gpcd) waterdemand is fulfilled by the year 2035. Water conservation decreasesthe energy requirements by 16.5%, as compared to the status quoscenario. This corresponds to as much as 0.22 million MWh/y ofenergy consumption, adequate for nearly 20,000 US homes for ayear, or 0.14 million metric tons of CO2 per year.

3.4. Water reuse increase scenario

On average, 10% of the treated effluent from wastewatertreatment is reused. However, the reuse of treated effluent hasincreased from 25 MCM (18 mgd) in 2003 to nearly 30 MCM

0.4

0.5

0.6

0.7

0.8

0.9

1.0

2003

CO

2 em

issi

ons

(mill

ion

met

ric to

ns/y

)

YearStatus quo Change by ±0.5%

2013 2023 2033

reased or decreased by 0.5% in the Las Vegas Valley. (a) Energy and (b) Co2 emissions.

0.4

0.5

0.6

0.7

0.8

0.9

2003

CO

2 em

issi

ons.

(mill

ion

met

ric to

ns/y

)

YearStatus quo Conservation

2013 2023 2033

oor conservation scenario. (a) Energy and (b) Co2 emissions.

Page 8

Fig. 8. Energy and CO2 emissions when reuse is varied from 77 MCM reuse by 2020 to 100% reuse at an increase interval of 20%. (a) Energy and (b) Co2 emissions.

0.7

0.9

1.1

1.3

1.5

2003Year

Status quo (335 m (1099 ft) lake level)

At 320 m (1050 ft) lake level

At 350 m (1150 ft) lake level

0.4

0.5

0.6

0.7

0.8

0.9

2003

CO

2 em

issi

ons

(mill

ion

met

ric to

ns/y

)

YearStatus quo (335m (1099 ft) lake level)

At 320 m (1050 ft) lake level

At 350 m (1150 ft) lake level

Ener

gy(m

illio

n M

Wh/

y)

2013 2023 2033 2013 2023 2033

Fig. 9. Energy and CO2 emissions when Lake level is altered. (a) Energy and (b) Co2 emissions.

E. Shrestha et al. / Energy Policy 42 (2012) 201–212208

(22 mgd) in 2008; it is expected to reach 77 MCM (56 mgd) by2020 (CCN, 2000). Fig. 8 shows the energy requirements andassociated CO2 emissions for the cases due to change in reuserates. In 77 MCM reuse scenario (Fig. 8) it is assumed that thereuse rate will vary gradually from 30 MCM (22 mgd) in the year2009 to 77 MCM (56 mgd) by 2020 and remain constant onwards.This results in the decrease of energy use and associated CO2

emissions by nearly 3.6% by 2035. The energy use is decreased bynearly 0.05 million MWh/y, sufficient for nearly 4500 US residen-tial homes on average and associated CO2 emissions by nearly0.03 million metric t/yr. The other lines in Fig. 8 represent thescenarios in which the reuse of treated effluent is varied from 20%to 100%. For example, reusing 20% of the treated wastewater(nearly 127 MCM or 92 mgd) within the Valley can reduce theenergy requirements and the CO2 emissions by nearly 9% by 2035.This is a total decrease in energy consumption by 0.12 mil-lion MWh/y, enough to light 11,000 US homes on average for ayear and associated CO2 emissions by 0.08 million metric t/yr.

3.5. Change in the lake level scenario

The level of Lake Mead has been continuously declining since1997 and is expected to decline even more in coming years (Barnett

and Pierce, 2008; USBR, 2010). If the Lake level declines to 320 m(1050 ft), the level below which intake 1 will be out of operation(Feroz et al., 2007), the total energy requirements as compared tostatus quo (335 m or 1099 ft at Lake level) will increase by 3.3%. Also,the CO2 emissions will increase by the same rate. Likewise, the rise inLake Level to 350 m (1150 ft) will alter the energy requirement andCO2 emissions by same ratio (Fig. 9).

