www.epa.gov/researc
United States
Environmental Protection
Agency
    The Impact of Traditional and
    Alternative Energy Production
    on Water Resources:

    Assessment and Adaptation Studies

-------
                                     EPA/600/R-14/272
                                      September 2014
 The Impact of Traditional and Alternative Energy
Production on Water Resources: Assessment and
              Adaptation Studies

-------
                                 DISCLAIMER

       The U.S. Environmental Protection Agency, through its Office of Research and
Development, conducted, funded and managed the research described herein. The report "The
Impact of Traditional and Alternative Energy Production on Water Resources: Assessment and
Adaptation Studies", EPA/600/R-14/272, has been subjected to the Agency's peer and
administrative review and has been approved for external publication. Any opinions expressed in
this paper are those of the authors and do not necessarily reflect the views of the Agency,
therefore, no official endorsement should be inferred. Any mention of trade names or
commercial products does not constitute endorsement or recommendation for use.

-------
                                    ABSTRACT

       Water, fuel, and energy issues are intricately related and cannot be addressed in isolation.
With increasing population, increasing energy demand, continued migration towards and
population growth within water stressed regions of the U.S., and with the continuing impacts of
climate change on water availability, scarcity of freshwater will be an issue of paramount
importance. Finding alternative water resources to replace freshwater demand for thermoelectric
power generation and/or reducing water usage in cooling applications is inevitable and urgent.
Biofuel production for transportation also has significant water requirements, especially at the
cultivation stage.  The assessment and adaptation studies described here investigate and integrate
the current knowledge base of water usage in energy production industries. This report
documents the research results on the generation of electricity and the emerging production of
biofuels by assessing major trends in thermoelectric power generation and biofuels and
investigating future water availability and water allocation  for these energy production and
energy transformation processes. Its primary focus is on coal-fired and natural gas-fired electric
power plants, the  production of corn-starch-based ethanol and cellulosic biofuels, the production
of biodiesel, the impacts of all of these processes on water resources, as well as technologies
available to adapt these processes to reduce their impact, particularly in water-stressed regions of
the United States. The report includes detailed analyses and water resource adaptation strategies
for sustainable energy production, including a case study focusing on the water-stressed
southwestern U.S. using Las Vegas, Nevada as a specific example.
                                            11

-------
                                     PREFACE

       Water is essential to life. Uneven global distribution of population and water resources
has resulted in more than 1.1 billion people world-wide lacking access to clean drinking water
and 2.6 billion people living in regions with inadequate freshwater treatment. Today freshwater
is being consumed at an alarming rate and is almost doubling every 20 years. Global climate
change exacerbates this already stressed situation. Water availability is not only a problem for
developing countries, but also one facing developed nations now saddled with an aging water
infrastructure. Throughout history, civilizations have found innovative solutions to meet their
water resource needs and responded to evolving social and environmental conditions. This spirit
of adaptation continues to this date.

       Today, one of the most complex challenges facing our nation revolves around the water-
energy nexus. The linkages between water use, climate change and the production,
transformation and use of energy require an interdisciplinary, holistic approach to future water
management, both in quality and quantity. The energy sector is one of our nation's largest water
users of water. Large quantities of water are withdrawn and consumed every year to provide
electricity and liquid fuels and these amounts are expected to continue to grow unless steps are
taken to establish more efficient and renewable methods of generation and production.
Environmental conditions in the U.S. are becoming increasingly more important in making
decisions about the location, size, and type of energy production necessary to supply the vast
amounts of energy required to power our economy. Furthermore, the energy sector is in
transition. The U.S. is transitioning towards less carbon-intensive electric power generation in
order to reduce CO2 emissions and their contribution to climate change. Biofuels are increasingly
being used in the U.S. in response to policies to reduce petroleum imports. All of these factors
also have important consequences with respect to water withdrawal and consumption, especially
in  water stressed regions.

       This report presents a preliminary assessment of water used in two segments of energy
production: electric power generation and  the production of biofuels for transportation. It is
structured to address science and engineering questions pertinent to adaptation and to support
technical managers and other stakeholders facing the enormous complexity of climate
adaptation. This objective is accomplished by structuring individual chapters around stand-alone,
but interrelated, science and engineering subjects. After discussing the "big" picture of
adaptation needs, the report provides in-depth analyses of water use and adaptation strategies and
emerging technologies. As an initial first step, it is hoped that this effort marks a beginning of the
long march toward the goal of sustainable infrastructure adaptation to changing climate and
socioeconomic conditions.
Mr. Joseph McDonald, MSME                  Dr. Y. Jeffrey Yang, P.E.
National Risk Management Research Laboratory   National Risk Management Research Laboratory
U.S. EPA Office of Research and Development    U.S. EPA Office of Research and Development
Cincinnati, Ohio 45268                        Cincinnati, Ohio 45268
                                           in

-------
                           ACKNOWLEDGMENTS
This work was funded and managed by the U.S. Environmental Protection Agency - Office of
Research and Development as part of the Air, Climate and Energy (ACE) Research Program,
Project MA-4 "Sustainability, Interactions, and Co-benefits", Task 203 "Regional Adaptation
Case Studies for Sustainable Water Resources". Research in support of ACE MA-4/203 was
primarily through EPA Contract Number EP-C-11-006 with Pegasus Technical Services (PTS)
with additional research support provided to PTS by the University of Cincinnati and
Washington University in St. Louis. The authors acknowledge Dr. Michael Moeykens, Stephen
Wright, Sandra Savage and Candice Charlton of the U.S. EPA; and Dr. Karen Koran and
Thomas Davis of Pegasus Technical Services, Inc. for contract support of this research. The
authors acknowledge Melanie Prarat, Hodon Ryu, Robert Grosser, Colin White, and
Raghuraman Venkatapathy of Pegasus Technical Services, Inc. for their editorial assistance. The
authors also express their thanks to Dr. Thomas Speth, Director of U.S. EPA Water Supply and
Water Resources Division (WSWRD), Dr. Michelle Simon,  Chief of U.S. EPA WSWRD's
Urban Water Management Branch, and to the EPA Office of Water (EPA/OW) for their ongoing
support of research related to the water-energy nexus.

      Principle Investigators and Lead Authors
       Mr. Joseph McDonald, MSME - ORD/NRMRL
       Dr. Y. Jeffrey Yang, P.E., D.WRE - ORD/NRMRL
      Principle Authors and Contributors
      Dr. Timothy Keener, University of Cincinnati
      Dr. Mingming Lu, University of Cincinnati
      Dr. Pratim Biswas, Washington University in St. Louis
      Dr. Wei-Ning Wang, Washington University in St. Louis
      Dr. Marissa S. Liang, University of Cincinnati
      Dr. Joo-Youp Lee, University of Cincinnati
      Ms. Patcha Huntra, University of Cincinnati
      Ms. Bala Lingaraju, University of Cincinnati
      Ms. Sowmya Karunakaran, University of Cincinnati
      Mr. Vishnu Sriram, University  of Cincinnati
      Dr. Ying Li, University of Wisconsin-Milwaukee
      Mr. Qingshi Tu, University of Cincinnati
      Mr. Colin White, Pegasus Technical Services, Inc.
      Ms. Darsana Menon, Washington University in St. Louis
      Dr. William Flatten, Pegasus Technical Services, Inc.
      Mr. Patricio Pinto, Pegasus Technical Services, Inc.
      Mr. Christopher Holder, Pegasus Technical Services, Inc.
                                          IV

-------
Peer Reviewers
Mr. Paul Machiele, U.S. EPA/OAR/OTAQ/ASD/FC
Mr. Paul Argyropoulos, U.S. EPA/OAR/OTAQ
Dr. William Linak, U.S. EPA/ORD/NRMRL/APPCD
Dr. Chi Ho Sham, The Cadmus Group, Inc.
Dr. Jonathan J. Koplos, The Cadmus Group, Inc.
Dr. Richard A. Krop, The Cadmus Group, Inc.
Dr. Frederick Bloetscher, Florida Atlantic University

-------
                           TABLE OF CONTENTS
Disclaimer	i
Abstract	ii
Preface	iii
Acknowledgments	iv
List of Tables	xi
List of Figures	xiv
Abbreviation and Acronyms	xvii
EXECUTIVE SUMMARY	1
1   Introduction	6
    1.1  References	9
2   Trends in Energy Production	10
    2.1  Introduction	10
    2.2  Current Energy Production Methods	10
        2.2.1   Electrical energy demand and production	10
        2.2.2   Transportation fuels	11
    2.3  New Energy Technologies and Future Outlook	14
        2.3.1   Trends in advanced coal technologies for power production	14
        2.3.2   The emerging role of natural gas in power generation	15
        2.3.3   Other alternative energy developments	15
        2.3.4   Regulatory impacts on future electric power generation	16
        2.3.5   Future biofuel usage	17
    2.4  References	21
3   Climate Change Impacts on Water Availability for Energy Production	24
    3.1  Introduction	24
    3.2  Climate change projections and water availability outlook	25
        3.2.1   Climate systems controlling U.S. precipitation and water availability	25
        3.2.2   Observed climate changes and their impacts on water availability	30
        3.2.3   Changes in river flows and river basin hydrology	34
    3.3  Implications With Respect to Energy Production	39
        3.3.1   Climate considerations for thermal electric power generation	39
        3.3.2   Carrying capacity for thermal pollution  and nutrient loading	42
                                          VI

-------
    3.3.3  Implications for energy biomass production	43
    3.3.4  Non-point source nutrients in biomass production	43
3.4  Adaptation Potential and Conclusion	44
3.5  References	44
Water Impacts from Coal-Fired Electric Power Plants	50
4.1  Introduction	50
4.2  Water Impacts from Coal Mining and Processing	50
4.3  Water Impacts from Thermoelectric Power Systems	51
    4.3.1  Steam turbines	51
    4.3.2  Cooling tower operations	52
    4.3.3  Bottom ash	53
    4.3.4  Flue gas desulfurization	54
    4.3.5  Particulate matter	54
4.4  Water Reuse and Conservation Potential	55
    4.4.1  Advanced cooling technologies	55
    4.4.2  Power plant flue gas water capture	56
    4.4.3  Wet electrostatic precipitators	57
4.5  Trends in water withdrawal and water consumption in coal-fired power plants in the
     U.S	57
4.6  Power Plant Carbon Models to Predict Water Withdrawal Rate	60
    4.6.1  Description of models	60
    4.6.2  Flash™ Module Calculation Comparison to EIA Data	62
4.7  Summary	65
4.8  References	66
4.9  Appendix	69
    4.9.1  Flue gas desulfurization technology	69
    4.9.2  Flue gas water capture technologies	79
    4.9.3  Design optimization of air-cooled condensing wet ESP for flue gas water
           recovery	82
    4.9.4  Appendix references:	108
Natural Gas Electric Power  Generation and Water Usage	110
5.1  Introduction	110
5.2  Gas-fired Boiler Thermoelectric Plants	110
                                       vn

-------
    5.3  Gas Turbine and Boiler Cogeneration	Ill
    5.4  Water Usage	113
         5.4.1  Water usage for fuel extraction	113
         5.4.2  Water usage for electricity generation	115
    5.5  Summary	116
    5.6  References	117
6   Corn-Starch-Based Ethanol Production and Impacts on Water Resources	119
    6.1  Introduction	119
    6.2  Energy and Mass Balance Model	120
         6.2.1  Corn-to-ethanol production process overview	120
         6.2.2  Mass balance model	121
         6.2.3  Energy balance model	125
         6.2.4  Flash-based interactive model	126
    6.3  Pilot Scale Results	128
         6.3.1  Background	128
         6.3.2  Production process overview	128
         6.3.3  Wastewater sampling and analysis	128
    6.4  Full-Scale Plant Review	131
         6.4.1  Overview	131
         6.4.2  Full-scale plant description and operation	131
         6.4.3  Assessment	133
    6.5  Summary and Future Work	134
         6.5.1  Summary	134
         6.5.2  Future work	136
    6.6  References	136
7   Water Usage within Lignocellulosic Biomass and Cellulosic Biofuels Production	139
    7.1  Introduction	139
    7.2  Cellulosic Ethanol Process	141
    7.3  Biomass Harvesting and Biofuel Conversion	142
    7.4  Water Usage and Wastewater Generation	146
         7.4.1  Water quantity	148
         7.4.2  Water quality	157
    7.5  Technological Challenges and Opportunities for Water Reuse and Conservation .... 159

                                          viii

-------
    7.6  Summary	160
    7.7  References	161
8   Biodiesel Production and Impacts on Water Resources	167
    8.1  Introduction	167
    8.2  Water Consumption Estimates for the Soybean Oil FAME Biodiesel Process in
         theU.S	168
        8.2.1   Methodology	168
        8.2.2   Data sources	168
        8.2.3   Results	170
        8.2.4   Discussion	176
    8.3  New Trends in Biodiesel Research and Development	179
        8.3.1   Feedstock development	179
        8.3.2   Biodiesel and bio-based diesel fuel technology development	187
    8.4  Summary	189
    8.5  References	190
    8.6  Appendix	197
        8.6.1   Sample calculation for irrigation water consumption and normalized water
               intensity for Ohio	197
        8.6.2   Calculation of normalized water consumption and sample calculation for water
               consumption in the soybean crushing and processing stage for Ohio	199
        8.6.3   Sample calculation for normalized and total water consumption for biodiesel
               manufacturing in Ohio	200
        8.6.4   Water-saving technologies developed in the algae industry	200
        8.6.5   Appendix 8 References	203
9   Impacts of Electric Generation on Water Resources: Regional Assessment and
    Adaptation	205
    9.1  Introduction	205
    9.2  Regional Integrated Water-Energy Resource Management	205
        9.2.1   Case study:  Las Vegas	210
    9.3  Holistic Water Resource Adaptation	214
        9.3.1   Future energy production scenarios	214
        9.3.2   Water Demand Distribution and Water availability	218
        9.3.3   Adaptation of electricity generation water use	223
                                           IX

-------
    9.4   Summary of U.S. GAO Report on Climate Change Energy Infrastructure Risks and
         Adaptation	227
    9.5   Summary	230
    9.6   References	232
    9.7   Appendix	237
10  Conclusions and Recommendations	239
    10.1  Conclusions	239
    10.2  Recommendations and Future Work	248
    10.3  References	250

-------
                               LIST OF TABLES
Table ESI  Water-intensity on a withdrawal and consumptive basis for thermoelectric
           generation using different sources of energy and using recirculating cooling
           systems	2
Table ES2  Water intensities on a consumptive basis for producing different types of
           biofuels	4
Table 2.1   Fuels production water intensity	14
Table 2.2   Proposed volume for 2014 RFS	20
Table 4.1   Water-saving approaches in coal power plants outside the U.S	57
Table 4.2   Flash model validation for different types of coal seams	64
  Appendix Table 4.1  Summary of pollutants included in FGD wastewater	70
  Appendix Table 4.2  Summary of the reviewed wastewater treatments	75
  Appendix Table 4.3  Suggestions for selecting an appropriate CMD treatment technology .. 76
  Appendix Table 4.4  Advantages and disadvantages of treatment technologies that can be
                     used to remove TDS from CMD	77
  Appendix Table 4.5  Estimated fractions of cooling tower makeup water achievable,
                     assuming 100% water vapor capture	82
           Table 4.6  ESP channels and fin characteristics	87
           Table 4.7  Properties of coal and flue gas assumed	93
           Table 4.8  Cases of fin configuration evaluated varying spacing and thickness .... 94
           Table 4.9  Cases of fin depth evaluated	95
           Table 4.10 Wet ESP dimensions variation cases	96
Appendix
Appendix
Appendix
Appendix
Appendix
Appendix
  Appendix

  Appendix

  Appendix

Table 5.1
Table 5.2
Table 5.3
Table 4.11  System parameters and water recovery obtained from changing fin
           spacing and thickness	97
Table 4.12  Proposed cases for ESP dimensions and design parameters
           calculated	100
Table 4.13  Results for Case A (length = 50 feet, height = 40 feet,
           no. ESPs = 60)	101
Table 4.14  Results for Case B (length = 60 feet, height = 40 feet,
           no. ESPs = 60)	102
 System types by primary energy source in 2012	Ill
 Water requirements for natural gas power plant (gal MW"1 h"1)	115
 Annual average withdrawal rate  of cooling system for natural gas-fire steam
 turbine thermoelectric power plants	116
                                          XI

-------
Table 6.1    Corn composition	123
Table 6.2    Total energy consumption for corn-to-ethanol process	125
Table 6.3    Energy flow in corn-to-ethanol process	126
Table 6.4    Summary of operating parameters from a Midwest, full-scale ethanol production
            facility	132
Table 7.1    Literature review of water requirements for ethanol production	141
Table 7.2a   Literature review of moisture content during the biomass harvesting process	147
Table 7.2b   Literature review of water consumption for different feedstocks	147
Table 7.3    Four cases analyzed for water quality and quantity requirements	148
Table 7.4    Overall water balance	150
Table 7.5    Water balance in individual unit operations	155
Table 7.6    Water quality used for four cases	158
Table 8.1    Total annual water consumption (Wtotj) in top 10 soybean harvesting states	173
Table 8.2    Total annual water consumption (Wtotj) in top 10 biodiesel producing states	173
Table 8.3    Regional water consumption data	176
Table 8.4    Total annual water consumption (Wtotj) for soybean biodiesel production in the
            states in the water-stressed areas	177
Table 8.5    Literature summary of water consumption during different algal biodiesel
            production stages	183
  Appendix Table 8.1   Biodiesel plant capacity in each state	198
  Appendix Table 8.2  Water consumption data in biodiesel wash collected from biodiesel
                      manufacturers	201
  Appendix Table 8.3   Water consumption in cooling tower make-up	201
  Appendix Table 8.4  Summary of water-stressed states from literature	202
  Appendix Table 8.5   Key assumptions and parameters of the studies	202
Table 9.1    Range of fuel cycle water consumption and withdrawal estimates for selected
            generation technologies and sub-categories	207
Table 9.2    Range of power plant equipment water consumption and withdrawal estimates
            for selected generation technologies and sub-categories	208
Table 9.3    Power plant operations cycle water consumption and withdrawal estimates for
            selected generation technologies and sub-categories	208
Table 9.4    National electricity water crisis areas	209
Table 9.5    Electricity Modeling Scenarios	215
Table 9.6    The top 5 states with the highest thermoelectric-power water withdrawals for
            once-through cooling type in 2005	219
                                           xn

-------
Table 9.7    The top 5 states with the highest thermoelectric-power water withdrawals for
            recirculation cooling type in 2005	220
Table 9.8    Cooling-system types used to classify plant by cooling system technology	226
Table 9.9    Summaries of Selected Federal Roles in Energy Infrastructure for Water
            Management	229
  Appendix Table 9.1   Detailed summary of Southern Nevada Water Authority's incentive
                      programs	237
  Appendix Table 9.2   Detailed summary of Southern Nevada Water Authority's education
                      programs	238
Table 10.1   Water-intensity on a withdrawal and consumptive basis for thermoelectric
            generation using different sources of energy and using recirculating cooling
            systems	243
Table 10.2   Water-intensity on a withdrawal and consumptive basis for thermoelectric
            generation using different sources of energy and using once-through cooling
            systems	244
Table 10.3   Water intensities on a consumptive basis for producing different types of
            biofuels	246
                                          Xlll

-------
                               LIST OF FIGURES

Figure 2.1  Electricity generation and energy source contributions	11
Figure 2.2  Comparison of the use of gasoline and diesel transportation fuels to domestic
           production of biofuels	13
Figure 2.3  U.S. production of petroleum and other liquids by source, 2012-2040 (million
           barrels per day)	18
Figure 2.4  EISA 2007 RFS credits earned in selected years, 2012-2040	19
Figure 2.5  Projections of U.S. domestic gasoline and diesel fuel consumption through 2040 .. 19
Figure 2.6  Gasoline consumption and percentage ethanol share of gasoline	20
Figure 2.7  Consumption of biofuels in motor gasoline blends in the Reference case, 2012-40
           (million barrels per day)	21
Figure 3.1  Schematic illustration of major climatic systems interacting with local climatic
           factors producing precipitation characteristics in each of the hydroclimatic
           provinces	26
Figure 3.2  Multi-model dataset (MMD) bias compared to observed precipitation (Xie and
           Arkin, 1997; U.S. EPA, 2014a,b) in control runs (1980-1999) for three North
           America regions	27
Figure 3.3  Hydroclimatic provinces and extreme precipitation changes in the contiguous
           U.S	31
Figure 3.4  Spatial distributions of long-term precipitation changes and population change
           in the contiguous U.S	33
Figure 3.5  Temporal variations of (A) water volume in Lake Mead; (B) annual precipitation
           in the LVR basin	37
Figure 3.6  Locations of major coal-fired thermoelectric power plants superimposed upon a
           map of the hydroclimatic provinces of the U.S	40
Figure 3.7  Annual flow variation of Colorado River recorded at USGS gage station
           942400 at  Topock, Arizona	41
Figure 4.1  Annual electricity generation in coal plants (>100 MW) in the U.S	58
Figure 4.2  Percentage of electricity generated per cooling system type in 2010	59
Figure 4.3  Schematic diagram of a coal fired power plant with/without CO2
           capture/sequestration	61
Figure 4.4  Flash™ model of different storage method	62
Figure 4.5  IGCC Flash™ Model	63
   Appendix Figure 4.1  Schematic  of a two-stage TMC for power plant flue gas heat and
                       water recovery	80
                                          xiv

-------
   Appendix Figure 4.2  Condensing wet ESP system within the power plant and wet ESP unit
                       with external air-cooling	84
   Appendix Figure 4.3  System's design - a cluster of five ESP per FGD Unit	85
   Appendix Figure 4.4  Condensing wet ESP initial fin configuration	86
   Appendix Figure 4.5  FGD water recovery vs. fin-side fan power	86
   Appendix Figure 4.6  Effects of outside air temperature on water recovery	87
   Appendix Figure 4.7  Sections in the wESP channel	92
   Appendix Figure 4.8  Condensing wESP original fin configuration	94
   Appendix Figure 4.9  FGD water recovery % vs. fin-side fan power for the highlighted
                       cases	98
   Appendix Figure 4.10 Effect of fin-side fan power on FGD water recovery % for case
                       7w-2	98
   Appendix Figure 4.11 Effect of outside air temperature on FGD water recovery % for case
                       7w-2	99
   Appendix Figure 4.12 Water recovery vs. fin-side fan power for Case A9	103
   Appendix Figure 4.13 Water recovery percentage vs. cooling air temperature for
                       CaseA9	104
   Appendix Figure 4.14 Water recovery in gallons per min vs. cooling air temperature for
                       CaseA9	104
   Appendix Figure 4.15 Water recovery vs. fin-side fan power for Case C9	106
   Appendix Figure 4.16 Water recovery percentage vs. cooling air temperature for Case
                       C9	107
   Appendix Figure 4.17 Water recovery vs. cooling air temperature for Case C9	107
Figure 5.1  Schematic drawing of the gas turbine combined cycle power plant	112
Figure 5.2  Diagram of hydraulic fracturing of natural gas	114
Figure 6.1  System diagram of typical dry-mill corn-to-ethanol production process	121
Figure 6.2  Inputs and outputs at the system boundary	121
Figure 6.3  Mass block flow diagram of ethanol production process	122
Figure 6.4  Microsoft Excel™ spreadsheet worksheet for  the mass balance model	124
Figure 6.5  Energy block flow diagram of ethanol production process	126
Figure 6.6  Flash-based interactive model on energy and mass balance - calculation based
           on (a) corn feed and (b) ethanol plant capacity	127
Figure 6.7  NCERC process flow diagram	129
Figure 6.8  Process water tank flow scheme	130
Figure 6.9  Wastewater sampling locations	130
                                          xv

-------
Figure 6.10 Heat capture through heat exchangers greatly improves energy efficiency in the
           Illinois ethanol plant	133
Figure 7.1  Process schematic for cellulosic ethanol production	143
Figure 7.2  Unit operations involved in a biomass supply chain from field to bio refinery	145
Figure 7.3  Equipment used in the biomass harvesting process	145
Figure 7.4  Overall water balance of cellulosic ethanol plant	149
Figure 7.5  Process layout based on case 1 (hardwood + dilute acid pretreatment)	151
Figure 7.6  Process layout based on case 2 (corn stover + dilute acid pretreatment)	152
Figure 7.7  Process layout based on case 3 (switchgrass + dilute acid pretreatment)	153
Figure 7.8  Process layout based on case 4 (switchgrass + afex pretreatment)	154
Figure 8.1  Annual biodiesel production in U.S. (2003-2013)	167
Figure 8.2  Total annual water consumption of the soybean-to-biodiesel process (Wtotj)	174
Figure 8.3  Total water consumption of the soybean-to-biodiesel process on per gallon
           biodiesel basis (Ntotj)	175
   Appendix Figure 8.1   Irrigation water use at state level for soybean growth (Wl, million
                        gallons per year). Note that 35 out of 50 states have data	203
   Appendix Figure 8.2  The irrigation water intensity for soybean growth by state (Nl,
                        gallon water per gallon biodiesel)	203
Figure 9.1  Typical rate of water consumption for electricity generation (consumptive use)... 206
Figure 9.2  Lake Mead elevation from 1950 to 2012	210
Figure 9.3  Electricity Generation by Source Type for Nevada as of February 2014	211
Figure 9 4  Electricity generation by fuel, 1990-2040 (trillion kilowatt-hours)	214
Figure 9.5  National electricity generation by scenario	216
Figure 9.6  Electricity generation in the southwest by scenario	217
Figure 9.7  Electricity generation in the southeast by scenario	218
Figure 9.8  Thermoelectric-power water withdrawals 2005	219
Figure 9.9  Overview of water withdrawal factors by technology	221
Figure 9.10 Overview of water consumption factors by technology	221
Figure 9.11 ReEDS modeling regions including the Interconnect, Regional Transmission
           Organization (RTO), Power Control Authorities (PCA), and Wind/Concentrating
           Solar Power (CSP) Region	223
Figure 9.12 Operational water consumptive factors for electricity generating technologies	225
                                           xvi

-------
ABBREVIATION AND ACRONYMS
7Q10
ABMet
ABRWTF
ACC
AEO
AFEX
Al
A12O3
A-O
AMO
ANL
AO
AOGCM
AR
As
B5
B20
BIT
BFD
BOD
BTU
°C
Ca
CA
CA-MC
CAA
CaCO3
CAFE
CAN
CaO
Ca(OH)2
CapWa
CaSO3
CaSO4*2H2O
CCF
CCGT
CCS
CCS
Cd
CMD
CMIP5
CMW
Co
CO
Seven-day average of 10 year return flows
Advanced biological metals (removal process)
American bottoms regional wastewater treatment facility
Air-cooled condenser
Annual energy outlook
Ammonia fiber expansion
Aluminum
Aluminum oxide
Atmospheric-ocean
Atlantic multidecadal oscillation
Argonne National Laboratory
Arctic oscillation
Atmospheric-ocean general circulation model
Arkansas
Arsenic
5% biodiesel blended with 95% petroleum diesel
20% biodiesel blended with 80% petroleum diesel
Bituminous and anthracite coal
Block flow diagram
Biological oxygen demand
British thermal unit
Degrees Celsius
Calcium
California
Cellular automata Markov chain
Clean Air Act
Calcium carbonate
Corporate average fuel economy
Central North America
Lime (calcium oxide)
Slaked lime (calcium hydroxide)
Water capture
Calcium sulfite
Gypsum
Recirculation cooling using freshwater
Combined cycle gas turbine
Carbon capture and storage
Recirculation cooling using saline water
Cadmium
Coal mine drainage
Coupled model intercomparison project phase 5
Concentrated municipal wastewater
Cobalt
Colorado or carbon monoxide
                                 xvn

-------
CO2
CO2+H2
COD
Cr
CSP
CT
Cu
CWA
CWTS
DDGS
DJF
DOE
DS
E10
El 5
E85
EGS
EIA
EISA
ENA
ENSO
EPRI
EPS
ESP
ESRI
°F
F-T
FAME
Fe
FERC
FFA
FGD
FL
Fsoy
Fuse
GA
g
gal
gal/gal
GAO
GCM
GE
GHG
GJ
GTU
Carbondioxide
Syn gas
Chemical oxygen demand
Chromium
Concentrated solar power
Connecticut
Copper
Clean Water Act
Constructed wetlands treatment system
Dry distiller grains with solubles
Winter months (December, January, February)
Department of Energy
Dissolved solids
10% ethanol blended with 90% gasoline
1 5% ethanol blended with 85% gasoline
85% ethanol blended with 15% gasoline
Enhanced geothermal systems
Energy Information Administration
Energy Independence and Security Act
Eastern North America
El-Nino southern oscillation
Electric Power Research Institute
Extracellular polymeric substances
Electrostatic precipitator
Environmental Systems Research Institute
Degrees Fahrenheit
Percentage conversion from soy oil to biodiesel
Fisher-Troche (process)
Fatty acid methyl ester
Iron
Federal Energy Regulatory Commission
Free fatty acid
Flue gas desulfurization
Florida
Percentage oil content of soybeans
Percentage of soy oil produced used for biodiesel production
Georgia
Gram
Gallons
Gallons of water per gallon of (biofuel)
Government Accountability Office
General circulation model
General Electric
Greenhouse gas
Giga joules
Gas Technology Institute
                   xvin

-------
GW           Giga-watt
GWe          Giga-watt electrical
H             Hydrogen
H2O           Water
H2SO4         Sulfuric acid
HERO         High efficiency reverse osmosis
Hg            Mercury
hr             Hour
HRSG         Heat recovery steam generator
IA             Iowa
ICS            Intentional created surplus
IGCC          Integrated gasification combined cycle
IPCC          Intergovernmental Panel on Climate Change
ITS            Ice thermal storage
JJA            Summer months (June, July, August)
K             Potassium
Kg            Kilogram
KJ             Kilojoules
KOH          Potassium hydroxide
KS            Kansas
kWh           Kilowatt-hours
L              Liter
LA            Louisiana
Ib/hr           Pounds per hour
LCA           Life cycle assessment
LHV          Lower Heating value
LIG           Lignite coal
LMR          Lower Mississippi River or Little Miami River
LMRB         Lower Mississippi River Basin
LONE         Lower Mississippi - Ohio River Valley - New England Region
LVR           Lower Virgin River
LVRB         Lower Virgin River Basin
m2             Square meter
m3/sec         Cubic meters per second
m3/hr          Cubic meters per hour
MA           Massachusetts
MAM         Spring months (March, April, May)
MASS2        Modular aquatic simulation system 2-dimensional
MCE          Multi-criterion evaluation
MCL          Maximum contaminant level
MD           Maryland
mg/L          Milligrams per liter
|j,g/L           Microgram per liter
MgO          Magnesium oxide
MGPY         Million gallons per year
Mgy           Million gallons per year
                                  xix

-------
MIW          Mining impacted water
MJ            Megajoules
MMBtu/hr      One million BTUs
MMD          Multi-modal dataset
MMgpy        Million gallons per year
Mn            Manganese
MW           Megawatts
MWe          Megawatt electrical
MWh          Megawatt-hours
N2             Nitrogen
Nij            Normalized irrigation water used in soybean growth stage for each state
               (MGPY)
-A/2/            Normalized water use during soybean oil processing stage for each state
               (MGPY)
NSJ            Normalized water use in the biodiesel production stage for each
               individual stage (MGPY)
Ntot            Normalized total water consumption for all stages for each individual
               state (MGPY)
Na            Sodium
NAO          North Atlantic Oscillation
NaOCHa       Sodium methoxide
NaOH          Sodium hydroxide
NAP           National Academies Press
NBAA         National Business Aviation Association
NBB           National Biodiesel Board
NCERC        National Corn to Ethanol Research Center
NDPES        National Pollutant Discharge Elimination System
NE            Nebraska
NEMS         National Energy Modeling System
NERC          North American Electricity Reliability Council
NETL          National Energy Technology Laboratory
NGCC         Natural gas combined cycle
NHa-N         Ammonia as nitrogen
NJ             New Jersey
NNE-SSW      North-Northeast - South-Southwest
NO            Nitrous oxide
NO2-N         Nitrate as nitrogen
NOs-N         Nitrite as nitrogen
NOAA         National Oceanic and Atmospheric Administration
NOPA         National Oilseed Processors Association
NOx           Generic term for all nitrous oxides
NRC           National Research Council
NREL          National Renewable Energy Laboratory
NV            Nevada
NY            New York
O2             Oxygen
                                  xx

