Water Smart Power Full Report

7/27/2019 Water Smart Power Full Report

1/58

Water-Smart PowerS t r e n g t h e n i n g t h e U . S . e l ec t r i c i t y S y St e m

i n a W a r m i n g W o r l d

A Report of the Eer a Water i a Warmi Worl Iitiative


2/58

John Rogers

Krsen Aery

See Clemmer

Mchelle Das

Francsco Flores-Lopez

Doug Kenney

Jordan Macknck

Nada Madden

James Meldrum

Sandra Saler

Erka Spanger-Segred

Dad Yaes

eW3 S as c

Peer Frumho

George Hornberger

Rober Jackson

Robn Newmark

Jonahan Oerpeck

Brad Udall

Mchael Webber

A Repor o he

e W W W i

JULY 2013

Water-Smart PowerS t r e n g t h e n i n g t h e U . S . e l ec t r i c i t y S y St e m

i n a W a r m i n g W o r l d


3/58

CitAtiON:

Rogers, J., K. Aery, S. Clemmer, M. Das, F. Flores-Lopez, P. Frumho, D. Kenney, J. Macknck,

N. Madden, J. Meldrum, J. Oerpeck, S. Saler, E. Spanger-Segred, and D. Yaes. 2013.Water-smart power: Strengthening the U.S. electricity system in a warming world. Cambrdge, MA:

Unon o Concerned Scenss. July.

2013 Unon o Concerned Scenss

All rghs resered

the Unon o Concerned Scenss pus rgorous, ndependen scence o work o sole our

planes mos pressng problems. Jonng wh czens across he counry, we combne echncal

analyss and eece adocacy o creae nnoae, praccal soluons or a healhy, sae, andsusanable uure. For more normaon abou UCS, s our webse a www.ucsusa.org.

ths repor s aalable on he UCS webse (www.ucsusa.org/publications)

or may be obaned rom:

UCS Publcaons

2 Brale Square

Cambrdge, MA 02238-9105

Or, [email protected] or call (617) 547-5552.

COvER PHOtOS

top: Shuersock/zhangyang13576997233

Boom, le o rgh: Flckr/K Al, Flckr/james_gordon_losangeles,

Jenner Hepnsall, Flckr/LarryHB

ti tL E PAG E PH Ot O: Flckr/Dad Joyce

Energy and Water in a Warming World Initiative


4/58

Te Energy and Water in a Warming World initiative (EW3) is a collaborative eort

between the Union o Concerned Scientists and a team o independent experts to build

and synthesize policy-relevant research on the water demands o energy production in

the context o climate variability and change. Te initiative includes core research col-

laborations intended to raise the national prole o the water demands o energy, along

with policy-relevant energy development scenarios and regional perspectives.

Tis report is based primarily on the research o the EW3 energy-water utures collab-

orators listed below. Te research appears in a special issue oEnvironmental ResearchLetters: Focus on Electricity, Water and Climate Connections (ERL 2013). Tis report

is also available online at www.ucsusa.org/watersmartpower.

e W 3 e n e r g y - W a t e r F U t U r e S r e S e a r c h c o l l a b o r a t o r S

A multidisciplinary advisory committee composed o senior scientists provides

oversight and guidance or EW3:

e W 3 S c i e n t i F i c a d v i S o r y c o m m i t t e e

Peter Frumho (chair), Union o Concerned Scientists, Cambridge, MA

George Hornberger, Vanderbilt University, Nashville, TN

Robert Jackson, Duke University, Durham, NC

Robin Newmark, National Renewable Energy Laboratory, Golden, CO

Jonathan Overpeck, University o Arizona, Tucson, AZ

Brad Udall, University o Colorado, Boulder, CO

Michael Webber, University o Texas, Austin, TX

e W 3 P r o j e c t m a n a g e r S

Erika Spanger-Siegried, Union o Concerned Scientists, Cambridge, MA

John Rogers, Union o Concerned Scientists, Cambridge, MA

About EW3

Nadia Madden, Union o Concerned Scientists, Cambridge, MA

James Meldrum, University o Colorado, Boulder, CO

Sandra Sattler, Union o Concerned Scientists, Cambridge, MA

Erika Spanger-Siegried, Union o Concerned Scientists,

Cambridge, MA

David Yates, National Center or Atmospheric Research,

Boulder, CO

John Rogers, Union o Concerned Scientists, Cambridge, MA

Kristen Averyt, University o Colorado, Boulder, CO

Steve Clemmer, Union o Concerned Scientists, Cambridge, MA

Michelle Davis, Union o Concerned Scientists, Cambridge, MA

Francisco Flores-Lopez, Stockholm Environment Institute, Davis, CA

Doug Kenney, University o Colorado, Boulder, COJordan Macknick, National Renewable Energy Laboratory, Golden, CO

Water-Smart Power: Strengthening the U.S. Electricity System in a Warming World


5/58

Tis report is the product o active collaboration and

contributions rom people with diverse expertise related

to energy, water, and climate change.

For thoughtul comments on review drats o this

report, we thank Sara Barczak, Mandy Hancock, and

Ulla-Britt Reeves (Southern Alliance or Clean Energy);

Jan Dell (CH2M HILL); Laura Hartt (Chattahoochee

Riverkeeper); Mike Hightower (Sandia NationalLaboratories); om Iseman; Cindy Lowry (Alabama

River Alliance); Chris Manganiello (Georgia River

Network); odd Rasmussen (University o Georgia);

Stacy ellinghuisen (Western Resource Advocates); and

om Wilbanks (Oak Ridge National Laboratory).

We also appreciate the insights o participants in the

May 2013 workshop Energy and Water in a Warming

World: Reducing Water and Other Climate Risks,

which helped us develop the recommendations in

this report. Tey included Doug Arent, Vicki Arroyo,Sara Barczak, Rajnish Barua, Bruce Biewald, Lynn

Broaddus, Ken Colburn, Steve Fleischli, Guido Franco,

Gary Helm, Mike Hightower, Sarah Hoverter, om

Iseman, Mike Jacobs, Joe Kwasnik, Rob McCulloch,

Steve Rose, Keith Schneider, Alison Silverstein, Julie

aylor, and om Wilbanks.

For assistance with or input on the EW3 research and

manuscripts that underpin this report, we thank Corrie

Clark, Ethan Davis, imothy Diehl, Easan Drury, Etan

Gumerman, KC Hallett, Chris Harto, Garvin Heath,Al Hicks, Carey King, John (Skip) Laitner, Courtney

Lee, Anthony Lopez, Bob Lotts, Mary Lukkonen,

rieu Mai, Claudio Martinez, Steve Nadel, Syndi

Nettles-Anderson, Amanda Ormond, Martin (Mike)

Pasqualetti, Walter Short, Brad Smith, Samir Succar,

Vince idwell, Ellen Vancko, Laura Vimmerstedt, John

Wilson, and Phillip Wu.

For extraordinary editorial and graphical support, we

are deeply indebted to Sandra Hackman and yler

Kemp-Benedict. And or essential editorial guidance,

assistance, and insight, we thank Keith Schneider.

We also appreciate the assistance and input o Angela

Anderson, Eric Bontrager, Nancy Cole, Jef Deyette,

Brenda Ekwurzel, Lesley Fleischman, Lisa Nurnberger,

Megan Rising, Seth Shulman, and Bryan Wadsworth,all o the Union o Concerned Scientists.

And we remain indebted to the pioneers in exploring

and communicating energy-water-climate challenges

colleagues who have conducted important research on

problems and solutions and worked to broaden and

deepen understanding among policy makers and the

public. We are also grateul to those working to address

these challenges rom national and state perspectives,

and at the level o individual rivers and watersheds.

Te production o this report was made pos-

sible through the generous support o Te Kresge

Foundation, Roger and Vicki Sant, the Wallace

Research Foundation, and the Common Sense Fund.

NOTE: Employees o the Alliance or Sustainable Energy, LLC (Alliance), the

operator o the National Renewable Energy Laboratory (NREL) or the U.S.

Department o Energy (DOE), have contributed to this report. The views and

opinions expressed herein do not necessarily state or refect those o Alliance,

NREL, the DOE, or the U.S. government. Furthermore, Alliance, NREL, the DOE,

and the U.S. government make no warranty, express or implied, and assume no

liability or responsibility or the accuracy, completeness, or useulness o any in-

ormation disclosed herein. Reerence herein to any product, process, or service

by trade name, trademark, manuacturer, or otherwise does not constitute or

imply its endorsement, recommendation, or avoring by Alliance, NREL, the DOE,

or the U.S. government.