3.6. Combination scenario

The combination scenario involves water conservation to753 lpcd (199 gpcd) by 2035, reuse increase to 77 MCM (56 mgd)by 2020, and a change in fuel mix to 25% renewable resources. Thesescenarios are selected because these are the future policy goals set bySNWA (2009b), CCN (2000), and PUCN (2009), respectively. Fig. 10illustrates that the combination of these scenarios result in thedecrease of energy use by 20.7% (0.28 million MWh/y) and asso-ciated CO2 emissions by 29% (0.24 million metric t/yr), as comparedto the status quo scenario, which is adequate to light nearly 25,400U.S. homes on average for a year. The summary of results for abovementioned scenarios is presented in Table 2. The values reported arefor the year 2035. A comparison of energy savings, due to waterconservation and reuse, summarized from other studies is presented

Page 9

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

2003

Ener

gy(m

illio

n M

Wh/

y)

YearStatus quo Combination

0.3

0.4

0.5

0.6

0.7

0.8

0.9

2003

CO

2 em

issi

ons

(mill

ion

met

ric to

ns/y

)

YearStatus quo Combination

2013 2023 2033 2013 2023 2033

Fig. 10. Combination of scenarios—water conservation, increase in reuse of treated wastewater, and increase in use of renewable energy sources. (a) Energy and (b) Co2 emissions.

Table 2Summary of results.

Scenario Energy(millionMWh/y)

CO2

emissions(million -metric t/y)

Percentchange fromstatus quo (%)

Status quo 1.34 0.84

Change in estimated population growth rate

þ0.5% 1.53 0.96 712.8

�0.5% 1.18 0.74

Water conservation 1.12 0.71 �16.5

Water reuse increase

to 77 MCM by 2020

1.3 0.81 �3.6

Change in the lake level

þ15 m 1.3 0.82 73.3

�15 m 1.39 0.87

Change in resource

mix as 3:1

non-renewable to

renewable resource

1.34 0.76 (�10.4)

Combination scenario 1.07 0.6 �20.7 (�29)a

a The numbers in parenthesis are for CO2 emissions for respective scenario.

Table 3Comparison of energy saving due to conservation and reuse with other studies.

Study region Energy saving due to Source

Conservation(%)

Reuse

Las Vegas, Nevada, USA 16.5 3.6% Results from this study

San Diego, California, USA 12.7 8.9% NRDC (2004)

Florianopolis, Brazil 18.80 24.1% Proenca et al. (2011)

Albania 15–17 – Zavalani and Luga (2010)

Gauteng, South Africa 20 – Wyma (2008)

E. Shrestha et al. / Energy Policy 42 (2012) 201–212 209

in Table 3. It suggests that application of water conservation andreuse can save considerable amount of energy resulting in decreasein carbon footprint.

4. Discussion

A system dynamics model was developed to analyze theenergy requirements for water conveyance in the LVV and, as

its consequence, the carbon footprint of the system. This studyexplored the relationship of energy for water and associated CO2

emissions. The model simulations showed that a significantamount of energy is required to satisfy the water needs of theLVV; this need will increase substantially (nearly 58%) by the year2035, provided that the population grows as predicted by CBER. Ifenergy mix does not change, CO2 emissions will rise to 0.84million metric tons by 2035, a 58% increase. If the renewableenergy use increases to 25%, the increase in CO2 emission will be42%. A considerable amount of energy is required to pump waterfrom Lake Mead to water treatment plants; in fact, this energyrequires constitutes nearly 35% of the total energy requirementsfor water production in southern Nevada, as opposed to the U.S.average of 9% for pumping raw water to the treatment plant.However, a major portion of the total energy requirement isconsumed in moving treated water in the distribution system(65%). In California, water-related energy constitutes 19% of thestate’s total energy use, and includes energy for conveyance,storage, treatment, distribution, wastewater collection, treatment,and discharge (CEC, 2007).