-------
O&M          Operation and maintenance
OC/OTC       Once-through, with cooling ponds or canals
OF/OTF        Once-through, freshwater
OTS           Once-through, saline water
OH            Ohio
OK            Oklahoma
Ortho-P        Orthophosphate
OS            Once-through, saline water
Pb             Lead
PER           Photobioreactor
PCA           Power controlling authorities
PCMDI        Program for Climate Model Diagnosis and Intercomparison
PDO           Pacific Decadal Oscillation
PNNL          Pacific Northwest National Laboratory
POTW         Publicly owned treatment works
ppb            Parts per billion
PRB           Powder River Basin
PSD           Prevention of significant deterioration
PV            Photovoltaic
R              Normalized precipitation rate of change (cm/month)
R&D           Research and development
RC            Recirculating, with cooling ponds or canals
RCM           Regional climate model
ReEDS         Regional Energy Deployment System
RF            Recirculating, with forced draft cooling towers
RFS           Renewable Fuel Standards
RFS2           Revised Renewable Fuel Standards
RI             Recirculating, with induced draft cooling towers
RIN           Renewable Identification Number
RN            Recirculating, with natural draft cooling towers
RTO           Regional Transmission Organization
SDWA         Safe Drinking Water Act
SiO2           Silicon dioxide
SIUE           Southern Illinois University, Edwardsville
SNWA         Southwest Nevada Water Authority
SO2           Sulfur dioxide
SO3           Sulfite
SO4           Sulfate
SON           Autumn months (September,  October, November)
SOx           Generic term for sulfur oxides
SrO           Strontium oxide
SS             Suspended solids
SSCF           Simultaneous saccharification and co-fermentation
SST           Surface Sea Temperature
SUB           Sub-bituminous coal
TDS           Total dissolved  solids
                                  xxi

-------
TiO2
TKN-N
TMC
TMDL
TOC
TSS
TX
UIC
UNESCO
U.S.
USB
USDA
U.S. EPA
USCCSP
USGS
USHCN
VA
Vr
Wtot
WA
WC
WCG
wESP
WMO
WNA
WY
Zn
Titanium dioxide
Total Kjeldahl nitrogen as nitrogen
Transport membrane condenser
Total Maximum Daily Load
Total organic carbon
Total suspended solids
Texas
Underground injection control
United Nations Educational, Scientific and Cultural Organization
United States
United Soybean Board
U.S. Department of Agriculture
U.S. Environmental Protection Agency
U.S. Climate Change Science Program
United States Geological Survey
U.S. Historical Climatological Network
Virginia
Total volume of irrigation water for soybean cultivation in a state or
region (acre-feet)
Irrigation water used in soybean growth stage for each individual state
(MGPY)
Water use during soybean oil processing stage for each individual state
(MGPY)
Water use in the biodiesel production stage for each individual stage
(MGPY)
Total water consumption for all stages for each individual state (MGPY)
Washington
Waste coal
Waste coffee grounds
Wet electrostatic precipitators
World Meteorological  Organization
Western North America
Wyoming
Zinc
                    xxn

-------
EXECUTIVE SUMMARY

       Electricity and transportation are currently the two largest sources of energy demand in
the United States (U.S.). According to the U.S. Energy Information Administration's Annual
Energy Outlook 2014, coal was used to generate 37% of the total electricity, followed by natural
gas, nuclear power and renewable energy at 30%,  19%, and 12%, respectively, in 2012. Each
step in the production of fuel for energy, transportation, and electricity generation requires large
amounts of water that is either withdrawn or consumed. Water consumption refers to the loss of
water from a water source due to water evaporation, water use in processes or return to a
different source. Water withdrawal refers to water that is taken from a particular water source
without regard to the amount that is returned.
       Thermoelectric generation is responsible for a significant portion of total water
withdrawals in the U.S. According to the most recent data from the U.S. Geological  Survey,
thermoelectric power plants are responsible for approximately 40% of total freshwater
withdrawals, approximately 50%  of total water use, and approximately 4% of total freshwater
consumption. This will stress water availability particularly in drought-prone regions of the U.S.
such as in the Southwest. Biofuels from agriculture account for approximately 7% of transport
fuel consumption. Agricultural irrigation for biofuel production accounts for approximately 4%
of total freshwater consumption. Climate change models summarized within recent reports by
the United Nations Intergovernmental Panel  on Climate Change project 10% or more decreases
in precipitation in southwest U.S., and changes in precipitation from snow to rain, resulting in a
decrease in snow pack, and hence, seasonal stream flows. Simulations also project decreased
precipitation and increased drought in Southeast U.S., and increased precipitation and flooding
along the Great Lakes and Northeast U.S.
       The availability of water for power generation and biomass production for biofuels will
be affected by extraneous factors  such as climate change, population growth and redistribution,
domestic consumption, and land use. Furthermore, water availability also limits the potential of
production water discharge from thermoelectric power plants into streams and rivers to prevent
thermal pollution of the receiving water body. Climate change has been shown to cause
precipitation variations in its intensity, frequency,  seasonality, and amounts, leading to variations
in surface water flow and groundwater levels, which in turn can affect energy production
processes. Due to the uncertainty in climate change and the increasing demand for water for
electricity and biofuels, there will be substantial competition for water in many water-stressed
regions between energy, commercial, industrial  and residential sectors.
       Emission of greenhouse gases such as CO2 through the combustion of coal and other
fossil fuels is a leading contributor to global  warming.  Coal combustion is also a significant
contributor to air emissions of NOx, SOx, fine particulate matter, mercury and other hazardous
air pollutants. For coal to continue to drive electric power generation and economic expansion
across the globe in an environmentally neutral manner, technology must continue to be
developed in order to reduce coal plant emissions to near-zero levels. Natural-gas fired plants
offer an alternative to using coal as a fuel source for thermoelectric generation while also
providing advantages in terms of fuel costs, lower greenhouse gas emissions, potentially higher
efficiency, and potentially lower volumes of water withdrawals. There is also  greater
accessibility and abundance of natural gas in the U.S. due to the use of hydraulic fracturing.  The

-------
use of natural gas in electric power generation is expected to increase from 30% to
approximately 63% of total electricity production in the U.S. by 2040.
       Currently, a majority of the electricity in the U.S. is generated by combusting fossil fuels
in a boiler to produce steam and using the kinetic energy of the steam to generate electricity with
steam turbines. In addition to the water required to produce steam, other uses of water in power
plants include the water required to condense steam after it passes through the steam turbine,
water lost due to evaporation in cooling towers, water required for scrubbing flue gases to meet
Clean Air Act (CAA) regulations,  and water required to dispose fly ash, among others.
Considering water usage is critically important in areas with  significant levels of irrigation and
thermoelectric power generation (especially the amount of water required to condense steam)
and is especially important in areas effected by climate change. Various direct and indirect steps
can be taken to minimize water use in thermoelectric power plants, including using advanced
cooling technologies such as air or hybrid cooling, using supercritical or ultra-supercritical steam
turbines, use of advanced power generation units such as a natural gas combined cycle (NGCC)
turbine with heat recovery  steam generators (HRSG), use of non-traditional sources for process
and cooling waters, and use of advanced wastewater treatment technologies to treat process
water and wastewater from flue gas desulfurization (FGD) units that can  allow reuse of effluent
in the energy  generation process.
       A summary of the amount  of water withdrawn and the amount of water consumed during
electricity generation using coal (with and without integrated gasification combined cycle
(IGCC) and CO2 capture) and natural gas (Rankine and NGCC), nuclear  and concentrated solar
(thermal), all  with recirculating steam cooling systems are compared in Table ESI. Table ESI
also contains  references to specific chapters of this report where additional details can be found
regarding water use for thermoelectric generation. An interactive Adobe Flash™ based module
has also been developed and validated to relate power plant emissions to power plant ratings and
coal characteristics for current  and future coal-fired electric generation technologies as well as
provide estimates of water usage and price of electricity production.

     Table ESI  Water-intensity on a withdrawal  and consumptive basis  for thermoelectric
                generation using different sources of energy and using recirculating cooling
                systems
System
Coal
IGCC
Natural Gas
(Rankine/steam
turbine only)
NGCC
Nuclear
Concentrated Solar
Thermoelectric
Water withdrawals
gal/MWh*
500-1200
161-605
950-1460
150-283
793-2589
740-1110
10-4m3/MJ**
5.3-12.6
1.7-6.4
10.0-15.4
1.6-3.0
8.3-27.2
7.8-11.7
Water consumption
gal/MWh
201-1189
34-449
662-1170
130-300
581-898
555-1902
10-4m3/MJ
5.0-11.6
0.4-4.7
7.0-12.3
1.4-3.2
6.1-9.4
5.8-20.0
Chapter
4&9
9
5
5
9
9
 Note: Please refer to the individual chapters for information on sources of water use data.
       * English units of gallons per megawatt-hour
       ** SI units of cubic meter per mega-joule

-------
       According to the U.S. Energy Information Administration's Annual Energy Outlook
2014, petroleum-derived fossil fuels represented over 95% of all transportation energy consumed
in the U.S. in 2012.  Gasoline (63%), diesel fuel (22%) and jet fuel (15%) are the most widely
used types of petroleum-based transportation fuels. Biofuels contributed only 4.5% of the total
energy consumed for transportation in 2012, the majority of which (4.1% of the total
transportation energy) is ethanol blended into gasoline. Dependence on non-renewable fuels,
concern about global warming, and a push for greater energy independence led to adoption of the
Energy Policy Act of 2005, the Energy Independence and Security Act of 2007 (EISA) and the
introduction of the Renewable Fuel Standards (RFS) in the U.S. These laws and regulations
called for a reduction in annual petroleum consumption by at least 20%, an increase in the use of
alternate fuels/biofuels such as ethanol and biodiesel by 10% by 2015, and a four-fold increase
by 2022.
       In 2012, ethanol constituted 94% of all biofuel produced in the U.S., and was produced
primarily from corn. Other raw materials that can be used to produce ethanol include sugarcane,
sorghum, beverage waste, cheese whey, cellulose and hemi-cellulose, corn stover, hardwood,
and switch grass. Biodiesel, which in the U.S. is primarily made from soy oil, is another biofuel
that is widely available. With technology improvements, low quality feedstocks, especially
feedstocks from waste (trap grease from restaurants and sewer pipelines and other oil containing
wastes) are expected to be increasingly used for biodiesel production and serve to provide waste
reduction and renewable fuel production. Biodiesel can also be produced from algal lipids.
Irrigation for biomass cultivation consumes the largest quantity of water in the production of
biofuels. Water losses during cultivation also occur due to evaporation, evapotranspiration, and
wind loss. Water is used for washing biomass to remove contaminants after harvest. In the
production of ethanol, water is used for pretreatment, fermentation, and recovery processes,
which may include the use of fresh, recycled/treated, or recycled/untreated/carried-over waters.
Water is used to remove impurities during the processing of fatty-acid methyl esters that are used
as biodiesel. Water is also used during cooling operations required for both ethanol and biodiesel
production. The water used for processing biofuels is generally recycled, while the water used
for cooking processes for ethanol production exits the plant as water vapor from cooling towers.
Technology is available to build biofuel production plants capable of achieving zero discharge, if
necessary, and using lower quality surface or gray waters is possible, which  could play an
important role in water-stressed regions.
       In 2005, approximately 3% of the irrigation water used worldwide was used for the
production of biofuels; by 2030, this proportion is projected to grow to approximately 30%.
Additional technologies that can be used to minimize water usage include using pervaporation,
which uses membrane separation to separate biofuels from water  instead of steam distillation,
membrane solvent extraction,  which uses porous membranes to separate biofuel from the
fermentation broth using an extracting solvent, or thermophilic yeasts, which minimizes the
cooling required for the feed going into the fermentation unit. The amount of water consumed for
producing ethanol and biodiesel from different fuel sources is listed in Table ES2. For
comparison purposes, the amount of water consumed to produce gasoline and diesel from
petroleum is also listed. Further information regarding water use in corn ethanol and cellulosic
ethanol production can be found in Chapters 6 and 7, respectively, of this report. Further details
regarding water use for biodiesel production can be found within  Chapter 8 of this report.

-------
     Table ES2 Water intensities on a consumptive basis for producing different types of
                biofuels*
Fuel
Gasoline
Ethanol (corn)
Ethanol (cellulose/switch grass
+SRWC/corn stover)
Diesel Fuel Oil
Biodiesel (soy, current hydroxide TE)
Biodiesel (waste oil, acid-ester)
Biodiesel (algal, hydroxide TE)
Water used for Fuel
Processing*
1-2.5
2.7-40(13.4)1
12-172
1-2.5
0.3-0.5
0.3
1
Water used for Crop
Irrigation or Petroleum
Extraction*
0
15-934(113)1
27 - 691 (28)3
0
1-1059(62)4
0
40-1421 (554)4
 Note: * - Water intensity in gal hbO/gal fuel or m3 hbO/m3 fuel
       1 Approximate average value based on King and Webber (see Chapter 6 of this report)
       2 See Chapter 7 of this report
       3 Approximate average value based on Tidewell et al.'s (2011) projections for 2030 (see Chapter 7)
       4 Approximate average value based on literature discussed in Chapter 8 of this report
       Water, climate, electric generation and transportation fuel issues are closely interrelated
and cannot be adequately addressed in isolation. Water stressed regions of the U.S. are expected
to have continued population growth and energy demand. Additional stresses from climate
change upon these regions is expected to result in freshwater scarcity becoming an issue of
paramount importance. The majority of the water withdrawn and consumed by a thermoelectric
generation is for used cooling steam. Finding alternative water resources to replace freshwater
demand for thermoelectric cooling is both inevitable and urgent. Impaired waters and saline
waters are potential alternatives to freshwater sources that could be used to meet future
thermoelectric cooling needs. There is already some experience with the use of impaired water
for thermoelectric cooling. Examples include the use treated municipal wastewater and the use of
seawater in coastal areas.
       One issue that has not been addressed in this report is the amount of energy required to
transport and treat source water for power plant use as well as the energy required to treat
wastewater from power plants. Additional research related to energy for water will be necessary,
particularly for water-stressed regions of the U.S. Another topic that has not been addressed
within this report and that is a potential area  of future research is the use of waste biomass from
secondary wastewater treatment processes to produce electricity (e.g., using microbial fuel cells),
biofuels or methane gas, which can be used as a fuel source in wastewater treatment plants.
       Laws and regulations put in place in response to climate change and the need to reduce
U.S. dependence on foreign energy imports,  e.g., RFS, RFS2 and EISA, mandate reductions in
the use of petroleum-based fuels and increases in alternative fuels such as the biofuels ethanol
and biodiesel. Biofuels also have significant water requirements, especially at the cultivation
stage. Some of the same water-conserving techniques used by the energy industry such as
recycling process water or using treated municipal wastewater can also be used in the biofuel
industry. Additionally, more research is needed to assess the regional and local water impacts of

-------
different types of electricity generation and biofuel production, and to analyze the water impacts
of electricity-sector choices. More studies of viable energy resources and the impacts of
geographical limitations may be useful in adapting the use of water locally within the fuel and
energy generation sectors.

-------
 1   Introduction
       Timothy C. Keener'
       Freshwater availability and future water demands have become major concerns in the
U.S. when considering its historic economic growth and the ability to sustain its current quality
of life. Past reports on water usage trends have shown that water demand is growing. Meanwhile,
the  capacity for surface water storage is becoming increasingly more limited and ground water is
being depleted. Furthermore, population is rapidly increasing in water-stressed areas, especially
in arid regions in the southwestern U.S. As will be shown in Chapter 4, the energy sector
consumes massive amounts of water every year to provide electricity, and these amounts are
expected to continue to grow unless steps are taken to utilize more efficient and renewable
methods of generation and production. As described in Chapter 2, production of liquid fuels for
transportation is also becoming increasingly water-intensive as the use of agriculturally-derived,
renewable biofuels increases. Also, environmental conditions are increasingly becoming an
important factor to consider while making decisions about the location, size and type of energy
production methods necessary to supply the vast amounts of energy required to power our
economy. As will be seen in Chapter 3, overall water demands nationally have changed little
since the 1980s despite population growth on a national level.  Water demands have changed
within sectors and in different parts of the U.S. and often in places that are the most water
stressed. The U.S. energy sector is in transition; biofuels use is increasing in the U.S. in response
to policies that reduce transportation fuel imports. These factors also affect water use and
consumption, especially in water-stressed regions. Finally, the U.S. has initiated a program to
reduce greenhouse gas emissions from power plants and other sources in order to mitigate the
impacts of climate change, making energy adaptation critical.
       This report seeks to investigate and integrate the current knowledgebase of water usage in
energy production industries, including the generation of electricity and the emerging production
of biofuels. The report covers a discussion and explanation of the major areas of energy and
biofuel production processes, including: 1) trends in energy production, future water availability
and allocation for energy production, 2) the water resource impacts from  coal-fired and natural
gas-fired electric power plants, 3) the water resource impacts of production of corn-starch-based
ethanol, biodiesel, biomass and cellulosic biofuel, and 4) water resource adaptation strategies for
sustainable energy production using Las Vegas, Nevada as a case study.
       While it is important to introduce the topics covered in this report, it is equally important
to realize what this report does not address, as many areas of water usage in the energy sector
were beyond the scope of this report or represent topics for future research. For instance, this
report does not address in detail the substantial amounts of water  consumption from the nuclear
power industry, water consumed upstream of the energy production processes, such  as water
used in the mining of coal or the production of natural gas, the energy required to move the
massive amounts of water used from one place to another since this is so site specific, energy
required to desalinize the water that is used for producing steam or the energy required to reclaim
water.  The report does not directly address the impacts on water quality that the energy sector is
   Director, Air Quality Management/Air Pollution Control Program, University of Cincinnati
   Department of Biomedical, Chemical, and Environmental Engineering

-------
responsible for, nor the ecological effects such water quality may create. Such studies are
appropriate and necessary for future assessments of the overall impact of energy production on
water and the environment.
       Chapter 2 of this report discusses energy and energy production processes, which are an
essential part of life in the U.S. today. This chapter briefly reviews the current coal- and natural
gas-fired thermoelectric production methods (Rankine Cycle) and ethanol and biodiesel
transportation fuel types, as well as their respective contributions to the total energy output in the
U.S. and potential future energy production methods. The way that energy is generated, the
sources that provide the energy and the contribution of each energy source to the total energy
output are described. Future trends of each are discussed based on projections  factoring in  prior
production, existing infrastructure, government regulations and available natural resources.
       In Chapter 3, the impacts of climate change on water resources are discussed in the
context of future energy and transportation biofuel production. The chapter provides an overview
of attributes related to climate and population changes, and the degree of their impacts on water
availability for future energy production. The impacts on energy production are multi-
dimensional, affecting not  only current energy production, but also future energy choices and
overall makeup. To avoid redundancy with other recent reports (e.g., Tidwell et al., 2011; DOE,
2014), this chapter is only focused on the aspects of future water availability that could impact
thermoelectric power production and biofuel production.
       Chapter 4 discusses the water impacts from coal-fired thermoelectric power plants,
including water required for capturing by-products such as flue gas, fly  ash, and bottom ash.  The
term water consumption for these  and other plants refers to the loss of water from the water
source in a catchment area due to water evaporation, or return to another catchment area. The
term water withdrawal refers to water that is removed from a water body. The  main difference
between water withdrawal, which  is the traditional concept to measure water use, and water
footprint is that the former does not consider the amount of water that is returned to the
catchment area. Also, the water footprint concept includes the use of green and gray water in
addition to direct and indirect use  of water during the electricity production process.
       Significant water volumes  are required for the operation of thermoelectric plants as seen
from the fact that in 2005, thermoelectric generation accounted for the largest  percentage (41
percent) of all freshwater withdrawals in the U.S., with coal-fired power plants accounting for
67% of freshwater withdrawals among thermoelectric power plants. Depending upon the cooling
and steam generation technology used, water withdrawal by the thermoelectric sector is expected
to stay the same or decline slightly over the next 25 years. Nonetheless, water  consumption is
expected to increase from around 28 to nearly 50 percent on a national basis. It is projected that
older power plants, which mainly use once-through cooling systems with high water withdrawal
rates, are likely to be retired over the next  20 years. Facilities that have been built in the last two
decades, and the ones projected to be built in the coming years, are most likely to employ wet
recirculating cooling systems that  have low withdrawal rates but high water consumption values.
Other topics that are covered under this chapter include impacts of coal  mining, wastewater
treatment systems to treat flue gas desulfurization wastewater, potential to conserve and/or reuse
water, and models to predict carbon dioxide generation and water withdrawal rates.

-------
       Chapter 5 addresses water usage from natural gas-fired thermoelectric power plants.
Natural gas is the second largest fossil fuel used for electric power generation, accounting for
30% of total U.S. electricity production in 2012. However, this percentage is expected to
increase as the emissions of carbon dioxide from coal thermoelectric plants are regulated to
levels at or near those from natural gas thermoelectric plants. Due to the relatively low natural
gas prices and capital costs, a natural gas plant is more competitive and an alternative choice for
new generation capacity. According to EIA predictions in reference cases, natural gas power
plants will account for 63% of electricity capacity additions from 2012 to 2040 (EIA, 2013).
       Chapters 6, 7 and 8 discuss water consumption requirements for the production of
biofuels, including corn-starch based ethanol production (Chapter 6), biomass and cellulosic
production (Chapter 7) and biodiesel production (Chapter 8). The objective of the analysis
contained in Chapter 6 is to systematically evaluate water usage impacts of the  ethanol
production process by studying existing ethanol plants. As concerns about global warming and
dependence on fossil fuels grows, the search for renewable energy sources that reduce carbon
dioxide emissions has become a priority. Although biofuels offer a diverse range of promising
alternatives, ethanol constitutes 94% of all biofuels produced in the U.S. in 2012. At present,
virtually all U.S. fuel ethanol is produced from  the fermentation of corn in dry and wet milling
plants, most of which are located in the Midwestern states (Wang, 2005).  The methodologies
applied within the analysis are mathematical modeling of the energy balance (including thermal
energy and electricity) and the mass balance (including corn input, water usage, wastewater
discharge, co-products and CO2 emissions) with respect to the  various process system
boundaries.
       Cellulosic ethanol production is discussed in Chapter 7. In this chapter,  the water
requirements for biomass harvesting and cellulosic ethanol production via a fermentation
pathway are assessed on a volume-to-volume basis (i.e., gallons of water consumption per gallon
of ethanol production) using data reported in the literature. The water requirements have been
analyzed for a combination of three feedstocks  (e.g., hardwood, corn stover  and switch grass)
and two pretreatment technologies (e.g., dilute acid  and ammonia fiber expansion) under
different water network configurations with an  overall goal of zero wastewater  discharge. The
results indicate that the process water requirements are significantly dependent on the selection
of pretreatment processes and feedstocks, while effective cooling tower design  and operation of a
cooling tower offers an opportunity for saving water.
       Chapter 8 investigates water consumption from the processing of biodiesel. The
production of biodiesel has become a globally mature industry with many diesel vehicles now
capable of using higher percentages of biodiesel fuel, with many more in the design and
production stages. In the U.S., a record high of  1.1 billion gallons of biodiesel were produced
from 193 biodiesel manufacturers in 2012, compared to 28 million gallons in 2004.  The chapter
begins with a summary of the current status of the U.S. biodiesel industry, followed by an
estimate of water consumption for the processing of lipids into biodiesel. Currently, soy oil is the
primary source of lipids that are processed into  biodiesel  fuel in the U.S. The analysis of water
consumption from biodiesel production processes began with a survey of relevant literature, and
then determining characteristic allocation factors for each of the various stages' methods. Both
state-level estimates and national averages of water consumption have been  determined and

-------
compared with several relevant studies in detail. Water consumption patterns of water-stressed
areas have also been summarized. The chapter concludes with a discussion of future biodiesel
trends, including the use of new feedstocks and the use of new biodiesel production technologies
since these changes can also affect water use. Water use from algal biodiesel is also briefly
summarized.
       In Chapter 9, an assessment of the impacts of electric power generation on water
availability has been made for the water-stressed areas of Arizona, California, New Mexico,
Nevada, Texas and Utah. Thermoelectric power plants in the southwest will most likely face
significant challenges regarding water withdrawal and water consumption because they are
located in an arid region. The impacts of the water withdrawals are exacerbated by increased
population migration to the region and reduced regional precipitation. Water that is removed for
cooling purposes in thermoelectric plants and not available for other uses is an especially
important consideration for water-scarce regions, and is particularly relevant in future energy
resource development adaptation strategies. Changes in  energy regulations and policies as well
as shifts in the electricity generation portfolio toward implementing innovative technologies, can
therefore, be expected to have significant impacts on the management of local, regional and
national water resources.
       Finally, Chapter 10 provides a discussion about the future of a sustainable water-energy
nexus, including an assessment of the vulnerability of such a future with a discussion of how
adaptive engineering may be useful in overcoming obstacles. The chapter concludes with a
discussion of current knowledge gaps, and future research and development issues that may need
to be resolved to overcome these gaps.

1.1  References

Energy Information Administration (EIA). 2013. "Short-Term Energy Outlook, Number 21."
       http://www.eia.gov/todayinenergv/detail.cfm?id= 13891. Accessed July 8, 2014.

Tidwell, V., A.C.-T. Sun, and L. Malczynski. 2011. "Biofuel impacts on water." Sandia Report
       SAND2011-0375, 31p. http://prod.sandia.gov/techlib/access-
       control.cgi/2011/110375.pdf.

U.S. Department of Energy (DOE). 2014. "The Water-Energy Nexus: Challenges and
       Opportunities."
       http ://www.energy. gov/sites/prod/files/2014/07/fl 7/Water%20Energy%20Nexus%20Full
       %20Report%20Julv%202014.pdf. Accessed July 2014.

Wang, M. 2005. "Updated Energy and Greenhouse Gas  Emission Results of Fuel Ethanol." The
       15th International Symposium on Alcohol Fuels. 26-28 September 2005, San Diego, CA,
       USA. http://www.transportation.anl.gov/pdfs/TA/375.pdf. Accessed July 2014.

-------
2    Trends in Energy Production
       Marissa S. Liang,1 Qingshi Tu,1 Timothy C. Keener,1 J. McDonald,11 Pratim Biswas,111
       Wei-Ning Wang1", and William E. Flatten IIIiv

2.1  Introduction

       The total energy production in the U.S. in 2012 was 95 quadrillion BTUs (EIA, 2014a).
Electricity generation and transportation are two of the largest sources of energy demand in the
U.S, accounting for over half of the total energy production. This chapter discusses the current
electricity production methods and transportation fuel types, their respective contributions to the
total energy output in the U.S., and potential future energy production methods. The way that
energy is generated, the sources that provide the energy, and the contribution of each energy
source to the total energy output are described. Future trends of each are discussed based on
projections factoring in prior production, existing infrastructure, government regulations and
available natural resources.

2.2  Current Energy Production Methods

2.2.1   Electrical energy demand and production
       In 2012, total electricity demand was 3,826 billion kilowatt-hours (kWh) (13 quadrillion
BTUs) in the U.S. (EIA,  2014a). The largest consumption of electricity is for residential
applications (36%) such  as lighting, heating or cooling, home appliances and consumer
electronics, while the commercial and industrial sectors consume 35% and 26%, respectively
(EIA, 2014a).
       Electricity demand fluctuates with respect to many factors. Weather conditions, prices
and business cycles are the dominant factors that impact electric demand on a short-term basis.
There has been a steady increase in demand for electricity over the last century, although this
increase has slowed somewhat in the decades since  1950. The growth rate is predicted to be
0.8% per year through 2035, down from 9% per year in the 1950s and 2.5% per year in the
1990s, (EIA, 2014a).
       Most of the electricity in the U.S. is generated by fossil fuel combustion using steam
turbines.  Fossil fuels such as coal, natural gas, and petroleum oil are burned in a boiler furnace to
produce steam. The kinetic energy of the steam is converted to mechanical energy within  a steam
turbine to provide shaft-work to drive electrical generators.
       Among fossil fuels, coal and natural gas have been the most common fuels used to
generate  electricity. Historic and projected electrical production and the contributions of each
energy source are shown in Figure 2.1. Coal was used to generate nearly 37% of the 4 trillion
kWh (1.3 x 1016 BTUs) of electricity used in the U.S.  in 2012, followed by natural gas at 30%.
   University of Cincinnati, Department of Biomedical, Chemical, and Environmental
   Engineering
   U.S. Environmental Protection Agency - National Risk Management Research Laboratory
   Washington University in St. Louis
   Pegasus Technical Services, Inc., Cincinnati, OH

                                           10

-------
By comparison, nuclear power and renewable energy provided 19% and 12% of the total energy
output, respectively (EIA, 2014a).
                5
             c
             o
             =  2
                          History
                           —r~
            2012
              Projections
                1990
2000
2010
2020
2030
2040
Figure 2.1 Electricity generation and energy source contributions. Adopted from EIA
       (2014b).

2.2.2   Transportation fuels
       The majority of the transportation in the U.S. is done using motor vehicles utilizing liquid
fuels.  A total of 26.7 quadrillion BTUs of energy was used for transportation in 2012 (EIA,
2014a). Petroleum-derived fossil fuels represented over 95% of all transportation energy
consumed in the U.S. in 2012. Gasoline (63%), diesel fuel (22%) and jet fuel (15%) are the most
widely used types of transportation fuels (EIA, 2014a). These percentages also take into account
blends of crude-oil-derived gasoline and diesel fuel with biofuels such as ethanol and biodiesel.6
        Biofuels contribute only 1.2 quadrillion BTUs, or 4.5% of the total energy consumed for
transportation, the majority of which (4.1% of the total transportation energy) is ethanol blended
into gasoline. Jet fuel accounted for the majority of the approximately 15% remainder in
transportation fuel consumption in 2012, with other transportation fuels such as natural gas,
propane and hydrogen, or energy sources such as electricity for battery electric or plug-in hybrid
vehicles, contributing less than 1%.
       Overreliance on non-renewable fuels and a push for greater energy independence led to
the introduction of the Renewable Fuel Standards (RFS), which was established in 2005 as part
of the Energy Policy Act (EPAct), which was then amended in the 2007 Energy Independence
       6  Gasoline in the United States may be blended up to 10% with ethanol. The ASTM
         D975 specification for No. 2 diesel fuel oil includes blends of up to 5% biodiesel.
                                           11

-------
and Security Act (EISA). EPAct and then EISA were established to set up goals and
requirements to secure energy independence for the U.S. (U.S. DOE, 2010). The act defines the
goal of a reduction in annual petroleum consumption by at least 20%, and a 10% increase in
renewable and alternative fuel use by 2015 from the 2005 baseline. The renewable and
alternative fuel requirements have led to an increase in production of ethanol and biodiesel. U.S.
production of biofuels approximately tripled in volume between 2005 and 2012 (EIA, 2014c; see
Figure 2.2). Ethanol is the primary biofuel used in the U.S. today, and is primarily produced by
the conversion of sugars from grain into alcohol (ethanol), which is then purified through
distillation. The main source of ethanol is corn, but other sources include sorghum, barley,
sugarcane and other agricultural feedstocks.
       Almost all gasoline used in vehicles in the U.S. contains some percentage of ethanol,
typically 10% by volume (E10). In 2010 and 2011, EPA granted two partial waivers that allow,
but do not require, the use of El 5 in model year 2001 and newer light-duty motor vehicles,
subject to certain conditions (U.S. EPA, 2013a). Higher ethanol percentages can be used, but
only if the vehicle has been designed to handle the higher concentration of ethanol, which can
cause deterioration of fuel system components unless they are specifically designed for such use.
The most common high concentration blend is E85 (85% ethanol blend), which is used in flex-
fuel vehicles. E85 accounted for only 0.014 quadrillion BTUs of energy used for transportation
in 2012, or 1.3% of all ethanol-blended gasoline. Biodiesel is the next most common biofuel. It is
a mixture of fatty acid methyl  esters (FAME) produced primarily from a trans-esterification
reaction between triglycerides (oil) and methanol with alkali catalysts. Current biodiesel
feedstocks in the U.S. are primarily from seed-crop oils (e.g., soy oil, rapeseed oil) and animal
fats (e.g., tallow and lard). Biodiesel blends of 5% (B5) with petroleum diesel fuel are capable of
being used in most modern diesel engines without modification and blends up to 20% (B20) can
be used in several of the most recent engine manufacturer offerings, but biodiesel is not currently
produced in large quantities. Biodiesel accounted for approximately 1,159 and 1,244 trillion
BTUs (991 and 1,339 million gallons in 2012 and 2013, respectively) in 2012 and 2013,
respectively, or 10% and 14% (114 Btu in 2012 and  175 Btu in 2013) of all the transportation
energy produced from biofuels in 2012 and 2013, respectively (EIA, 2014d).
       Along with the growth of biofuel production, there has been a rapid expansion of the
feedstock market. However, due to an ongoing "food vs. fuel" debate (Canali and Aragrande,
2010; Cassman and Liska, 2007; Tilman et al.,  2009), high feedstock costs (Haas, 2011) and
concerns over sustainability that include water use (King and Webber, 2008, Dominguez-Faus et
al., 2009} and land use (Rathmann et al., 2010, Achten et al., 2011), renewable fuel producers
are also actively investigating alternative feedstocks. For example, Table 2.1 shows a summary
of water consumption for gasoline, petroleum diesel, E85  from corn grain (starch) and corn
stover (cellulose), and soybean biodiesel from the literature. The water consumption for
producing the biofuels is significantly higher than for producing petroleum fuels, mostly due to
irrigation requirements. The potential for new biofuel feedstocks is discussed further within
Chapter 8 of this report.
                                           12

-------
              Million Barrels per Day Consumed for Transportation
             10.0	
              7.5-
                                                         Gasoline
                                                         • Diesel Fuel Oil
              5.0-
              2.5-
                       2006    2007     2008    2009    2010     2011    2012
              Million Barrels of Ethanol per Day (U.S. Production for Domestic Use)
             1.00	
0.75.