Acknowledgments

iv Energy and Water in a Warming World Initiative


6/58

Contents

About EW3

AcknoWlEdgmEnts

FigurEs And tExt boxEs

1 ExEcutivE summAry

2 Te Challenges We Face

3 Change Is Under Way

3 Decisions in the Power Sector Matter

4 oward a Water-Smart Energy Future

chAptEr 1

6 Ee, Wae, a ca: i

7 Evaluating Our Options

chAptEr 2

8 EeWae c

8 Te Power Sector and Water Risks

9 Winners and Losers in Water Collisions

10 Climate Complications oday and omorrow

chAptEr 3

13 p p u.s. pwe

13 Change Is Under Way

15 Building the Electricity System o thewenty-rst Century

16 Focusing on wo Vulnerable Regions

17 Our Innovative Approach to Modeling

chAptEr 4

19 F: te ia pwe pawa

Wae

19 Business as Usual: Good, Bad, and Ugly

24 A Better Pathway: Curbing Carbon EmissionsandWater Use

28 How the Costs Add Up

29 Other Pathways, Other Outcomes

chAptEr 5

31 ma lwca, Waesa

Ee ce ta

31 Securing Our Energy Future

32 Moving Decisions oday

33 Water-Smart Criteria: Best Practice oday,Standard Practice omorrow

33 Who Makes Water-Smart Decisions in the Real World?

36 Promoting Integrated Decision Making

36 Conclusion

37 rEFErEncEs

AppEndicEs

43 APPENDix A. U.S. Electricity Mix under Four Scenarios

44 APPENDix B. U.S. Power Plant Water Useunder Four Scenarios, 20102050

45 APPENDix C1. U.S. Power Plant Water Withdrawalacross Scenarios, by State, 20102050

46 APPENDix C2. U.S. Power Plant Water Consumptionacross Scenarios, by State, 20102050

EW3 biogrAphiEs47 Energy-Water Futures Research eam

48 Scientic Advisory Committee

50 About ucs



7/58

Fe

9 FiGURE 1. Energy-Water Collisions

11 FiGURE 2. Dry imes, Present and Future

12 FiGURE 3. Water Supply versus Water Demand in the Colorado River Basin

13 FiGURE 4. Electricity Sector in ransition: Te U.S. Electricity Mix and Retiring Coal Plants

15 FiGURE 5. A Carbon Budget or the U.S. Electricity Sector

16 FiGURE 6. echnology argets or the Electricity Sector

17 FiGURE 7. Electricity-Water Chal lenges in wo Regions

20 FiGURE 8. U.S. Electricity Mix under Business as Usual, 20102050

21 FiGURE 9. Power Plant Water Use under Business as Usual, 20102050

24 FiGURE 10. U.S. Electricity Mix under the Renewables-and-Eciency Scenario

25 FiGURE 11. Power Plant Water Use under the Renewables-and-Eciency Case, 20102050

26 FiGURE 12. Regional Variations in Power Plant Water Use

27 FiGURE 13. Te Impact o Electricity Choices on Reservoir Levels in Lake Mead and Lake Powell

28 FiGURE 14. Groundwater Savings in the Southwest across Scenarios

28 FiGURE 15. Te Impact o Electricity Choices on Coosa River emperatures

29 FiGURE 16. Electricity Prices, Electricity Bills, and Natural Gas Prices under wo Scenarios

te be

14 BOx 1. Energy echnology ransitions

23 BOx 2. Te Impact o Hydroracking on Water

32 BOx 3. Te Impact o Coal Retirements on Water Use

35 BOx 4. Improving the Water Use o Energy echnologies

Figures and Text Boxes

Energy and Water in a Warming World Initiative


8/58

Executive Summary

We can, however, use uel and technology options

available now to design an electricity uture that begins

to shed some o these risks. We can also expand our

options by making strategic investments in energy and

cooling technologies. Te key is to understand what

a low-carbon, water-smart electricity uture lookslikewhich electric sector decisions best prepare us to

avoid and minimize energy-water collisions, and to cope

with those we cannot avoidand to make decisions

that will set and keep us on that path.

Tis report is the second rom the Energy and

Water in a Warming World Initiative (EW3), organized

by the Union o Concerned Scientists to ocus on the

water implications o U.S. electricity choices. Te rst,

the heat waves and drought that hit the United

States in 2011 and 2012 shined a harsh light on

the vulnerability o the U.S. electricity sector

to extreme weather. During the historic 2011 drought

in exas, power plant operators trucked in water rom

miles away to keep the plants running, and disputes

deepened between cities and utilities seeking to con-

struct new water-intensive coal plants. In 2012, heat

and drought orced power plants, rom the Gallatincoal plant in ennessee to the Vermont Yankee nu-

clear plant on the Connecticut River, to reduce their

output or shut down altogether. Tat summer, amid

low water levels and soaring water temperatures, op-

erators o other plantsat least seven coal and nuclear

plants in the Midwest alonereceived permission

to discharge even hotter cooling water, to enable the

plants to keep generating. Tese consecutive summers

alone revealed water-related electricity risks across

the country.

Te power sector has historically placed largedemands on both our air and water. In 2011, electric-

ity generation accounted or one-third o U.S. heat-

trapping emissions, the drivers o climate change.

Power plants also accounted or more than 40 percent

o U.S. reshwater withdrawals in 2005, and are one

o the largest consumers o reshwaterlosing water

through evaporation during the cooling processout-

side the agricultural sector.

Te electricity system our nation built over the

second hal o the twentieth century helped uel the

growth o the U.S. economy and improve the qualityo lie o many Americans. Yet we built that system

beore ully appreciating the reality and risks o

climate change, and beore converging pressures cre-

ated the strain on local water resources we see today

in many places. Tis system clearly cannot meet our

needs in a uture o growing demand or electricity,

worsening strains on water resources, and an urgent

need to mitigate climate change.

WkmedaComm

ons/ArnoldPaul



9/58

Freshwater Use by U.S. Power Plants, documented the

energy-water collisions already occurring because o

the dependence o U.S. power plants on water. In thatresearch, we ound that past choices on uel and cooling

technologies in the power sector are contributing to

water stress in many areas o the country.

Like the rst report, this one stems rom a collabo-

ration among experts rom universities, government,

and the nonprot sector. Water-Smart Powerrefects

comprehensive new research on the water implications

o electricity choices in the United States under a range

o pathways, at national, regional, and local levels. Te

report aims to provide critical inormation to inorm

decisions on U.S. power plants and the electricity sup-ply, and motivate choices that saeguard water resourc-

es, reduce carbon emissions, and provide reliable power

at a reasonable priceeven in the context o a changing

climate and pressure on water resources.

te caee We Fae

Our examination o todays electricity-water landscape

reveals prominent challenges:

Energy-water collisions are happening now.

Because o its outsized water dependence, the U.S.

electricity sector is running into and exacerbat-ing growing water constraints in many parts o

the country. Te reliance o many power plants on

lakes, rivers, and groundwater or cooling water can

exert heavy pressure on those sources andleave the

plants vulnerable to energy-water collisions, partic-

ularly during drought or hot weather. When plants

cannot get enough cooling water, or example,

they must cut back or completely shut down their

generators, as happened repeatedly in 2012 at plants

around the country.

As the contest or water heats up, the power

sector is no guaranteed winner. When the water

supply has been tight, power plant operators have

oten secured the water they need. In the summer

o 2012, or example, amid soaring temperatures in

the Midwest and multiple large sh kills, a hand-

ul o power plant operators received permission

to discharge exceptionally hot water rather than

reduce power output. However, some users are

WaerkeeperAllance

2 Energy and Water in a Warming World Initiative


10/58

pushing back against the power sectors dominant

stake. In Utah, or example, a proposal to build a

3,000-megawatt nuclear power plant ueled graveconcerns about the impact o the plants water use.

And in exas, regulators denied developers o a pro-

posed 1,320-megawatt coal plant a permit to with-

draw 8.3 billion gallons (25,000 acre-eet) o water

annually rom the states Lower Colorado River.

Climate change complicates matters. Energy-

water collisions are poised to worsen in a warming

world as the power sector helps drive climate

change, which in turn aects water availability and

quality. Climate change is already constraining or

altering the water supply in many regions by chang-

ing the hydrology. In the Southwest, or example,

where the population is growing rapidly and water

supply is typically tight, much o the surace water

on which many water users depend is declining.

Scientists expect rising average temperatures, more

extreme heat, and more intense droughts in many

regions, along with reductions in water availability.

Tese conditionsheightened competition or water

and more hydrologic variabilityare not what our

power sector was built to withstand. However, to beresilient, it must adjust to them.

cae i ue Wa

Building an electricity system that can meet the chal-

lenges o the twenty-rst century is a considerable

task. Not only is the needed technology commercially

available now, but a transition is also under way that is

creating opportunities or real system-wide change:

Te U.S. power sector is undergoing rapid trans-

ormation. Te biggest shit in capacity and uel inhal a century is under way, as electricity rom coal

plants shrinks and power rom natural gas and renew-

ables grows. Several actors are spurring this transition

to a new mix o technologies and uels. Tey include

the advanced age o many power plants, expanding

domestic gas supplies and low natural gas prices,

state renewable energy and eciency policies, new

ederal air-quality regulations, and the relative costs

and risks o coal-red and nuclear energy.

Tis presents an opportunity we cannot aord

to miss. Decisions about which power plants to

retrot or retire and which kind to build have both

near-term and long-term implications, given the

long lietimes o power plants, their carbon emis-

sions, and their water needs. Even a single aver-

age new coal plant could emit 150 million tons o

carbon dioxide over 40 yearstwice as much as a

natural gas plant, and more than 20 million cars

emit each year. Power plants that need cooling

water will be at risk over their long lietimes rom

declining water availability and rising water tem-

peratures stemming rom climate change, extreme

weather events, and competition rom other users.

And power plants, in turn, will exacerbate the water

risks o other users.

de e pwe se mae

Choices, however, are important only i they lead to di-

erent outcomes. o analyze the impact o various op-

tions or our electricity uture on water withdrawals and

consumption, carbon emissions, and power prices, under

this new research we ocused on several key scenarios.