Population growth rate change scenario indicated that thechange in population growth rate by even 0.5% (70.41 million)can change the energy and CO2 emissions by 12.8% as comparedto status quo. Likewise, a change in the Lake level did not changethe energy requirements and CO2 release by any significantamount. However, conserving water resulted in 16.5% reductionin energy consumption and associated CO2 emissions. Also,reducing water use can lower energy consumption by significantamount. For instance, the Natural Resources Defense Council(NRDC) (2004) reported that water conservation measuresapplied in San Diego could save enough energy to provideelectricity for 25% of all of the households in San Diego. Applyingconservation measures, the estimated energy saving in Texas is330–859 million kWh per annum and reduction in CO2 emissionsby 0.17–0.42 million metric tons each year (Stillwell and Webber,2010).

Increasing the reuse rate of treated wastewater effluent withinthe Valley can lower the energy requirements and associated CO2

emissions needed to move water in the LVV by considerableamount. However, the increase in reuse to 77 MCM (56 mgd) by2020 within the Valley only lowers the energy use by only 3.6%,sufficient to light approximately 4500 US homes for a year inaverage. Reusing water is far less energy intensive than transport-ing water from distant source locations. A water recycling systemin Orange County in California uses only half the amount ofenergy required to transport the same volume of water from

Page 10

E. Shrestha et al. / Energy Policy 42 (2012) 201–212210

northern California (NRDC, 2004). This results in the reduction ofCO2 emissions by 79%, which is equivalent to taking nearly 500cars off the road for a year (Taffler et al., 2008). Stillwell andWebber (2010) have reported that if 12% of the total waterdemand in Texas is filled by reuse water, the estimated decreasein energy consumption is 73–310 million kWh annually resultingin associated yearly CO2 emissions decrease by 0.04–0.16 millionmetric tons.

The combination of multiple scenarios—including water con-servation, increase in reuse of treated wastewater within theValley and increase in the use of renewable sources decreased theenergy requirements by nearly 20.7% and associated CO2 emis-sions by about 29%. This reduces energy and associated CO2

emissions by approximately 0.28 million MWh/y and 0.24million metric t/yr, respectively, when compared with the statusquo scenario. The combination scenario appears to be the mostenergy efficient scenario.

Another factor that was not directly considered in this modelis the impact of climate variability and change on the watersupply. Climate variability and change impacts both precipitation(Kalra and Ahmad, 2011) and streamflow (Kalra and Ahmad,2009) in the Colorado River Basin. The reduction in runoff in theColorado River Basin due to human induced climate change isestimated to be 10–30% resulting in water delivery shortfalls by0–20% (Barnett and Pierce, 2009). These shortfalls althoughsubstantial could be managed by applying demand side manage-ment measures such as water conservation, water reuse, andother measures (Barnett and Pierce, 2009). Even from the supplyshortfall point of view, the combination scenario is mostpreferred.

This study focuses mainly on the energy consumption and, as aresult, CO2 emissions, used in moving water in the LVV. Due tolack of data availability, some of the parameters are not includedin the study. For instance, in this study, the flow in each of thepumping stations is based on the water demand, capacity ofwater treatment plants and capacity of reservoirs in the distribu-tion system. The accurate prediction of energy requirements ineach of the pumping stations could have been achieved if thewater flow equations were developed based on the historical oractual flow at these stations. Also, the total dynamic headcalculation required for the power calculation included only headloss due to friction. Minor losses were ignored.