0.25-
0.00.
                                                         Ethanol
                       2006    2007    2008    2009    2010     2011    2012
              Million Barrels of Biodiesel per Day (U.S. Production for Domestic Use)
             0.08	==
             0.06-
             0.04.
             0.02.
             0.00.
                                                         Biodiesell
                       2006
                  2007
2008
2009    2010
2011    2012
Figure 2.2   Comparison of the use of gasoline and diesel transportation fuels to domestic
       production of biofuels. Note the difference in scales for the charts of petroleum,
       ethanol and biodiesel fuels, respectively. Gasoline consumption is dominated by light-
       duty vehicle use and in 2012, ethanol accounted for -10% of light-duty fuel
       consumed. Adopted from EIA (2014c).
                                            13

-------
     Table 2.1  Fuels production water intensity
Transportation Fuel
Gasoline from Liquid Petroleum
Diesel from Liquid Petroleum
E85from Irrigated Corn Grain
E85from Non-Irrigated Corn Grain
E85from Irrigated Corn Stover
E85from Non-Irrigated Corn Stover
Biodiesel from Irrigated Soy
Biodiesel from Non-Irrigated Soy
Consumption
(gal/gal)
1-2.5
1-2.5
20-936
2-5
39-695
4
15-617
0.3-0.5
            Source: King and Webber (2008).
2.3  New Energy Technologies and Future Outlook

2.3.1   Trends in advanced coal technologies for power production
       Because of the increasing problem of CO2 and other greenhouse gas (GHG) emissions,
the large fraction of electricity currently produced from fossil fuel power plants, Clean Air Act
compliance requirements, and the predicted lack of alternative energy sources, there is a need to
develop next-generation technologies that enable reliable, low-cost, low-carbon energy from
fossil fuels, particularly coal. While natural gas is expected to displace some coal-fired electric
generation capacity, it must be noted that coal-fired power plants offer benefits that include low
cost and affordable power rates in many parts of the world. Advanced coal technology
development efforts in the U.S., for example the U.S. DOE Clean Coal Research and
Development Program, have focused on developing and demonstrating novel concepts of power
generation, carbon capture, utilization, storage, and conversion technologies for existing facilities
and new fossil-fueled power plants while increasing overall system efficiencies and reducing
capital costs (U.S. DOE, 2014). In the near-term, advanced technologies that increase the
efficiency of power generation for new plants, and technologies to capture carbon dioxide (CO2)
from new and existing industrial and power-producing plants, are being developed with the
assistance of U.S. DOE funding (U.S. DOE, 2014). Carbon capture, however, consumes energy
and reduces the overall efficiency of the plant. In the longer-term, there will be continued focus
on increasing energy plant efficiencies, and reducing both the energy and capital costs of CO2
capture and storage from new, advanced coal plants and existing plants (U.S. DOE, 2014). These
strategies will ensure a diverse future energy portfolio for electric power generation.
                                           14

-------
       One approach to making the capture of carbon dioxide in exhaust gas cost-effective is to
combust the coal in oxygen-enriched air (by removing nitrogen) and recycle the exhaust gas to
serve the role of the diluent. This is known as oxy-combustion. Such a configuration would
result in a carbon dioxide concentrations of up to 95% 7, making it more feasible to capture CO2
(Suriyawong et al., 2006; Wang et al. 2012).  Other advantages include reduced volume of flue
gas to exhaust, potential for increasing boiler thermal efficiency, elimination of thermal NOx
(due to removal of nitrogen from the air stream), decreasing the conversion ratio of fuel-N to
exhaust NOx (nitrogen in the fuel that is converted to NOx during combustion), and increasing
the reduction of NOx to N2 (removal of NOx before exiting the plant).
       A variant of the oxy-combustion currently being tested at the pilot-scale level is the
chemical looping combustion process (Fan, 2011; Feeley, et al., 2007). Chemical looping splits
combustion into separate oxidation and reduction reactions. A metal (e.g., iron, nickel, copper or
manganese) oxide releases oxygen in a reducing atmosphere, which can then react with the fuel
to provide energy. The metal is then recycled back to a zone where the metal oxide is
regenerated in contact with air. The two sections used for the combustion process allow CO2 to
be concentrated, and when the water produced during combustion is removed, it is not diluted
with the nitrogen from air.  The advantage of this process is that no air separation units or
external CO2 separation equipment is needed (Feeley, et al., 2007).
2.3.2  The emerging role of natural gas in power generation
       Electric power generation  and the industrial sector are the two largest sectors of natural
gas consumption in the U.S. (EIA, 2014e). Due to the increased use in these two sectors, natural
gas consumption is predicted to grow by about 0.6 percent per year from 2011 to 2040. Natural
gas accounted for 30% of total electricity generation in 2012.  It is projected to increase to 35% in
2040 (Figure 2.1). Due to the domestic production of natural gas from shale deposits in North
America, the price of natural gas is likely to remain under the levels observed in 2005-2008.
Compared with coal-fired power plants, the relative low cost of natural gas makes the operation
of existing natural gas-fired power plants increasingly competitive with coal-fired power plants
(EIA, 2012). In addition, comparatively low capital costs make natural gas power plants an
attractive alternative choice as future decisions are made to bring new electrical generation
capacity online (EIA, 2014e).
2.3.3  Other alternative energy developments
       Nuclear and renewable energy provided, respectively, 19% and 12% of the total
electricity generation in the U.S. in 2012. Renewable energy sources include hydroelectric
power, wind, biomass wood and waste, geothermal and solar. The largest share of renewable-
generated electricity was from hydroelectric power. It generated 56% of the total electricity from
renewable sources, followed by wind (28%), biomass wood (8%), biomass waste (4%),
geothermal (3%) and solar (1%) (EIA, 2014f).
       7 95% CO2 represents a theoretical limit. In practice, CO2 concentrations are lower,
primarily due to air-leakage.

                                           15

-------
       In 2011, a total of 104 commercial nuclear power plants were in operation and generated
886 billion kWh (3.02 quadrillion BTUs) of electricity. This share is projected to increase until
2025, and to eventually reach 1000 billion kWh (3.41 quadrillion BTUs). After a projected
decline in 2036, mainly due to plant retirements, newly built nuclear power plants are expected
to bring nuclear capacity back up to 991 billion kWh (3.38 quadrillion BTUs) in 2040. However,
the share of nuclear power in the U.S. energy profile is expected to decrease as the increases of
generation capacity from renewable sources and  natural gas outpace the growth of nuclear
generation (EIA, 2014f).
       The use of renewable energy in electricity generation depends on both the availability of
resources and the generation capacity that is required by the applications. Nearly all current
hydroelectric power plants were built in the mid-1970s. They were mostly built at dams operated
by federal agencies. Most of the wood biomass is used as an energy source by  lumber and paper
mills to generate steam and electricity on-site for their own needs. As of 2012, there were 13
solar thermoelectric plants8 that generated electricity by concentrating solar energy to heat fluids
that power a steam turbine generator. Solar photovoltaic cells can be used on a relatively small
scale for individual buildings or on a much larger scale as part of "solar farms" that provide
electricity to specific industrial applications (e.g., Microsoft, Facebook, Amazon or Google data
server farms) or directly to the electric grid. Solar energy generated at small-scale installations,
such as on individual building roof-tops, generated 9.86 billion kWh of energy in the U.S. in
2013 (EIA, 2014a).
       Government policies also affect the use of renewable energy for electricity generation.
For example, state and regional programs and federal financial  incentives have driven an
increasing share of wind generation during the past decade. Wind and solar power generation are
expected to lead the renewable energy category,  increasing from 59 and 8 GW, respectively, in
2012 to 87 and 48 GW, respectively, in 2040  (EIA, 2014g). Other sources, such as biomass and
geothermal, are expected to continue to increase, but at a much slower pace, while hydropower
will remain almost constant over the entire period.
2.3.4  Regulatory impacts on future electric power generation
       U.S. EPA has published a set of regulations under the Clean Air Act (CAA) for stationary
source GHG mitigation. In 2010, the U.S. EPA issued the Greenhouse Gas Tailoring Rule to
address GHG emissions from stationary sources  under CAA permitting programs (U.S. EPA,
2010a). These  regulations set thresholds for GHG emissions that define when permits under the
New Source Review Prevention of Significant Deterioration (PSD) and title V Operating Permit
programs are required for new and existing industrial facilities. This final rule  "tailors" the
requirements of these CAA permitting programs to limit what facilities will be required to obtain
PSD and title V permits. Facilities responsible for nearly 70% of the national GHG emissions
from stationary sources will be subject to permitting requirements under this rule. This includes
the nation's largest stationary GHG emitters—electric power plants, refineries and cement
       8 There were eleven solar thermoelectric plants in CA, one each in FL and NV, and one
in FL that provides supplemental steam for an existing oil and gas-fired power plant.

                                           16

-------
production facilities. Emissions from small farms, restaurants and all but the very largest
commercial facilities are not covered under these regulations.
       In September 2013, the U.S. EPA proposed a Carbon Pollution Standard for New Power
Plants (U.S. EPA, 2013b). This program would establish new national limits on the amount of
carbon pollution emitted by future electric power plants.  The proposed standards would apply
only to new fossil-fuel-fired electric utility generating units. EPA proposed CO2 standards of
performance for sources within the following sub categories:

•   1,000 Ib CO2/MWh for natural gas-fired stationary combustion turbines with a heat input
    rating to the turbine engine that is greater than 850 MMBtu/hr;
•   1,100 Ib CO2/MWh for natural gas-fired stationary combustion turbines with a heat input
    rating to the turbine engine that is less than or equal to 850 MMBtu/hr, and
•   1,100 Ib CO2/MWh for all other fossil fuel-fired boilers and IGCC units
       The proposed standards would not apply to plants currently operating or to newly
permitted plants that begin construction during the 12 months following adoption of the
regulation. For existing power plants, EPA issued proposed Carbon Pollution Standards on June
2, 2014 (U.S. EPA, 2014a) that reduces power plant GHG emission for each state by 30%
relative to 2005 by 2030.
       Nearly all (95%) of the natural gas combined cycle (NGCC) units built since 2005 would
meet the  proposed GHG standards; so,  it is expected that any new NGCC power plants would be
able to meet the proposed standards without additional emission controls. New power plants that
are designed to use coal or petroleum coke would need to incorporate technology such as carbon
capture and sequestration to reduce CO2 emissions sufficiently to meet the proposed standards.
       The capture and injection of CO2 produced by human activities for storage via long-term
geologic  sequestration is one of a portfolio of options that are expected to reduce CO2 emissions
to the atmosphere from large stationary sources of GHG emissions. Geologic sequestration that
may occur from future carbon pollution stationary-source standards under the authority of the
C AA must be performed in a manner that safeguards underground sources of drinking water,
such as aquifers, as required by the Safe Drinking Water Act (SDWA). In December 2010, the
U.S. EPA finalized "Federal Requirements under the Underground Injection Control for Carbon
Dioxide Geologic Sequestration Wells" (U.S. EPA, 2010b) under the authority of the SDWA's
Underground Injection Control (UIC) Program. These requirements, also known as the Class VI
rule, are designed to protect underground sources of drinking water from CCh-injection related
activities. The Class VI rule builds on existing UIC Program requirements, with extensively
tailored requirements that address CO2 injection for long-term storage to ensure that wells used
for geologic sequestration are appropriately sited, constructed, tested, monitored, funded and
closed. The rule also affords owners or operators injection depth flexibility to address injection
in various geologic settings in the U.S.  where geologic sequestration may occur, including very
deep formations, and oil and gas fields  that are transitioned for use as CO2 storage sites.
2.3.5  Future biofuel usage
       The "Annual Energy Outlook 2014" (AEO 2014) published by the Energy Information
Administration (EIA, 2014a) shows the predicted total petroleum and other liquid fuel


                                           17

-------
production from 2012 through 2040 (Figure 2.3). It should be noted that these predictions are
based on available information at the time of its publication (April 2014) and may vary
significantly from the actual future values. Market forces, technological advancement, and
regulations all would have a significant impact on these predictions. The AEO 2014 report
predicts flat production of biofuels over the next three decades, largely due to difficulties with
increasing the use of ethanol. Ethanol production will remain relatively flat or decrease slightly
through 2040 due to a decrease in gasoline usage, the limited availability of flex fuel vehicles
capable of operation on ethanol blends above El 5, the limited availability of retrofitted filling
stations capable of dispensing higher ethanol blends The production of biodiesel is projected to
be constant as well, based on the assumption that the required volume under RFS for biomass-
based diesel will remain at 1.28 billion gallons. The outlook also indicates that biofuel
consumption  may fall short of the EISA2007 goal of 36 billion RFS credits (Figure 2.4). The
major reason  is the decrease in gasoline consumption due to the recently enacted Light-Duty
Vehicle Greenhouse Gas Emissions and Corporate Average Fuel Economy (CAFE)  Standards
(U.S. EPA, 2010c, 2012) and updated expectations for the sales of E85 compatible vehicles. For
example, the  data shows that the projection of E10 and E15 demand drops from 8.7 million
barrels per day in 2012 to 7.9 million barrels per day in 2022 and 6.7 million barrels per day in
2040 (Figure  2.5).
               Million Barrels per Day
               15
               10
                 2012   2015
                                 2020
                                          2025
                                                    2030
                                                              2035
Figure 2.3   U.S. production of petroleum and other liquids by source, 2012-2040 (million
       barrels per day). Adopted from EIA (2014b).

       The proposed standards in the 2014 RFS, which are still under consideration and subject
to change before final approval, are structured to ensure continued growth of renewable fuels
while recognizing the practical limits on ethanol blending known as the ethanol "blend wall"
(U.S. EPA, 2013c). The blend wall refers to the difficulty in incorporating increasing amounts of
ethanol into the transportation fuel supply at volumes exceeding those achieved by the sale of
nearly all gasoline as E10. The proposed standards cover both ethanol and non-ethanol biofuels,
                                           18

-------
and would establish a total annual volume of 15.21 billion gallons of renewable fuels (see Table
2.2). As shown in Figure 2.6, the proposed ethanol shares of the total gasoline pool (green line)
for both 2013 and 2014 are almost constant at approximately 10% due to the reduction in
gasoline consumption (blue line) in the U.S. This percentage is less than the amount anticipated
when the U.S. Congress established the program in 2007.
             Billion Credits
              40
                                        EISA 2007 Renewable Fuel Standard
              30
              20
              10
                                          Drop-in Biofuels
                                             Biobutanol
                                             Biodiesel
                 2014
                              2020
                                          2025
                                                     2030
                                                                 2035
                                                                           2040
Figure 2.4  EISA 2007 RFS credits earned in selected years, 2012-2040. Adopted from EIA
       (2014b).
              10
               0  r
                                        Motor Gasoline Consumption
                                           (includes E10 & E15)
                                            Product Exports
                2012    2015
                                 2020
                                            2025
                                                       2030
                                                                  2035
                                                                            2040
Figure 2.5   Projections of U.S. domestic gasoline and diesel fuel consumption through 2040.
       Exports of refined products are also shown. Adopted from EIA (2014b).
                                             19

-------
     Table 2.2  Proposed volume for 2014 RFS
Category
Cellulosic biofuel
Biomass-based diesel
Advanced biofuel
Renewable fuel
Volume (% of All Fuels)
17 million gallons (0.010%)
1.28 billion gallons (1.1 6%)
2.20 billion gallons (1.33%)
15.21 billion gallons (9.20%)
              Source: U.S. EPA (2014c)
       The daily supply of ethanol in the projected period (2012-2040) is in the range of 0.83 to
0.95 million barrels, with an annual growth rate of 0.5%. Both ethanol blending into gasoline at
15% or less and E85 consumption for energy use are also essentially flat with an annual growth
rate of 0.6% throughout the projection period (2012-2040) as a result of declining gasoline
consumption and limited penetration of flex-fuel vehicles capable of operation on fuel blends of
up to 85% ethanol (EIA, 2014h). Flex-fuel vehicles are projected to represent only 11% of all
new light-duty vehicles sales in 2040. The wholesale price of ethanol for transportation was
projected to be near constant from 2012 to 2040, with an average price of $2.5 to 2.6/gal.
          Gasoline Consumption and Ethanot Share of Gasoline
                2004  2005  2006 2007 2008 2009 2010 2011 2012 2013 2014
Figure 2.6 Gasoline consumption and percentage ethanol share of gasoline. Adopted from
      EIA (20141).

      As shown in Figure 2.7, consumption of E15 and E85 fuels is predicted to increase at the
expense of E10, but total ethanol biofuel consumption is expected to decrease as a whole by
2040. Starting in 2020, El5 is expected to slowly penetrate the motor gasoline market, as blend
wall issues are assumed to be resolved over time and, by 2040, will make up approximately 40%
                                           20

-------
of the total motor gasoline market. The increase in consumption of El 5 is based on an
assumption that consumers, refiners and vehicle manufacturers will transition to E15 from E10,
choose El 5 over E85 and other blends, and that infrastructure constraints will be resolved
gradually over time (EIA, 2014h).
                          2012       2020        2030       2040
Figure 2.7   Consumption of biofuels in motor gasoline blends in the Reference case, 2012-
      2040 (million barrels per day). Adopted from EIA (2014b).
2.4  References

Achten, W.M.J., andL.V. Verchot. 2011. "Implications of biodiesel-induced land-use changes
       for CO2 emissions: Case studies in tropical America, Africa, and Southeast Asia."
       Ecology and Society, 16(4): 14.

Canali, M., and M. Aragrande. 2010. "Biofuels versus Food Competition for Agricultural
       Resources: Impacts on the EU Farming Systems." In: The Economic Impact of Public
       Support to Agriculture, pp. 191-209.

Cassman, K., and AJ. Liska. 2007. "Food and fuel for all: realistic or foolish?" Biofuels,
       Bioproducts and Biorefining, 1(1): 18-23.

Dominguez-Faus, R., S.E. Powers, J.G. Burken, and PJ. Alvarez. 2009. "The water footprint of
       biofuels: A drink or drive issue?" Environmental Science & Technology, 43(9):3005-
       3010.

Energy Information Administration (EIA). 2014a. Annual Energy Outlook 2014.
       http://www.eia.gov/forecasts/aeo/index.cfm. Accessed June, 2014.
                                          21

-------
EIA. 2012. "Annual Energy Outlook 2012 with Projections to 2035."
       http://www.eia.gov/forecasts/aeo/pdf/0383%282012%29.pdf. Accessed August, 2014.

EIA. 2014b. "Annual Energy Outlook 2014."
       http://www.eia.gov/forecasts/aeo/MT_liquidfuels.cfm. Accessed August, 2014.

EIA. 2014c. "Short Term Energy Outlook." http://www.eia.gov/forecasts/steo/query/. Accessed
       June, 2014.

EIA. 2014d. "Monthly Energy Review." U.S. Energy Information Administration, July 2014.
       Office of Energy Statistics, U.S. Department of Energy, Washington, DC. DOE/EIA-
       0035 (2014/07).

EIA. 2014e. "Natural Gas Explained. Use of Natural Gas."
       http://www.eia.gov/energyexplained/index.cfm?page=natural_gas_use. Accessed June,
       2014.

EIA. 2014f. "Electricity in the United States - Basics."
       http://www.eia.gov/energyexplained/index.cfm?page=electricity in the united states.
       Accessed June, 2014.

EIA. 2014g. "Energy in Brief."
       http://www.eia.gov/energy in brief/article/renewable  electricity.cfm. Accessed June,
       2014.

EIA. 2014h. "Annual Energy Outlook 2014."
       http: //www. ei a. gov/forecasts/aeo/MT_li qui dfuel s. cfm. Accessed April, 2014.

EIA. 20141. "Today in Energy, November 21, 2013. EPA Issues proposed rule for the 2014
       Renewable Fuel  Standard." http://www.eia.gov/todayinenergy/detail.cfm?id= 13891.
       Accessed April, 2014.

Feeley III, T.J., T. E. Fout, A. P. Jones. 2007. "DOE'S Carbon Capture and Sequestration R&D
       Program."  Proceedings of the 2007 POWER-GEN International Conference, New
       Orleans, LA, USA.

Fan, L.-S. 2011. Chemical Looping Systems for Fossil Energy Conversions. John Wiley & Sons.

Haas, M. J. 2011. Extraction and use of DDGS lipids for biodiesel production. In Distillers
       grains: production, properties, and utilization. CRC Press, Baca Raton, FL. pp 487-502.

King, C.W. and M.E. Webber. 2008. "Water intensity of transportation." Environmental Science
       & Technology, 42 (21):7866-7872.

Rathmann, R., A.  Szklo, and R. Schaeffer. 2010. "Land use competition for production of food
       and liquid biofuels: An analysis of the arguments in the current debate." Renewable
       Energy, 35(1): 14.
                                          22

-------
Suriyawong, A., M. Gamble, M.-H. Lee, R. Axelbaum, and P. Biswas. 2006. "Submicrometer
       particle formation and mercury speciation under Ch-CCh coal combustion." Energy &
       Fuels, 20(6):2357-2363.

Tilman, D., R. Socolow, J.A. Foley, J. Hill, E. Larson, L. Lynd, S. Pacala et al. 2009. "Beneficial
       biofuels—the food, energy, and environment trilemma." Science, 325(5938):270.

U.S. DOE. 2010. "Energy independence security and act."
       http://wwwl.eere.energy.gov/femp/regulations/eisa.html. Accessed June, 2014.

U.S. DOE. 2014. "Clean Coal Research." http://www.energy.gov/fe/science-innovation/clean-
       coal-research.

U.S. EPA. 2010a. "Final Rule: Prevention of Significant Deterioration and Title V Greenhouse
       Gas Tailoring Rule." http://www.epa.gov/nsr/documents/20100413fs.pdf.

U.S. EPA. 201 Ob. "Federal Requirements under the Underground Injection Control (UIC)
       Program for Carbon Dioxide (CO2) Geologic Sequestration (GS) Wells; Final Rule."
       http ://www. gpo. gov/fdsvs/pkg/FR-2010-12-10/pdf/2010-29954.pdf

U.S. EPA. 2010c. "Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate
       Average Fuel Economy Standards; Final Rule." http://www.gpo.gov/fdsys/pkg/FR-2010-
       05-07/pdf/2010-8159.pdf.

U.S. EPA. 2012. "2017 and Later Model Year Light-Duty Vehicle Greenhouse Gas Emissions
       and Corporate Average Fuel Economy Standards; Final Rule."
       http://www.gpo.gov/fdsvs/pkg/FR-2012-10-15/pdf/2012-21972.pdf.

U.S. EPA. 2013a. "E15 (a blend of gasoline and ethanol)"
       http://www.epa.gov/otaq/regs/fuels/additive/e 15/index.htm. Accessed August 20, 2014.

U.S. EPA. 2013b. "2013 Proposed Carbon Pollution Standard for New Power Plants."
       http://www2.epa.gov/carbon-pollution-standards/2013-proposed-carbon-pollution-
       standard-new-power-plants. Accessed June, 2014.

U.S. EPA. 2013c. "2014 Standards for the Renewable Fuel Standard Program." U.S. Federal
       Register, 78 (230):71732-71784. November 29, 2013.
       http://www.epa.gov/otaq/fuels/renewablefuels/regulations.htm. Accessed June, 2014.

U.S. EPA. 2014a. "Clean Power Plan Proposed Rule." http://www2.epa.gov/carbon-pollution-
       standards/clean-power-plan-proposed-rule. Accessed June, 2014.

Wang, W.-N., W.-J.  An, B. Ramalingam, S. Mukherjee, D.M. Niedzwiedzki, S. Gangopadhyay,
       and P. Biswas. 2012. "Size and structure matter: Enhanced CO2 photoreduction
       efficiency by size-resolved ultrafine Pt nanoparticles on TiO2 single crystals."  Journal of
       the American Chemical Society,  134(27):11276-11281.
                                          23

-------
3    Climate Change Impacts on Water Availability for Energy
     Production

       Y. Jeffrey Yang1

3.1  Introduction

       Climate change affects water availability. At the same time, population growth and
redistribution affect the degree and spatial distribution of water demand. This further exacerbates
the stress that climate change places upon water resources. Most of all, the Nation's regulatory
programs, such as those impacting ecological stream flow and Section 305 of the Clean Water
Act (CWA), are overriding considerations for sustainable water resource development, especially
in water-stressed regions. The confluence of these factors can impact the relationship between
water and energy and affect development of sustainable energy production in the future.
       Climate is a statistical term that describes the mean and standard deviation of "weather"
over a period of time - usually 30 years as defined by the World Meteorological Organization
(WMO). This time frame is compatible with long-term planning for energy production, and
therefore, the impact of climate change on water resources is a necessary element to consider
during such planning. Two terms are pertinent to this discussion and require clear definitions
given the similarity of their meanings: climate variability and climate change. Climate variability
refers to temporal variation of the climate mean and other statistics due to internal variability of
the climate system, or the external variability in external forcing. Climate  change is defined as a
statistically significant difference in the climate mean or variation over an extended period of
time as a result of natural or man-made climate forcing. In other words, climate variability
pertains to a hydroclimatic state (mean and variation) over a time period, whereas climate change
is related to a significant change of this state. Both terms are relevant and  should be
distinguished for the analysis of impacts of water resources on energy and transportation fuel
production.
       Climate change affects the interaction among airshed and watershed processes, exerting
impacts on water quantity and quality, air quality, and consequently influencing the ways in
which energy production will adapt in the future. There are numerous publications  in the
literature that document the nature of potential impacts, specifically in watershed hydrology
sectors such as stream flow, storm runoff, and water quality. Impacts on ambient air temperature
and ecological conditions in wetlands and coastal regions have been documented in prior EPA
and other assessment reports (USCCSP, 2008; U.S. EPA, 2009).
       This  chapter begins with a general assessment of hydroclimatic changes for which the
historical precipitation records spanning approximately  100 years were analyzed and the results
of climate change projections are summarized. Details of this historical precipitation analysis and
climate model projections can be found in a companion EPA report (U.S.  EPA, 2014a,b).
Subsequently, a section of this chapter is devoted to describing population change and spatial
shifts in the U.S. that have  resulted in significant land use changes. The combined effects from
climate change and land  use changes affect surface water flow and water availability at
   U.S. Environmental Protection Agency - National Risk Management Research Laboratory


                                           24

-------
watershed scales. Three examples are used to illustrate the projected future watershed river flow
changes that should be considered for electric power production. They include the Colorado
River basin and its Lower Virgin River tributary, the Mississippi River basin and the agricultural
Little Miami River watershed. As shown in these examples, climate-induced hydrological
changes appear to have significant potential to limit energy production in at least two ways: 1)
stream flow decreases in the U.S. will have implications with respect to cooling water discharge
from thermal electric power plants, and 2) shifting regional precipitation has an effect on
domestic biomass production for transportation fuels. Details of thermoelectric power production
and biofuel production processes are described along with quantification of their impacts on
water resources in subsequent chapters of this report.

3.2   Climate change projections and water availability outlook

3.2.1   Climate systems controlling U.S. precipitation and water availability
       Two major aspects of climate dynamics are affecting continental precipitation and water
variability in the contiguous U.S. One is the manner in which ocean-origin climate systems are
coupled to inter-annual, decadal and multi-decadal precipitation variability. These systems
operate on both continental and global scales. The other is the manner in which regional and
local factors are responding to planetary boundary feedbacks, including the effects of large
topographic features and land use in North America. In the next 30 to 50 years, local climate
forcing from topography, surface water bodies, and major categories of land use will remain
relatively unchanged, given no significant and disruptive human-environment interactions. The
interactions will continue into the future, and will superimpose onto the effects of large-scale
climatic systems.
       The resulting precipitation variations are shown through variables such as precipitation
intensity, frequency, seasonality, and total amounts. These variables all affect water availability
in surface water flow and groundwater levels, timing of snow melt, and are thus  consequential to
energy production processes. Precipitation variability can be analyzed mathematically by using
wavelet patterns and analysis methods (Torrence,  1998; Keener, 2010) in order to reveal the
spatiotemporal properties of the precipitation. Details of a precipitation variability wavelet
analysis and analytical results will be presented in the following sections of this chapter. The
major climate systems affecting continental precipitation are summarized in the following
subsections.

   3.2.1.1  Continent-Ocean interactions

       The schematic in Figure 3.1 illustrates major climate systems and regional/local climate
factors that influence synoptic-scale precipitation in the contiguous U.S. In a physiographic
setting, the U.S. is bounded by the Pacific Ocean, the Atlantic Ocean, and the Gulf of Mexico.
As such, synoptic precipitation is influenced by large-scale climatic systems, including the El-
Nino  Southern Oscillation (ENSO), Pacific Decadal Oscillation (PDO), North Atlantic
Oscillation (NAO) and Atlantic Multidecadal Oscillation (AMO) systems  (IPCC, 2007; Durkee
et al., 2007; USCCSP, 2001). Arctic Oscillation (AO), Aleutian Low, surface albedo, and other
                                           25

-------
                                  P-V
                                   P-IV IVb
                                     '
P-lll, 1Mb
                                                                     P-VI
                                                                                               P-ll, lib
                                                                                                             P-l
                        ENSO, PDO, etc.
                   w
                                                      ENSO, PDO, etc.
                                                                                                     NAO/AMD
/ TV
Pacific
W
T x^
	 — - >
ENSO
Convective / /
storm r
if
Lake effect
1*
^y ii .
Great Lakes
7
f ENSO, Gulf
moisture stream
Orographic
precipitation

                                                                                                                 Atlantic
to
Figure S.frefidoletnatic illustration of major climatic systems interacting with local climatic factors producing precipitation
       characteristics in each of the hydroclimatic provinces. Note: Ocean-originated synoptic climate systems are marked
       by blue arrow lines, while land and topography-related precipitation is marked by brown arrows.