Tese included business as usual and three scenarios

based on a strict carbon budgetto address the power

sectors contributions to global warming. wo o those

G

.Garca-Roge

rs



11/58

three scenarios assumed the use o specic technologies

to make those signicant cuts in carbon emissions.

o explore the outcomes o these scenarios we usedtwo models: the Regional Energy Deployment System

(ReEDS) and the Water Evaluation and Planning

(WEAP) system. With these two models and our set o

scenarios, we analyzed the implications o water use in

the power sector under dierent electricity pathways or

the entire nation, or various regions, and or individual

river basins in the southwestern and southeastern

United States.

Our distinctive approach and new researchalong

with previous workshows that our electricity choices

will have major consequences over the coming decades,especially in water-stressed regions. Trough this re-

search, we have learned that:

Business as usual in the power sector would ail

to reduce carbon emissions, and would not tap

opportunities to saeguard water. Because such a

pathway or meeting uture electricity needs would

not cut carbon emissions, it would do nothing to

address the impact o climate change on water.

Changes in the power plant feet would mean that

water withdrawals by power plants would drop, yet

plants water consumption would not decline or

decades, and then only slowly. Te harmul eects

o power plants on water temperatures in lakes and

rivers might continue unabated, or even worsen.

Greater extraction o ossil uels or power plants

would also aect water use and quality.

Low-carbon pathways can be water-smart. A

pathway ocused on renewable energy and energy

eciency, we ound, could deeply cut both carbon

emissions and water eects rom the power sec-

tor. Water withdrawals would drop 97 percentby 2050much more than under business as usual.

Tey would also drop aster, with 2030 withdrawals

only hal those under business as usual. And water

consumption would decline 85 percent by 2050.

Tis pathway could also curb local increases in water

temperature rom a warming climate. Meanwhile

lower carbon emissions would help slow the pace and

reduce the severity o climate change, including its

long-term eects on water quantity and quality.

However, low-carbon power is not necessarily

water-smart. Te menu o technologies qualiying as

low-carbon is long, and includes some with substan-tial water needs. Electricity mixes that emphasize

carbon capture and storage or coal plants, nuclear

energy, or even water-cooled renewables such as some

geothermal, biomass, or concentrating solar could

worsen rather than lessen the sectors eects on water.

Renewables and energy efciency can be a win-

ning combination. Tis scenario would be most

eective in reducing carbon emissions, pressure

on water resources, and electricity bills. Energy

eciency eorts could more than meet growth

in demand or electricity, and renewable energy

could supply 80 percent o the remaining demand.

Although other low-carbon paths could rival this

one in cutting water withdrawals and consump-

tion, it would edge ahead in reducing groundwater

use in the Southwest, improving river fows in the

Southeast, and moderating high river temperatures.

Tis scenario could also provide the lowest costs to

consumers, with consumer electricity bills almost

one-third lower than under business as usual.

twa a Waesa Ee Fe

Water-smart energy decision making depends on under-

standing and eectively navigating the electricity-water-

climate nexus, and applying best practices in decision

making:

We can make decisions now to reduce water and

climate risk. Fuel and technology options already

available mean we can design an electricity system

with ar lower water and climate risks. Tese in-

clude prioritizing low-carbon, water-smart options

such as renewable energy and energy eciency,

upgrading power plant cooling systems with those

that ease water stress, and matching cooling needs

with the most appropriate water sources.

Electricity decisions should meet water-smart

criteria. Tese criteria can point decision makers to

options that reduce carbon emissions andexposure

to water-related risks, make sense locally, and are

cost-eective.



12/58

and extremes, allowing planners to consider low-

probability but high-impact events. And scientists

and engineers can improve the eciency and reducethe cost o low-water energy options.

Understanding and addressing the water impact o

our electricity choices is urgent business. Because most

power sector decisions are long-lived, what we do in the

near term commits us to risks or resiliencies or decades.

We can untangle the production o electricity rom the

water supply, and we can build an electricity system

that produces no carbon emissions. But we cannot wait,

nor do either in isolation, without compromising both.

For our climateand or a secure supply o water andpowerwe must get this right.

Actors in many sectors have essential roles to

play. No single platorm exists or sound, long-term

decisions at the nexus o electricity and water, butthose made in isolation will serve neither sector.

Instead, actors across sectors and scales need to

engage. For example: plant owners can prioritize

low-carbon options that are water-appropriate or

the local environment. Legislators can empower

energy regulators to take carbon and water into

account. Consumer groups can ensure that utili-

ties do not simply pass on to ratepayers the costs o

risky, water-intensive plants. Investors in utilities

can demand inormation on water-related risks and

seek low-carbon, water-smart options. Researcherscan analyze uture climate and water conditions

JennerHepnsall



13/58

chAptEr 1

Electricity, Water, and Carbon: Introduction

hotter water can disrupt local ecosystems (Stewart et al.

2013; EPA 2011a; GAO 2009; Langord 2001).2

Newer power plants tend to use recirculating cool-

ing systems, which withdraw much less water. Tose

systems consume much o it through evaporation

during cooling, however, and also require more initial

investment than once-through cooling systems and are

less energy-ecient (GAO 2009).

Either type o cooling poses risks or the powerplants that use them. Developers must site the plants

near major sources o water. And the plants are exposed

to risks when water is too scarce or too hot to allow

the generators to operate saely or eciently (Spanger-

Siegried 2012).3

Droughts and heat waves already aect power

generation in the United States, particularly in summer,

when demand or electricity is highest. Our changing

climate means that such events are becoming more

requent in many parts o the country. Energy acili-

ties that need cooling water will ace risks during theirlong lietimes rom any drops in water availability rom

climate change or competition or water sources. Te

acilities will also be at risk rom increases in cooling

water temperatures. And they will exacerbate the water-

supply risks o other users (Averyt et al. 2011).

Te power sector, meanwhile, is undergoing an

unprecedented level o change, too. A newound abun-

dance o natural gas in the United States and histori-

cally low pricescombined with increases in renewable

Power plants that generate steam to make electric-

ityall coal and nuclear plants, many natural

gas plants, and some renewable energy acili-

tiestypically use water to cool and re-condense that

steam or reuse, and oten in large quantities (DOE

2006). Such thermoelectricpower plants are responsible

or the largest share o reshwater withdrawals in the

United States: more than 40 percent in 2005 (Kenny et

al. 2009). Tey are also one o the largest non-agricul-tural consumerso such water, through the evaporation

that serves to remove the excess heat during the cooling

process (Solley, Pierce, and Perlman 1998).1

Power plants that require cooling use dierent

technologies, each with advantages and disadvantages.

Some use once-through cooling systems, which with-

draw enormous amounts o water rom lakes, rivers, or

streams, use it once, and return it to the source. Once-

through systems are the least capital-intensive, and lose

less water to evaporation, but discharge much-hotter

water. Te water withdrawals and the discharge o

Power plants and cooling water. Pw ps us w

k pps s.

S ww us w u pu s

ku . os ww u ss u

su (p) s . e p

ps sks ps us w sus

w p.

Phoocoure

syDomnon,w

ww.d

om.c

om

1 Whle agrculure accouns or 84 percen o waer consumpon, he power

secor consumes 20 percen o he remander (3.3 percen oerall), second only

o household (domesc) use. Daa are rom 1995, he year o he mos recen

surey o waer consumpon by he U.S. Geologcal Surey.

2 Power plan whdrawals kll sh hrough mpngemen (rappng agans a

screen), and sh larae hrough enranmen (pullng hrough he coolng pro-

cess). the dozens o power saons ha whdraw Grea Lakes waer or coolng,

or eample, kll an esmaed 100 mllon sh and more han a bllon laral sh

annually (Kelso and Mlburn 1979; also see EPA 2001).

3 Whle hs repor ocuses on waer scarcy and hgh waer emperaures,

many power plans are a rsk o oodng, gen ha hey are oen locaed near

major bodes o waer. Freezes can also nerere wh power plans nake o

coolng waer. See, or eample, Webber 2012.



14/58

energy and energy eciency and new ederal air-quality

regulationshave challenged coals dominance in the

electricity sector. ens o thousands o megawatts o

U.S. coal power capacity are slated or retirement in the

next several years, and many other coal plants are eco-

nomically vulnerable, given pressure to upgrade pollu-

tion controls and competition rom other power sources

(Cleetus et al. 2012).Whats more, decisions about which power plants

to retrot or retire and which new ones to build will

themselves have enormous bearing on both near-term

and long-term climate changeincluding its impact on

water resources. Te power sector is the largest single

contributor to U.S. carbon emissions33 percent,

largely because o coal plants (EPA 2013a).4 And because

they will last decades, new power plants have long-term

implications or carbon emissions and water use.

In the ace o overwhelming evidence o human-

induced climate change, the challenge is not whetherto address it but how best to both limit urther climate

change and adapt to what is already coming. While

meeting those challenges will require changes all across

the economy, the power sectors massive carbon emissions,

and the many lower-carbon options or producing electric-

ity, mean that our electricity choices will play a major

role in our ability to mitigate climate change.