Electricity mix for the state of Nevada, which was consideredin determining the energy source, is composed of 85% non-renewable resources and 15% renewable resources. According tothe RPS, the percent share of renewable energy by 2025 should be25% of the total energy use in Nevada (PUCN, 2009). This can beachieved by developing renewable resources that include, but arenot limited to, biomass, fuel cells, geothermal energy, solarenergy, hydropower, and wind. However, the switch to suchrenewable resources as solar energy, which uses water as acooling agent, results in increased stress to water scarce regionslike the arid American Southwest. Hence, the actual source ofenergy to be used in the water conveyance system needs to beconsidered along with possible consequences; this will providemore accurate estimate of the CO2 emissions. Moreover, thisstudy considers only operational energy requirements. A com-plete life cycle energy analysis is beyond the scope of thisresearch. The evaluation of life cycle energy requirements willresult in a more accurate emission analysis, given that emissionsare differentiated as both direct and indirect. Direct emissions arethose that are released during the operational phase of the plantlife cycle, and indirect emissions are those that are emitted duringthe non-operational phase of the plant life cycle. The life cycleenergy analysis for power plant sector includes energy associatedin the extraction, processing and transportation of fuels, building

of power plants, production of electricity, waste disposal, andfinally, decommissioning of the plant at the end of its life.

5. Conclusions

Water management decisions should consider energy use toimprove the resource management. Consideration of the criticallink between water and energy during water planning and policymaking can lead to significant energy saving as well as, reductionsin the associated CO2 emissions. Water production requiresenergy and energy production contributes to the carbon footprint.Climate change, in turn, has greater potential to affect watersupply. In Nevada, climate change may lead to greater risk ofdrought or water shortages. Thus, the integration of energy issuesinto water policy decision making is important.

The conveyance of treated water in the distribution lateralsdominates the energy use for water provision in the LVV. Savingwater can be an excellent way to save energy and reduce CO2

emissions. Conservation eliminates the energy required to pump,move, and treat fresh water from the source. It also eliminates theenergy required to collect wastewater, treat, and dispose or reuseit. In addition, the reuse of treated wastewater effluent within theValley also appear to be an energy efficient water source becauseit eliminates the water transport energy requirements fromsource to the reuse points.

Acknowledgments

This work was funded by the United States Department ofEnergy Award DOE DE-EE-0000716 and National Science Founda-tion (NSF) Award CMMI 0846952.

References

Ahmad, S., Prashar, D., 2010. Evaluating municipal water conservation policiesusing a dynamic simulation model. Water Resource Management 24 (13),3371–3395.

Ahmad, S., Kalra, A., Stephen, H., 2010. Estimating soil moisture using remote sensingdata: a machine learning approach. Advances in Water Resources 33 (1),69–80.

Ahmad, S., Simonovic, S.P., 2000. System dynamics modeling of reservoir opera-tions for flood management. Journal of Computing in Civil Engineering 14 (3),190–198.

Ahmad, S., Simonovic, S.P., 2001. Integration of heuristic knowledge with analy-tical tools for selection of flood control measures. Canadian Journal of CivilEngineering 28 (2), 208–221.

Ahmad, S., Simonovic, S.P., 2004. Spatial system dynamics: new approach forsimulation of water resources systems. Journal of Computing in Civil Engi-neering 18 (4), 331–340.

Ahmad, S., Simonovic, S.P., 2005. An artificial neural network model for generatinghydrograph from hydro-meteorological parameters. Journal of Hydrology 315(1-4), 236–251.

Ahmad, S., Simonovic, S.P., 2006. An intelligent decision support system formanagement of floods. Water Resource Management 20 (3), 391–410.

Bakhshi, A., Demonsabert, S., 2009. A GIS methodology for estimating the carbonfootprint in municipal water and wastewater in Fairfax County, Virginia.Energy Engineering: Journal of the Association of Energy Engineering 106 (5),7–24.

Barnett, T.P., Pierce, D.W., 2008. When will Lake Mead go dry? Water ResourcesResearch 44, 3.

Barnett, T.P., Pierce, D.W., 2009. Sustainable water deliveries from the ColoradoRiver in a changing climate. Proceedings of the National Academy of Sciences.