-------
climate variables further add complexity and variability in temporal and spatial precipitation
distributions.
       These climate systems operate in different regions of the U.S. (Figure 3.2) and produce
periodic variations (e.g., 3-5 year cycle, multi-decadal) in precipitation intensity, water
availability, and to some degree, drought occurrence and  frequency. A series of climate studies
using atmosphere-ocean (A-O) general circulation model (AOGCM) simulations investigated
how these A-O interactions will change under future anthropogenic emissions scenarios (IPCC,
2007; IPCC, 2013). As shown in Figure 3.2, the average bias in mountainous western North
America is >60% on an annual average basis and nearly 200% during the summer (JJA). The
spread  of the bias and probability curves are the smallest  for the central North American region
that has uniform topography and consistent precipitation wavelet spectra. DJF, MAM, JJA, SON
are month abbreviations denoting the winter, spring, summer and autumn seasons, respectively.




^p
^
g
CD
s
D.



^^
o^
•- — '
-^
3
CD
"§
D.





^P
?_,
>,
=
CD
2
D.


1 UU


80


60
40
20
g

— • — DJF
— •- - MAM

— •-- JJA
— •-— SON
9 • O •^^-»

/ /^ -^ — ^^

0^ O' 9^^^
/ / / (
— O — Annual / 1 /
_
_
/Jl/
•I •Pfr
/ / /. MMD Bias
,'l/_/;#^ Western North America
-100 -50 0 50 100 150 2C
1 00
80


60

40

20
g
—m — DJF
- — •— MAM
— •-- JJA
,lf/
fl/Y
rvr
— •- SON 1 I jl
— D— Annual / Jill
	 T^Wr 	
/

/ ']/ 1
Tfi
' MMD Bias
J /^u Eastern North America
-100 -50 0 50 100 150 2C
1 00

80


60

40

20
g
-^~ DJF f ft ^
—a- - MAM
/ ,

— ^ — ^^^ Br £vM
— B— SON jl
- O — Annual //fl //
jL' M
fil
1 Jl
Eb OQ
/
n
MMD Bias
Central North America
-100 -50 0 50 100 150 2C
B\as %
Figure 3.2   Multi-model dataset (MMD) bias compared to observed precipitation (Xie and
      Arkin, 1997; U.S. EPA, 2014a,b) in control runs (1980-1999) for three North America
      regions.
                                          27

-------
       In the long term, elevated green-house-gas (GHG) levels are projected to intensify A-O
interactions and produce greater impacts on climatic systems such as ENSO. For example, an
enhanced ENSO system would further decrease water availability in an expanded area of the
U.S. southwest, increase the frequency of high-intensity precipitation in U.S. northeast and
Midwest, and reduce snow pack in U.S.  northwest while consequently increasing drought
prevalence in the region (IPCC, 2007; IPCC, 2013).

   3.2.1.2 Regional and local factors

       While synoptic precipitation variations respond to larger-scale climate dynamics or
climate state (e.g., IPCC, 2007; Barsugli et al., 2009), many precipitation changes tangible to
water resources are related to local and regional factors. Examples can be observed in convective
precipitation in the Great Plains (Weaver and Nigam, 2007), dynamic uplifting and rain shadows
in the interior of South Carolina (Konrad II, 1997; Changnon, 2006), and regional synoptic
patterns of short-duration (e.g., 24 hours) precipitation due to orographic uplifting in the coastal
region of the state of Washington (Wallis et al., 2007). Yang et al. (2008) showed a localized
increase of high-intensity, 75% upper-percentile, 24-hour precipitation in the Lower Mississippi
River Basin (LMRB), and attributed it to topographic influence. Details of these findings will be
summarized in a companion EPA report (U.S. EPA 2014b).
       The IPCC (2007) stated that AOGCM models are of value in representing the general
features of future continental precipitation regimes. However, AOGCM projections have
substantial uncertainty at local watershed scales. Examples include poor model performance with
respect to high altitude orographic precipitation in the Appalachian Mountains (McKenney et al.,
2006; Konrad II, 1997), convective precipitation in the Great Plains (Ruiz-Barradas and Nigam,
2005; Higgins et al.,  1998), albedo effects of snow, forecast, aerosols, and other local factors
(Roesch, 2006; Nijssen et al., 2003), and the timing and spatial distribution of climate systems
related to sea surface temperature (SST) anomalies (IPCC, 2007 Chapter 8, and references
therein).  These difficulties are in part due to insufficient spatial resolution of the AOGCM model
grids in fully representing the topographic forcing, high-altitude surface albedo of snow packs as
well as other climate mechanisms such as soil moisture contribution, low-level clouds  and
aerosol radiative forcing.
       To improve general circulation model (GCM) predictability, individual climate model
runs can be combined into averages. The 21 AOGCM model ensemble, referred to as the multi-
model dataset (MMD), hosted at the Program for Climate Model Diagnosis and Intercomparison
(PCMDI) has been used to assess future changes in precipitation. The MMD still has significant
uncertainties in projection, as shown by  the model bias. The model bias is the percentage
difference between the GCM calculated  precipitation and the observed precipitation data of Xie
and Arkin (Xie and Arkin, 1997) for western North America (WNA), eastern North America
(ENA) and central North America (CAN). It reflects how GCM simulations are capable of
representing the regional and local climate factors. In Figure 3.2, the bias distribution clearly
shows that large model  over-predictions of 28 to 93% in mean precipitation occur for all  four
seasons within the WNA model domain. This regional model limitation is also shown by the
large spread of probability curves for all four seasons. The discrepancy can be much larger for
daily or monthly precipitation at a single location than for the seasonal average over the entire
                                           28

-------
WNA model domain. Comparatively, the MMD outputs are best for the CAN model domain,
with the average model mean bias ranging from 8% to 16%. Less robust are the model
predictions with a bias of-4% to 21% in ENA (Figure 3.2), which encompass the Appalachian
Mountains and the Northeast U.S. These results, as summarized by the IPCC (IPCC 2007),
reflect model inadequacy for a full representation of ENSO, PDO and NAO periodicity and
magnitude variations, as well as the Hudson Bay and Canadian Archipelago system, tropical
cyclones and landfall precipitation connections to the Labrador-Arctic climatic system in
northeastern U.S., snow albedo feedbacks and climatic variations in high-altitude mountain
regions.

   3.2.1.3  Generalized precipitation changes in regional scales

       Given the magnitude of GCM model bias, there have been extensive efforts to improve
the reliability of climate projections. Such efforts are important to hydrological applications,
including water resource planning for energy and fuel production. Among several approaches for
model improvement, dynamic climate downscaling in regional climate modeling (RCM)
incorporates regional and local climate factors in climate simulations. However, because of the
computational intensity of dynamic RCM simulations, significant, widely applicable
breakthroughs or improvements are unlikely to occur over the next 10 to 15 years (Barsugli,
2009).
       Another technical approach for future climate projection is statistical downscaling. The
GCM output of a future climate  state is converted to regional-scale conditions using a statistical
converter to correct the GCM bias against known local precipitation records. The bias-corrected
RCM results are often used for water resource planning (e.g., large scale applications in the U.S.
northwest and California). Statistical analysis of long-duration historical records often allows one
to define specific regional and local climate factors that control precipitation variability. The
results can further help define statistical downscaling and help predict precipitation and
hydrological changes over the next 30 years. One such investigation has been carried out at the
EPA for the contiguous U.S., as described below in  Section 3.2.2.
       There are several generalized trends found in projected precipitation changes due to
climate change that can be recognized from the investigations conducted so far. A summary
review of climate model simulations can be found within IPCC reports (IPCC 2007, 2013).
Below are notable features in the model outputs related to precipitation changes in the U.S.:

•  Large degrees of precipitation decreases over 10% are projected for the U.S.  southwest,
   including Arizona, New Mexico, western Texas, Nebraska and Nevada. Snow precipitation
   will likely change to rainfall, resulting in large decreases in snow pack and consequently
   stream flow and seasonal variation. The decrease in the  flow of the Colorado River, which
   started in the last century, will continue and intensify in future decades  according to work by
   Woodhouse et al. (Woodhouse, 2006) using tree ring data.

•  Continued precipitation decreases and intensified droughts are projected for Florida, Georgia
   and the neighboring states further north along the Atlantic sea board.

•  By contrast, precipitation in  the form of increased downpours and flash floods are very likely
   in the U.S. northeast. Similar trends are likely in the Great Lakes region and the  Ohio River
   valley.


                                           29

-------
•   Changes in seasonal precipitation and forms of precipitation in the U.S. north plains will
    likely continue. The changes will likely impact the hydrology and pollutant transport of the
    Mississippi River system (e.g., Scavia, 2003; Jager, 2014). This trend is already evident
    within Lower Mississippi River hydrological studies (Yang, 2008) that are detailed in the
    accompanying National Assessment of Water Infrastructure Adaptation report (U.S. EPA,
    2014a).

3.2.2  Observed climate changes and their impacts on water availability

   3.2.2.1  Changes in historical precipitation

       A detailed analysis of historical precipitation records was conducted for 1,207 monitoring
stations within the U.S. historical climatological network (USHCN) over the entire contiguous
U.S. The investigation was based on a total dataset of 129,288 station-years, or an average record
length of 107 years per station. The long time-scales of the datasets were necessary to ensure that
historical records were long enough to capture climate variation and climate change. Such
datasets can be then used for statistical downscaling and analysis of long-range variation trends.
       The historical precipitation frequency, intensity, spatial distributions and their variations
through time have been examined by using linear regressions and frequency-domain de-noise
operations10 (Torrence, 1998; Keener, 2010). The frequency and variation pattern analysis led to
the delineation of six hydroclimatic areas across the U.S. (Figure 3.3): Florida and the Southeast,
the Lower Mississippi - Ohio River valley - New England region (LONE), the Great Plains and
Midwest, the Basin-and-Range region, the West Coast and the Great Lakes provinces. In
addition to the spectrum analysis, linear regression of the monthly average precipitation against
time yielded a rate of change for each station. The obtained slope, a, was further normalized by
the precipitation in 1950 at the same location. The normalized precipitation change rate, R
(cm/month), was then calculated, and was used to compare the precipitation changes for all
stations of various hydroclimatic settings.
       Overall, historical precipitation increased slightly to 0.083% per year over the entire
contiguous U.S., or 0.079% per year relative to 1950.  The change varies by over one order of
magnitude among parts of the country. Specifically, each hydroclimatic province has unique
precipitation variability and unique long-term changes:

•   Province One is located in Florida and the Southeastern U.S. This province has strong
    decadal and multi-decadal precipitation variability. Its long-term precipitation changes are
    the smallest among all of the analyzed regions, with a markedly slight historical precipitation
    decrease of -0.004% per year relative to 1950.

•   Province Two consists of two sub-regions: P-II and P-IIb (Figure 3.3). P-IIb covers the lower
    portion of the LMRB. Precipitation has increased there over the past several decades
10 The frequency-domain de-noise operation is a type of wavelet spectrum method for
identification of underlying trends in noisy datasets. It is based on frequency variations in
datasets to reduce unimportant noise levels. See references cited for more details.

                                            30

-------
                             m.K*. t*    \  ^T-W
                              * •     l^ZtS""" -
                                        •-
        NHCN Station
        Hydroclimatic Province
       Change Rate (cm/yr)
  Decreases         Increases
                   •   0.001
                   •   0.005
 •   0.001
 •   0.005
 •   0.01
•  0.05
      0.1
0.01
0.05

0.1

                                                                                        0  250  500
                                                                                        I—(—i—i—I—i
Figure 3.3. Hydroclimatic provinces and extreme precipitation changes in the contiguous U.S. Adopted from U.S. EPA
       (2014a,b). Note that the spatially consistent large precipitation increases in the 90% percentile are congregated in the
       Mississippi River valley and northern California Central Basin. Precipitation decreases are primarily in the Great
       Plains, Basin and Range region, U.S. southwest, the Appalachian mountain area and Florida. All analysis was based
       upon long-duration climate records.

-------
   at an overall rate of 0.121% per year for 90% of the stations. The rest of the P-II region
   covers much of the U.S. east of the Mississippi River, including the LONE region.

•  Precipitation in this area is enhanced by moisture movement from the Gulf of Mexico, and
   the intensity of precipitation will likely increase as temperatures warm in the future.

•  Province Three is a large region that includes the Great Pains and the Midwest (Figure 3.3).
   Historical precipitation shows characteristics of the ENSO-enhanced three to five year
   periodic variations with PDO-related decadal variation. In addition to the ENSO and PDO
   related variability, several regions within Province Three show long-term precipitation
   increases, possibly as a result of convective precipitation and moisture movement from the
   Gulf of Mexico. On the contrary, a large region in Province Three centered at the Iowa-
   Nebraska region has had a persistent decrease in precipitation. Similarly large precipitation
   decreases have been recorded for the southwest U.S. in the transition zone known as P-IIIb.

•  Province Four covers the basins and range region of the U.S. that includes most of the
   noncoastal western U.S., the Rocky Mountains and adjacent areas (Figure 3.3). Precipitation
   in the region is affected by PDO teleconnection and topographic forcing, creating local
   variability. Historically, precipitation has increased in local water bodies such as the Great
   Salt Lake and along moisture channels such as the Snake River Valley. Other parts of the
   province have experienced a steady decline in precipitation and persistent drought conditions.
   Woodhouse et  al. (2006) conducted a detailed study of the historical river flow in the
   Colorado River using a tree ring based paleohydrological reconstruction method. The results
   indicated that river flow decreased by 75% following a preceding peak flow; river flow is
   currently following a flow-decreasing trajectory.

•  Province Five consists of the U.S. west coast region (Figure 3.3), and is distinguished by a
   persistent decrease in precipitation in the south and an  increase in the U.S. northwest in
   recent decades. Precipitation variability is marked by strong 12 to 15 year cycles, while high-
   intensity precipitation is consistent with the ENSO cycles of the three to five year time
   interval. The frequency  and intensity of ENSO-related downpours have shown a steady
   increase in recent years.

•  Province Six covers the Great Lakes region, which has the largest rates of precipitation
   increase since 1950.  The rate of long-term increase is 0.13±0.183 %yr-1 (fn ± la), and the
   trimmed mean  is 0.126 %yr"1. Like P-IIb, the province is adjacent to large water bodies.

   3.2.2.2   Extreme precipitation changes

       As shown in Figure  3.4A, many stations have reported extremely large rates of
precipitation decrease and increase. This subset of changes found from within datasets of over a
100-year period is  significant as they reflect the underlying causes with respect to climate
factors. Furthermore, those within the 90% and 10% population percentile represent the
extremely high and the extremely low precipitation within each hydroclimatic province,
respectively. The extreme precipitation  areas delineated from all datasets are shown in Figure
3.3.
California and U.S. Southwest - Synoptic-scale precipitation extremes can be identified in
several large areas in P-IV and P-V (Figure 3.4A). Extreme precipitation in the 90th percentile


                                           32

-------
increased in areas of the Central Basin, northern California, Oregon and the Sierra Nevada. To
the southeast, a large rate of precipitation decrease is identified within the 10th percentile
distribution of the data. This area, vulnerable to drought occurrence, includes southern
California, most of Arizona, western and northern New Mexico, part of Nevada and Utah. In
                      ™.~
                            .//".
               (A)
               Observed Long-term
               Precipitation Changes
                                   .   llib
Long-term precipitation increase
Long-term precipitation decrease
10th percentile increased rate of change
10th percentile decreased rate of change
Boundary
(B)
2012-2017 Projected
Population Change
Source: ESRI, 2013
• > 1.5 %
• 1.0-1.5%
0 - 0.9%
<0%
'Vv-b Boundary
Figure 3.4   Spatial distributions of long-term precipitation changes and population change in
       the contiguous U.S. (A). Areas of long-term precipitation decrease (red) and increase
       (blue) delineated from spatial aggregation of changes in Figure 3.3. Detailed
       information on hydroclimatic provinces is available in a separate EPA report (U.S.
       EPA, 2014a);  (B). The ISRI population data was for 2009-2014 on a county scale. Red
       lines mark the boundaries of hydroclimatic provinces.
                                            33

-------
addition, high altitudes stations in the mountain ranges such as the Wasatch Range also fall into
the 10th percentile displaying a high rate of precipitation decrease.
Great Plains and Gulf Coast - In P-III and P-II, a large-scale distribution of extreme rates of
precipitation change is observed in Texas, Oklahoma, the Great Plains and the Midwest. In
Figure 3.4A, the 90th percentile for a high rate of precipitation increase is found in areas
extending from the Texas Gulf coast near Houston, and north-northwestward into Oklahoma and
central Kansas. The area's northern extension is limited to immediately south of an elevated
topography in the Smoky Hills and by the Smoky Hill River. Further to the north, another area of
the 90th percentile precipitation increase extends from Lake Superior westward into southern
Wisconsin, northern Iowa, South Dakota and northwestern Nebraska (Figure 3.4A). Between the
two areas are the stations, mostly in Nebraska, that display high rates of precipitation decrease in
the 10th percentile. This area encompasses northern Kansas and eastern Nebraska along the Platte
River in Omaha and the surrounding vicinity.
Eastern U.S. o f P-II and P-III - The eastern U.S., including the Northeast, only contains climatic
stations of regional distribution marked with large rates of precipitation decrease in the 10th
percentile (Figure 3.4A). One prominent case is centered within the Appalachian Mountains in
the Blue Ridge region of South Carolina and North Carolina, extending to the Appalachian
foothills of Tennessee, Kentucky and northern Georgia. To the north in New York, northeastern
Pennsylvania and New Jersey, a cluster of stations with the rate of precipitation decrease fall into
the 10th percentile. This area includes the Catskill Mountains—a headwater region for water
supply for the City of New York. At station NY306774 in Port Jervis (NY) south of the Catskill
Mountains, for example,  the normalized rate of precipitation  change is -0.131% yr"1 over the past
120 years. In the Catskill region, Burns et al. conducted a detailed hydrological investigation
using statistical analysis of air temperature, precipitation and stream flow measurements (Burns
et al., 2007). Their data segment covers the period 1952-2005, for which the analysis clearly
showed increased precipitation under a warming local climate.
Florida and Southeast Coast - In P-I, extreme monthly precipitation shows no spatial
association. Most stations in either the 90th or 10th percentile  range are scattered all over the
province (Figure 3.4A). This absence of geographic association over a large area is consistent
with the relatively homogeneous topographic terrain of the province (See Figure 3.3),  which is
affected by climate systems that originate from the Atlantic Ocean and the Gulf of Mexico.

3.2.3  Changes in river flows and river basin hydrology

   3.2.3.1  General understanding of river flow changes and water availability

       Climate model ensembles such as MMD provide general trends in precipitation changes
and their spatial distribution. These results are in agreement with statistical analyses of historical,
long-duration hydroclimatic records, while the latter provides more specific information on the
regional and local characteristics with quantifiable uncertainties for water resource planning.
Overall, likely future changes follow a general trend that regions with drier climates will
experience more intensified drought, while the spatial distribution will likely expand and the
intensities of precipitation events will likely increase. More specific climate model projections
                                            34

-------
are available from the CMIP5 project11, and is compiled by the National Oceanic and
Atmospheric Administration (NOAA) from sources both within NOAA and from other U.S.
Federal agencies12. Furthermore, the long-duration hydroclimatic data described in Section 3.2.2
can be used for climate model verification, and in some cases may be useful for local climate
model downscaling. More information on model projections and how the information can be
used for climate change adaptation planning and design can be found in the accompanying EPA
reports on water infrastructure adaptation (U.S. EPA, 2014a,b).
       Changes in regional precipitation, and to a lesser extent changes in regional temperatures,
occur as a result of climate change. These changes, in turn, lead to changes in regional
hydrological cycles, such as stream and river flows and the levels within water reservoirs, and
therefore changes in water availability. Resulting changes in water availability can limit the
choice of energy and fuel production because of constraints on 1) direct water usage for energy
production such as evaporative cooling in thermoelectric power generation, and 2) water
requirements for biomass growth within bioenergy production processes such as the production
of ethanol and biodiesel transportation fuels. The later sections of this report specifically describe
these water-energy relations. With respect to hydrological responses to future climate and
precipitation changes, there exists a wealth of literature that includes quantitative analysis in
watershed and river basin scales. Examples of the studies conducted by EPA are described in
subsections below. General trends are:

•  Colorado River and its river basin in the U.S. southwest - As precipitation is expected to
   decrease substantially, a large decrease in river flow is very likely. The hydrological
   modeling based on the A1B emission scenario (TPCC, 2007) and projected land use changes
   in the Lower Virgin River Basin (LVRB) shows a likely 35% river  flow decrease by 2050.
   The Lower Virgin River is a large tributary of the Lower Colorado River that provides water
   inflow into Lake Mead. Because LVRB has similar land use and physiographical properties
   to the Colorado River basin, projected river flow changes in LVRB are likely indicative of
   general trends in river flow and, generally, in water availability in the basin.
          This water-poor region will likely experience increasing competition for water
   allocation to meet environmental requirements for minimum ecological stream flow, often
   quantified  on a basis of a seven day average of 10 year return flows (7Q10). Water
   withdrawal for municipal usage and agricultural irrigation are among the largest water
   withdrawals in the U.S.  The land use and population changes in the Colorado River basin and
   the U.S. southwest will further limit water availability for meeting energy production needs.
   These land use and  population changes will be described in further  detail in section 3.3.

•  Great Plains and the U.S. northwest - These large geophysical provinces host several river
   basins with concentrated power and energy biomass production. These include the Snake
   River - Columbia River basin and the upper Mississippi river basin. Significant future
   changes include the seasonality and timing of river flows due to the change of precipitation
   forms (e.g., snow versus rain), the timing of snow melt during the spring season, and the
   increased intensity of rainfall precipitation. All these factors can affect water availability.
       11 http://cmip-pcmdi.llnl.gov/cmip5/data_portal.html
       12 www.climate.gov
                                           35

-------
   Ohio River basin and lower Mississippi river basin - River flows in these two river basins are
   likely to remain stable or increase. The precipitation in the Lower Mississippi River (LMR)
   region has increased in the past 60 years, leading to an increase of river flows in major LMR
   tributaries such as the Yazoo river basin in Mississippi, Alabama, Texas and Oklahoma.

   River systems in Florida and U.S. southeast coastal region - The river systems of the Florida
   Peninsula, those  in the Atlantic coast of Georgia and South Carolina, and those along the
   Florida Gold coast are likely to experience declines in base flow. Water availability in the
   region is likely to decrease as this hydroclimatic province is the only one showing a
   consistent, decades-long decrease in precipitation.
   3.2.3.2
Examples: Stream flow change in Lower Virgin River
       The Lower Virgin River (LVR) watershed region has a typical semi-arid climate in the
U.S. west. The river flows into Lake Mead at the confluence of the Colorado River. Lake Mead
is the second largest human-made freshwater reservoir in the U.S., and is an indispensable source
of freshwater supply for millions of people in the American southwest. The lake provides 90% of
the freshwater supply for the Las Vegas metropolitan area. Overall, precipitation in the LVR
watershed varied with no clear trend for the past 50 years. Water discharge from the LVR
watershed and the main stem of the Colorado River has decreased downstream into Lake Mead.
The decrease accelerated after 1999, resulting in large decreases in the water volume of Lake
Mead (Figure 3.5). The amount of water in Lake Mead in 2010 reached its lowest level since
1940. The largest drop in water volume occurred from 1998 to 2010 (from 30.79 to 13.27 km3).
       Future river flows have been examined by an integrated hydrological simulation that
combines land use modeling and climate change for the LVR watershed under future emission
storyline IPCC Bl (IPCC, 2007). In this quantitative study, land cover variation of the river basin
in 2030 and 2050 are projected using a cellular-automata Markov Chain (CA-MC) land use
model. A cell-based hydrological model  in high spatial resolutions (Chen et al., 2014) is used to
project river flows using future precipitation and land use as input parameters. Model-simulation
results  clearly suggest large and differential changes in river discharge in the future. There will
be different watershed hydrologic responses between summer dry seasons and winter wet
seasons, and among the climate and land cover change scenarios. Under the IPCC Bl emission
scenario, future temperatures will increase both in summer and in winter. The projected
precipitation will increase in summer but decrease in winter. When only future climate change is
considered, the projected total discharge of the LVR for the three winter months (December,
January and February) would be 6.74 m3/sec in 2029-2030, and 5.98 nvVsec in 2049-2050. This
represents a decrease of 64.82% and 68.79%, respectively, from the 2009-2010 levels. In
summer, except for the month of August, the projected discharge will increase, and the rate of
increase will decline in the two decades from 2030-2050 when compared to the preceding two
decades.
       When the combined effects of climate and land cover changes are considered together,
the amount of river discharge will decrease in the winter. The largest decrease may occur in
January 2050 by as much as 75.4% (Chen et al., 2014). The river will most likely be drier in
winter, and the problem of water shortage will likely be aggravated in the region.
                                          36

-------
       The model projections are consistent with the paleoclimatological investigation results.
Woodhouse et al. reconstructed historical flow variation in the upper Colorado River at several
locations, including one at Lees Ferry, Arizona, approximately 150 km east of the LVR
watershed (Woodhouse, 2006). Their results showed large variability of yearly Colorado River
flow in the last 50 years. In the latest cycle prior to 1970, the river flow decreased by 80.5% in
approximately 40 years. This rate is comparable to a maximum of 75.4% reduction in LVR
discharge that our cell-based model has simulated under climate change conditions of the next 40
years.
               Os ON  ON ON  ON ON  O\ ON O> ON O\  ON ON  ON ON ON ON ON  ON O\  O O  O O O
                                             Year
                                                           ON ONONONOOOOOO
                                             Year

Figure 3.5 Temporal variations of (A) water volume in Lake Mead; (B) annual precipitation
       in the LVR basin.

-------
   3.2.3.3 Examples: Water quality changes in the Little Miami River basin

       The Little Miami River (LMR), a major tributary of the Ohio River, originates southeast
of Springfield in southwestern Ohio. It flows 169.78 km from Clark County through several
steep-sloped, forested gorges to join the Ohio River at its confluence in Hamilton County, near
the eastern edge of Cincinnati. Draining an area of 5840 km2, the watershed covers Clark,
Greene, Warren, Clermont and Hamilton Counties, as well as portions of Montgomery, Clinton,
Brown, Highland and Madison Counties in Ohio (long et al. 2012). Agricultural farm lands and
forests are the two largest land use types.
       A hydrological simulation conducted by EPA (long et al., 2012) evaluated the future
stream flow and water quality changes in the LMR through 2050. The simulation is based on
land use projections in the region using the CA-MC land use model that also considers
population growth as a driving force for urbanization and suburban sprawl. The population for
year 2050 was estimated using a geometric growth model, and the population density for each
census block group in the study area was calculated. These results were then incorporated into a
suitability map for urban areas through the Multi-Criterion Evaluation (MCE) simulation module
in GIS (Tong et al., 2012) to produce a population filter, a surrogate depicting the impacts of
population growth on land use changes.
       The simulation was focused on both changes in river flow and the concentrations of
nutrients (e.g., phosphorus and nitrogen). In the  LMR watershed, the calibrated model yielded
projections for large changes in river flow and water quality for the analyzed precipitation
scenarios. Notable conclusions from 2012 EPA/Tong et al. LMR study include):

•  The combined climate and land use changes could significantly impact LMR flow. With no
   climate change, the flow is projected to increase from 23.3 to 41.7 m3/s, or by 29.09%. In wet
   and dry seasons, the combined effect on river flow could be an increase of 16.22% to 43.83%
   and a decrease of 43.92% to 53.08%, respectively.

•  Water quality will likely deteriorate in the future. Total phosphorus and total nitrogen could
   increase in all of the analyzed land use and climate change scenarios. Maximum changes in
   phosphorus and nitrogen are projected to be  11.99% and 7.22%, respectively.

   3.2.3.4 Examples: Mississippi River flow and nutrient loading

       The Lower Mississippi River basin (LMRB)  covers an 181,600-km2 area consisting of
Louisiana, Mississippi, Arkansas, Missouri, Kentucky and Tennessee. The basin consists of nine
sub-basins following its tributaries including the Atchafalaya River, the Arkansas River, the
Ouachita River, the Red River, the Yazoo River and the Big Black River. The area,
geographically known as the Mississippi Embayment (van Arsdale and Cox, 2007), is
characteristic of the low-lying, NNE to SSW trending basin that traces the nearly 1600-km long
Lower Mississippi River. The embayment topomorphology has a U-shaped cross-section and
gentle slopes along the center of LMR valley. Because of the geophysical characteristics, the
small longitudinal hydraulic gradient in the river has caused frequent channel migration,
numerous oxbow lake formations and  flood occurrences (Smith,  1997).
       Comprehensive hydrological studies on precipitation and flood occurrences have shown
the prevalence of local precipitation variability in the LMRB, and the synoptic ENSO-related
                                           38

-------
hydroclimatic effects in the form of disruptive meteorological events, river flows and nutrient
fluxes into the Gulf of Mexico. Wavelet-reconstructed river flow variations and the time-
frequency spectrum13 (Keener, 2010; Torrence,  1998) exhibit strong perennial flow variations
with three to five year short-term, high-frequency periodicity, and two distinct periods of
different variability characteristics in the records since 1930s. The three to five year periodicity
in river flow is prevalent and characteristic of the ENSO variability reported previously for rivers
in North America (George, 2006; Makkeasorn et al., 2008; Qian et al., 2007; Zhang and
Schilling, 2006; Hereford et al., 2006).
       The embayment along the LMR valley induces strong north and northeast moisture
movement from the Gulf of Mexico. As a consequence, increased precipitation over the last
several decades and frequent flooding events responding to disruptive meteorological events like
hurricanes is likely. This long-term change is expected to intensify from the warm air mass
convection driven by southwestern winds because of greater heat content in the atmosphere and
stronger A-O interactions as a result of climate change.
       Unlike the water stressed Basins-and-Range and Great Plains in the U.S. west, the LMRB
is facing different climate change impacts that can potentially place constraints on power and
energy biomass production. Water availability is very unlikely a problem, but water quality
changes are likely a concern for future planning. High-intensity precipitation and flash floods are
known as the leading cause for a high degree  of surface water quality variations. Precipitation-
induced areal floods (Lecce, 2000; Patterson,  1964) have been reported to be responsible for
propagation of water-born biological contaminants (Furey et al., 2007; Muirhead et al., 2004;
Few et al., 2004; Barry, 2002; Curriero et al.,  2007; Dortch et al., 2008; Borchardt et al., 2004)
and high levels of chemical contaminants and nutrients (Donner and Scavia, 2007; Aulenbach et
al., 2007). Therefore, nutrient runoff from  energy biomass production in agricultural cultivation
is particularly problematic because surface water bodies will already be under stress with respect
to water quality.

3.3  Implications With  Respect to Energy Production

3.3.1  Climate considerations for thermal electric power generation

  3.3.1.1   Water availability from mismatched spatial distribution

       Water competition for power generation - Thermal  electric power generation consumes a
large quantity of water (See Chapter 4, and related  sections). This water demand may not be met
because of changes in precipitation and limited water availability due to climate change. The
constraints are particularly acute in the hydroclimatic province P-IV and the southwestern
portion of P-IIIb. Figure 3.6 shows the locations of existing coal-fired power plants in relation to
the hydroclimatic provinces.  It is reported that more thermoelectric power plants are being
planned in the Great Plains and the Basins-and-Range region. There are ten existing coal-fired
thermoelectric power plants located in the  P-IV province along the Colorado River (Figure 3.6).
These are all equipped with closed loop cooling systems, consuming water at a combined rate of
100.4m3/sec.
       13 The analysis combined both the frequency and time domains

                                           39

-------
Figure 3.6 Locations of major coal-fired thermoelectric power plants superimposed upon a map of the hydroclimatic
      provinces of the U.S. Note: Water consumption rates (m3/sec) are for plants in the Lower Colorado River basin.
      Data are from IEA 2005 and 2007 surveys (http://www.iea.org/etp/tracking/figures/power/).