Eaa o oTis report is the second rom the Energy and Water in

a Warming World initiative (EW3), organized by the

Union o Concerned Scientists to analyze the implica-

tions o U.S. electricity choices or our water supply and

water use.5 Produced by a team o experts rom univer-

sities, government, and nonprot organizations, this re-

port refects comprehensive new research on the impact

on water o a range o electricity choices at national,

regional, and local levels.Te report aims to inorm our choices so we saeguard

our water resources while obtaining reliable electricity at a

reasonable price, strengthening the economy, and reduc-

ing the carbon prole o our electricity supply. oward

that end, we explore options or cutting carbon emissions

rom power plants signicantly and reducing water with-

drawals and consumption and related risksincluding in

the nations driest and astest-growing regions.

Chapter 2 describes current energy-water colli-

sions, and climate change dimensions that are likely to

exacerbate those over the next ew decades. Chapter 3describes changes in uel costs, environmental regula-

tions, and technologies that have produced a pivot point

in the U.S. electricity sector. Tat chapter also explains

the distinctive approach developed by our research team

to produce new ndings on water and other implica-

tions o various electricity pathways.

Chapter 4 compares the carbon emissions, water

use and impact, risks, and costs o a range o scenarios

or the electricity sector, including business-as-usual and

low-carbon cases, drawing on both our own research and

that o others. Chapter 5 suggests strategies or incorpo-rating water more ully into decision making on which

new power plants to build or existing plants to retire, to

reduce the sectors water-related impact and strengthen its

resilience in the ace o a changing climate.

Compounding actors. m pw ps

s w uss. t s pupssu w sus w u

s s w.

WaerkeeperAllance

4 in hs repor, carbon emssons reers o carbon dode equalen, akng no accoun he poency o arous hea-rappng gases such as mehane and nrous

ode and conerng her oal global warmng mpac o an equalen mass o carbon dode, he mos prealen hea-rappng gas.

5 the rs EW3 repor was Aery e al. 2011.



15/58

KEy FIndIngs

Energy-water collisions are happening now.

Because o its outsized water dependence, the U.S.

electricity sector is running into and exacerbat-

ing growing water constraints in many parts o

the country. Te reliance o many power plants on

lakes, rivers, and groundwater or cooling water can

exert heavy pressure on those sources andleave theplants vulnerable to energy-water collisions, partic-

ularly during drought or hot weather. When plants

cannot get enough cooling water, or example,

they must cut back or completely shut down their

generators, as happened repeatedly in 2012 at plants

around the country.

As the contest or water heats up, the power sector

is no guaranteed winner. When the water supply has

been tight, power plant operators have oten secured

the water they need. In the summer o 2012, or ex-

ample, amid soaring temperatures in the Midwest andmultiple large sh kills, a handul o power plant op-

erators received permission to discharge exceptionally

hot water rather than reduce power output. However,

some users are pushing back against the power sectors

dominant stake. In Utah, or example, a proposal to

build a 3,000-megawatt nuclear power plant ueled

grave concerns about the impact o the plants water

use. And in exas, regulators denied developers o a

proposed 1,320-megawatt coal plant a permit to with-

draw 8.3 billion gallons (25,000 acre-eet) o water

annually rom the states Lower Colorado River.

Climate change complicates matters. Energy-

water collisions are poised to worsen in a warming

world as the power sector helps drive climate change,

which in turn aects water availability and quality.

Climate change is already constraining or alter-

ing the water supply in many regions by chang-

ing the hydrology. In the Southwest, or example,

where the population is growing rapidly and water

chAptEr 2

Energy-Water Collisions

supply is typically tight, much o the surace water

on which many water users depend is declining.

Scientists expect rising average temperatures, more

extreme heat, and more intense droughts in many

regions, along with reductions in water availability.

te pwe se a Wae r

Power plants are aected by water quantity, water qual-ityparticularly temperatureor both. Intake water

that is too hot can reduce the eciency o a power

plant, or even make it unsae to operate (UCS 2007).

And hot water exiting a plant can place it out o compli-

ance with temperature limits set to prevent harm to eco-

systems, leading to sh kills and other eects (Madden,

Lewis, and Davis 2013).

When plants cannot get enough cooling water,

operators must reduce power production or completely

shut down the generators, as happened repeatedly in

2012 at plants around the country. For example, opera-tors o the Powerton coal plant in central Illinois had to

temporarily shut down a generator during peak summer

heat, when water in the cooling pond became too warm

or eective cooling (Bruch 2012; Schulte 2012).

Operators o the 620-megawatt Vermont Yankee

nuclear plant cut power production by up to 17 per-

cent in July that year, because o high water tempera-

tures and low fows in the Connecticut River (Harvey

2012; Nuclear Regulatory Commission 2012a). In

Connecticut, operators shut down one o two reac-

tors o the Millstone nuclear plant in mid-July becausewater in Long Island Sound was too warm to cool the

plant (Wald 2012). And operators o the Gallatin and

Cumberland coal plants in ennessee had to limit power

output because o high river temperatures (VA 2012).

Te 2012 drought was severe, but the power plant

problems it provoked were hardly unique. In 2011, or

example, during the historic drought in exas, plants

had to cut back their operations and truck water in to

address the lack o cooling water (Averyt et al. 2011).



16/58

Over roughly the past decade, a range o water-relatedissues have cropped up around the country, aecting a

variety o power plants (Figure 1).

Wnns and Loss n Wat Collsons

When water supply is tight and users are in competi-

tion, power plants oten win. During the 2011 drought,

or example, the Texas Commission on Environmental

Quality elected not to suspend power plant water rights

because o saety concerns (Ickert 2013).

However, the impact on the losing side can be con-

siderable. In 2012, or example, at least seven coal andnuclear plants in Illinoisincluding the Will County

and Joliet coal plants and the Braidwood and Dresden

nuclear acilitiesreceived state waivers to discharge

water hotter than their permits allow. Regulators ap-

proved the thermal variances even though hot water

in rivers and streams was already causing extensive sh

kills across the Midwest (Spanger-Siegried 2012).

Given heightened confict over water resources,

some states and communities have pushed back

Figure 1. Energy-Water Collisions

Power plant dependence on water can create a range o problems, including or the plants themselves. Plants

have recently run into three kinds o challenges: incoming cooling water that is too warm or efcient and sae

operation, cooling water that is too hot or sae release into nearby rivers or lakes, and inadequate water sup-

plies. In response, operators must reduce plant output or discharge hot water anyway, at times when demand

or electricity is high and rivers and lakes are already warm.

Turning up the heat in local waters. When power plants dis-

charge hot cooling water back into lakes and rivers, they can

raise water temperatures, disrupting local ecosystems. Dur-

ing the extensive drought and heat o summer 2012, the Braid-

wood nuclear plant in Illinois was one o at least seven coal

and nuclear plants to receive state permission to release hot-

ter water than normal, so they could keep producing power.

With worse summer heat projected or the Midwest and much

o the nation in coming decades, water-dependent plants

could require more such thermal variancesstressing lakes,

rivers, wildlie, and the millions o people who count on them.

Flickr/NuclearRegulatoryCommission

Source: Spanger-

Siegfried 2012.

Note: Selected

events, 20062012.



17/58

including exas and Colorado (Ceres 2013a). And

even in water-rich Pennsylvania, the Susquehanna

River Basin Commission temporarily suspended waterwithdrawals or hydroracking o natural gas in summer

2012 because o low stream levels (SRBC 2012).

cae ca ta a twAn assortment o meteorological conditions led to

2012s punishing heat, drought, and storms. But many

o the incidents bear connections to longer-term climate

change trends. Climate change projections point to

increases in average temperatures as well as extreme heat

in most regions, an intensication o droughts, espe-

cially in the Southwest and Great Plains, and reductionsin water availability in the Southwest and Southeast

(Dai 2013; Kunkel et al. 2013a; Hoerling et al. 2012a).6

Meanwhile some regions such as the Northeast and

Midwest are expected to see more precipitation, deliv-

ered through extreme rainall events that occur more

oten (Bales et a l. 2013; Kunkel et al. 2013b).

Regional trends toward higher temperatures,

more intense precipitation, longer and more persistent

drought, and other extremes oer strong evidence that

the climate is already changing, both globally and in

the United States (IPCC 2012; Karl et al. 2009). Terst decade o the twenty-rst century was the hot-

test on record globally. Te 10 hottest years on record

worldwide have all occurred in the last 15 years (NCDC

2013; NOAA 2012). Average temperatures have risen

0.3 F to 0.45 F (0.17 C to 0.25 C) each decade since

the late 1970s (EPA 2013b).

Nationally, the 2012 drought was the worst in hal

a century, with more than 60 percent o the continental

United States suering rom moderate to exceptional

drought (Freedman 2012) (see Figure 2). In the sum-

mer o 2011, exas suered rom the driest 10 monthssince recordkeeping began in 1895 (LCRA 2011). And

research shows that human activities have already in-

creased the probability o extreme heat events like that

o 2011, and exacerbated drought intensity (Hoerling

et al. 2012b; Weiss, Overpeck, and Cole 2012).

against or rejected proposals to build power plants that

would require too much water. A proposal to build a

3,000-megawatt nuclear power plant on Utahs Green

River, or example, ignited erce opposition centered on

its proposed water use (Hasemyer 2012; NoGRN 2012;HEAL Utah 2011). exas regulators denied a request by

developers o the proposed $2.5 billion, 1,320-megawatt

coal-red White Stallion Energy Center to withdraw

8.3 billion gal lons (25,000 acre-eet) o water annu-

ally rom the states Lower Colorado River (SCO

2011). And Arizonas public utility commission ruled

that a proposed 340-megawatt concentrating solar

power plant in Mohave County must use dry cooling

or treated wastewater rather than 780 million gallons

(2,400 acre-eet) o groundwater annually rom the

Hualapai Valley Aquier (ACC 2010). We can expect tosee more such collisions.