California Energy Commission (CEC), 2007. Water-related Energy Use in California.From: /http://www.energy.ca.gov/2007publications/CEC-999-2007-008/CEC-999-2007-008.PDFS (retrieved on August 17, 2010).

Centre for Business and Economic Research (CBER), 2009. 2009 Clark CountyPopulation Forecast: 2000–2050. From: : /http://www.accessclarkcounty.com/depts/comprehensive_planning/demographics/Documents/FinalPopforecast.pdfS (retrieved on January 13, 2010).

Clark County, Nevada (CCN), 2000. Area Wide Reuse Study Las Vegas Valley StudyArea. From: /http://www.accessclarkcounty.com/depts/daqem/epd/waterquality/Documents/AreaWideReuseStudy.pdfS (retrieved on May 07, 2010).

Clark County Sewage and Wastewater Advisory Committee (SWAC), 2009.

Page 11

E. Shrestha et al. / Energy Policy 42 (2012) 201–212 211

Darwish, M.A., Al-Awadhi, F., Darwish, A.M., 2008. Energy and water in Kuwaitpart I. A sustainability view point. Desalination 225 (1-3), 341–355.

Electric Power Research Institute (EPRI), 2002. Water and Sustainability: U.S.Electricity Consumption for Water Supply and Treatment—The Next HalfCentury. EPRI, Palo Alto, CA: 2000. 1006787.

Eltawil, M.A., Zhengming, Z., Yuan, L., 2009. A review of renewable energytechnologies integrated with desalination systems. Renewable SustainableEnergy Reviews 13 (9), 2245–2262.

Evans, A., Strezov, V., Evans, T.J., 2009. Assessment of sustainability indicators forrenewable energy technologies. Renewable Sustainable Energy Reviews 13 (5),1082–1088.

Feroz, M., Jensen, M., Lindell, J.E., 2007. The Lake Mead Intake 3 Water Tunnel andPumping Station, Las Vegas, Nevada, USA. In: Paper Presented at the RapidExcavation and Tunneling Conference—2007, pp. 647–662.

Forsee, W., Ahmad, S., 2011. Evaluating urban stormwater infrastructure design inresponse to projected climate change. Journal of Hydrologic Engineering.doi:10.1061/(ASCE)HE.1943-5584.0000383.

Forster, P., Ramaswamy, V., Artaxo, P., Berntsen, T., Betts, R., Fahey, D.W.,Haywood, J., Lean, J., Lowe, D.C., Myhre, G., Nganga, J., Prinn, R., Raga, G.,Schulz, M., Van Dorland, R., 2007. Changes in atmospheric constituents and inradiative forcing. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M.,Averyt, K.B., Tignor, M., Miller, H.L. (Eds.). Climate Change 2007: The PhysicalScience Basis. Contribution of Working Group I to the Fourth AssessmentReport of the Intergovernmental Panel on Climate Change. Cambridge Uni-versity Press, Cambridge, United Kingdom and New York, NY, USA.

Gastelum, J.R., Valdes, J.B., Stewart, S., 2010. A system dynamics model to evaluatetemporary water transfers in the Mexican Conchos basin. Water ResourceManagement 24 (7), 1285–1311.

Griffiths-Sattenspiel, B., Wilson, W., 2009. The Carbon Footprint of Water. A RiverNetwork Report. From: /http://rivernetwork.org/sites/default/files/The%20Carbon%20Footprint%20of%20Water-River%20Network-2009.pdfS (retrieved onDecember 01, 2009).

Intergovernmental Panel on Climate Change (IPCC), 2000. Good Practice Guidanceand Uncertainty Management in National Greenhouse Gas Inventories. NGGIPPublications. From: /http://www.ipcc-nggip.iges.or.jp/public/gp/english/S(retrieved on August 10, 2010).

Jones, G.M., Sanks, R.L., Tchobanoglous, G., Bosserman II, B.E., 2008. PumpingStation Design. Butterworth-Heinemann.