-------
       The water consumption rate is remarkably large given the decreasing flow in the
Colorado River. Water flow in Colorado River decreases along its course through the Basins and
Ranges. Downstream at Topock, Arizona, the low river flow is, at most times, less than the water
consumption from upstream power plants. Figure 3.7 shows historical river flow measurements
from 1917 to 1982 (USGS, 2014). Maximum flows (Qmax) during the observation period were
-1000 m3/sec, and reached nearly 5000 m3/sec in the summer. On the other hand, the minimum
flows (Qmin) during the period averaged 171.9 m3/sec in spring, largely reflecting the snow-melt
release of runoff from the river's source water region. In the fall and winter seasons, the average
low flow was only 65 m3/sec, smaller than the 100.4 m3/sec water consumption rate for upstream
power generation. It is clear from this comparison that the ten power plants' consumption
amounts to a large fraction of river discharge during low flow conditions.
      o
      
-------
plants are located in the water-rich Ohio River basin and Upper Mississippi River basin. Coal-
fired power generation in other regions are also vulnerable to climate changes. These areas
include Florida and the coastal region of Georgia in P-I (Figure 3.6).
       In the western coast P-V and P-IVb hydroclimatic provinces, 2007 IEA data shows only
two coal-fired power plants. The power plants in these regions have been converted to natural
gas, and are therefore not shown in Figure 3.6. As discussed later in this report, natural gas
power plants tend to have smaller water footprints than their coal-fired counterparts. Even with a
small water footprint, however, water impacts on southern California cannot be neglected
because of the expected persistent decrease in future precipitation in the region.
       Competition for water - Additional stress on water availability comes from population
changes and regulation-mandated minimum  ecological  stream flows. These two factors compete
with water usage for thermoelectric power plants.
       In the last five years (2009-2014), the U.S. population has witnessed spatial  redistribution
(Figure 3.4b). The largest population increase occurred in the water-poor regions of P-IV and P-
V, with projected precipitation decreases in the future due to climate change. In Florida and
further to the north in the metropolitan areas of Atlanta, Washington, D.C., and New York City,
the population has also increased markedly. These are areas within the water poor P-I
hydroclimatic province, or areas of extreme precipitation decrease.
       In Figure 3.4b, it is clear that the changes in precipitation and population are mismatched
with respect to water availability in the future. Several regions where precipitation has long
decreased, and will likely continue to decrease, host the largest population growth. Such
mismatch is particularly  acute in the Lower Colorado River basin of the U.S. southwest, southern
California and the central Appalachian regions from Atlanta to New York. Population growth in
these regions can result in greater demands for both water and energy, and both can further stress
the local  water resources unless  long-distance electric power transmission and water conveyance
are available at the expense of further energy consumption.
       Ecological flow is the other competing factor for water, and is placing constraints on
thermoelectric power generation, particularly for water-poor regions under likely future climate
scenarios. For example, large decreases in average  stream flow were projected in the Lower
Virgin River of the Colorado River basin, and also  in the Little Miami River basin of the Ohio
River valley (See Section 3.2.3). Similar reductions in river flows have also been projected for
the U.S. northwest region. The decrease in snow pack at high altitudes has been found
responsible for decreasing river base flows, and consequently, the river flows diverted for power
generation year-round become vulnerable.

3.3.2  Carrying capacity for thermal pollution and nutrient loading
       The most direct impacts  of the present and future changes in water availability are made
on the ability of power plants to discharge cooling water and other process water into surface
water (e.g., rivers, streams and lakes) without violation of National  Pollutant Discharge
Elimination System (NPDES) under the CWA. Similarly, energy production operations, such as
hydraulic fracturing for natural gas, may be constrained by reduced flow and assimilation
capacity  of receiving streams under future climatic conditions.
                                           42

-------
       Under Section 305 of the CWA, EPA's NPDES and the Total Maximum Daily Load
(TMDL) programs set up thresholds for thermal and nutrient loading into a segment of U.S.
water. In streams and rivers, this threshold is based on 7Q10 stream flows close to the base flow
of a river or stream. As described earlier in Section 3.2.3, there are many watersheds and
waterways that will very likely experience a flow reduction, nutrient loading increase or both.
The combination can make the water discharge from power and energy plants problematic in the
future. Considering this likelihood, EPA is conducting research and rule-making analyses on the
potential of incorporating climate change effects into water discharge programs.

3.3.3   Implications for energy biomass production
       Energy biomass production for transportation biofuels has its own challenges with
respect to the water-energy nexus that differ somewhat from the challenges of thermo-electric
power generation. Cultivation of biomass crops is land- and water-dependent, while biofuel
production tends to be located in geographic proximity to biomass production. Biomass crop
cultivation is concentrated in the Great Plains and the Upper Mississippi River basin. This spatial
association is a basic  consideration in evaluating climate change effects on the sustainability of
biofuels production.
       Water usage in biomass cultivation has been investigated by EPA using MARKAL
modeling (Dodder, 2011; Dodder, 2014). A separate study of the water needs within this report
includes water demand  from both biomass cultivation and bioenergy production. The largest
water consumption rates for biofuel production are located in Nebraska,  Kansas, Missouri,
Arkansas and Mississippi. Similar but more detailed analyses and conclusions were reported by
Sandia National Laboratory (Tidwell et al., 2011).
       Compared to the precipitation changes in Figure 3.4A, the biomass cultivation areas in
two Great Plains states  (NE and KS) are located in  a region of long-term precipitation decrease.
In this region, the  productive artesian Ogallala aquifer has long been used for agricultural
irrigation (Sophocleous and Marriam, 2012; Tidwell et al., 2011). Precipitation in the aquifer
recharging area in the Rocky Mountains (CO  and WY) has been decreasing (See Figure 3.3). As
a result, the groundwater level has experienced rapid decline. This water availability stress is
likely to worsen in the future. Areal precipitation will very likely decrease, while water loss from
crop irrigation  will increase along with increased biomass production. For these likely future
conditions, the water  constraints on bioenergy production have not been investigated according
to the data and information examined by this study.
3.3.4   Non-point source nutrients in biomass production
       Crop irrigation methods common in the Great Plains and the U.S. Midwest are known to
release nutrients, mainly nitrogen and phosphorus, into water and sediments by non-point source
pollution (e.g., Cunha, 2014;  Yang, 2008, and references therein). Increased biomass cultivation
for biofuels production  is expected to increase nutrients released from agricultural fields. Jager et
al. demonstrated the linkage of energy  crop production to river water quality deterioration within
a major Lower Mississippi River tributary (Jager, 2014).
                                           43

-------
       As described in preceding Section 3.2.3.4, the LMR flows and nutrient loadings varied in
relation to different climate systems. It is known that nutrient loading from LMR contributes to
severe nutrient enrichment and hypoxia in the Gulf of Mexico (Scavia, 2003). From this review
and analysis, biomass cultivation for biofuels will further increase nutrient loads contributing not
only to nutrient levels in the Mississippi River system, but hypoxia in the Gulf of Mexico.

3.4   Adaptation Potential and Conclusion

       In this chapter, the impacts of climate change on water resources were discussed in the
context of future energy  and transportation fuel  production. It provided a brief overview of
attributes related to climate and population changes and the degree of their impacts on water
availability for future energy production. It is noteworthy that the impacts on energy production
are multi-dimensional and  affect not only current energy production, but also future energy
choices and overall makeup. A comprehensive investigation on the energy-water nexus can be
found within recent Department of Energy reports (Tidwell, 2011; DOE, 2014).
       The review and analysis  described herein show that there are substantial, regional climate
change impacts upon precipitation resulting in critical anticipated effects on water availability in
the contiguous U.S. Thermoelectric power generation and biomass cultivation for biofuels will
further stress water availability, particularly in specific regions. This effect is expressed in the
form of water consumption from limited water resources and from competition between energy
production, domestic consumption and ecological flow in rivers and streams. Furthermore, water
availability also limits potential water discharge from power plants into streams and rivers.
Thermal  and nutrient pollution from production water discharge will likely be a concern with
regards to CWA regulations and thus may potentially affect the sustainability of thermoelectric
power and biofuels production.
       It was found from this  review and analysis that climate change effects on water
availability are region and location specific. These effects are further amplified by changes in
watershed hydrology and by population and land use changes. Therefore, adaptation requires
consideration of local and regional conditions for conducting energy production planning and
analysis.

3.5   References

Aulenbach, B.T., H.T. Buxton, W.A. Battaglin,  and R.H. Coupe. 2007. "Streamflow and nutrient
       fluxes of the Mississippi-Atchafalaya River Basin and sub-basins for the period of record
       through 2005." U.S. Geological Survey Open-File Report 2007-1080.
       http://toxics.usgs.gov/pubs/of-2007-1080/index.html. Accessed June 2014.

Barry, J.M. 2002. "The 1927 Mississippi River flood and its impact on U.S. society and flood
       management strategy." Geological Society of America Annual Meeting, Denver, CO.

Barsugli, J., C. Anderson, J. B. Smith, and J. M. Vogel. 2009. "Options for improving climate
       modeling to assist water utility planning for climate change." Water Utility Climate
       Alliance, San Francisco, CA. 144p.
                                           44

-------
Borchardt, M.A., N. L. Haas, and R. J. Hunt. 2004. "Vulnerability of drinking-water wells in La
       Crosse, Wisconsin, to enteric-virus contamination from surface water contributions."
       Applied and Environmental Microbiology, 70(10):593 7-5 946.

Burns, D. A., J. Klaus, and M. R. McHale. 2007. "Recent climate trends and implications for
       water resources in the Catskill Mountain region, New York, USA." Journal of
       Hydrology,  336:155-170.

Changnon, D. 2006. "Regional and temporal variations in heavy precipitation in South
       Carolina." InternationalJournal of Climatology, 14(2): 165-177.

Chen, H., S.T.Y. long, H. Yang, and Y.J. Yang. 2014. "Simulating the hydrologic impacts of
       land cover and climate changes in a semi-arid watershed." Hydrological Science
       Journal, DOT: 10.1080/02626667.2014.948445

Cunha, D.G.F.C., M. Calijuri, and W.K. Dodds. 2014. "Trends in nutrient and sediment retention
       in Great Plains reservoirs (USA)." Environmental Monitoring and Assessment, 186:
       1143-1155.

Curriero, F.C., J. A. Patz, J. B. Rose, and S. Lele. 2007. "The association between extreme
       precipitation and waterborne disease outbreaks in the United States, 1948-1994."
       American Journal of Public Health, 91 (8): 1194-1199.

Dodder, R.S. 2014. "A review of water use in the U.S. electric power sector: insights from
       systems-level perspectives."  Current Opinion in Chemical Engineering, 5:7-14.

Dodder, R., W. Yelverton, T. Felgenhauer, and C. King. 2011. "Water and greenhouse gas
       tradeoffs associated with a transition to a low carbon transportation system." Proc.
       IMECE2011, Denver, CO.

Donner, S.D., and D. Scavia. 2007. "How climate controls the flux of nitrogen by the Mississippi
       River and the development of hypoxia in the Gulf of Mexico."  Limnology and
       Oceanography, 52(2):856-861.

Dortch, M.S., M. Zakikhani, S-C Kim, and J. A. Stevens. 2008. "Modeling water and sediment
       contamination of Lake Pontchartrain following pump-out of Hurricane Katrina
       floodwater." Journal of Environmental Management, 87(3):429-42.

Durkee, J. D., J. D. Frye, C. M.  Fuhrmann, M. C. Lacke, H.  G. Jeong,  and T. L. Mote. 2007.
       "Effects of the north Atlantic oscillation on precipitation-type frequency and distribution
       in the eastern United States." Theoretical and Applied Climatology, doi:
       10.1007/s00704-007-0345-x.

Few, R., M. Ahern, F. Matthies, and S. Kovats. 2004. "Floods, health and climate change: A
       strategic review." Tyndall Centre for Climate Change Research Working Paper 63,
       England.
                                          45

-------
Furey, J.S., H. Fredrickson, C. Foote, and M. Richmond. 2007. "Post-Katrina fecal
       contamination in violet marsh near New Orleans." InternationalJournal of
       Environmental Research and Public Health, 4(2): 84-92.

George, S.S. 2006. "Streamflow in the Winnipeg River basin, Canada: Trends, extremes and
       climate linkages."  Journal of Hydrology, 332:396-411.

Hereford, R., R.H. Webb,  and C.I. Longpre. 2006. "Precipitation history and ecosystem response
       to multidecadal precipitation variability in the Mojave Desert region, 1893-2001."
       Journal of Arid Environments., 67:13-34.

Higgins, R.W., K.C. Mo, and Y. Yao. 1998. "Interannual variability of the U.S. summer
       precipitation regime with emphasis on the Southwestern monsoon." Journal of Climate.,
       11:2582-2606.

Intergovernmental Panel on Climate Changes (TPCC). 2007. "Chapter 8, Climate models and
       their evaluation."  In Climate Change 2007 - The Physical Science Basis: Fourth
       Assessment Report of the Intergovernmental Panel on Climate Change, edited by S.
       Solomon, D. Qin, M. Manning, M. Marquis, K. Averyt, M.M.B. Tignor, H. L. Miller, Jr.,
       and Z. Chen, p.589-662. Cambridge University Press, New York.

Intergovernmental Panel on Climate Changes (TPCC). 2013. Climate change 2013, the physical
       science basis. Edited by T.F. Stacker, D. Qin, G.-K. Plattner, M.M.B. Tignor, S.K. Allen,
       J. Boschung, A. Nauels, Y. Xia, V. Bex, and P.M. Midgley. Cambridge University Press,
       New York.

Jager, H., L.M. Baskaran,  P.E. Schweizer, A.F. Turhollow, C.C. Brandt, and R. Srinivasan.
       2014. "Forecasting changes in water quality in rivers associated with growing biofuels in
       the Arkansas-White-Red River drainage, USA." GCB Bioenergy.
       doi:10.1111/gcbb.l2169.

Keener, V.W.,  G.W. Feyersisen, U. Lall, J.W. Jones, D.D. Bosch, and R. Lowrance. 2010. "El-
       Nino/Southern Oscillation  (ENSO) influences on monthly NOs load and concentration,
       stream flow and precipitation in the Little River Watershed." Journal of Hydrology, 381:
       352-363.

Konrad II, C.E. 1997. "Synoptic-scale features associated with warm season heavy rainfall over
       the interior southeastern United States."  Journal of Climate, 12:557-571.

Lecce, S.A. 2000. "Spatial variations in the timing of annual floods in the southeastern United
       States." Journal of Hydrology, 235:151-169.

Makkeasorn, A., N.B. Chang, and X. Zhou. 2008. "Short-term stream flow forecasting with
       global climate change implications - A comparative study between genetic programming
       and neural network models." Journal of Hydrology, 352:336-354.

McKenney, D.W., J.H. Pedlar, P. Papadopol, and M.F. Hutchinson. 2006. "The development of
       1901-2000 historical monthly climate models for Canada and the United States."
       Agricultural and Forest Meteorology, 138:69-81.
                                          46

-------
Muirhead, R.W., R.J. Davies-Colley, A.M. Donnison, and J.W. Nagels. 2004. "Fecal bacteria
       yields in artificial flood events: quantifying in-stream stores."  Water Research, 38:1215-
       1224.

Nijssen, B., L.C. Bowling, D.P. Lettenmaier, D.B. Clark, M. El Maayar, R. Essery, S. Goers,
       Y.M. Gusev, F. Habets, B. van den Hurk, J. Jin, D. Kahan, D. Lohmann, X. Ma, S.
       Mahanama, D. Mocko, O. Nasonova, G.-Y. Niu, P. Samuelsson, A.B.  Shmakin, K.
       Takata, D. Verseghy, P.  Viterbo, Y. Xia, Y. Xue, and Z.-L. Yang. 2003. "Simulation of
       high latitude  hydrological processes in the Torne-Kalix basin: PILPS Phase 2(e) 2:
       Comparison of model results with observations." Global Planetary Change, 38, 31-53.

Patterson, J.L. 1964. "Magnitude and frequency of floods in the United States Part 7, Lower
       Mississippi River basin." U.S. Geological Survey, Water-Supply Paper 1681.
       Washington,  DC. 55p.

Qian, T., A. Dai, and K.E. Trenberth. 2007. "Hydroclimatic trends in the Mississippi River basin
       from 1948 to 2004" Journal of Climate, 20:4599-4614.

Roesch, A. 2006. "Evaluation of surface albedo and snow cover in AR4 coupled climate
       models." Journal of Geophysical Research, 111, D15111, doi: 10.1029/2005 JD006473.

Ruiz-Barradas, A., and S. Nigam. 2005. "Warm season rainfall variability over the U.S. Great
       Plains in observations, NCEP and ERA-40 reanalyses, and NCAR and NASA
       atmospheric model simulations." Journal of Climate, 18:1808-1830.

Scavia, D., N.N. Rabalais, R.E.  Turner, D. Justic, and W.J. Wiseman, Jr. 2003. "Predicting the
       response of Gulf of Mexico hypoxia to variations in Mississippi River nitrogen load."
       Limnology and Oceanography, 48(3), 951-956.

Smith, L.M.  1997. "Fluvial geomorphic features of the Lower Mississippi alluvial valley."
       Engineering Geology, 45(1-4): 139-165.

Sophocleous, M., and D. Merriam. 2012. "The Ogallala formation of the Great Plains in central
       U.S. and its containment of life-giving water." Natural Resources Research, 21(4): 415-
       426.

Torrence, C., and G. P. Compo. 1998. "A practical guide to wavelet analysis." Bulletin of the
       American Meteorological Society, 79 (1): 61-78.

Tidwell, V., A.C.-T. Sun, and L. Malczynski. 2011. "Biofuel impacts on water."  Sandia Report
       SAND2011-0375, 31p. http://prod.sandia.gov/techlib/access-
       control.cgi/2011/110375.pdf

Tong, S.T.Y., Y. Sun, T. Ranatunga, J. He, and  Y.J. Yang. 2012. "Predicting plausible impacts
       of sets of climate and land use change scenarios on water resources." Applied
       Geography, 32:477-489.
                                          47

-------
U.S. Climate Change Science Program (USCCSP). 2001. Climate change impacts on the United
       States: The potential consequences of climate variability and change. Cambridge
       University Press, New York.

U.S. Climate Change Science Program (USCCSP). 2008. Preliminary Review of Adaptation
       Options for Climate-Sensitive Ecosystems and Resources - Final Report, Synthesis and
       Assessment Product 4.4.

U.S. Department of Energy (DOE). 2014. The Water-Energy Nexus: Challenges and
       Opportunities.
       http ://www.energy. gov/sites/prod/files/2014/07/fl 7/Water%20Energy%20Nexus%20Full
       %20Report%20Julv%202014.pdf. Accessed July 2014.

U.S. EPA. 2009. Synthesis of Adaptation Options for Coastal Areas. EPA 430-F-08-024.

U.S. EPA. 2014a. National Water Infrastructure Adaption Assessment, Part I and II: Climate
       Change and Hydrologic Impact Quantification, and Adaptation Readiness Assessment.
       (in print).

U.S. EPA. 2014b. National Water Infrastructure Adaption Assessment, Part III: Modeling and
       Monitoring Climate Change Impacts at Watershed, (in print).

U.S. Geological Service (USGS). 2014. National Water Information System: Web Interface.
       http://nwis.waterdata.usgs.gov/nwis/dv?cb_00060=on&cb_80154=on&cb_80155=on&fo
       rmat=html&begin date=1935-02-14&end date=1935-02-
       14&site  no=09424000&referred module=sw. Accessed June 2014.

van Arsdale, R.B., and R.T. Cox. 2007. "The Mississippi's curious origins." Scientific American,
       January: 76-82B.

Wallis, J.R., M.G. Schaefer, B.L. Barker, and G.H. Taylor. 2007. "Regional precipitation-
       frequency analysis and spatial mapping for 24-hour and 2-hour durations for Washington
       State." Hydrology andEarth System Science, ll(a):415-442.

Weaver, S.J., and S. Nigam. 2 007. "Variability of the Great Plains low-level jet: Large-scale
       circulation context and hydroclimate impacts."  Journal of Climate., 21:1532-1551.

Woodhouse, C.A., S.T. Gray, and D.M. Meko.  2006. "Updated streamflow reconstructions for
       the Upper Colorado River Basin."  Water Resources Research, 42:W05415.

Xie, P. and P. A. Arkin. 1997. "Global precipitation: a 17-year monthly analysis based on gauge
       observations, satellite estimates, and numerical  model outputs." Bulletin of the American
       Meteorological Society, 78:2539-2558.

Yang, Y.J., R. C. Haught, J. Neal, K. Carlton-Perkins, T. C. Keener, and J.-Y. Lee. 2008.
       "Variations in high-intensity precipitation under climate changes in the LMRB and
       implications for drinking water supply security." Proc. ASCE/EWRI2008 Water
       Environmental and Water Resources Congress, Hawaii.
                                          48

-------
Yang, Y. 2010. "Topographic Factors in Precipitation Dynamics Affecting Water Resource
       Engineering in the Contiguous U.S.: Notes from Hydroclimatic Studies." Proc. ASCE
       World Environmental and Water Resources Congress., Providence, Rhode Island, May
       16-20, 2010. pp. 164-170. doi: 10.1061/41114(371)19

Zhang, Y.-K., and K.E. Schilling. 2006. "Increasing streamflow and baseflow in Mississippi
       River since the 1940s: Effect of land use change." Journal of Hydrology, 324:412-422.
                                          49

-------
 4   Water Impacts from Coal-Fired Electric Power Plants
       Timothy C. Keener,1  Marissa S. Liang,1 Jhosmar Sosa Pieroni,1 Darsana Menon,11 Pratim
       Biswas," Wei-Ning Wang,11 William E. Flatten III,111 Patricio X. Pinto,111 Christopher D.
       Holder1"

4.1  Introduction

       Coal is the most commonly used fossil fuel for power generation, accounting for
approximately 37% of the total electricity produced in the U.S. in 2012 (EIA, 2014a). The U.S.
electric power sector consumed approximately 232 million short tons of coal in the first quarter
of 2014 (EIA, 2014b). Coal consumption by the electricity power sector in the U.S. is projected
to decrease from 45% in 2010 to approximately 35% in 2040, with a corresponding decrease in
electricity generation from 317GW in 2010 to a projected capacity of 278GW in 2040, mainly
because of the need to meet emission control requirements (EIA, 2014c).
       Generating electricity in coal-fired electric power plants involves mining coal, cleaning
or washing the coal to remove impurities, transporting the coal to the power plants and burning
the coal in power plants to produce steam, which is then run through a steam turbine to generate
electricity. Once energy from the steam is recovered, the steam is cooled and condensed back to
water, while other waste products from the combustion process, including bottom ash, fly ash
and flue gas, is prepared for disposal. In this chapter, water usage in each of these processes and
their impact on water is briefly described.

4.2  Water Impacts  from Coal Mining and Processing

       The majority of the coal produced in the U.S. comes from Wyoming (41%), West
Virginia (12%), Kentucky (10%) and Pennsylvania (5%). Coal is obtained through underground
mining or through surface mining techniques  such as open pit or open top, high wall or strip
mining (NMA, 2012). Water is typically not used during coal mining operations. However, water
may flow into mines from groundwater seepages, surface water intrusion or precipitation, which
then dissolves organics or inorganics that are present in the mines, and results in water that is
acidic and has elevated levels of total dissolved  solids (TDS), iron, aluminum, sulfate and other
dissolved metal ions (DOE, 2009). In addition, water that contains elevated TDS may also drain
from active and inactive mines into surface water, ground water or other water bodies.
Depending on the nature  of the ore body and the geology of the mining site, water from active,
inactive or abandoned coal mines, often referred to as coal mine drainage (CMD) or mining
impacted water (MIW), often contains significant concentrations of metals (such as aluminum,
arsenic, boron, cadmium, calcium, chromium, cobalt, copper, dysprosium, gadolinium,
germanium, iron, lanthanum, lead, magnesium, manganese, nickel, potassium, scandium, silica,
samarium, sodium, titanium, yttrium and zinc), oxyanions (such as chromate and arsenites) and
salts (such as chloride, nitrates and sulfate) (U.S. EPA, 2013). This water, which can have TDS
that can vary from very low (<200 mg/L TDS) to very high (>500,000 mg/L TDS), has to be
  University of Cincinnati, Department of Biomedical, Chemical, and Environmental
  Engineering
  Washington University in St. Louis
  Pegasus Technical Services, Inc., Cincinnati, OH

                                          50

-------
treated to meet certain water quality standards (typically, 250 or 500 mg/L TDS) before it can be
further treated or discharged into receiving water bodies such as rivers or streams.
       Multiple treatment technologies, usually classified as active or passive, are available for
the removal of TDS from CMD. Examples of passive treatment technologies, which typically
use natural processes to treat contaminated water, include lagoons, aerobic and anaerobic
wetlands, limestone ponds and drains, and diversion wells. Examples of active treatment
technologies, which typically require continuous input of chemicals or energy to treat the water,
include aeration tanks, neutralization ponds, evaporation, ion exchange, distillation,
electrodialysis, filtration and adsorption. Many of these techniques often use energy and water,
or impact water in some way (e.g., reverse osmosis typically purifies approximately 80% of the
waste stream, while the remaining 20% is concentrated brine that has to be disposed). A
discussion of these impacts is beyond the scope of this report.
       Mined coal is typically cleaned or washed at the mining site or at some intermediate
facility to remove impurities before being transported to power plants. Washing coal facilitates
waste material removal from the mined coal, which, in addition to lowering transportation costs,
can lower its ash content when combusted in power plants. Typically, coal is washed using
density or gravity separation methods that may or may not use  water. Once coal is separated,
water is removed by passing the slurry through dewatering screens, filters, centrifuges or
thickeners, and the recovered water (blackwater) is treated and reused. The amount of water
required for cleaning coal depends on the type of coal, the amount of impurities that are present,
and its intended use, among other factors. Currently, there is no information in the literature on
the amounts of water used for density separation, or the amount of reusable water that is
recovered from blackwater. However, the U.S.  Department of Energy estimates that
approximately 70 - 260 million gallons of water per day is used for coal mining, including the
water used for washing coal and cooling drilling equipment (DOE, 2006). After washing, most
coal is transported to power plants by truck, rail or barges. In a few cases,  however, finely
ground coal is transported via pipelines as a slurry, which involves the use of large amounts of
water. Meldrum et al. estimates that transportation of coal via pipelines involves the use of
hundreds of gallons of water per megawatt-hour (MWh) of electricity produced (Meldrum,
2013).

4.3   Water Impacts from Thermoelectric Power Systems

4.3.1   Steam turbines
       The amount of water a thermoelectric power plant requires can be  significant.
Thermoelectric generation relies on the use of a fuel source to convert water into steam, which is
then transferred to a turbine-generator where electricity is produced. The steam exhausted from
the turbine is  then passed to a condenser to remove the remaining heat, and the water is returned
to the boiler (Miller, 2010). The condenser is usually a shell and tube heat exchanger through
which cooling water runs to cool the steam passed in the  shell.  The cooling water is returned to a
cooling distribution basin to eject the heat into the atmosphere. In cases where cooling water
towers are used, the heat is transferred from the water in the tower by means of evaporation and
convection mechanisms, and the water is sent back to the condenser in a continuous circuit
(Kroger, 2004). Cooling water mass flow rate requirements can reach values >50 times the steam
mass flow rates depending on the  allowable temperature rise of the water (DOE, 2010).
                                           51

-------
       Conversion of thermal to electrical energy relies on the use of the Rankine cycle. In the
Rankine cycle, high pressure steam is produced in a boiler and a superheater is used to further
increase the steam pressure. In the condenser of the turbine, the steam is condensed to create a
vacuum pressure. The difference between the high pressure  steam and the condensed  steam
creates the driving force for energy conversion. Essentially,  the steam is sped up in the turbine
due to the pressure difference between the superheater and the condenser (Mortensen, 2009;
Prisyazhniuk, 2006; Perry et al., 1998). The low pressure achieved in the condenser is therefore
critical to avoid increased backpressure, and thus, a decrease in process efficiency. The low
pressures at the exit of the turbine are achieved by the use of cooling water (Mortensen, 2009).
The process efficiency can be increased by increasing the steam pressure and temperature within
the boiler, leading to plant operations under supercritical (>3200 psi, >1000 °F) or ultra-
supercritical (>3500 psi, > 1100 °F) conditions. Increases in plant efficiency by increasing
temperature and pressure can have a significant effect on water consumption, and it has been
estimated that the water volume used by supercritical plants is about 13 percent less than that by
subcritical plants (<2400 psi, <1000 °F). The lower steam pressure in the subcritical plants
translates into less energy transferred from the boiler to the turbine; therefore, a higher steam
flow,  and thus more cooling water, is required  to generate the same amount of electricity (DOE,
2010).
4.3.2   Cooling tower operations
       Cooling water is used to cool the steam and condense it back into water before it can be
sent to the furnace to produce steam. The heat  acquired by the cooling water is transferred to the
atmosphere or a receiving water body through  heat rejection systems such as cooling towers,
which can be classified into two main types: wet and dry.  In the U.S. electric power generation
industry, wet-cooling systems are the most common type, and  can be designed as once-through
or recirculating systems; the former have higher water withdrawal and the latter have  higher
water consumption. Cooling systems can also be designed to either freshwater or saline water.
Due to regulations regarding water discharge temperatures,  once-through cooling systems are no
longer being built, and closed-loop recirculating heat rejection systems are the most commonly
used systems in the U.S. Among these, cooling ponds or lakes are typically associated with older
plants, whereas wet cooling towers are more common in newer power plants. As of 2012,
approximately 40% of coal plants in the U.S. used wet-recirculating cooling towers, while 53%
used once-through cooling towers (UCS, 2014a). Smart and Aspinall (2009)  estimate that a
typical conventional coal-fired power plant withdraws between 20,000 - 50,000 gallons of
water/MWh and 500 - 1,200 gallons of water/MWh for once-through and recirculating cooling
systems, respectively. Their estimates for water consumed were between 100 - 317 gal/MWh
and 480 - 1,100 gal/MWh for the two systems, respectively.

   4.3.2.1  Wet cooling towers

       Wet cooling towers are direct contact heat exchangers. Warm water comes into contact
with relatively dry air, and energy is transferred by means of sensible heat loss to the  air and
latent heat loss with the evaporated water. The greater the water-air contact area and the longer
the residence time, the greater the heat transfer (Stultz and Kitto, 1992).  Depending on the
mechanism of air flow in the tower, wet cooling towers can  be subdivided into natural draft or
induced draft depending on whether mechanical fans are used  to move the air inside the tower
(Miller, 2010).

                                           52

-------
       Even though wet cooling towers are among the most energy-efficient and cost-effective
technologies for rejecting waste heat, they also require high water volumes. Water is lost in
cooling towers by means of drift, evaporation, and blowdown. Drift is a consequence of the tight
contact between the air flow and the water flow which produces droplets or mist, creating a
water loss that can range from 0.001% to 0.3% of tower flow rate, depending on the quality of
drift reducers or mist eliminators (Sovocool, 2008; Aquaprox, 2009). Additionally, it has been
estimated that 1% of the tower flow rate is evaporated for every 10°F drop in temperature
(Sovocool 2008; Perry et al.,  1998). Due to water evaporation, the concentration of IDS and
other particulates increases and generates conditions that contribute to scaling, corrosion and
biofouling. To maintain water quality, a fraction of the water is continuously removed from the
circuit in the form of blowdown and make-up water is introduced to sustain the tower's flow rate
(Sovocool, 2008; Aquaprox, 2009). A parameter known as cycles of concentration determines
the amount of blowdown required for a specific cooling system. The cycles of concentration are
the ratio of dissolved solids in the circulating water to that in the make-up water. An increase in
the cycles of concentration causes a decrease in the volume of blowdown and make-up water;
therefore, water savings can be associated with elevating this parameter (Sovocool, 2008;
Aquaprox, 2009).