Te contest over water extends to the production

o uels or power plants as well (see Box 2, p. 23). In

Colorado and elsewhere, the purchase by the hydraulic

racturing (hydroracking) industry o water rights or

gas and oil extraction has prompted concern about the

impact on armers (Burke 2013; Finley 2012; Healy

2012). Almost hal o all hydroracking is occurring

in regions with high or extremely high water stress,

Stakeholders step in. txss s 2011 u z

w uss pps W S, pps

p su us s s w p. a lw c r au

w qus 2011, ps pps

, w wu u w us s ps

pp sss. t ps s sus

p p (gw 2013; h 2013).

MchaelSraaoorTheTexasTribune

6 the changes oulned n hs secon are projeced o be smlar oer he ne

seeral decades under boh hgher and lower greenhouse gas emssons sce-

naros; greaer derences beween scenaros become apparen n he second

hal o he cenury (Kunkel e al. 2013a).



18/58

Stream fows in major Southwest rivers were

5 percent to nearly 40 percent lower rom 2001 to 2010

than average twentieth-century fows (Hoerling et al.

2012c; Rousseau 2012). Indeed, tree ring data suggest

that the western United States had the driest conditions

in 800 years over the last decade (Schwalm et al. 2012).

Meanwhile the regions population is growing rapidly.

In the Southwest, the vast majority o water with-

drawn is used to irrigate arid agricultural lands (Kenny

Water pressure in the Southeast. i Sus,

p s ( ) us, p w

ppu w s s s

u u s. r ws

ups, Sus ss s,

ss us . i s x, ss

w squs, u s k

a, w ps w gs lk l,

sw u 20072008 u. cs u

u w lk l

wsus w up Sus ss.

Flckr/Seenv

7 Scenss hae arbued up o 60 percen o he change n arral me o

rsng concenraons o hea-rappng gases n he amosphere (Garn 2012).

8 Recen warmng n he Souheas (2 F [1.1 C] snce 1970) ollows a cool

perod n he 1960s and 1970s, and subsanal arably n he rs hal o he

weny-rs cenury (Kunkel e al. 2013a).

et al. 2009). Te regions water supply depends heavily

on snowpack, which melts in spring, supplying water to

streams and reservoirs. Yet snowpack and stream fow

are declining (Hoerling et al. 2012c; Overpeck and Udall

2010). Snowpack has been melting ever earlier over the

past 50 years, so most o each years stream fow is arriv-

ing earliera shit attributed partly to climate change

(Garn 2012; Hidalgo et al. 2009; Pierce et al. 2008;

Stewart, Cayan, and Dettinger 2005).7

In the Southeast, average annual temperatures have

been rising steadily in recent decades, with 2001 to2010 the warmest on record. 8 Summers in the region

have shited toward the hydrological extremes: either

very dry or very wet compared with the middle o the

twentieth century (Kunkel et al. 2013a; Wang et al.

2010). In a region where thermoelectric power plants

account or more than two-thirds o water withdrawals,

states are in continual confict over water use, creat-

ing demand-driven drought conditions in some areas

(Georgakakos, Zhang, and Yao 2010; Kenny et al.

2009). And unlike in the Southwest, Southeast reser-

voirs typically have the capacity to store just a singleyears water use, making the water supply vulnerable to

both short-term and long-term changes (Ingram, Dow,

and Carter 2012).

Source: Drough Monor

map rom Augus 21, 2012.

the U.S. Drough Monor

s produced n parnershp

beween he Naonal

Drough Mgaon

Cener a he Unersy

o NebraskaLncoln, he

Uned Saes Deparmen

o Agrculure, and

he Naonal Oceanc

and Amospherc

Admnsraon. Map

couresy o NDMC-UNL. SaulLoeb/AFP/Geyimages

FigurE 2. d ts, Ps Fuu

S xp u sss s s su 2012, wp ss

uu . W u p s s , w p u

s s.



19/58

Given continued high carbon emissions, Caliornias

snow water equivalentthe depth o water i snowpack

were meltedis projected to drop nearly 60 percent by2099. And scientists expect Arizonas snow water equiv-

alent to decline by nearly 90 percent, and Colorados

which supplies water to much o the regionby more

than 25 percent.11 Along with rising population, those

changes would deeply compromise the ability o the

water supplyalready scarce and overallocatedto meet

the needs o power plants as well as Southwest cities,

agriculture, and ecosystems.

19301920 2000 2010199019801970196019501940

Volume(MillionAcre-feet)

25

20

15

10

5

0

Water Supply(10-year Running Average)

Water Use(10-year Running Average)

Compared with recent averages, temperatures

are projected to rise another 2.5 F to 5.5 F (1.4 C to

3.1 C) in most regions o the United States by mid-century (Kunkel et al. 2013a). In that time rame,

parts o the Southeast, Southwest, and South Central

U.S. can expect an additional 25 days above 95 F each

year, on average (Kunkel et al. 2013a). As warming

shits historic patterns o precipitation, hydrology will

become more variable and prone to extremes. Extreme

heat events and droughts are expected to become more

intense in many regions, especially the Southwest and

Great Plains (Cayan et al. 2013; Dai 2013; Kunkel et al.

2013a; Hoerling et al. 2012a; Hoerling et al. 2012c).

Rising temperatures and changes in precipitationmean that less water would be available in the Southwest

and Southeast over the longer term (Kunkel et al. 2013a;

Caldwell et al. 2012). In the Southeast, scientists expect

the net water supply to decline by 2060 while popu-

lation and demand rise, worsening water stress and

aecting wildlie in some o the nations most sensitive and

biologically diverse rivers (Kunkel et al. 2013a; Caldwell

et al. 2012; U.S. Census Bureau 2010).9

For many major Southwest cities, water supply

challenges are highly likely in the decades ahead even

withoutclimate change (Figure 3).10 Yet a changing cli-mate is expected to intensiy Southwest drought and lower

both surace and groundwater levels signicantly (Kunkel

et al. 2013a; Overpeck and Udall 2010). Climate

scientists also project urther drops in late winter and

spring snowpack and subsequent reductions in runo

and soil moisture, which are vital to regional reservoirs

(Cayan et al. 2010; Cayan et al. 2008; Christensen and

Lettenmaier 2007).

FigurE 3. W Supp sus W d c r bs

o s u, u fw c r

s u 16 (5 s)

p . hw, w us s s s ,

w w supp s pp us u.

rs w s u wws

w s ss su s lk m lk

Pw.12 Source: USBR 2012.

9 these declnes are epeced o be sronges n he subregon spannng Georga, Alabama, tennessee, Msssspp, Lousana, and teas. Pars o he Eas Coas may

see some ncreases n precpaon (Kunkel e al. 2013a; Caldwell e al. 2012). Durng perods o ereme hea and drough, hgh rer emperaures can combne wh

hermal dscharges o reduce dssoled oygen and he capacy o sreams o absorb wase (Kunkel e al. 2013a).

10 the probably o concs reec s seeral acors, ncludng populaon growh and he waer requremens o endangered speces. Mulple locaons hroughou

he Souhwes are consdered subsanally lkely or hghly lkely o see waer concs by 2025, een whou he eecs o clmae change (USGCRP 2009; USBR 2005).

11 Calorna , Neada, Uah, Colorado, Arzona, and New Meco are projeced o see marked reducons n snow waer equalen. Declnng precp aon s he cause n

some cases, and a sh oward more ran and less snow n ohers (Cayan e al. 2013).

12 toal use o waer hroughou he basn ncludes agrculural, muncpal , ndusral, and oher consumpe uses (ncludng ows o Meco), plus use by egeaon

and losses hrough eaporaon a mansream reserors. Naural ow s used o esmae waer supply n he basn. in he curren naural ow record, hsorcal nows

based on U.S. Geologcal Surey gauged records are used o esmae he naural ow or he Para, Lle Colorado, vrgn, and Bll Wllams rers, whou adjusng or

upsream waer use. Howeer, he Gla Rer s no ncluded n he naural ow record. thereore, he use repored here ecludes consumpe uses on hese rbuares.



20/58

KEy FIndIngs

Te U.S. power sector is undergoing rapid

transormation. Te biggest shit in capacity and

uel in hal a century is under way, as electricity

rom coal plants shrinks and power rom natural

gas and renewables grows. Several actors are spur-

ring this transition to a new mix o technologies

and uels. Tey include the advanced age o manypower plants, expanding domestic gas supplies and

low natural gas prices, state renewable energy and

eciency policies, new ederal air-quality regula-

tions, and the relative costs and risks o coal-red

and nuclear energy.