Kalra, A., Ahmad, S., 2009. Using oceanic–atmospheric oscillations for long leadtime streamflow forecasting. Water Resources Research 45, W03413.

Kalra, A., Ahmad, S., 2011. Evaluating changes and estimating seasonal precipita-tion for Colorado River Basin using stochastic non-parametric disaggregationtechnique. Water Resources Research 47, W05555.

Khan, S., Yufeng, L., Ahmad, A., 2009. Analysing complex behaviour of hydrologicalsystems through a system dynamics approach. Environmental Modelling andSoftware 24 (12), 1363–1372.

Lampe, L.K., Adams, L.M., Jensen, D.G., 2009. Water supply and energy generation.In: Paper Presented at the World Environmental and Water ResourcesCongress 2009, Great Rivers, May 17–21, 2009, vol. 342, pp. 5202–5211.

Las Vegas Valley Watershed Advisory Committee (LVVWAC), 2009. RegionalWater Quality Plan. From: /http://www.snwa.com/assets/pdf/wq_regional_plan.pdfS (retrieved on January 25, 2010).

Las Vegas Wash Comprehensive Adaptive Management Plant (LVWCAMP), 1999.From: /http://www.lvwash.org/html/resources_library_lvwcamp.htmlS(retrieved on January 25, 2010).

Lenzen, M., 2008. Life cycle energy and greenhouse gas emissions of nuclearenergy: a review. Energy Conservation and Management 49, 2178–2199.

Morrison, J., Morikawa, M., Murphy, M., Schulte, P., 2009. Water Scarcity andClimate Change: Growing Risks for Businesses and Investors. A Ceres Report.From: /http://www.pacinst.org/reports/business_water_climate/full_report.pdfS (retrieved on January 20, 2010).

Mosquera-Machado, S.C., Ahmad, S., 2007. Flood hazard assessment of Atrato Riverin Colombia. Water Resource Management 21 (3), 591–609.

Natural Resources Defense Council (NRDC), 2004. Energy Down the Drain. TheHidden Costs of California’s Water Supply. From: /http://www.nrdc.org/water/conservation/edrain/edrain.pdfS (retrieved on June 08, 2010).

Pelli, T., Hitz, H.U., 2000. Energy indicators and savings in water supply. Journal ofthe American Water Works Association 92 (6), 55–62.

Proenca, L.C., Ghisi, E., Tavares, D.F., Coelho, G.M., 2011. Potential for electricitysavings by reducing potable water consumption in a city scale. ResourceConservation and Recycling 55 (11), 960–965.

Public Utilities Commission of Nevada (PUCN), 2009. Senate Bill No. 358-Committee on Energy, Infrastructure and Transportation. From: /http://www.leg.state.nv.us/75th2009/Bills/SB/SB358_EN.pdfS (retrieved on Novem-ber 10, 2011).

Puri, S., Stephen, H., Ahmad, S., 2011. Relating TRMM precipitation radar landsurface backscatter response to soil moisture in the Southern United States.Journal of Hydrology 402, 115–125.

Qaiser, K., Ahmad, S., Johnson, W., Batista, J., 2011. Evaluating the impact of waterconservation on fate of outdoor water use: a study in an arid region. Journal ofEnvironmental Management 92 (8), 2061–2068.

Reiling, S.J., Roberson, J.A., Cromwell III, J.E., 2009. Drinking water regulations:estimated cumulative energy use and costs. Journal of the American WaterWorks Association 101 (3), 42–53.

Rio Carrillo, A.M., Frei, C., 2009. Water: a key resource in energy production.Energy Policy 37 (11), 4303–4312.

Shrestha, E., Ahmad, S., Johnson, W., Shrestha, P., Batista, J.R., 2011. Carbonfootprint of water conveyance versus desalination as alternatives to expandwater supply. Desalination 280 (1–3), 33–43.

Simonovic, S.P., Ahmad, S., 2005. Computer-based model for flood evacuationemergency planning. Natural Hazards 34, 25–51.