   4.3.2.2  Dry cooling systems

       Dry cooling systems eliminate evaporation losses due to the use of convective heat as the
cooling mechanism, and can be classified as direct or indirect dry cooling systems. In direct dry
cooling systems, also known  as air-cooled steam condensers, the heat rejected from the steam is
absorbed in the form of sensible heat gain in the ambient air (Stultz and Kitto, 1992). Indirect dry
cooling systems operate with a conventional condenser that uses cooling water to condense the
steam turbine exhaust; however, the heat is transferred from the cooling water to the ambient air
through a closed, air-cooled heat exchanger. The dry-cooling towers operate  on the basis of
sensible heat transfer by the use of dry surface coil sections, where there is no direct contact
between the air and water, thus eliminating evaporation (Hensley, 2009). Dry cooling systems
typically result in  significant water savings by eliminating water losses due to evaporation;
however, these systems are not as efficient at rejecting heat,  which lowers the process efficiency.
In addition, the process efficiency is highly dependent on ambient air temperatures.
4.3.3   Bottom ash
       Coal contains non-combustible residue, which at the end of the combustion cycle in a
power plant's furnace, is collected in a water-filled hopper at the bottom of the furnace. Bottom
ash typically consists of 20% of the unburned material. The remaining 80% of the unburned
material is captured in paniculate control devices as fly ash.  The role of the water-filled  hopper
is to quench and store the hot ash. Crushers then crush the big particles into small pieces, after
which the slurry is pumped into an ash disposal pond or it is dewatered and shipped for disposal.
More  modern systems adopt a continuous heavy duty chain conveyor belt that is submerged in
water below the furnace. The bottom ash is quenched as it falls from the furnace, and the wet ash
is removed continuously up a dewatering slope, after which the ash is discharged into a storage
silo. In all cases, the water that comes into contact with bottom ash typically has the same
characteristics of CMD, and has to be treated to remove TDS and heavy metals before it can be
disposed or reused.
                                           53

-------
4.3.4  Flue gas desulfurization
       After coal combustion in a furnace, exhaust gases and fly ash are released into a flue,
after which the gas is treated prior to being released into the atmosphere via a smoke stack. In
addition to fly ash, flue gas typically contains nitrogen from the air used for combustion, carbon
dioxide, excess oxygen, water vapor, particulate matter, carbon monoxide, nitrous oxides,  and
sulfur oxides. Flue gas is typically treated with scrubbers and other chemical processes to
remove the pollutants prior to discharge into the atmosphere.
       Flue Gas Desulfurization (FGD) units are used to reduce 862 emissions, and are
significant contributors to the wastewater discharge emanating from coal-fired power plants.
Concerns have been raised about their use due to the volume of water required and the presence
and quantity of pollutants in the generated wastewater. The use of pollution control equipment in
power plants typically produce approximately 200,000 tons of FGD wastewater per year for a
500 MW power plant (UCS,  2014b). Generally, the pollutants present in FGD wastewater
include chlorides, mercury, arsenic, boron, aluminum, copper, selenium, and other toxic metals
and metalloids, as well as dissolved and suspended solids. Information on FGD technology, its
wastewater composition, the most widely applied FGD wastewater treatment technologies, and
several alternative technologies that might be of potential interest is presented in the Appendix to
this chapter.
4.3.5  Particulate matter
       Electrostatic precipitation technology has been widely used to control particulate matter
in power plant operations. For a long time, dry electrostatic precipitators (ESPs), used as
particulate control devices for industrial gas streams, offered removal efficiencies exceeding
99%, easy operation and reliability; however, the challenges for controlling air pollution in some
industrial applications have surpassed the capabilities of these devices (Bayless et al. 2004;
Khang et al., 2008). One of the issues associated with dry ESPs is that the charging mechanism
limits the size of particles to  a range of 0.1 to 2.0 |im, reducing the collection efficiency. This is
due to the lower corona power exerted as a consequence of the resistivity of the ash layer
accumulated in the collecting surfaces. Also, re-entrainment losses due to rapping17 result in the
non-desired emission of fine particles. Moreover, because aerosol formation does not occur due
to high operation temperatures, the removal of acid aerosols is not achieved in dry precipitators
(Bayless et al., 2004; Khang  et al., 2008)
       Wet precipitators (wESPs) have exhibited high collection efficiencies for fine
particulates, due to the avoidance of back corona and re-entrainment losses. In wESPs, a water
film flows down the walls of the collecting electrodes, where the high degree of adhesion
between the water and the collected particles prevent re-entrainment from occurring. Back
corona is prevented because flowing water constantly washes out the resistive ash. The collection
of acid gases is also enhanced due to the  lower operation temperatures that lead to acid
condensation and acid aerosol formation (Bayless et al., 2004; Khang et al., 2008). Even though
there are clear advantages for the use of wet precipitators, their use at large-scale coal-fired
power plants is still under development. The high particulate loads in the flue gas potentially lead
to the formation of dry spots in the collection surfaces, which reduces the  collection efficiency.
       17 Rapping involves imparting a physical force into an ESP collector plate or discharge
electrode in order to discharge deposits.

                                           54

-------
The transport and disposal of the particulate-saturated water flowing out of the ESPs creates a
challenge as well. As a result, wet precipitation is generally recommended to be used
downstream of other particulate collection devices (Bayless et al., 2004; Khang et al., 2008).

4.4  Water Reuse and Conservation Potential

       Electricity demand is expected to increase throughout the U.S. Coal-fired power plants
are expected to remain the single largest source among all sources of electricity production in the
U.S. (EIA, 2013), though coal's share of electricity production is less than 50% of total
electricity generated in the U.S. and continues to decline. Moreover, populations are projected to
rapidly increase in many areas where freshwater availability is severely limited. Consequently,
industries and government are searching for ways to reduce freshwater consumption at coal-fired
power plants (DOE, 2010). Several approaches are under investigation in the U.S. and around the
world to address challenging water demand issues (DOE, 2010). These approaches can be
classified into direct or indirect, depending on whether their implementation's main purpose is to
directly reduce freshwater consumption or to indirectly contribute to more sustainable water
usage (DOE, 2010). A brief review of research projects in the U.S., and some of the solutions
being implemented outside the U.S., are discussed below.
4.4.1   A dvanced cooling technologies
       In the U.S.,  the Department of Energy (DOE) through the National Energy Technology
Laboratory (NETL) has been conducting research and development on new water-energy
technologies in the  following areas: advanced cooling technologies, water reuse and recovery,
use of non-traditional sources for process and cooling water, and advanced water treatment and
detection technologies (EIA, 2011).
       Advanced cooling technologies include solutions designed to improve performance and
reduce costs associated with wet, dry, and hybrid cooling systems (Carney et al., 2014).
Developing technologies and strategies to minimize evaporative water losses and reduce the
water blowdown requirements that result  from these losses is of primary importance. For
instance, some of the current research is considering the implementation of condensing modules
within the cooling towers (Air2Air® by SPX Cooling Technologies, Inc.), and the application of
filtration methods to prevent scaling and increase the cycles of concentration (Pulsed Electrical
Fields and Pulse Spark Discharges for Advanced Cooling by Drexel University) (SPX Cooling
Technologies, Inc.,  2008; Carney et al., 2014).
       Outside the  NETL water-energy program, much effort has been put into the design of
hybrid  and dry cooling systems (Wurtz and Nagel, 2010; Schimmoller, 2007). Methods to reduce
water consumption in coal-fired power plants include the installation of an ice thermal storage
(ITS) system to cool the intake-air to gas turbines, (ITS Technology by University of Pittsburgh)
and the use of an air cooled condenser (ACC)—a dry cooling technology (ACC by SPX Cooling
Technologies, Inc.) (Mortensen, 2009).
       The water reuse and recovery component of the NETL water-energy program focuses on
the potential use of power plant cooling water and its associated waste heat. A study at Lehigh
University proposes the use of the hot cooling water returned from the condenser to heat ambient
air that will later be used to dry the coal. The evaporation losses in the cooling tower are
minimized due to the reduction in temperature of the return cooling water. Also, by drying the
                                           55

-------
coal prior to combustion, the plant heat rate is improved, and thus, the overall air emissions can
be reduced (Carney et al., 2014). Other projects promoted by Great River Energy make use of
the coal drying method to enhance the cycle efficiency through coal heat rate improvements, and
as a consequence, cooling water requirements and emissions of CO2, NOx, 862, Hg and
particulate matter per unit of energy produced are reduced (DOE, 2010).
       The last component of the NETL water-energy program deals with the potential use of
non-traditional sources of process and cooling water. For instance, researchers at the University
of Pittsburgh and Carnegie Mellon University are identifying a variety of impaired waters for
cooling water make-up and assessing secondary treated municipal wastewater, passively treated
coal mine drainage and ash pond effluent. Researchers from West Virginia University, in
partnership with the Water Research Institute and the National Mine Reclamation Center,  have
focused on evaluating the technical and economic feasibility of using water from abandoned
underground coal mines. EPRI has been evaluating the feasibility of using produced water to
meet make-up cooling water requirements (EPRI, 2007). Furthermore, the Nalco Company, the
Argonne National Laboratory, Clemson University, and GE Global Research are considering the
development of water treatment technologies to facilitate the use of impaired waters in cooling
towers (Carney et al., 2014;  DOE, 2010).
       Outside of the U.S., approaches for reducing freshwater consumption in coal-fired power
plants range from fuel enhancement and plant efficiency improvements to the use of dry cooling
systems, desalination, and the reuse  and recycling of wastewater.  Countries facing severe water
shortages such as China, Australia, South Africa, and India are also the countries with the highest
percentages of electricity generated by coal. In these countries, efficiency improvements and
plant retrofits are used as an indirect method for reducing water consumption. The primary
motivation for increasing plant efficiency in these countries is the need to reduce carbon
emissions in response to climate change. In addition, increasing electricity demand in some
highly populated countries such as China and India has exposed the need for larger and more
efficient plants (DOE, 2010). Many  countries have established policies encouraging or
mandating direct and/or indirect water reduction. China's energy strategy plan specifies that in
order to optimize production, they must promote the growth of large-scale and higher efficiency
power plants and implement air-cooling technologies in water-stressed regions (DOE, 2010).
Table 4.1 presents a summary of the water-saving and recovery approaches being implemented
outside the U.S.

4.4.2   Power plant flue gas water capture
       During the last few decades,  many studies have focused on recovering usable water from
alternative sources such as water from the flue gas emitted by coal-fired power plants. The water
vapor in flue gas comes from moisture content of the fuel, water vapor formed by the oxidation
of hydrogen in the fuel and moisture content in the combustion air (Levy et al. 2008, 2011). For
instance, a typical 400 MWe power plant burning coal and equipped with a FGD system could
release about 150 m3/h of water vapor through the flue stack (Levy et al., 2008) into the
atmosphere. Compression systems used for oxy-combustion require also require the removal of
nearly all flue gas moisture.  The recovery of flue water could be a valuable source for power
plants located in regions facing water shortages. The captured water, depending on its quality
and acidity level, could be used in boilers as feed water, cooling water, or FGD makeup water,
                                           56

-------
among other options. A brief review of various water vapor capture technologies is presented in
the Appendix to this chapter.
     Table 4.1 Water-saving approaches in coal power plants outside the U.S. (Adopted
                from DOE, 2010)
     Country
  Share of
 Electricity
Generated by
  Coal (%)
Water Scarcity Drivers
  Reducing Water Consumption Approaches

            (Direct and Indirect)
   China
   Australia
     80
     70
Third driest country in
the world.
Many areas subjected to
severe drought, and
groundwater use is
restricted.
Indirect: replacement/retrofit of small, inefficient
plants. Increase supercritical and ultra-supercritical
units, and exploration of Integrated Gasification
Combination Cycle (IGCC) technology.
Direct: use of dry cooling and desalination
systems.

Indirect: supercritical steam cycles, coal drying,
turbine upgrades.
Direct: dry cooling and on site water-recycling.
   South Africa
   India
     85
     70
Coal resources and
power plants located in
dry regions.


Water issues are not a
driver for power plant
improvements.
Indirect: supercritical technologies.
Direct: air-cooled condensers, advanced control
systems, and desalination.
Water infrastructure development.

Indirect: efficiency improvements (supercritical
steam parameters). Replacement of old plants.
IGCC research and development.
Direct: reuse and recycle of wastewater.
4.4.3  Wet electrostatic precipitators

       The use of wet precipitation for air pollution control has been widely studied; however,
its application for simultaneous flue gas water recovery and pollution control requires further
exploration. Khang et al. have proposed a flue gas water recovery system based on an air-cooled
condensing wESP to be implemented in coal-fired power plants equipped with FGD units
(Khang et al., 2008). A preliminary analysis of a proposed wESP, which is air-cooled through
fins attached to its external walls, is also provided in the Appendix to this chapter.

4.5   Trends in water withdrawal and water consumption in coal-fired power
      plants in the U.S.

       Information on coal-fired power plants was obtained through the U.S DOE/NETL
website (DOE, 2007). It includes data from the EIA-767 database, which contains information on
coal power plant generation, average water withdrawal and consumption, cooling water source,
                                             57

-------
type of cooling water system, type of boiler and type of FGD system. The data was organized by
the type of cooling systems used and electricity generation size bins. Furthermore, the water-
related information was arranged by types of cooling systems because water withdrawal volumes
vary according to the type of system. Additional coal-fired power plant data was obtained from
the 2010 EIA database (EIA, 2012). The DOE and EIA data were then cross-checked to verify
the validity of data on individual power plants
      Power plants selected for analysis included those with boilers that utilized coal as the
main fuel source. Boilers were further sorted based on the different types of coal used:
bituminous and anthracite (BIT), lignite coal (LIG), sub-bituminous coal (SUB) and waste coal
(WC). The boiler data was then linked to the generator data and cooling systems of each coal-
fired power plant. Only those systems with an annual electricity generation higher than 100 MW
were considered. Because the cooling system information is reported in the  database as monthly
averages, annual averages were calculated to facilitate its management and the identification of
water-related trends. The systems were also sorted as either a once-through  or recirculating type
of cooling system. The once-through systems were classified according to the EIA database as
systems using cooling ponds, freshwater and/or saline water while recirculating systems were
divided into units with cooling ponds, natural draft cooling towers or mechanical draft cooling
towers (forced and induced). After sorting the data  and eliminating those units with incomplete
cooling water information, the number  of cooling systems analyzed totaled 416 units, of which
198 corresponded to once-through systems and 218 were recirculating cooling systems.
      The annual electricity generation for the U.S. coal-fired power plants generating 100 MW
or more for recirculating and once-through systems is shown in Figure 4.1.
                       Total Generation
Recirculating
  Systems
Once-through
  Systems
Figure 4.1 Annual electricity generation in coal plants (>100 MW) in the U.S. Data from DOE
       (2007) and EIA (2012).

       A 10 percent decrease in total annual electricity generation in coal-fired power plants
occurred between 2005 and 2010 (Figure 4.1). In 2005, the annual electricity generation of the
power plants included in the study was 196 GW, of which about 115 GW, or 58.6%, came from
generators operating with recirculated cooling systems, and 80 GW (41.2%) used once-through
cooling systems (Figure 4.1). In 2010, the total electricity generation was 177 GW, with 103 GW
(58.2%) generated using recirculating cooling systems and 73 GW (41.6%) generated using
                                           58

-------
once-through cooling systems. The increased use of recirculating cooling systems could be
related to the implementation of more stringent regulations on water withdrawals and water
discharges imposed by National Pollutant Discharge Elimination System (NPDES) permits18.
According to the databases, only two coal power plants were operating with other types of
cooling systems such as dry cooling systems that do not use water to cool steam. The coal plants
are located in Wyoming and Illinois, and their combined electricity generation in 2010 was about
436 MW.
       In the U.S., once-through cooling systems use either freshwater or saline water. Some
power plants are equipped with cooling ponds through which the water is withdrawn from the
source, utilized in the condenser and then sent to the ponds to lower the temperature of the water
so that the thermal impacts on the source can be minimized during discharge. As seen in Figure
4.2, in 2010, 22 (16.7%) of the once-through cooling systems contained cooling ponds. The total
number of freshwater once-through cooling systems (with or without installed cooling ponds) in
2010 was 176 units, comprising 92.1% of the total electricity generation from plants with
installed once-through cooling systems (See Figure 4.2). The units using saline water once-
through cooling systems represent only  7.9% of the total electricity generated in 2010. These
units were mostly located on the East Coast (VA, MD, MA, CT, NJ, etc.) and at Florida's coal-
fired power plants.
              Once-Through Cooling Systems     Recirculating Cooling Systems
                                                  527%
                 ^^ Cooling Pond (22 Units)            ^^ Cooling Pond (39 Units)
                 ^— Fresh Water (154 Units)            <-~ Natural Draft (47 Units)
                      Saline Water (23 Units)              • Forced/Induced Draft (135 Units)

Figure 4.2   Percentage of electricity generated per cooling system type in 2010. Adopted
       from EIA (2012).

       In addition to information regarding the use of cooling ponds, recirculating cooling
systems in operation within coal power plants in the U.S. are also classified in the EIA databases
according to their use of natural and mechanical  draft (forced and induced) cooling towers. Most
of the power plants operate mechanical draft cooling towers. In 2010, the amount of electricity
generated by mechanical draft cooling towers was about 52.7% of the total electricity generated
using recirculating cooling systems. Natural draft cooling accounted for about 47 units and
approximately 39 units installed cooling ponds (Figure 4.2).
       18 More information on NPDES can be found at: http://cfpub.epa.gov/npdes/
                                           59

-------
4.6  Power Plant Carbon Models to Predict Water Withdrawal Rate

4.6.1   Description of models
       An interactive Adobe Flash™ model has been developed to predict technology
development trends of CO2 capture in coal-fired power plants (Wang et al., 2014) and is
schematically shown in Figure 4.319. The model is designed to investigate and compare various
methods of CO2 sequestration. Inputs include power plant capacity (power generated, MW), coal
parameters and CO2 capture method. Outputs include necessary amounts of air for combustion
(tons air/hour), CO2 released per hour, ash produced per hour and cost of electricity per kWh.
Water volumes (tons FbO used per hour) are also calculated based on an empirical model.
       Three different storage methods for CO2 are shown schematically in Fig. 4.4, including
enhanced oil recovery, depleted oil/gas reservoir and injection into saline aquifer, with costs of
$4.87, $3.82 and $2.93 per ton CO2, respectively (Bock, 2003). The price point outputs will
change as these unit costs vary with time. For all three storage methods, the Flash™ Model
predicts that production of 500 MW of power utilizing conventional coal combustion without
CO2 capture would require  182 tons coal/hour, 1385.5 tons air/hour and 950 tons water/hour.
Consequently, 27 tons ash/hour would be produced, while the amount of tons of CO2
released/hour would depend on the storage method used for CO2.
       A Flash™ model  was also developed for Integrated Gasification Combined Cycle
(IGCC) power plants to predict carbon capture and water usage in the boiler and is schematically
shown in Figure 4.520. In a typical IGCC power plant, coal is burned with air and steam in  a
gasifier to form syngas (CO+Fb) at 2450°F and 615 psi. Bottom  slag is also produced as a  result
and collected at the bottom  of the gasifier. The syngas is sent to a particle capture device where
fly ash is removed. The syngas, which is free of fine particles, then moves to the shift reactor
where it is acted upon by steam and a catalyst to produce CO2 and Fb. In the next step, sulfur is
removed and collected for industrial purposes. In the CO2 absorber, the CO2 is separated by
absorption. In the CO2 desorber, the CO2 is stripped off the absorbent, and the absorbent is
regenerated for further use.  Meanwhile, the Fb that exits the CCh absorber is used to  generate
electricity in a gas turbine/steam turbine combined cycle. The input and output parameters  are
shown in Figure 4.5.
       19 An demonstration of this interactive model can be accessed at the following Internet
URL: http://www.aerosols.wustl.edu/education/energy/CoalCO2/index.html
       20 An demonstration of this interactive model can be accessed at the following Internet
URL: http://www.aerosols.wustl.edu/education/energy/IGCC/index.html

                                          60

-------
           ( Oi Capture Method:
                        Input
                                                                               "Select Storage
                                                                               *.  Method
                           182 tons coal
                                                                    400 tons CO> capture^ / hr
          Coal parameters:      TO
                    | on|^        *
                Output
1095 tons
    Nj/hr
Air Separation:
COa Compression:
Cd Transportation:
COj Stc*age:

Total Cost Inciva;*,:
                                                             $0.0026 per kWh
                                                             S0.0463perkWh
                                             S0.0953 per kWh
Figure 4.3  Schematic diagram of a coal fired power plant with/without CO2 capture/sequestration. Adopted from Wang
      (2014).

-------
      Select CO2 Storage Method
                                          Forest Managmcnt Sink
                                          Enhancement: S2-15/tonCO:
       Depleted Oil/lias
       Reservoir:S3.82 / ton CO
     (;<> Back
                    Enhanced Oil
                    Recovery: S4.87/ton CO:
                       Select
                      Method
                                                                Injection into
                                                                Saline Aquifer:S2.93 / ton CO2
                                                                     Select
                                                                    Method
Figure 4.4   Flash™ model of different storage method. Adopted from Wang (2014).
       Utilizing this model for a power plant producing 600 MW and CO2 capture efficiency of
90%, using coal with 60% carbon content, 15% ash content, energy content of 30 KJ/Kg and 8%
sulphur content, approximately 218 tons coal/hr, 349 tons Ch/hr and 1136 tons steam/hr is
required. This produces 33 tons ash/hr, 17 tons sulphur/hr and 479 tons CCh/hr, of which 90% is
captured and stored.  The total cost of electricity depends on which storage method is chosen. For
example, if saline aquifer injection is chosen, the total cost of electricity will be $0.1081 per
kWh.

4.6.2  Flash™ Module Calculation Comparison to EIA Data
       The Flash™ model was validated by comparing the calculation results with EIA data
(2010). The following equation, developed based on the classical energy balance principle, was
used to calculate net power generation of coal-fired power plants:
                                          P(MVK)
                                                       	X3600
               Mcoal-in   -  =
                               Energy Content I -r^- I x 0.33
                                            62

-------
       Here MCoai-in is the coal feeding rate (kg/hr), 0.33 is the typical thermal efficiency for
this plant type, P and Energy Content are the plant size in MW and coal lower heating value
(LHV) in MJ/Kg, respectively.
Plant Size (MW) :

Carbon Capture Efficiency ((



v.udi rataiucccia
Energy Content (Kl/kg)
/Ov Sulphur Content (%)
Carbon Content (%)
Ash Content (%)











Calculate |

              Coal
                                                                    Select COz Capture &
                                                                    Sequestration Method
   Steam -

  Oxygen-
1
Guifier
Slag
CO 4- H.


Paniculutc
Rcmmcr
	 1 	
1
Fly Ash
« () - H:

1
W «n Bill
Shift
H.a.t.ll

<0:-lt
W

Sulphur
Riratn ti
Sulphur
                                                                      CO2 Produced
                                             Steam Turbine & CeiH.Titor
Figure 4.5  IGCC Flash™ Model. Adopted from Wang (2014).

       Ten power plants across the country were selected for the validation, representing four
major coal  seams: Bituminous (low, medium and high volatile), Lignite (North Dakota and
Texas) and Sub-Bituminous. The calculated annual fuel (coal) consumption rate (short tons/y)
and the net power generation (MW) were compared with the corresponding EIA data (EIA,
2010). As shown in Table 4.2, the Flash™ model predictions agreed with the EIA data, with
<10% error. The least error was achieved with power plants using Sub-Bituminous coal.
Relatively large errors were found for the Texas power plants using Lignite coal, suggesting that
modification of the Flash™ model is necessary depending on the coal type.
       Water withdrawal rate values for specific power plants from the EIA database are
reported in  Table 4.2. The water use rate in a boiler based on the heating value and power rating
of the facility (as calculated by the Flash™ model developed by Wang et al.) is also listed in
Table 4.2 (Wang etal., 2014).
                                           63

-------
     Table 4.2 Flash model validation for different types of coal seams*
Plant
ID
6019
26
1378
2817
6030
6146
6146
6076
6257
6204
Boiler #
1
5
3
2
1
3
1
4
4
3
Fuel Type
Bituminous
-High
Volatile
Bituminous
-Low
Volatile
Bituminous
- Medium
Volatile
Lignite -
North
Dakota
Lignite -
North
Dakota
Lignite -
Texas
Lignite -
Texas
Sub
Bituminous
Sub
Bituminous
Sub
Bituminous
Lower
Heating
Value
(KJ/Kg)
26151
33818
30108
14804
14804
14601
14601
19738
19738
19738
Total Fuel
Consumption
(Short tons/y,
ElAdata)
3980956
2306545
2599143
1831851
3835876
3685483
3655803
3668968
3211671
2757718
Total Fuel
Consumption
(Short tons/y,
Calculated)
3961873.15
2004924.90
2342358.04
2042946.99
3800084.62
5170338.90
4949412.36
3589536.55
3321917.29
2784781.99
Error
(%)
0.48
13.08
9.88
-11.52
0.93
-40.29
-35.39
2.16
-3.43
-0.98
Annual Net
Generation
(MWh, EIA
data))
9700403
5638631
5864655
2515030
4678211
6277820
6009568
5891809
5302541
4570897
Total MW
(Converted
from EIA
data)
1107.35
643.68
669.48
287.10
534.04
716.65
686.02
672.58
605.31
521.79
Total MW
(Cal.)
988.28
740.48
742.88
257.44
539.07
510.83
506.72
687.46
601.78
516.72
Error
(%)
10.75
-15.04
-10.96
10.33
-0.94
28.72
26.14
-2.21
0.58
0.97
Water
Withdrawal
Rate (tons/hr,
converted
from EIA
data)
716.55
182.47
NA
1936.94
NA
NA
28316.30
132.52
190.88
87.16
Boiler
Water Use
Rate
(tons/hr,
Calculated)
2.096E+03
1.218E+03
1.267E+03
5.433E+02
1.011E+03
1.356E+03
1.298E+03
1.273E+03
1.146E+03
9.875E+02
Note: *LHVs (energy content) are taken from DOE/NETL report: Detailed Coal Specifications, DOE-401/012111, January 2012.
                                                                 64

-------
4.7  Summary

       This chapter provides a comprehensive overview of coal-fired electric power generation
processes, and related water usage and wastewater generation. In the coal-based electric power
plant, water withdrawal and water usage vary primarily depending on cooling technologies used
to maintain the overall thermal system efficiency. There are two types of wet cooling systems,
i.e., once-through cooling systems and recirculated cooling systems. In 2010, the total electricity
generation was 177 GW, where 103 GW or 58.2 percent was generated at plants incorporating
recirculating systems, and 73 GW or 41.6 percent was generated through the use of once-through
cooling systems.
       Based on  the source of water withdrawal, once-through and recirculating cooling systems
can be further identified as cooling ponds, freshwater, and saline water systems. For the once-
through cooling systems, no significant differences were found among these three subcategories.
Water withdrawal volumes range from 0.001 to 0. lm3/MJ, based on data from 2010. The water
withdrawal volume for once-through cooling systems is higher than that for the recirculating
systems. The average water withdrawal rate for once-through cooling systems was 4.4E-02 ±
2.3E-02 m3/MJ, and 4.2E-03 ± l.OE-02 m3/MJ for recirculating units in 2005. The water
consumption, however, is expected to be higher in recirculating systems due to evaporation
losses occurring in cooling towers. The water consumed in once-through cooling systems
represented about one percent of the total water withdrawn, while for the recirculating systems,
the water consumed was roughly 95 percent of the water withdrawn. The use of recirculating
cooling systems is higher than that of once-through systems,  which in the U.S. is related to the
implementation of more stringent regulations on water withdrawals and water discharges
imposed through the National Pollutant Discharge Elimination System (NPDES) permits.
       One unique process in coal-fired power plants is the operation of Flue  Gas
Desulfurization (FGD) units. This air abatement process is commonly employed to reduce SO2
emissions, a by-product of which is a significant contributor to the wastewater discharge
emanating from coal-fired power plants. FGD designs include wet scrubbing,  dry scrubbing, and
dry sorbent injection. Wet scrubbing is being used in 85% of the FGD systems in the U.S., while
dry scrubbing and dry sorbent injection represent only 12% and 3%, respectively, of total use in
the U.S. For FGDs, wastewater generation, treatment, and disposal is the most pertinent issue
impacting water use. Wet scrubbers require a constant purge  flow to remove contaminants and
maintain proper performance. This wastewater stream contains high concentrations of gypsum.
Methods for treating FGD wastewater can be grouped into three categories: mechanical
separation, chemical treatment or biological treatment.
       An interactive module based on systems level modeling was developed using Adobe
Flash™ to relate  power plant emissions to power plant ratings and coal characteristics and water
withdrawal for the different technologies. The model was validated by comparison with EIA
power plant data  reported by ten power plants. The program provides a comparison of three
technologies, and also has baseline economic data on the resultant price of electricity. The
module also reports water use rates in the boiler; however, actual water withdrawal rates, which
are system specific, cannot be evaluated. Newer technologies such as supercritical oxy-coal with
staged combustion that result in enhanced overall efficiencies were not evaluated in this work.
                                          65

-------
4.8  References

Aquaprox. 2009. Fundamental Principles of Cooling Towers, Treatment of Cooling Water. 87-
       100: Springer Berlin Heidelberg.

Bayless, D. I, M. K. Alam, R. Radcliff, and J. Caine. 2004. "Membrane-Based Wet Electrostatic
       Precipitation." Fuel Processing Technology, 85(6):781-798.

Bock, B. 2003. "Economic Evaluation of CO2 Storage and Sink Enhancement Options." TVA
       Public Power Institute: Muscle Shoals, AL.

Carney, B., T. Feeley, and T. McNemar. 2014. "A. Power Plant-Water R&D Program."
       Department of Energy, NETL,
       http://www.netl.doe.gov/technologies/coalpower/ewr/water/pdfs/NETL%20Paper%20Un
       esco%20Conference.pdf. Accessed July 2014.

DOE. 2006. "Energy Demands on Water Resources: Report to Congress on the Interdependency
       of Energy and Water." Washington, DC.

DOE. 2007. http://www.netl.doe.gov/energy-analyses/hold/technology.html. Accessed June
       2013.

DOE. 2009. "Use of Non-Traditional Water for Power Plant Applications: An Overview of
       DOE/NETL R&D Efforts."
       http://seca.doe.gov/technologies/coalpower/ewr/water/pdfs/Use%20of%20Nontraditional
       %20Water%20for%20Power%20Plant%20Applications.pdf. Accessed August 2014.

DOE. 2010. DOE/NETL report: "Estimating Freshwater Needs to Meet Future Thermoelectric
       Generation Requirements". DOE/NETL-400/2010/1339, September 2010.

EIA. 2011. "Annual Energy Outlook 2011  with Projections to 2035." Washington DC: Energy
       Information Administration, United States Department of Energy.

EIA. 2012. "Independent Statistics and Analysis - Electricity."
       http://www.eia.gov/electricity/data.cfm. Accessed June 2014.

EIA. 2013. "Annual Energy Outlook 2013  with Projections to 2040" DOE/EIA-0383(2013).

EIA. 2014a. "Annual energy outlook 2014." http://www.eia.gov/forecasts/aeo/index.cfm.
       Accessed July, 2014.

EIA. 2014b. "Quarterly coal report." http://www.eia.gov/coal/production/quarterly/. Accessed
       July, 2014.

EIA. 2014c. "International Energy Outlook 2013." http://www.eia.gov/forecasts/ieo/coal.cfm.
       Accessed July, 2014.

Electric Power Research Institute (EPRI). 2007. "Treatment Technology Summary for Critical
       Pollutants of Concern in Power Plant Wastewaters." EPRI (1012549).
                                         66

-------
Hensley, J.C. 2009. Cooling Tower Fundamentals. SPX Cooling Technologies Inc., Overland
       Park, Kansas, USA.