Tis presents an opportunity we cannot aord

to miss. Decisions about which power plants to

retrot or retire and which kind to build have both

near-term and long-term implications, given the

long lietimes o power plants, their carbon emis-

sions, and their water needs. Even a single aver-age new coal plant could emit 150 million tons o

carbon dioxide over 40 yearstwice as much as a

natural gas plant, and more than 20 million cars

emit each year. Power plants that need cooling

water will be at risk over their long lietimes rom

declining water availability and rising water tem-

peratures stemming rom climate change, extreme

weather events, and competition rom other users.

And power plants, in turn, will exacerbate the water

risks o other users.

cae i ue WaIn 2007, the U.S. electricity sector still pointed

strongly toward more coal-red power. Te industry

was proposing to construct 159 new coal-red plants,

or 96,000 megawatts o new capacitya 29 percent

increase over existing coal capacity (Shuster 2007). Te

Energy Inormation Administration (EIA) projected

that electricitys uel mix would persist largely un-

changed through 2035 (EIA 2007).

But the picture quickly began to change dramati-

cally. Coal shrank rom ueling nearly hal o U.S.

power production in 2008 to 37 percent in 2012(Figure 4) (EIA 2013a). By early 2013, plant owners

had announced plans to retire almost 50,000 megawatts

o coal plants14 percent o the U.S. coal feetand

another 52,000 megawatts were economically vulner-

able (UCS 2013).

Meanwhile, electricity ueled by natural gas rose

rom 21 percent to 30 percent o the U.S. mix rom

2008 to 2012, while power rom non-hydro renew-

ables such as wind and solar grew rom 3.1 percent to

chAptEr 3

Pivot Point for U.S. Power

FigurE 4. e S ts: tU.S. e mx r c Ps

gw us pw u s

ws w us pw ps s U.S.

x ( xs). i 2008, supp s

U.S. . b 2012, s pp 37 p

, w u s w sup

p 35 p. ts uss ws

s s , u p

( xs). (tW = wus,

wus; gW = ws, uss ws)

2008 2012201120102009

2008 2015201420132012201120102009

Generation(TW

h)

CumulativeCoalCapacityRetirements(GW)

2,500

2,000

1,500

1,000

500

0

0

5

10

15

20

25

30

35

40

45

Coal

Non-Hydro Renewables

Natural Gas

Non-Hydro Renewables+ Natural Gas

Sources: EiA 2013a;

SNL 2013.



21/58

5.4 percent (EIA 2013a). Tese changes are being driven

by the advanced age o many power plants; a signicant

expansion o U.S. natural gas production; low natural

gas prices; state renewable energy, energy eciency, and

climate policies; new ederal air-quality regulations;

and the costs and risks o coal-red and nuclear energy

(see Box 1). Coal s dominance seems likely to continue

to wane as operators retire more coal plants, given

pressure to upgrade pollution controls and competition

rom other power sources (Cusick 2013).

box 1. Eer Techolo Traitio

prces hae recenly dropped, coal producers usng

mounanop remoal are acng more srngen regula-

ons desgned o proec local sreams, waer qualy,

and publc healh (EiA 2013d; H 2013; U.S. ACE 2012;

EPA 2011b). Meanwhle he aerage age13 o he U.S. coal

ee s 42 years, and many older plans are necen (SNL

Fnancal 2013). the ederal goernmen has also aken

seps o cu ar polluon and reduce he publc healh e-

ecs rom coal-red plans, and carbon dode emssons

rom new power plans, so operaors are acng new coss

(EPA 2012a; EPA 2012b).

nea. Whle calls or more low-carbon elecrcy se-

eral years ago led o predcons o a nuclear renassance,

he secor has sruggled. Hgh coss hae plagued nuclear

energy or decades (McMahon 2012; Madsen, Neumann,

and Rusch 2009). the 2011 Fukushma Dach dsaser n

Japan, leaks o radoace seam ha led o he shudown

o Calornas San Onore plan, and oher ncdens also

dampened enhusasm or a nuclear renassance (Lee

2013; Nuclear Regulaory Commsson 2012b). Few nuclear

projecs are mong orward n 2013, and some plans are

beng rered, ncludng San Onore (EiA 2013e; SCE 2013).

naa a. A sgncan epanson o U.S. naural gas

producon s drng changes n he m o uels used o

generae power. indusrys use o horzonal drllng and

hydraulc racurng (hydrorackng) o ap he gas n deep

shale ormaons had begun o epand domesc supples

by 2010. By 2012, U.S. naural gas producon had clmbed

34 percen rom 2005 leels (EiA 2013b). tha ncrease

along wh weaker demand and some warmer wners

brough naural gas prces o near-record lows (EiA 2013c).

reewae ee. Sae and ederal renewable elecrc-

y polces, and a recen declne n he cos o wnd and

solar power, hae acceleraed he growh o renewable

energy. the U.S. wnd energy ndusry had nsalled more

han 40,000 urbnes, capable o supplyng 60,000 mega-

was, by he end o 2012 (AWEA 2013). Wnd power

accouned or more han 35 percen o all new capacy

nsalled rom 2008 o 2012more han nuclear and coal

combned (AWEA 2013). Meanwhle he U.S. solar ndusry

nsalled 76 percen more capacy n 2012 han had n

2011, and oal solar capacy epanded by a acor o e

rom 2009 o 2012 (SEiA 2013).

ca. Whle naural gas coss plummeed, coal prces

rose by 31 percen rom 2007 o 2011 because o rsng

producon and ransporaon coss (EiA 2012). Alhough

Energy-water-climate tradeos. t ks

, w, px. ru

sss qu u ss w u sss,

qus w. y ss

, s s sss

u . nu ps sss u

ws. c pu s ps, s

b p m, al, pu , s

s sss ps p u

u. hw, s ks pw p

u ss , ss w sup

45 p 90 p (mkk . 2012).WkCommons/Alarsar

13 tha s, he capacy -weghed mean.



22/58

We analyzed three other scenarios that refect the

power sector component o an economy-wide carbon

budget to achieve 80 percent lower emissions in 2050

than in 1990, as a budget that has a reasonable chance

o limiting global heat-trapping emissions to 450 parts

per million (National Research Council 2010).15

Tree-quarters o cuts in such emissions would come rom the

electricity sector, according to studies by the Stanord

Energy Modeling Forum and National Research

Council, because it has more near-term opportunities

(Figure 5) (Clemmer et al. 2013; Fawcett et al. 2009).

We ocused two o our three carbon budget sce-

narios on particular energy technologies. One assumed

aggressive deployment o renewable energy and energy

eciency technologies. Another assumed high levels o

coal use with carbon capture and storage (CCS)to

reduce heat-trapping emissions rom coal plantsandnuclear energy. (Te third scenario included the carbon

budget but did not speciy particular technologies. We

ocused much o our analysis on the rst two.)

b e Ee se etwe ce

Tese rapid changes give us the opportunity to putthe United States on a pathway that lowers carbon

emissions rom the electricity sector while curbing

energy-water collisions. Power plant owners, developers,

regulators, and legislators are making critical choices

now about our nations electricity mix in coming

decades. For improved resilience o long-lived power

projects and the sector as a whole, these decision mak-

ers will need to consider the impact o climate change,

greater hydrologic variability, higher peak electricity

demand, and the need to switly and deeply cut carbon

emissions.o analyze the impact o various options or our

electricity uture on water withdrawals and consump-

tion, carbon emissions, and electricity and natural gas

prices, we ocused on several key scenarios. Te rst

business as usualassumed an electricity mix based

on existing state and ederal policies and the costs o

various technologies.14

2010 2020 205020402030

2,500

2,000

1,500

1,000

500

0

Heat-trappingEmissions

(MtCO

2eqperYear)

Carbon Budget

Business as Usual

FigurE 5. a c bu U.S.e S

U ussssusu pw, U.S. pw

p sss wu s s s

s. U w u u

sss 80 p 2050, s u

u qus us 2010

2050, ps (n rs

cu 2010; Fw . 2009). (mco2q = s

x qu)

Source: Clemmer e al. 2013.

How renewable energy stacks up, water-wise. S w

s, su s w us s

ps, us ss w pu

. os, su s , ss, s pw ps, us s psss

, us w s. Su ps

us , u, pu w

us ().

Flckr/PhooMojoMke

14 We paerned hs scenaro aer he reerence case n he EiAsAnnual Energy

Outlookor 2011 (EiA 2011).

15 Four hundred and y pars per mllon s he leel projeced o prode

a roughly 50 percen chance o keepng he global aerage emperaure rom

rsng more han 3.6 F (2 C) aboe pre-ndusral leels (Luers e al. 2007).



23/58

We chose these energy pathways to accentuate di-

erences in water withdrawals and consumption. Under

the business-as-usual case, the electricity sector remains

heavily dependent on ossil uels, which tend to use

water or cooling. Under the renewables-and-eciency

scenario, while some renewable sources such as geo-

thermal, biomass, and concentrating solar can be

water-intensive, most o the electricity demand could

be met with technologies that use little or no water.Under the CCS-and-nuclear scenario, electricity comes

mostly rom technologies that require large amounts

o water.

Our carbon budget scenarios include technology

mixes that are within the range o those o other analyses,

although at or near the upper end or each technology

(Figure 6) (Clemmer et al. 2013). We drew assumptions

about the cost and perormance o dierent technologies

primarily rom the EIAsAnnual Energy Outlook, updated

with data on recent projects (Clemmer et al. 2013). We

also used the EIAs projections or growth in electricity

demand by 2035, extrapolated to 2050.