Simonovic, S.P., Fahmy, H., 1999. A new modeling approach for water resourcespolicy analysis. Water Resources Research 35 (1), 295–304.

Southern Nevada Water Authority (SNWA), 2009a. Water Resource Plan 09. From:/http://www.snwa.com/assets/pdf/wr_plan.pdfS (retrieved on January 26,2010).

Southern Nevada Water Authority (SNWA), 2009b. Conservation Plan 2009–2013.From: /http://www.snwa.com/assets/pdf/cons_plan.pdfS (retrieved on Feb-ruary 08, 2010).

Southern Nevada Water Authority (SNWA), 2010a. Water Resources. From:/http://www.snwa.com/html/wr_index.htmlS (Retrieved on May 28, 2010).

Southern Nevada Water Authority (SNWA), 2010b. Water Treatment Facilities.From: /http://www.snwa.com/html/wq_treatment_facilities.htmlS(retrieved on May 28, 2010).

Spang, N., 2007. Stakeholder participation: a bridge between integrated waterresource management and system dynamics. IDEAS Journal: InternationalDevelopment, Environment And Sustainability From: /http://fletcher.tufts.edu/ierp/ideas/pdfs/issue3/SpangNed_IWRMandSystemDynamics.pdfS(retrieved on January 21, 2010).

Stave, K.A., 2003. A system dynamics model to facilitate public understanding ofwater management options in Las Vegas, Nevada. Journal of EnvironmentalManagement 67 (4), 303–313.

Stephen, H., Ahmad, S., Piechota, T.C., Tang, C., 2010. Relating surface backscatterresponse from TRMM precipitation radar to soil moisture: results over a semi-arid region. Hydrology and Earth System Sciences 14, 193–204.

Sterman, J.D., 2000. Business Dynamics: Systems Thinking and Modeling for aComplex World. Irwin/McGraw-Hill.

Stillwell, A.S., Webber, M.E., 2010. Water Conservation and Reuse: A Case Study ofthe Energy–Water Nexus in Texas. In: World Environmental and WaterResources Congress 2010: Challenges of Change—Proceedings of the WorldEnvironmental and Water Resources Congress, RI.

Stokes, J.R., Horvath, A., 2009. Energy and air emission effects of water supply.Environmental Science and Technology 43 (8), 2680–2687.

Strutt, J., Wilson, S., Shorney-Darby, H., Shaw, A., Byers, A., 2008. Assessing thecarbon footprint of water production. Journal of the American Water WorksAssociation 100 (6) 80–91þ10.

Taffler, D., Lesley, D., Zelenka, A., 2008. Hidden potential. recycled water and thewater–energy–carbon nexus. Water Environment and Technology 20 (11),34–41.

Tidwell, V.C., Passell, H.D., Conrad, S.H., Thomas, R.P., 2004. System dynamicsmodeling for community-based water planning: application to the Middle RioGrande. Aquatic Sciences—Research Across Boundaries 66 (4), 357–372.

U.S. Bureau of Reclamation (USBR), 2010. Lake Mead at Hoover Dam, Elevation(feet). From: /http://www.usbr.gov/lc/region/g4000/hourly/mead-elv.htmlS(retrieved on May 28, 2010).

U.S. Census Bureau (USCB), 2010. World Population: 1950–2050. From: /http://www.census.gov/ipc/www/idb/worldpopgraph.phpS (retrieved on May 28,2010).

U.S. Department of Energy (USDOE), 2006. Energy Demands on Water Resources.Report to Congress on the Interdependency of Energy and Water. From:/http://www.sandia.gov/energy-water/docs/121-RptToCongress-EWwEIAcomments-FINAL.pdfS (retrieved on January 15, 2010).