Khang, S. J., T. C. Keener, and J. Y. Lee. 2008. Water Recovery and Removal of Residual
       Pollutants from Flue Gas using Air-Cooled Condensing Wet ESP. April,  2008.

Kroger, D.G. 2004. Air-Cooled Heat Exchangers and Cooling Towers. Vol. 1. PennWell Books.

Levy, E., H. Bilirgen, and J. DuPont. 2011. Recovery of Water from Boiler Flue  Gas using
       Condensing Heat Exchangers. Final Project Report DOE/NETL Project DE-NT0005648.

Levy, E., H. Bilirgen, K. Jeong, M. Kessen, C. Samuelson, and C. Whitcombe. 2008. Recovery
       of Water from Boiler Flue Gas. DOE Award Number DE-FC26-06NT42727, Energy
       Research Center, Lehigh University, Bethlehem, PA.

Mel drum, J., S. Nettles-Anderson, G. Heath, and J. Macknick. 2013. Life Cycle Water use for
       Electricity Generation: A Review and Harmonization of Literature Estimates.
       Environmental Research Letter. 8 doi:l 0.1088/1748-9326/8/1/015031.

Miller, E.G. 2010. Clean Coal Engineering Technology. Elsevier. ISBN: 978-1-85617-710-8.

Mortensen, K. 2009. "Use of Air2AirTM Technology to Recover Fresh-Water from the Normal
       Evaporative Cooling Loss at Coal-Based Thermoelectric Power Plants." Contract No.
       DEFC2605NT42725, National Energy Technology Laboratory, Pittsburgh, PA.

NMA. 2012. "U.S. coal production by state by rank." National Mining Association.
       http://www.nma.org/pdf/c_production_state_rank.pdf. Last accessed July, 2014.

Perry, R.H., D.W. Green, and J.O. Maloney. 1998. Heat Transfer Equipment. Perry's Chemical
       Engineers' Handbook (7th Edition), Section 11.

Prisyazhniuk, V.A. 2006. "Pressure Drop as a Criterion of a Steam-Turbine plant's Efficiency."
       Journal of Energy Engineering, 132(1):37-41.

Schimmoller, B. 2007. "Wet, Dry and in between." Power Engineering, March 1, 2007.
       http://www.power-eng.com/articles/print/volume-lll/issue-3/departments/industry-
       watch/wet-dry-and-in-between.html. Accessed August 2014.

Smart, A.,  and A. Aspinall. 2009. Water and the Electricity Generation Industry: Implications of
       Use. Australian National Water Commission.

Sovocool, K. 2008. "Maximizing Cooling-Tower Water Efficiency." Heating/Piping/Air
       Conditioning Engineering: HP AC 80(2): 18.

SPX Cooling Technologies, Inc. "Introducing Air2Air Cooling Technology."  SPX,
       www. spxcooling. com.

Stultz, S.C., and J.B. Kitto. 1992. Steam: Its Generation and Use. Babcock & Wilcox Company,
       New York.
                                         67

-------
UCS. 2014a. "UCS EW3 Energy-Water Database V.I.3." www.ucsusa.org/ew3database.
       Accessed August, 2014.

UCS. 2014b. "Based on a 500 MW coal power plant."
       www.ucsusa.org/clean_energy/coalvswind/c02d.html. Accessed August, 2014.

U.S. EPA. 2013. "Critical Review of Treatment Technologies for Removal of Salts [Ca, HCO3,
       K, Mg, Na, 864] from Water: Potential for Application to Coal Mining Impacted Water."
       National Risk Management Research Laboratory, Cincinnati, OH.

Wang, W.-N., D. Menon, Y., Li, Y.J., Yang, and P. Biswas. 2014. "Web-based Educational
       Module Development for Water and Carbon Footprints Tracking." Submitted, 2014.

Wurtz, W., and P. Nagel. 2010. "The Case for Dry Cooling Technologies." www.spx.com.
                                         68

-------
4.9  Appendix

4.9.1   Flue gas desulfurization technology

   4.9.1.1  The FGD process, types of systems, and wastewater generation

       FGD is a process by which SO2 gas, emitted during the burning of coal, is removed from
the emissions of the plant. SO2 is a major air pollutant, contributing to the acidification of rain,
and is regulated by the EPA under the Clean Air Act (CAA). In FGD systems, flue gas comes
into contact with (CaO) or limestone (CaCCb) to form calcium sulfite (CaSCb), which is usually
converted into gypsum (CaSO4«2H2O) by further processing, and accompanied by carbon
dioxide (CCh) release. The generated gypsum, along  with the other flue gas components such as
particulate matter, heavy metals, trace elements, and  other harmful compounds, is captured and
transferred  into  an absorbent slurry (liquid phase), while the CO2 and other gases are released
into the atmosphere.
       Several FGD designs exist, the three most popular of which are wet scrubbing, dry
scrubbing and dry sorbent injection. In the U.S., 85% of the FGD systems utilize wet scrubbing,
while dry scrubbing and dry sorbent injection represent only 12% and 3% of the FGD systems,
respectively (Higgins et al., 2009). Wet scrubbers are preferred in areas where the coal is high in
sulfur content because they have a higher removal rate (>90%). Dry  scrubbers, which only
achieve -80% removal, are used if the sulfur content is low (Heimbigner, 2007; Higgins et al.,
2009).  In a  wet scrubber FGD system, particulate control devices such as an electrostatic
precipitator are installed before the wet scrubber to remove most of the fly ash upstream of the
scrubber. The flue gas enters the scrubber and is mixed with a spray  of the treating solution,
causing the SO2 to solidify and stay in the solution. For dry scrubbers, this process is reversed.
The dry or nearly dry treating chemicals are added into the flue gas,  causing the SO2 to form a
solid dry  particle. The flue gases then pass through the particulate control devices, removing the
solid particles alongside the other particulates  and ash. Wet scrubbers ultimately capture the 862
in an aqueous form while dry scrubbers capture it in a solid form.
       Wet scrubbers require a constant purge flow of water to remove the contaminants and
maintain  proper performance. This wastewater stream contains high  concentrations of gypsum,
contaminants  from the coal, lime or scrubbing solution and any pollutants introduced during
processes that precede the scrubber.  This wastewater has a variable pH range  (4-9.6) and high
concentrations of suspended and dissolved solids, heavy metals, trace elements, chlorides and
other compounds (Higgins  et al., 2009). Typically, the chloride concentration is one of the most
important parameters in the wastewater because high chlorides can cause corrosion of the
scrubber, water tanks, and gas pipes, and are the determining factor in how the scrubber is
operated. The exact composition of the wastewater is difficult to predict, and is likely to change
over time because of variations in coal and limestone sources. Plant design, plant operations, and
scrubbing method will also affect the composition of the wastewater.

   4.9.1.2  FGD wastewater composition and pollutants

       TSS (total suspended solids) and TDS  are major components of wastewater generated by
coal-fired power plants. In a typical FGD system, solids usually consist of bottom or fly ash,
other particulates, raw materials (e.g., limestone) and gypsum. Heavy metal and other trace
                                          69

-------
element pollutants, including selenium, arsenic, mercury, antimony, beryllium and thallium, are
present in potentially high quantities in power plant wastewater. Most of the pollutants originate
from coal combustion, while aluminum compounds come primarily from the limestone used
within the FGD. All of the listed pollutants are subject to ppb concentration standards due to
their toxicity. Appendix Table 4.1 summarizes the national drinking water maximum
contaminant level (MCL) standards compared with typical pollutant concentrations in the FGD
wastewater along with potential health effects.  The typical chloride limit for wastewater in many
existing FGD systems is 12,000 mg/L to avoid metal corrosion. But, when corrosion-resistant
metals are used, chloride concentrations as high as 35,000 mg/L are allowed. Chlorides also have
disrupting effects on several wastewater treatment methods.

     Appendix Table 4.1 Summary of pollutants included in FGD wastewater
Pollutant
Aluminum
Arsenic
Boron
Copper
Mercury
Selenium
Chloride
Total dissolved
solids
Potential health effect(s)
Nerve damage
Skin and circulatory
system damage, risk of
cancer
Stomach, liver and kidney
damages
Gastrointestinal irritation,
liver or kidney damage
Kidney damage
Hair and fingernail loss,
numbness, circulatory
problems
Damage to respiratory
system
N/A
National drinking water
standards (MCLs) (|jg/L)
50-200
(Secondary MCL)
10
n/a
1,300
2
50
250 mg/L
(Secondary MCL)
500 mg/L
(Secondary MCL)
Typical FGD wastewater
(M9/L)
87,478
211
262,550
2,153
56
1,485
14,592 mg/L
31, 025 mg/L
   Note:  Adopted from EPRI (2007a)

      Nitrogen pollutants, including nitrites, nitrates and ammonia, are products of coal
combustion processes. The fate of nitrogen compounds can vary depending on different
upstream combustion temperatures. Most of the nitrates will be captured in a particulate-control
device, but a significant portion will also be collected within the FGD. Nitrogen is a major
component of other wastewater streams, particularly municipal wastewater, and each form has its
own maximum allowable release standard. Depending on concentration, they can disrupt the
treatment of other components as well.
      Calcium and sulfates are other potential components of the wastewater.  While not
necessarily pollutants, they can be complicating factors during wastewater treatment. Calcium
can lead to scale buildup, and sulfates can interfere with treatment processes (EPRI, 2007b).
                                           70

-------
   4.9.1.3  Current FGD wastewater treatment systems

       The combination of the described contaminants creates wastewater that requires careful
handling. Generally, FGD wastewater is kept separated from other wastewater streams and
treated independently. Methods for treating FGD wastewater can be grouped into three
categories: mechanical separation, chemical treatment or biological treatment. Depending on the
wastewater composition and plant needs, these processes can be combined or used in series to
target different components of the wastewater with specialized processes.

•  Mechanical Separation
       Mechanical solid separation techniques such as a dedicated gravity settling basin,
wetland, or dewatering system are the most widely used methods to separate suspended solids.
The rich slurry containing suspended solids, after the FGD scrubber, will first pass through a
mechanical separator to settle  out some of the solids. The purged slurry then continues to flow
into a hydrocyclone. This unit is designed as a retroflexed cone in order to settle out most of the
gypsum. Meanwhile, the overflow from the hydrocyclone is introduced into downstream
wastewater treatment units. After most of the gypsum and other solids are squeezed from the
bottom of the hydrocyclone, a filtration process is used to dewater the gypsum, forming a solid
gypsum cake.
       Evaporative Processes: The evaporation method is a mechanical process currently used in
many plants. It involves separating the water from  the chemical residues in the wastewater via
water evaporation, generating  two products: distilled water  and separated solids. The water can
be captured and reused, or allowed to evaporate naturally in detention ponds. The separated
solids are disposed of in a landfill or through further waste treatment. For  example, ash ponds are
used for the disposal of wet fly ash slurry. The ponds are constructed with low permeability clay
layers and cut-off walls to prevent groundwater pollution. After evaporation, the solid fly  ash is
disposed. Advantages of the evaporation method include simplicity of the process, low
concentrations of pollutants that meet discharge requirements, and zero discharge wastewater
potential. This method is disadvantageous because of the large energy input required to heat the
wastewater, separate it from the  solids and then recapture it using condensers. Large detention
ponds are required to use solar evaporation. Also, some of the contaminants can be volatile and
may continue to be present in the distillate.
       Membrane Separation - HERO Process: The HERO (high efficiency reverse osmosis)
process is a membrane separation procedure that handles high silica content and treats cooling
tower blowdown. The membrane is capable of operation at  high pH (11.5), which is significant
because silica is more soluble  and the membrane fouling is  reduced at high pH. Of particular
interest is boron removal through HERO, because boron is not removed effectively by other
technologies. Using the high pH capabilities of HERO, up to 99% of boron can be removed from
a system  (EPRI, 2007b). The disadvantages of the  HERO process are the same as any membrane
system: scaling and fouling remain major hurdles, and it requires pretreatment and possibly
water softening.

•  Chemical Treatments
       Chemical treatment is used to precipitate calcium and magnesium ions as well as a
portion of heavy metal and trace elemental pollutants via a hydroxide precipitation method (lime
                                           71

-------
or sodium hydroxide). The post-treated water can be reused in the FGD plant. Additional
chemical treatments such as iron co-precipitation and organic or inorganic sulfide precipitation
can be employed to reduce the level of heavy metal pollutants.
       Conventional chemical precipitation: lime precipitation for metal hydroxides: The most
common chemical method for the removal of soluble metal ions from FGD wastewater is
lime/hydroxide precipitation. In this process, metal ions are precipitated as metal hydroxides.
Regulation of pH is a critical component of the precipitation. Since FGD wastewater can be
acidic, the pH is often increased using alkaline chemicals like lime or sodium hydroxide.
Because each metal has a unique solubility point, the rate at which they are removed from FGD
wastewater varies from one plant to the next. Most heavy metal ions have lower solubility at
higher pH. The lowest solubility for Cr is around 0.1 mg/L when pH is 7.4, Zn is 0.1 mg/L at pH
10 and Pb is 8 mg/L at pH 10. At pH 6, Cu has a solubility of 18 mg/L, but this number can go
down to 0.05 mg/L if the pH is raised to 8. Ferric chloride and aluminum sulfate are usually
added to accelerate heavy metal ion precipitation processes. Certain heavy metal ions such as
copper,  zinc and cadmium can form metallic complexes with ammonia. The complexes cause
these heavy metals to stay soluble under high pH conditions. Adding soluble sulfide ions to the
solution is the most widely used method to destroy the complexes, allowing the precipitation of
the metallic ions and reducing the concentration of heavy metal ions to acceptable levels.
Conventional lime precipitation can lower the concentration of heavy metal ions present in FGD
wastewater. However, due to the lower threshold levels for the discharge of certain metal and
elemental ions, additional enhanced chemical precipitations such as iron coprecipitation and
inorganic metal sulfide precipitation may also be required.
       Enhanced chemical precipitation: iron co-precipitation: In coal-fired power plants, the
iron co-precipitation method is  one the most promising technologies to lower heavy metals
concentrations to ppb levels (EPRI, 2007b). It can effectively remove arsenic, selenium,
cadmium, copper, silver, chromium and zinc; it is especially useful for arsenic and  selenium
removal. A ferric salt, typically ferric chloride or ferric sulfate, is added to the FGD wastewater
stream,  and reacts with water to form ferric hydroxide. Ferric hydroxide will quickly precipitate
due to its low solubility. Ferric  hydroxide has an amorphous structure that can absorb and bind
the dissolved ions as well as suspended solids, and then eventually precipitate. Compared to
hydroxide precipitation, iron co-precipitation removes a wider range of ions, resulting in overall
lower final concentrations. However, since chloride and sulfate ions are introduced into the FGD
wastewater, the iron co-precipitation treatment requires corrosive-resistant materials, and it will
also increase the overall wastewater TDS concentration.
       Enhanced chemical precipitation: Inorganic metal sulfide precipitation: Chemical
precipitation using  sulfide salts (sodium, ferrous or calcium sulfide) can achieve significantly
lower metal concentrations in the treated water, often reducing residual metals 100- to 1000-fold,
compared to hydroxide precipitation levels. Further advantages of sulfide precipitation are that,
unlike hydroxides, metal sulfides are not amphoteric; hence, they do not resolubilize with
changing pH, and residual metal sulfide sludge volumes are smaller. Sulfide precipitation
effectively reduces most metals to very low levels, including copper and mercury. The
disadvantage of using sulfide precipitation is the potential for forming hydrogen sulfide gas at
low pH, and the effluent may have to be oxidized to reduce the dissolved sulfide residual after
precipitation, depending on the type of sulfide salt used. While the sulfide process is suitable for
                                           72

-------
the treatment of aluminum, copper and mercury, it is only marginally effective for arsenic and
selenium (Merrill et al., 1987).
       Enhanced chemical precipitation: Organosulfide precipitation with TMT 15: TMT 15 is a
commercial organosulfide compound that is widely used in wastewater treatment in various
industry areas, including combustion plants (like coal-powered plants), incineration plants, and
the metal and mining industries. TMT 15 has two major applications: removing heavy metal ions
(e.g., copper, lead, nickel, cadmium and silver) from wastewater, and reducing mercury
emissions in wet FGD systems by using TMT 15 in the alkaline scrubber or incinerator scrubber.
TMT 15 is a 15% aqueous solution of the trimercapto-s-triazine, tri-sodium salt developed to
work similarly to sulfide precipitation, but without the drawbacks of using sulfides. The
chemical is stable, can be stored safely, and produces a stable, non-toxic residual. It produces
similar removal efficiencies to sulfide precipitation.

•  Biological treatment
       Biological treatment allows specific bacteria to attack and degrade pollutants via aerobic
or anaerobic processes. A biological reactor in a coal-fired power plant can remove nitrogen,
phosphorus, and some metals by precipitation. However, previous research shows that some
compounds, including selenate, cannot be effectively treated. In general, biological treatment
consists of nitrification and denitrification, but anaerobic processes can also be used. Ammonia
and organic matter are converted to nitrate, biomass and CO2 (gas) via nitrification, while nitrate
and nitrite are converted to N2 (gas) via denitrification. Nitrification is a strictly aerobic process
while denitrification is an anoxic (no oxygen, but not yet anaerobic) process. Anaerobic
processes can be used to produce methane, remove nitrogen and organic matter, and to capture
metals, phosphorus, and  other elements from the wastewater. The biomass absorbs the
compounds through a variety of mechanisms, and they can then be captured by separating the
biomass from the water.  Any combination of these methods can be used, depending on the needs
of the wastewater. Typically, biological treatments require some form of pre- or post-treatment
to work effectively.
       Compared with traditional chemical and physical wastewater treatment methods,
biological methods are an effective means for the removal of elemental pollutants such as arsenic
and selenium. Conventional biological treatments such as anaerobic biofilm reactor processes
have shown an ability to remove heavy metals from FGD wastewater. A new process, called
Advanced Biological Metals Removal Process (ABMet), developed by GE (Sonstegard and
Pickett, 2009) has shown success in removing heavy metals, inorganic compounds and
metalloids. A Constructed Wetlands Treatment System (CWTS, consisting of biological and
physical adsorption processes) is another cost-effective and low-energy wastewater pollutant
removal method.
       ABMet Process: The ABMet process utilizes biofilms attached to activated carbon in an
up-flow bioreactor to treat FGD wastewater. Within this reactor, a specialized mixture of
microbes can reduce and precipitate most of the pollutants of concern from the solution or
transform them into harmless formations. Metals, arsenic, selenium and nitrates can all be
reduced to ppb levels using the process. For the microbes employed in this system, little attention
is required.  The reactor needs only a stream of nutrients, which can be fed automatically at a
scheduled time. This can result in a highly effective, precisely controlled biological system.
                                           73

-------
       Constructed Wetlands: Constructed wetlands simulate a natural wetland. They utilize the
natural filtration and pollution removal capabilities of sediments and plants to remove nitrogen,
phosphorus and trace elements from a wastewater as it permeates through the wetland. They are
a low cost, low maintenance method for treating wastewater. However, they have been shown to
provide little removal of some of the target species in FGD wastewater, particularly selenate
(EPRI, 2007b). They also require a considerable footprint, depending on the wastewater flow
rate, and will require a pretreatment step to reduce the solids content.

   4.9.1.4 Alternative treatment systems

       Single-use Sorption Media Processes: Two critical pollutants found in FGD wastewater
are arsenic and selenium, which are required to be controlled at the ppb level. For this, metal-
based adsorbents (granular ferric oxide or hydroxide and titanium oxide) have proved to be the
most effective  single-use sorption media. A case study examining the performances of these
metal-based adsorbents was carried out by Anderson et al. (2003). The authors employed
aluminum oxide, spinel and titanium dioxide as single-use sorption media to remove  arsenic
from groundwater. The results indicated that TiO2 effectively converted As(III) to As(V), which
is the easiest and most widely used way to remove arsenite. However, a major drawback of this
method is that  a complex nano-particle (TiO2 in the form of nano-suspension) removal must be
performed after the adsorption process. In the Anderson et al. study, a heterogeneous adsorbent,
AhOs/TiCh, was employed instead of TiO2.  Thus, no separation process was required after
adsorption.  This heterogeneous adsorbent also contained a photocatalyst acting as an oxidizer,
which can convert As(III) to As(V). The results of different experiments carried out by the
authors indicate that the single-use sorption media along with photocatalytic adsorption could be
an effective method for As removal without additional separation processes. Furthermore, mixed
sorption metal  media have demonstrated the capability to remove As more effectively than either
pure TiO2 or AhCb.
       Selective Ion Exchange: Selective ion exchange is widely employed in general
wastewater treatment applications.  However, only a few power plants have adopted this
technology, likely because the ion exchange resins media can be easily plugged, especially by
FGD wastewater with high concentrations of sulfate and dissolved solids. Similar to traditional
ion exchange resins, selective resins can also be regenerated. Resins may contain different
functional groups that can be used for  selective ion exchange. Thus, specific ion exchange resins
are being designed and tested for FGD wastewater treatments, such as removal of heavy metals
and boron. Kabay conducted the first boron removal research at a power plant using a selective
ion exchange method (Kabay and Yilmaz, 2004). Two resins called N-glucamine-type chelating
resin Diaion CRB 02 and weak base resin WA 30 were packed into a stainless steel column.
Wastewater was passed through the column, reducing the boron concentration from 20 mg/L to 0
mg/L (Kabay and Yilmaz, 2004). The  total dissolved and undissolved SiO2 concentration from
power plant wastewater is high enough to easily plug the ion exchange column, and thus, a TDS
pre-removal step is typically recommended before an ion exchange system.
       Electro-coagulation and electrowinning: An alternative to the selective metal ion
exchange method is the application of electro-coagulation and electrowinning treatments.  The
only difference between electro-coagulation and chemical precipitation is the presence of an
electrode. In addition, electro-coagulation treatment may have a higher TSS removal  efficiency.
Chemical precipitation typically employs expensive aluminum, whereas the price of the iron

                                           74

-------
electrode is a major benefit of electro-coagulation treatment. Less chemical reagents are required
for electro-coagulation treatment compared to chemical precipitation, generating smaller
amounts of sludge. Drawbacks of electro-coagulation treatment include high energy
consumption and the necessity of electrode replacement. Although electro-coagulation has not
yet been utilized using FGD wastewater, it has been successfully applied in industrial wastewater
treatment.
       Electrowinning technology could also be considered for FGD wastewater treatment. This
technology focuses on the removal of metals, and operates on a principle similar to electro-
coagulation. Essentially, the metals are plated out on the electrowinning electrodes. The
electrodes will accumulate metal deposits that eventually need to be scraped off; alternatively,
the electrode may also be replaced. This end point condition is noted by a drop of amperage
across the plates. The scraped off metal or the removed electrodes can typically be disposed via a
waste hauler for metal recovery or at a landfill as a non-hazardous material. The electrowinning
technology is used primarily in metal plating and finishing processes. Although few applications
for electrowinning can be found  in wastewater treatment, it could recover nearly 100% of heavy
metals from FGD wastewater.

   4.9.1.5  Treatment technology selection

       Overall, there is no single treatment method that can treat FGD wastewater at all power
plants. A number of factors play a role in determining the most effective treatment, including
location, coal type, plant design and discharge limits.  Appendix Table 4.2 summarizes the
advantages and disadvantages of wastewater treatment technologies that might be applicable to
FGD wastewater.  Appendix Tables 4.3 and 4.4 provide additional information from EPA reports
on the treatment of CMD. While not identical, treatment of CMD requires many of the same
considerations that apply to FGD wastewater treatment. The tables provide approximate cost,
performance and advantage/disadvantage comparisons that can be used to guide the investigation
of a treatment system for a given power plant and FGD wastewater.
     Appendix Table 4.2 Summary of the reviewed wastewater treatments

Capital
O&M
Health issues
Time required
Sustainability
Disposition
Public perception
Permanence
Evaporation
High
Moderate
Yes
Long term
Moderate
On site and Off
site
No
Yes
Ion exchange
Moderate
Low
No
Short term
High
On site
Yes
Yes
Sorption
Moderate
Low
No
Short term
High
On site
Yes
Yes
ABMet
High
Low
Yes
Short term
High
On site and
Off site
N/A
No
TMT15
N/A
Low
Yes
Short term
High
On site and
Off site
N/A
No
      Note: Adopted from Cheng et al. (2011).
                                           75

-------
  Appendix Table 4.3   Suggestions for selecting an appropriate CMD treatment
             technology
Selection Criteria
Capital cost (per
m3/day capacity)
Chemical dosing
Energy efficiency
Operations &
maintenance cost
($ per m3 treated)
Potential
byproducts
recovery
Proven technology
on commercial
scale
Performance
Specialized
application
Waste sludge/brine
production
Water recovery
Coal Mine Drainage Treatment Technology
Chemical
Precipitation
$300-1,250
High
Moderate
$0.2-1.5/m3
Gypsum
Proven, with large
commercial plants
Robust process
General application
to high metals, high
S04 mine water
Large waste sludge
production
High water
recovery > 95%
Membrane
Treatment
$500-1,000
Limited
High
$0.5-1.0/m3
Potential, but not
demonstrated
Proven, with several
large commercial
plants
Process good
performance, but
sensitive to pre-
treatment
General application,
but with appropriate
pre-treatment
Sludge and brine
production
High water
recovery > 90%
Ion Exchange
N/A
High
Moderate
N/A
Gypsum
Demonstrated on pilot
scale, no large
commercial plants
IX process
performance and resin
recovery subject to
interference
Demonstrated for IDS,
with appropriate pre-
treatment
Medium waste sludge
production
High water recovery >
95%
Biological Sulphate
Removal
$800-1,500
Variable
Low - Moderate
$0.7-1.5/m3
Sulfur
Proven, with a limited
number of commercial
plants
Sensitive to toxics,
fluctuating feed water
quality and environmental
conditions
Specialized application to
high S04 mine waters
Small waste sludge
production
Very high water
recovery > 98%
Note: Adopted from U.S. EPA (2013).
                                         76

-------
Appendix Table 4.4 Advantages and disadvantages of treatment technologies that can
          be used to remove TDS from CMD
Treatment
Technology
Adsorption
Alkalinity Producing
Systems
Anoxic Limestone
Drains
Bioremediation
Chemical
Coagulation
Chemical
Precipitation
Electrochemical
methods
Electrodialysis
Ion Exchange
Lime Treatment
Advantages
Commonly available technology
Can handle different flow rates, acid and
metal loading rates, effective sulfate removal,
uses limestone
Useful for coal acidic drainage, decreased
overall rate of reaction, longer residence time
provides better neutralization
Less expensive to install, natural process to
treat contaminants, remediation not restricted
to treatment zone
Sludge settling, dewatering compacts waste
Simple, inexpensive, removes most metals
Metal selective, no chemicals consumed, can
recover pure metals
High recovery ratio, treats highly concentrated
feeds, minimum pretreatment requirements,
membranes not subject to scaling or fouling,
field applicable
High regeneration efficiency, metal selective,
portable, field applicable
Most proven method for sulfate removal, can
treat concentrated acidic CMD, most cost-
effective for large flows or highly
contaminated water
Disadvantages
Most TDS ions of interest are adsorbed to a
limited extent, especially in the presence of other
ions such as Fe, Mn, Zn, Pb, Co, etc.
Periodic exchange of substrate required,
occasional clogging, requires steep slope and
large land area
High Al content armors limestone, tough to
maintain anoxic conditions, does not handle
volume and water quality fluctuations well, can
only remove sulfate among the ions of interest,
requires other metals such as Fe for effective
sulfate removal
Some TDS components may not be amenable to
biodegradation, site characterization and
optimization needed for each site
High cost, consumes lots of chemicals, may not
remove all TDS components
Large amounts of sludge produced, sludge
disposal problems, may not remove all TDS
components
High capital, operating and maintenance costs,
highly dependent on initial solution pH and
current density
Frequent membrane leaks, bacteria, non-ionic
chemicals and turbidity may affect treatment,
requires a source of energy in the field
High cost, requires a constant source of energy
in the field
Requires frequent monitoring and sludge
management
                                    77

-------
Appendix Table 4.4 (continued)
Limestone
Ponds/Channels
  Relatively inexpensive, low maintenance,
     minimal pretreatment requirements
  Armoring of limestone reduces effectiveness,
 precipitated sludge requires management, may
       not remove all IDS components
Distillation
 Minimal pretreatment, high quality product,
   minimal operational requirements, high
    production capacity,  low capital cost,
     concentration-independent energy
  requirements, treats highly concentrated
                  feeds
Requires vapor cooling, low recovery ratio, water
              quality variations
    :i-Stage Flash
Distillation
   Can handle large capacities, minimum
   pretreatment requirements, high quality
 product water, can be combined with other
               technologies
  Expensive to build and operate, low recovery
   ratio, requires lots of technical knowledge
Permeable Reactive
Barriers
 Can remove sulfates, extremely stable, can
  treat both surface water and groundwater
     Uncertain life time, may need periodic
  replenishment of media, cost dependent on
sulfate reduction rates, may not remove all IDS
                components
Reverse Osmosis
Can remove IDS components of interest, low
 energy requirements, high production ratio,
  can be used with most other technologies
 High membrane and operating cost, affected by
 suspended particles, requires pretreatment and
 post treatment, relatively low flow rates, cannot
   operate under highly acidic, basic or high
           chlorine concentrations
Sulfate Reducing
Bacteria
Availability of inexpensive organic substrates,
  minimal power requirements, can produce
           alkalinity to raise pH
 Effluent sulfate concentrations may still exceed
 limits, occasional clogging, longevity dependent
  on carbon availability, requires large surface
                    area
Ultrafiltration
    Less solid waste produced, negligible
    chemical consumption, high efficiency
 High initial, operating and maintenance costs,
   low flow rates, efficiency decreases in the
           presence of other metals
Wetlands
   Easy to implement, low capital cost, low
     operational and maintenance costs
     relatively low flows, large area and flat
topography required, periodic sediment removal
   required, difficult to control metal migration,
         sensitive to low temperatures
   Note: Adopted from U.S. EPA (2013).
                                                   78

-------
4.9.2   Flue gas water capture technologies
       Technologies to capture water from flue gas are described in the following subsections.

   4.9.2.1  Liquid sorption

       Liquid sorption is based on the preferential absorption of a gas mixture component by a
liquid substance that is introduced into the gas stream. The driving force for the dehumidification
system is the difference between the partial pressure of water in the gas stream and the partial
pressure of water in the sorption substance or desiccant (Daal, 2012). Folkedahl et al. (2006)
developed a liquid desiccant-based flue gas dehydration process to reduce the amount of water
coal-fired power plants consume. The desiccant technology uses chemicals that extract water
vapor through strong physicochemical affinity. No additional heating or cooling is required to
capture the water because the process uses heat from absorption and vaporization. Furthermore,
minimum pressure drops can be achieved through engineering design optimization (Folkedahl,
2006). More importantly, the process produces water of high purity, making it feasible to use the
water in the steam cycle (Daal, 2012).
       The dehumidification system involves the operation of absorption columns, where the
desiccant flows downward and the flue gas flows upwards. After leaving the column and before
being sent to the stack, the flue gas is passed through a demister to remove any entrained
desiccant. The water-rich desiccant is heated and sent to a regenerator where water is separated
from the desiccant solution using pressure differences (Daal, 2012). Different scale tests have
been performed by Folkedahl et al. demonstrating that the technology exhibits up to a 70% water
capture rate when using a calcium chloride solution (Folkedahl, 2006). Concerns regarding this
technology include process safety due to the  nature of the desiccant and corrosion caused by
calcium chloride. The  liquid sorption technology is already used in gas dehydration and in
building dehumidification and cooling, but its use in coal-fired power plants is still considered to
be in the demonstration phase. Economical and technical viability of the desiccant-based
dehumidification system for coal  power plants has been demonstrated for regions where the price
of water is high (Daal, 2012).