Focusing on Two Vulnerable Regions

Beyond our national and general regional analyses, we

also explored how various power plant choices inter-

act with water supply and demand in particular water

basins (river systems) in the Southeast and Southwest.

Both regions have seen energy-water conficts, though

or dierent primary reasonswater scarcity in

the Southwest, and high water temperatures in the

Southeast.

As noted, the Southwest is acing rapid popula-

tion growth and rising electricity demand while water

resources are declining (Kunkel et al. 2013a). In that re-

gion, our modeling ocused on the surace and ground-

water systems in the Colorado River Basin and related

areas, including the Upper and Lower Colorado rivers,

the Rio Grande, andgiven long-distance transport o

water in the regionNorthern and Southern Caliornia

(Figure 7A).

Te Southeast is also seeing rapid population

growth, and is vulnerable to rising temperatures

and declining water availability in coming decades

(Ingram, Dow, and Carter 2012). In that region, we

ocused on the Alabama-Coosa-allapoosa (AC) and

the Apalachicola-Chattahoochee-Flint (ACF) basins

in Georgia, Alabama, and the Florida Panhandle

(Figure 7B). Tose states have ought over water rom

the two basins, particularly in times o water stress

(Yates et al. 2013a).

2010 2020 205020402030

100%

80%

60%

40%

20%

0

Penetration

Carbon Capture andStorage

Renewables

Energy Eciency

Nuclear

FIGURE 6. Technology Targets or the ElectricitySector

Our modeling included aggressive targets or technologies

that could provide the largest cuts in carbon emissions over

the next 40 years, according to numerous studies. Under a

scenario that emphasized renewable energy and efciency,

or example, we assumed that energy-efcient technolo-

gies and buildings would reduce U.S. electricity use by about

1 percent per year on average, and that electricity genera-

tion rom renewable energy technologies would grow rom

about 10 percent in 2010 to 50 percent in 2035 and 80 percent

by 2050.

Source: Clemmer et al. 2013.

For improved resilience o long-lived

power projects, decision makerswill need to consider the impact o

climate change, greater hydrologic

variability, higher peak electricity

demand, and the need to switly and

deeply cut carbon emissions.



24/58

o iae Aa meOur innovative approach to exploring the water im-

plications o electricity choices entailed pairing two

modelsone on electricity and one on waterand

eeding in our range o scenarios. Te rst model is

the Regional Energy Deployment System (ReEDS) o

the National Renewable Energy Laboratory. ReEDS

allowed us to analyze power generation by uel type or

134 regions around the countrya much ner degreeo geographic resolution than other models.

Te second model is the Water Evaluation and

Planning (WEAP) system o the Stockholm Environ-

ment Institute, which allowed us to analyze water

withdrawals and consumption in the power sector based

on results rom ReEDS. WEAP uses climate-driven

simulations o water supply and detailed descriptions

o water demand to capture both basin-wide and local

tradeos amid changing water conditions.

In both the WEAP analyses and higher-level (nation-

al and regional) results, we based water use on published

inormation or various combinations o power plant

uels and cooling technologies (Macknick et al. 2012a).16

For our two key regions, we ed results romReEDS into new WEAP-based models o the target

river basins (Sattler et al. 2013). We based precipitation

on dry sequences o years rom recent history (Flores-

Lopez and Yates 2013; Yates et al. 2013b). We also as-

sumed that air temperatures would rise by 3.6 F (2 C)

FigurE 7. eW cs tw rs

ew s . i Suws (a), u c r bs, ppu

s u w sus . i Sus (b), u acs

tps (act) bs apcF (acF) bs, s p ppu

w, s pus, s w. a, g, F s ss w,

pu u s w sss.

16 We used he medan alues or whdrawal and consumpon or each combnaon (Macknck e al. 2012a). We ncorporaed only waer use a power plans, chey

or coolng. We dd no nclude waer whdrawals or consumpon assocaed wh hydroelecrc acles, whch can be more challengng o calculae because o, or

eample, he mulple uses o reserors n addon o power generaon (Aery e al. 2011). the ReEDS model coers he 48 conguous saes, so calculaons do no

nclude Alaska and Hawa.

A b



25/58

We assumed that the water eciencies o various

technologiesthat is, water use per unit o electric-

itywould not change over time. In reality, theeciencies o established ossil uel, nuclear, and

renewable technologies, and newer technologies

such as CCS, may improve.

Our modeling does not take into account that other

water users may change their behavior in response

to changing water conditions, or that users o elec-

tricity may change their habits based on higher or

lower prices.

Our modeling provides monthly averages or water

use and eects. However, changes over shorter peri-ods o time could prove important in energy-water

collisions. We also used average gures or power

production at various points during each year, which

may not capture potentially important periods o

peak demand. In analyzing water temperatures, we

applied any average/monthly result below a certain

threshold (90 F, or 32 C) to the whole month, and

any result above that threshold to the whole month.

between 2010 and 2050 (Flores-Lopez and Yates 2013;

Yates et al. 2013b), consistent with projections used in

the National Climate Assessment (Kunkel et al. 2013a).Our approach also included several other notable

aspects:

We ocused on the impact o water use or cooling

power plantsonly a subset o the water implica-

tions o electricity choices. Other aspects o these

choices could also have important water-related

eects, such as the use o hydraulic racturing to

extract natural gas.

Climate change will likely increase peak power

demand, as hotter days drive more use o air

conditioning, or example. Te EIA projections or

electricity demand that we draw on do not take that

into account.

While decision makers shaping our electricity

uture should consider the availability o water

or power, our modeling treated water dimensions

solely as outputs, not inputs.

Linking electricity modeling

and water modeling.

ou pp

z

w ks

s

ss

w psp:

ss suss.

t c r, pu

, s ss

w .

Flckr/james_gordon

_losangeles



26/58

However, low-carbon power is not necessarily

water-smart. Te menu o technologies qualiy-

ing as low-carbon is long, and includes some with

substantial water needs. Electricity mixes that em-

phasize carbon capture and storage or coal plants,

nuclear energy, or even water-cooled renewables

such as some geothermal, biomass, or concentrating

solar could worsen rather than lessen the sectorseects on water.

Renewables and energy efciency can be a win-

ning combination. Tis scenario would be most

eective in reducing carbon emissions, pressure

on water resources, and electricity bills. Energy

eciency eorts could more than meet growth

in demand or electricity, and renewable energy

could supply 80 percent o the remaining demand.

Although other low-carbon paths could rival this

one in cutting water withdrawals and consump-

tion, it would edge ahead in reducing groundwateruse in the Southwest, improving river fows in the

Southeast, and moderating high river temperatures.

Tis scenario could also provide the lowest costs to

consumers, with consumer electricity bills almost

one-third lower than under business as usual.

be a ua: g, ba, a uEven under the business-as-usual pathway, we ound the

electricity mix would change drastically over the next

several decades, consistent with the rapid transorma-

tion now under way. Power production rom coal wouldshrink signicantly, based on plant retirements that

have already been announced and continued pressure

rom low natural gas prices, ederal air-quality and

other regulations, and clean energy policies.17

KEy FIndIngs

Business as usual in the power sector would ail

to reduce carbon emissions, and would not tap

opportunities to saeguard water. Because such a

pathway or meeting uture electricity needs would

not cut carbon emissions, it would do nothing to

address the impact o climate change on water.

Changes in the power plant feet would mean thatwater withdrawals by power plants would drop, yet

plants water consumption would not decline or

decades, and then only slowly. Te harmul eects

o power plants on water temperatures in lakes and

rivers might continue unabated, or even worsen.

Greater extraction o ossil uels or power plants

would also aect water use and quality.

Low-carbon pathways can be water-smart. A

pathway ocused on renewable energy and energy

eciency, we ound, could deeply cut both carbon

emissions and water eects rom the power sec-tor. Water withdrawals would drop 97 percent by

2050much more than under business as usual.

Tey would also drop aster, with 2030 withdrawals

only hal those under business as usual. And water

consumption would decline 85 percent by 2050.

Tis pathway could also curb local increases in wa-

ter temperature rom a warming climate. Meanwhile

lower carbon emissions would help slow the pace

and reduce the severity o climate change, including

its long-term eects on water quantity and quality.

chAptEr 4

Findings: The Impact of Power Pathways

on Water

t s sus qu

ussssusu pw s

wu u

pw ss sss. 17 Ecep where oherwse noed, elecrcy resuls are drawn rom Clemmer eal. 2013 and relaed research.



27/58

2010 2020 205020402030

ElectricityGeneration(TWhperYear)

6,000

5,000

4,000

3,000

2,000

1,000

0

2050, according to our modeling. Te EIA similarly

projects that electricity-related carbon emissions under

a business-as-usual scenario would not drop, and wouldindeed rise 12 percent above 2012 levels by 2040 (EIA

2013). Climate change would continue relatively un-

abated, with commensurate eects on water availability,

air and water temperatures, and demand or water and

electricity.

A business-as usual pathway would have some posi-

tive eects on water use by power plants but many neg-

ative eects. Virtually all plants cooled by once-through

systems would be among the plants that our modeling

predicted would be retired based on costs. New power

acilities would be more ecient, and would use recir-culating or dry cooling. In act, reshwater withdrawals

would drop more than 80 percent rom todays levels by

2050, and water consumption more than 40 percent,

under our business-as usual case.19 Reductions on that

scale would, at ace value, strengthen the power sectors

ability to cope with changes in water availability and

temperature while easing pressure on water resources.