U.S. Energy Information Administration (USEIA), 2009. Consumption, Price andExpenditure Estimates. From: /http://www.eia.doe.gov/emeu/states/sep_sum/html/pdf/sum_btu_eu.pdfS (retrieved on May 28, 2010).

U.S. Energy Information Administration (USEIA), 2010. Frequently askedQuestions—Electricity. From: /http://tonto.eia.doe.gov/ask/electricity_faqs.asp#electricity_use_homeS (retrieved on May 28, 2010).

U.S. Environmental Protection Agency (USEPA), 2009a. Energy and Water. From:/http://www.epa.gov/waterinfrastructure/bettermanagement_energy.htmlS(retrieved on May 28, 2010).

U.S. Environmental Protection Agency (USEPA), 2009b. Inventory of U.S. Green-house Gas Emissions and Sinks: 1990–2007. From: /http://www.epa.gov/climatechange/emissions/downloads09/GHG2007entire_report-508.pdfS(retrieved on May 28, 2010).

U.S. Environmental Protection Agency (USEPA), 2010. Greenhouse Gas Emissions.From: /http://www.epa.gov/climatechange/emissions/S (retrieved on Janu-ary 27, 2010).

Vedwan, N., Ahmad, S., Wilhelm, F.M., Broad, K., Letson, D., Podesta, G., 2008.Institutional evolution in lake Okeechobee management in Florida: charac-terstic, impacts, and limitations. Water Resource Management 22, 699–718.

Venkatesan, A.K., Ahmad, S., Johnson, W., Batista, J.R., 2011a. System dynamicsmodel to forecast salinity load to the Colorado River due to urbanizationwithin the Las Vegas Valley. Science of the Total Environment 409 (13),2616–2625.

Venkatesan, A.K., Ahmad, S., Johnson, W., Batista, J.R., 2011b. Salinity reductionand energy conservation in direct and indirect potable water reuse. Desalina-tion 272 (1–3), 120–127.

Vieira, F., Ramos, H.M., 2009. Optimization of the energy management inwater supply systems. Water Science and Technology: Water Supply 9 (1),59–65.

Von Uexkull, O., 2004. Energy and water: the ignored link. Refocus 5 (2), 40–44.

Page 12

E. Shrestha et al. / Energy Policy 42 (2012) 201–212212

Weisser, D., 2007. A guide to life-cycle greenhouse gas (GHG) emissions fromelectric supply technologies. Energy 32 (9), 1543–1559.

Wiedmann, T., Minx, J., 2008. A Definition of ‘Carbon Footprint’. In:. Pertsova C.C(Ed.), Ecological Economics Research Trends. Nova Science Publishers, Haup-pauge NY, USA, pp. 1–11 (Chapter 1). /https://www.novapublishers.com/catalog/product_info.php?products_id=5999S.

Winiwarter, W., Rypdal, K., 2001. Assessing the uncertainty associated withnational greenhouse gas emission inventories: a case study for Austria.Atmospheric Environment 35 (32), 5425–5440.

Winz, I., Brierley, G., Trowsdale, S., 2009. The use of system dynamics simulationin water resources management. Water Resource Management 23 (7),1301–1323.

Wyma, R., 2008. Case study: Water savings on government sites in Gauteng, SouthAfrica. In: Proceedings of the tenth Annual Water Distribution SystemsAnalysis Conference, WDSA 2008.

Zavalani, O., Luga, Y., 2010. Energy and water saving possibilities in publicfacilities in Albania. In: Proceedings of the UKSim fourth European ModellingSymposium on Computer Modelling and Simulation, EMS2010.

Zhang, X., Feng, H., Mao, X., Zheng, M., Wang, R., 2009. A dynamic water resourcesmanagement approach in Beijing using system dynamics model. In: PaperPresented at the International Conference on Management and ServiceScience, MASS 2009, September 20– 22, 2009, Communication University ofChina; Wuhan University; James Madison University; Institute of Policy andManagement, Chinese Academy of Sciences; IEEE Wuhan Section.