However, a business-as-usual trajectory would

bring amiliar problemsand new onesat national,

regional, and local levels.

National. Water withdrawals or power plants would

drop under a business-as-usual scenariobut only

slowly until around 2030.20 Water consumption would

stay basically unchanged or two decades beore nally

dropping. Neither trajectory would position the power

plant feet to perorm well i a deep drought hit beore

2030. Even by 2050, water consumption would all by

less than hal, prolonging the power sectors exposure to

water risks (Figure 9).

We prioritized retirement o plants with once-

through cooling, so our modeling produced relativelysteep reductions in water use. Other analyses project less

encouraging water trends. A study rom the National

Energy echnology Laboratory, or example, ound that

reshwater consumption in the power sector would rise

Nuclear-powered electricity would stay near todays

levels or two decades, then steadily all to nearly zero

by 2050, as existing nuclear plants reach the end o

their assumed 60-year lives and new nuclear reactors

would be unable to compete economically. Natural gas

would dominate the electricity mix, supplying almost

60 percent o the nations power by 2050 (Figure 8).18

Te most serious critique o a business-as-usualpathway is that it would do little or nothing to reduce

the power sectors carbon emissions, because o con-

tinued use o ossil uels and rising demand. Emissions

would stay within 5 percent o todays levels through

FigurE 8. U.S. e mx u bussss Usu, 20102050

t x wu k x s

s u ussssusu pw,

p s u w. c pw wu

p s, s p s

u, pssu w u s

ps, s ps p pu

. nu pw wu sp

p, s xs ps s w

s wu u p . nu

s wu x, supp s60 p U.S. pw 2050. (Pv = s ps;

cSP = s pw) Source: Clemmer e al. 2013.

18 We used EiA assumpons abou naural gas supply and prces, erapolaed o 2050 (Clemmer e al. 2013).

19 Waer resuls here and ollowng consder he use o reshwaer sources, dened n hs repor as all non-ocean sources.

20 Ecep where oherwse noed, large-scal e waer resuls are drawn rom Macknck e al. 2013b and relaed research.

WindOshore

WindOnshore

PV

CSP

Geothermal

Hydro

Biopower

Coal with CCS

Coal

Gas with CCS

GasCoal with CCS

Coal

Gas

Coal

Nuclear



28/58

hal as much in the Southwest, though, because o con-

tinued use o existing coal plants.

A study o power plant water use in the Great Lakes

region ound that a business-as-usual pathway would

actually lead to a 10 percent increase in water withdraw-

als and consumption (Moore, idwell, and Pebbles2013). And power plants already account or 76 percent

o withdrawals and 13 percent o consumption in that

region, according to that study.

Local. o gauge the local impact o power production,

we looked at how oten river temperatures might exceed

a 90 F (32 C) threshold or thermal pollution in select

locations.21 On the Coosa River, power plants above

Weiss Lake on the Alabama-Georgia bordersuch as

the 950-megawatt Plant Hammond coal acility, which

uses once-through coolingaect the temperature othe river.22 Under a business-as-usual scenario, river

temperatures rom 2040 to 2049 would exceed 90 F

(32 C) 18 days per year, on averagethree times the

number rom 2010 to 2019.23

between 16 percent and 29 percent rom 2005 to 2030

(NEL 2009a).

Another study projected that water consumption

by power plants would increase between 36 percent and

43 percent rom 1995 to 2035 (idwell et al. 2012). Tat

growth would occur chiefy in water basins with rapidlygrowing demand outside the power sector. And 10 percent

to 19 percent o all new thermoelectric power production

would likely occur in watersheds with limited surace

and/or groundwater availability, the study reported.

Regional. Under the business-as-usual case, withdraw-

als in the Southeast, Midwest, and Northeast would

largely track the national decline, as their larger num-

bers o once-through-cooled plants are retired. In the

Southwest, exas, and the Great Plains, withdrawals

would drop much less than at the national level by2050in the Southwest by only one-third.

In the Southeast and elsewhere, water consump-

tion by power plants would eventually dropin the

Southeast by a third. Consumption would drop only

FigurE 9. Pw P W Us u busss s Usu, 20102050

W us pw ps wu sus s s 2050 u usss s usu,

s x. Wws () wu p 80 p, sup

() 40 p. y wws wu sw u u 2030,

sup wu s s u s. as kp w ppu

w ss, s s u , . (1 s = 3

) Source: Macknck e al. 2012b.

2010 2020 205020402030

50

40

30

20

10

0WaterWithdrawal(TrillionGallonsperYear)

2010 2020 205020402030

1.6

1.4

1.2

1

0.8

0.6

0.4

0.2

0WaterConsumption(Trillion

GallonsperYear)

Noe: Projecons shown

here and n laer guresncorporae medan

waer-use alues rom

Macknck e al. 2012a,

and nclude reshwaer

(non-ocean waer) only.

21 A leas 14 saes prohb waer dsc harges aboe 90 F (32 C) o aod harm o sh and oher wl dle (Madden , Lews, and Das 2013; EPA 2011d; Benger e al. 1999).

22 Ecep where oherwse noed, Souheas waer resuls are drawn rom Yaes e al. 2013a and relaed research.

23 We assumed no changes n operaon o he power plans or managemen o he rer because o hgher emperaures. Such changes could nclude he use o por-

able coolng owers, whch occurred a Plan Hammond n 2007 (Cheek and Eans 2008).



29/58

I power plant operators do not cut back production

in the ace o drops in the water supply, and they com-

mand priority access to the water, the amount available

or other uses will drop. Tis eect is most evident in

agriculture. Under business as usual in the South PlatteBasin that includes the Denver metropolitan area, or

example, 16 percent less water would be available or

agriculture in summer rom 2040 to 2049 than rom

2010 to 2019, on average.24

Continued reliance on water-using power plants

could also steepen declines in the amount o water

stored in reservoirs and fowing in rivers. Under our

business-as-usual case, the amount o water stored in

Lake Mead in Nevada and Arizona, and Lake Powell in

Utah and Arizona, or example, would be only 50 per-

cent o the long-term historical average (19712007)by 2050, and the 20402049 average would be one-

third below the historical average (Yates et al. 2013a;

NRCS 2008).25 Average annual stream fow in the

Winners and losers. i pw ps w su ( p , xp), u

us w supps p, u uss su s uu w sk. tk cs Su P s.

U usss s usu, 16 p ss w wu uu su 20402049 sus

2010 2019w u us. a x uss ss w u u s s sss. t

ws s u p s s (s uss ) w

uu 2040 2049.

Mighty reservoirs ace mighty strain. lk m, s

s s, azn , lk

Pw, az U, p u s Su

ws . t qu

s fw c r, p u w s ss (n rs cu

2007). y u w k

s ss. i n 2010, lk m ws us

41 p u, 4.5 w s (14

) 1999, w s u.

Flckr/dsearls

Flckr/SusanSharplessSmh

J.CarlGaner/CrcleoBlueww

w.c

rcleoblue.o

rg

24 Agrculural users oen hold he mos senor waer rghs n saes wh pror appropraon laws, whch nclude mos wesern saes. Under such laws, waer s

scarce, hose wh more junor rghs, ncludng many power plan operaors, mus lease or purchase waer rom users wh senor rghs. Farmers oen do sell waer o

plan operaors n such suaons, because dong so s lucrae or hem.

25 Ecep where oherwse noed, Souhwes waer resuls are drawn rom Yaes, Meldrum, and Aery 2013 and relaed research.



30/58

box 2. The Impact of Hrofracki o Water

Naural gas combned-cycle power plans are much more

hermally ecen han coal or nuclear plansmeanng

hey need less waer or coolng. Such plans also hae

much less o an mpac on he qualy o he local waer

supply han coal or nuclear plans usng he same coolng

echnologes.

Howeer, connued ramp-up o hydrorackng could

grealy dmnsh he ne waer adanages o power plans

ha use naural gas. Whle power plan waer use s much

larger per un o elecrcy poenally generaed usng

naural gas rom hydrorackng, waer quanyand

qualyssues are sll mporan o consder, parcularlyn he cny o hydrorackng operaons (Meldrum e al.

2013; Cooley and Donnelly 2012).

For eample, he U.S. Enronmenal Proecon Agency

(EPA) esmaes ha some 35,000 hydrorackng wells used

70 bllon o 140 bllon gallons o waer n 2011 (EPA 2011c).

Dependng on he ype o well and s deph and loca-

on, a sngle well can requre 3 mllon o 12 mllon gallons

o waer when s rs drlled and rackedmany mes

he amoun used n conenonal ercal drllng (COGA

2013; Brelng Ol and Gas 2012; NEtL 2009b). And opera-

ors use smlar amouns o waer each me hey ge awell a work-oer o manan pressure and gas produc-

on. A ypcal shale gas well wll undergo wo work-oers

durng s le span (NEtL 2012).

Whdrawng hese amouns o waer oer a shor

perod o me can sran local wae

Date post:	02-Apr-2018
Category:	Documents
Upload:	rachel-treichler
View:	217 times
Download:	0 times

Water Smart Power Full Report

Documents