   4.9.2.2  Membranes

       Gas separation membranes have been studied as a potential solution for flue gas water
capture in power plants. In the U.S., the Gas  Technology Institute (GTI) developed a water vapor
extraction technology that separates water and latent heat from flue gas to later return them to the
steam cycle. The technology, Transport Membrane Condenser (TMC), consists of water vapor
passing through a  nano-porous ceramic separation membrane and its further condensation by
means of direct contact with a lower temperature water stream (Wang et al., 2012). The
selectivity of the membrane recovers high quality water, inhibiting the transport of contaminants
such as CO2, O2, NOx and 862. Industrial demonstration scale and commercial laundry
applications have already proven  the technology's effectiveness, and TMC is currently being
commercialized for industrial boiler waste heat and water recovery; however, for coal power
plant flue gas applications, GTI developed a two-stage design-tailored TMC technology intended
to reach maximum heat and water recovery that can be used for boiler makeup water, FGD
makeup water or other plant uses (Wang et al., 2012).
                                           79

-------
       The two-stage TMC unit (Appendix Figure 4.1) uses two separate cooling water streams.
On the flue gas side, the membrane condenser is located between the FGD unit and the stack. On
the water side, the inlet water for the first stage TMC is obtained from condensed steam, and the
water recovered and associated latent heat is sent to the de-aerator for later use as a boiler water
makeup. The second  stage TMC inlet water carries part of the condenser cooling water stream,
and the outlet water is sent back to the cooling water stream with extra water recovered from the
flue gas (Wang et al., 2012). The results of a pilot-scale test indicated that the membrane
performance, evaluated in terms of water and heat recovery, increased when the moisture content
in the flue gas increased, which matches the results obtained in laboratory-scale experiments.
Furthermore, the impact of SO2 concentration in the flue gas was evaluated.  It was reported that
typical values of SO2 concentration in the flue gas had no significant effect on the water and heat
recovery capacity of the TMC unit. Regarding the quality of the condensed water, the results
showed that small amounts of SO2 and CO2 were dissolved in the water coming out of the
membranes. Nonetheless, analyses demonstrated that the impact on water treatment needs would
be minimal because the amounts of both compounds were very low as a result of the high
selectivity of the membranes. Overall, the tests demonstrated a good heat and water recovery
performance of the TMC technology for coal-fired power plant applications, especially for those
that use high-moisture coals and/or FGD systems (Wang et al., 2012).
                                         Steam
                                                               Boiler
               Steam Turbine
                                                                      n
                                                                      CD
                                                 Water
                        Economizer
                                         Dearator
                        Stage #M

                        Steam bleed


                       Stage #N
                    Steam bleed
High Pressure
Water Heater
 Cooling Water Out
                 Condenser
Low Pressure
Water Heater
                            Air Preheater
                              Flue Gas
                           Desulphurization
                                Condensate
                                              Recovered water
                            Cooling Water to TMC
                            TMC/Stage 1
                                                                 TMC/Stage 2
               Cooling Water In
Appendix Figure 4.1 Schematic of a two-stage TMC for power plant flue gas heat and water
       recovery. Adopted from Wang et al. (2012).

       Another membrane technology is being developed by KEMA (Daal, 2012), an energy
services firm, through the CapWa project. This project aims to develop a technology capable of
                                           80

-------
increasing location flexibility for the operation of power plants by recovering evaporated water.
Results of the gas separation membrane tests in industrial plants in the Netherlands and Germany
have revealed that at least 40% of the water contained in flue gases can be recovered. The quality
of the water also is supposed to meet the boiler feed water quality requirements (Daal, 2012).
      Limitations of the membrane technology include the energy demand of the water capture
system due to the flue gas pressure drop, which is compensated for through the use of a vacuum
force. Furthermore, the power requirement of air-cooled condensers has to be considered because
cooling energy is needed to condense the captured water (Daal, 2012). The membrane
technology for coal power plant applications has only been tested on a small  scale; further
studies need to be done before it can be applied on an industrial scale.

  4.9.2.3  Condensing heat exchangers

      The basic principle behind water vapor condensing by heat exchanger technology is
reducing  the temperature of the flue gas to the water dew point. Currently, a project by Levy et
al. is under development  (Levy et al., 2008; 2011), and the technology has proven to be effective
on a pilot scale (Daal, 2012). According to Levy et al., the water vapor content in the flue gas
highly depends on the coal rank. The flue gas water rate of a bituminous pulverized coal power
plant generating approximately 600 MW can reach up to 200,000 Ib/hr, while a power plant of
the  same  size firing high  moisture lignites could have a flue gas moisture flow of 600,000 Ib/hr.
On  the other hand, the cooling makeup water requirements for the same power plant using
cooling towers is 2.1 million Ib/hr,  which means that recovered water from flue gas could
provide from 10% (bituminous coal) to 18% (Powder River Basin [PRB] coal), and up to 29%
(high moisture lignites) of the makeup cooling water (Levy et al., 2008; 2011).
      The authors have  estimated a capture efficiency of up to 70% depending on different
factors, such as flue gas water content (fuel type, wet FGD equipment), cooling water
temperature, equipment design and flow rates of flue gas and cooling water (Levy et al., 2008;
2011). The maximum water capture-to-makeup water ratio for different types of coal is presented
in Appendix Table 4.5 (Levy et al., 2011). It can be  observed that the higher water capture-to-
makeup water ratio is achieved when the condensing heat exchangers are located downstream of
the  wet FGD scrubber or when the  coal moisture content is higher.
      Other advantages of condensing heat exchangers and the cooling of the flue gas include:

•   Reduced CO2 Emissions: Latent and sensible heat recovered from the flue gas could be used
    to reduce the unit heat rate, thereby reducing CO2 emissions.
•   Reduced Acidity: Controlled acid condensation would provide environmental, operational
    and maintenance benefits due to the reduction of acid in the flue gas.
•   Enhanced Mercury Removal: The lower temperature of the flue gas would allow for
    improved mercury removal.
•   Reduced Cost of CO2 Capture:  Due to the reduced water and acid content in the flue gas, the
    cost of removing CO2 would decrease (Levy et al., 2008; 2011).
      One of the biggest disadvantages of this technology is the  potential for corrosion and
fouling problems caused  by reducing the flue gas temperature (Levy et al., 2008; 2011). The
common  operating temperature for the flue gas is 300°F, which prevents acid condensation and
provides  a buoyancy force to assist in the transport of flue gas up  the stack. The dew point of the
                                          81

-------
sulfuric acid depends on SOs concentrations in the flue gas, and it is known that for
concentrations up to 35 ppm, sulfuric acid begins condensing at temperatures between 250°F and
310°F. The dew point of the water vapor, considering a flue gas moisture content of 6 to 17.5
volume percent, is in the range of 100°F to 135°F. The dew points of other acids present in the
flue gas, such as hydrochloric and nitric acids, have a similar temperature range as that of water
vapor (Levy et al., 2011). To achieve water recovery in the condensing heat exchangers, the flue
gas is cooled down to the water vapor dew point, inevitably causing co-condensation of acids
and water. As a result, water treatment and acid trap installation should be considered to avoid
corrosion in the heat exchanger tubes and damages to the cooling system.
     Appendix Table 4.5 Estimated fractions of cooling tower makeup water achievable,
               assuming 100% water vapor capture
Case
Bituminous (Unscrubbed)
Bituminous (Wet FGD)
PRB (Unscrubbed)
High Moisture Lignite
(Unscrubbed)
Lignite (Wet FGD)
Inlet Flue Gas Moisture Fraction (%
Volume)
6-8
16-17
10.5-12
15.5-16.5
17.5
Maximum hhO
(Capture/Makeup H20)
0.10-0.13
0.30 - 0.33
0.19-0.22
0.29-0.31
0.33-0.34
       Note: Adopted from Levy et al. (2011).
       As previously mentioned, one of the benefits of lowering the flue gas temperature is the
increased potential for mercury removal. Mercury measurements by Levy et al. snowed that
vapor phase mercury decreased by 60% between the inlet and exit of the heat exchanger system
when the flue gas was extracted downstream of an ESP, and the reduction was between 30% and
80% when the flue gas was extracted downstream of a wet FGD (Levy et al., 2011). It was also
noted that the percent captured increased as the flue gas exit temperature decreased. Analyses of
the technology suggest that the installation of condensing heat exchangers downstream of the
wet FGD systems would be cost-effective, where the benefits include water recovery from flue
gas for use within the power plant and increase in net unit power output (Levy et al., 2011).

4.9.3  Design optimization of air-cooled condensing wet ESP for flue gas water recovery
       One of the objectives of the present study was to review different approaches for
freshwater consumption reduction within coal-fired power plants. Khang et al. (2008) proposed
the use of an air-cooled wet electrostatic precipitator (wESP) for the simultaneous removal of
water and pollutants from the flue gas of coal power plants equipped with flue gas
desulfurization (FGD) systems (Khang et al., 2008). Water recovery systems using  wESPs  as
flue gas water condensers are still in their early stage of research and development,  and many
design parameters are yet to be optimized. Therefore, during this study, a heat-transfer
optimization of the system's design was performed. The first design presented by the authors is
briefly described below. Even though the authors found high water capture efficiencies, further
                                          82

-------
estimations regarding heat transfer optimization were needed to advance the development of the
water saving system being proposed. The estimations carried out attempted to obtain an
optimized ESP design, as well as a fin configuration, where not only the water recovery would
be maximized, but the power requirements would be reduced as much as possible.

   4.9.3.1 Air-cooled condensing wet ESP initial proposal

       As mentioned previously, a system has been proposed that is intended for the
simultaneous recovery of water and removal of residual pollutants from flue gas through the use
of an air-cooled wESP. The design of the system consists of a cluster of five condensing wESPs
installed after each FGD unit. Water vapor present in the flue gas is condensed out in the inner
walls of the ESP, where the wall temperature is maintained lower than that of the flue gas by
external heat transfer fins. The amount of condensation depends on the external heat transfer
through the fins, and the collected water is recycled back to the wet FGD unit; thus, no additional
water treatment facilities are required. Preliminary schematics of the proposed system are
presented in Appendix Figures 4.2 and 4.3.
       Previous heat transfer estimations of the proposed system have shown the following
potential advantages:

•   Significant water recovery. Considering a typical 500 MW  power plant at 80°F air
    temperature, it has been estimated that it is possible to recover FGD makeup water in the
    amount of 9,670 gallons/hr (43% of the total) or 14,700 gallons/hr of water (65% of the total)
    using a 0.5MW or 2.5MW air fan, respectively. The amount of water recovered will highly
    depend on surrounding temperature and fan power, as will be shown later. Additionally, cost
    savings from the recovered water is sufficient to cover the cost of extra electric power
    requirement for an air fan.

•   Water condensation and corona wind lead to a high heat transfer coefficient in the unit.

•   Low pressure drop through the unit due to the wide and empty channels of the ESP.

•   Traditional mist eliminators in a wet FGD unit are no longer needed since fine particles are
    effectively removed in the proposed wESP unit, which saves approximately 0.5 inch-FbO
    pressure drop.

•   There is no need for extra water to clean the collection walls since the condensed water can
    serve this purpose.

•   Additional removal of residual SOX, NO2 and oxidized mercury due to the high mass transfer
    coefficient achieved by corona wind and electron attachment mechanism.
       Heat transfer coefficients and water recovery efficiencies were estimated for a typical 500
MW coal-fired power plant based on an initial fin configuration for the wESP (See Appendix
Figure 4.4). The preliminary results, as observed in Appendix Figure 4.5, show the dependency
of water recovery on the fan power (cooling air velocity).
                                           83

-------
                                                      FGD Make-Up
                                                          Water
           Flue Gas
           @~125'F
                                FGD Slurry Treatment
oo
 Proposed Condensing Wet ESP
(Water recovery, Hg Oxidation and
 Removal, Ultra-Cleaning ofNOx,
        SOx, and PM
          Corona Wires
Aluminum Fins
                                                                                            Rectangular Flow
                                                                                            Channel, Water
                                                                                            Condensation on
                                                                                            Inside Walls
                                                                                                   Water Condensate
             Appendix Figure 4.2  Condensing wet ESP system within the power plant and wet ESP unit with external air-cooling.

-------
                        Corona
                         Wires
                 Fins
        Flue Gas
       From FGD
        ~125 °F
           Water
           Sump
                            Cooling
                           Air Intake
                            ~80°F                 *-.^  FGD water
                                                         recovered
Appendix Figure 4.3  System's design- a cluster of five ESP per FGD Unit. Adopted from
       Khang et al. (2008).
       Also, from Appendix Figure 4.6, it is possible to observe the influence of cooling air
temperature on water recovery efficiency. In both figures, the FGD water recovery percentage
refers to the percentage of FGD makeup water that can be captured and fed to the system.
       Furthermore, it has been estimated that the proposed configuration requires ten charging
wires of 18 ft in length per unit, which results in a consumption of 2.1 kW for each wESP unit.
Therefore, the total wESP power consumption is estimated to be 52.5 kW for a 500 MW plant
with 25 wESP units (this can vary depending on particulate loading in the flue gas). Nonetheless,
the power consumption by the ESP is still much lower than the power required for external air
cooling, indicating that the major power consumption for the proposed system comes from the
air fan for outside cooling. For this reason, future work regarding this part of the project will
focus on heat transfer optimization, which involves defining a fin configuration to achieve the
maximum fin heat transfer (highest water condensation rate), while minimizing fin-side fan
power.
                                          85

-------
                                            1/8"—*1
                                            •Flue Gas-
                                                               1/4"
Top View of
Wet ESP Aluminum
Fin
^
0.5" 304 Stainless Steel
Condensing Wall ^2-
%
„>
-z-
f
.---
^
—
— —
i~
1
1 '
r
H
r

Appendix Figure 4.4 Condensing wet ESP initial fin configuration.


           120

                                   Outside Temperature  =60"F
              0         1X106        2X106        3X106
                                  Fin-side Fan Power, Watts

Appendix Figure 4.5 FGD water recovery vs. fin-side fan power.
4X106
5X106
                                           86

-------
            120
            100
                             70
80
90
                                                                             100
                                     Outside Temperature, °F
Appendix Figure 4.6 Effects of outside air temperature on water recovery.


   4.9.3.2  Heat transfer optimization methodology

       The original proposed system's configuration and fuel information was used as a base for
the optimization. All the information and design specifications were taken from the Project
Narrative prepared by Khang et al. (2008). The design parameters originally presented by the
authors are indicated in Appendix Table 4.6.


     Appendix Table 4.6 ESP channels and fin characteristics.
Fins
Thickness, in
Width, in
Length, ft
Spacing between fins, in
Number of fins on each side of the channels
Thermal conductivity (aluminum), Btu/ft.h.°F
1/8
12
20
1/4
640
136
Channels
Length, ft
Width, ft
Wall thickness, in
Thermal conductivity (stainless steel), Btu/ft.h.T
20
1.5
1/2
25.9
            Note: Data source from Khang et al. (2008).
       The heat-transfer simulation for the wESP channels were carried out with the use of a
calculus model developed in Mathematica v.8 for Students (Wolfram, Boston, MA). The
simulation code was validated by comparison with preliminary results from Khang et al. (2008).
                                           87

-------
The most important equations used for the heat transfer estimations are introduced in Section
1.9.3.3. The calculation process is described later on.
   4.9.3.3  Fundamental equations
       The equations used to define the required variables within the estimation process are
presented in the international system of units (SI). The calculation model included the conversion
equations for the variables calculated; so, the results were presented in the English units
according to the design specifications (See Appendix Table 4.6).
Water Vapor Pressure Correlation (Perry et al., 1998)
                          c
       Pw(r) =  XP  *82——                                                     (4.9-1)
       where,
       Pw(T) = Water vapor pressure, Pa
       T = Flue gas temperature, K
       ,4 = 77.3450
       5 = 0.0057
       C = 7235.0

Water vapor density (Perry et al.,  1998)
       P(T} = ™~™v>                                                        (4.9-2)
       where,
       p(T) = Water vapor density, Kg/m3
Heat of vaporization (Perry et al., 1998)
                              /l_T^C2+C3.T/Tr-C4-T/Tr-T/Tr
       Hv(T) = 0.1292 • d  •     -                                               (4.9-3)
       where,
       Hv(T) = Heat of vaporization of water, KJ/Kg
       Ci = 5.2053 x 107
       C2 = 0.3199
       Cj = -0.2120
       C4 = 0.25795
       7> = 647.13
                                           88

-------
Number of fins
                   2 -ESPLength                                                 ,
       no Fins =	                                               (4.9-4)
                Thickness+Spacing
       where,
       noFins = Number of fins per ESP channel
       Thickness = Thickness of the fins, m
       Spacing = Space between fins, m
f A ^ r\
(4.9-5)
Heat transfer coefficient calculations (Perry et al, 1998; Bayless et al, 2004)
       7     NU 'fc
       ho = —
             ueq
       where,
       ho = Heat transfer coefficient, Watts/ m2- K
       Nu = Nusselt number
       k = Thermal conductivity of air, Watts/ m-K
       Deq = Equivalent diameter, m
       Nu = 0.0265 -Re0-s-Pr°-3                                              (4.9-6)
       where,
       Re = Reynolds number
       Pr = Prandtl number

       Re = ^^                                                            (4.9-7)
               M
       where,
       v = Velocity of air on fin side, m/s
       p = Density of air, Kg/m3
       ju = Dynamic viscosity of air, Kg/m-s
       _       Channel Area                                                     ,     x
       Dpn =	                                                   (4.9-8)
        eq   Wetted Perimeter                                                   ^     '
       where,
                                          89

-------
       Channel Area = Area of the channel created by the fins and the ESP channels wall (it is
two times the fin width times the spacing between fins), m2
       Wetted Perimeter = Perimeter of the fins channel where the air passes through (it is twice
the fin spacing plus four times their width, because it is calculated on both walls of the ESP
channel).
Air temperature on the fin side
       The air temperature on the fin side was calculated considering the temperature gradient
along the fin length. This was done by using the efficiency of the fin (EIA, 2014)
       ml  = 1-222-  . w                                                       (4.9-9)
             AJ fc/in -t                                                           v     '
           _ ^— V'^                                                       ,4 9_1QN
        7 '        ml                                                           ^      '
       where,
       eff = Efficiency of the fins
       mL = Fin efficiency parameter
       kfin = Thermal conductivity of the fin (aluminum), Watts/m-K
       w =Fin width, m

       The temperature of the air along the fin length is then found by performing an energy
balance.
       Tair —Tw        14 ~ef f • hn ~w 'L\
       	= Exp[———	                                            (4.9-11)
        To-Tw       r  \ Qair • Cp .   )                                            ^      '
       where,
       Tair = Temperature of the air on the fins side, K
       Tw = Temperature of the wall of the wESP channel (assumed to be the same of the flue
gas), K
       To = Initial temperature of air, K
       Qair = Air flow rate,  kg/s
       L =Fin length, m
       Cp =Heat capacity of air, kJ/kg-K
                                          90

-------
Heat transfer from the fin

       q =  Qair • Cpair(Tair ~ To~)                                               (4.9-12)

       where,
       q = rate of heat transfer from the fin, kW


Water condensation rate
       Wf =    v(r)                                                           (4.9-13)

       where,
       Wf = Water condensation rate, kg/s


Fan power equations (Perry et al,  1998)

       The fan power was calculated by multiplying the pressure drop by the air flow rate.
       where,
       AP = Pressure drop, Pa
       f = Fanning friction factor, which is calculated by using the Blasius equation for turbulent
flow
       f - 0.079 /                                                              fAQ 1C\
       / -      /Re0-25                                                        (4.9-15)

              Then, the total fan power per each wESP channel is:
       / = AP • Channel Area • Total no. Channels • v                          (4.9-16)


   4. 9. 3. 4 Heat transfer calculation process

       The amount of water from the flue gas that could be condensed within one wESP channel
under given circumstances was estimated through an iterative process. One of the most important
aspects to consider was that while the flue gas is passing along the x-axis of the wESP channel,
its temperature is continuously decreasing due to the water condensation; thus, changing the ESP
wall temperature. Khang et al. (2008) demonstrated that the ESP wall temperature is the same as
the bulk flue gas temperature because of the significantly large value of the condensing  heat
transfer coefficient on the inside of the walls. For calculation simplicity, the flue gas temperature
change or the wall temperature change was calculated for each section of the ESP channel,
creating the iterative process. A section, in this study, corresponds to the wESP channel region
delimited by two fins on each wall of the channel, as shown in Appendix Figure 4.7.
                                          91

-------
                              Corona Wires
                   Aluminum Fins
                       Section 1

                  Rectangular Flow
                  Channel, Water
                  Condensation on
                  Inside Walls
                          Water Condensate
                    Appendix Figure 4.7 Sections in the wESP channel.
       The fin-side heat transfer coefficient, h0, was calculated using Equation 5 for a specific
set of conditions (air velocity on the fin side and fin configuration). For the first section of the
wESP, the fin efficiency from Equation 10 is used to determine the change in the cooling air
temperature (Equation 11), which represents the heat transferred from the fin-side to the wESP
wall (Equation 12). The water condensed in the section is calculated using Equation 13. The
water contained in the flue gas was determined based on the power plant efficiency and the type
of coal burned. For this study, a typical power plant with an electricity generation capacity of
500 MW was assumed. The properties of the coal and flue gas used for the calculations are
presented in Appendix Table 4.7.
       The water removed from the flue gas in the section is then subtracted from the total water
contained in the flue gas. The condensation of the water results in a change in the water vapor
partial pressure in the flue gas; thus, the flue gas temperature changes  according to Equation 1.
This new temperature is used to consecutively calculate the water removed in each section until
the nth section. The summation of the water removed in the sections from the 1st to the nth gives
the total water removed in the wESP channel.
       Another important factor to consider was the power usage or parasitic power in the
system. The heat transfer on the fin-side of the wESP is enhanced by forced convection, which is
achieved through the use of fans installed to force the cooling air through the fin channels. The
velocity of the cooling air on the fin side is then directly proportional to the fan power exerted or
vice versa. For the system proposed,  different air velocities were assumed and the fan power was
calculated using the Blasius correlation for turbulent flow, as previously shown in Equations 14
and 15. For each air velocity assumed, the water removal was calculated.
                                           92

-------
     Appendix Table 4.7 Properties of coal and flue gas assumed
Coal Characteristics
Coal heating value, Btu/lb
weight % carbon
weight % hydrogen
weight % oxygen
weight % sulfur
weight % nitrogen
weight % ash
weight % moisture
11,150
61.2
4.3
7.4
3.9
1.2
12.0
10.0
Flue Gas
Excess combustion air, %
Amount of dry exhaust gas, scf /1 06 Btu of fuel
C02 in dry exhaust gas, volume %
02 in dry exhaust gas, volume %
Molecular weight of dry exhaust gas
20
11,554
15
3.4
30.7
              Note: Source from Khang et al. (2008)
   4.9.3.5 Heat transfer optimization process

       The heat transfer optimization of the air-cooled condensing wESP was carried out in two
parts. The first approach was to simulate different fin configurations based on the original
proposal written by Khang et al. (2008), and then identify how the various configurations
influence the heat transfer, and thus, the water removal. As mentioned before, the goal of the
heat transfer optimization was to achieve a higher water removal while minimizing the fan power
requirements. In this first part, two parameter variations were analyzed.

Varying fin spacing and thickness

       To determine how changing the fin configuration in the wESP based on fin spacing and
thickness (number of fins per ESP channel) would influence the heat transfer, nine cases besides
the original, or initial case proposed, were evaluated (see Appendix Table 4.8). The original
configuration is presented in Appendix Figure 4.8. The only parameters changed during this part
of the optimization were the fin thickness and  spacing.
                                           93

-------
     Appendix Table 4.8   Cases of fin configuration evaluated varying spacing and
               thickness
Case
Original
1
2
3
4
5
6
7
8
9
No. Fins
1280
1600
2000
2500
1280
1280
1600
2000
1000
1000
Spacing+t
(ft)
0.031
0.025
0.020
0.016
0.031
0.031
0.025
0.020
0.040
0.040
Spacing
(ft)
0.021
0.015
0.010
0.006
0.026
0.016
0.020
0.015
0.030
0.035
t
(ft)
0.010
0.010
0.010
0.010
0.005
0.015
0.005
0.005
0.010
0.005
     Top View
    of Wet ESP
     0.5" 304 Stainless Steel
        Condensing Wall
                 	Flue Gas
                                          Depth
         Corona Wires
 Aluminum Fin:
Rectangular Flow
Channel, Water
Condensation on
Inside Walls
                                                                Water
                                                              Condensate
Appendix Figure 4.8 Condensing wESP original fin configuration.
Varying fin depth

       All the previous cases were evaluated again by varying the fin depth, which was
originally one foot, as shown in Appendix Figure 4.8. The different depths evaluated are
presented in Appendix Table 4.9 with their respective case label.
                                            94

-------
     Appendix Table 4.9 Cases of fin depth evaluated
Fin Depth (ft)
1.0
1.5
2.0
3.0
Case Label
—
w1
w2
w3
       The second part of the heat transfer optimization process involved the definition of ESP
design parameters that directly influence the ESP dimensions, and thus, influences the heat
transfer phenomena occurring. These design parameters are briefly described below (Holman,
1997).

   4.9.3.6  Specific collection area (SCA)

       It is the ratio of the collection surface area to the gas flow rate into the ESP channel. The
SCA represents the ratio A/Q in the Deutsch-Anderson equation; therefore, it is an important
parameter in the determination of the ESP collection efficiency.
       SCA =
 Total Collection Surface (ft2)

   Gas Flow Rate (1000 2—)
               v    mmj
(4.9-17)
       Increasing the SCA in the design generally leads to an increase in the collection
efficiency of the precipitator. A general range of SCA is between 200 and 800 ft2 per 1000 acfm
depending on the design conditions and desired efficiency.

Aspect ratio (AR)

       The AR is the ratio of the length of the ESP to its height. This parameter is important
when considering dust re-entrainment at the moment of trapping. Effective plate lengths are at
least 35 to 40 ft, which helps reduce the amount of collected dust that comes out of the ESP.
       AR =
Effective Length (ft)
Effective Height (ft)
(4.9-18)
       For high efficiency ESPs, the AR is usually between 1.0 and 1.5, and sometimes it can
reach values close to 2.0.
                                           95

-------
Gas flow distribution

       The gas velocity inside the ESP influences the particle collection. The gas velocity into
the ESP is generally reduced to 2 - 8 ft/sec for adequate collection. For aspect ratios of 1.5, the
optimum gas velocity is typically between 5 and 6 ft/sec.
       Considering the ESP design parameters introduced above, new cases for the wESP
dimensions were proposed and analyzed, including the original case presented in the previous
section. The new dimensions proposed for analysis are presented in Appendix Table 4.10.
     Appendix Table 4.10 Wet ESP dimensions variation cases
Case/Dimension
ESP Length (ft)
ESP Height (ft)
ESP Channel Width (ft)
No. ESP
Original
20
20
1.5
25
A
50
40
1.3
60
B
60
40
1.3
60
C
50
40
1.3
40
D
40
20
1.5
40
       All the cases presented in Appendix Table 4.10 were simulated in order to determine the
best water recovery potential. For each case, different fin configurations were evaluated
considering the results obtained in the first part of the optimization process. The flue gas water
removal percentages and fan power requirements were estimated for each case in order to
determine the optimum water recovery system design.

   4.9.3.7 Heat transfer optimization results

       The first attempt for the optimization process consisted of changing the fin configuration
of the ESP to determine the factors influencing the water removal efficiency and the power
requirements. The second part focused on the design parameters of the ESP and the further
required changes in the fin configuration that would allow the optimization of the heat transfer
within the system. The results of both parts are presented below.

   4.9.3.8 Fin configuration

       The original fin configuration was varied by changing the spacing and thickness of the
fins and by changing the depth of the fins. Nine different cases, the details of which are
presented in the Methodology section, were simulated in order to evaluate how the spacing
between the fins and their thickness influence the water removal. The results are presented in
Appendix Table 4.11.
                                           96

-------
     Appendix Table 4.11 System parameters and water recovery obtained from changing
               fin spacing and thickness
Case
Original
1
2
3
4
5
6
7
8
9
Reynolds
Number
18328.1
12977.6
8673.3
5214.3
22584.7
14050.4
17260.6
12977.6
25763.4
29983.5
ho (BTU/ °F-
ft2-hr)
22.8
24.4
26.5
29.3
21.8
24.0
23.1
24.5
21.3
20.7
Fin Efficiency
0.173
0.167
0.160
0.152
0.125
0.206
0.121
0.118
0.179
0.128
FGD Make Up
(%)
54.8
55.1
50.3
39.8
50.4
49.7
56.3
60.5
50.6
43.1
Total Water (gallon/hr)
per ESP
494.3
496.7
454.0
358.9
454.6
448.4
507.4
545.4
456.8
389.1
       As indicated in Appendix Table 4.11, four configurations, whose results are highlighted
in blue, exhibited the highest water recoveries from the flue gas. In these cases (cases 1, 6 and 7),
compared to the original case, increasing the number of fins, which means decreasing the fin
spacing, and in some cases their thickness, results in a higher percentage of water removal (fin
spacing and thickness were given in Appendix Table 4.8).
       When it comes to the fin-side fan power requirements, a slightly lower fan-power
requirement is obtained when evaluating cases 6 and 7 in comparison to case 1, as presented in
Appendix Figure 4.9. However, no improvements regarding power usage are obtained when
compared to the original case. For instance, with the original configuration (See Appendix
Figure 4.3), if a water removal of 40 percent of the FGD makeup water is desired, the fin-side
fan power required would be around 2 MW (for cases 6 and 7), whereas for case 1, the fan power
needed is around 2.8 MW. From Appendix Table 4.11, water recovery improvements are
observed when comparing cases 6 and 7 to the original case; thus,  as a preliminary result of this
first part of the optimization process, cases 6 and 7 can be considered to offer the best fin
configurations.
       The previous estimations were conducted maintaining the original fin depth (1.0 ft.). The
second part of the optimization process involved changing the depth of the fins from 1.0 ft. to
1.5, 2.0 and 3.0 ft. The analysis of the results indicated that increasing the depth of the fin
enhances the heat transfer in the system; thus, the amount of water removed from the flue gas is
higher. Also, by analyzing the fan-power results for those cases where the water recovery
percentage was higher, it was observed that the fin-side fan power requirements decrease when
increasing the depth of the fins, which is a desirable condition as part of the optimization
process.
       From the results obtained in the first part of the optimization process, case 7w-2 was
chosen for further analysis due to the higher water capture efficiency and lower power
                                          97

-------
requirements exhibited. Appendix Figure 4.10 shows the effect of fin-side fan power on water
recovery for different cooling air or outside air temperatures.
  100
                                                                  Case 1
    0     1X 106   2 X Id6   3 X 106   4 X 106  5 X106
              Fin-side Fan Power, Watts
                                                  0
                                                        1X106   2X106   3X106   4X106  5 X106
                                                              Fin-side Fan Power, Watts
  100

=S  go
u
8  60
I
S  40
ro

g  20
                      Case 6
  100

^_ 80


j"
£ 40
i
Q ,n
U ^u
                                                                   Case 7
          1 X 106   2 X 106   3 X 106  4 X 106  5 X 106     °
              Fin-side Fan Power, Watts
                                                         1 X 106  2 X Id6   3 X 106   4 X 106  5 X 106
                                                              Fin-side Fan Power, Watts
Appendix Figure 4.9 FGD water recovery % vs. fin-side fan power for the highlighted cases.

               100 „-
                           1X106
                                       2 X 106       3 X 106
                                       Fin-side Fan Power, Watts
                4X106
5X106
Appendix Figure 4.10 Effect of fin-side fan power on FGD water recovery % for case 7w-2.

       As expected, increasing the fin-side fan power results in a higher amount of water
recovered. Also, the lower the cooling air temperature, the better the heat transfer that was
achieved, and thus more water is condensed, as observed in Appendix Figure 4.11. Here, the
                                             98

-------
effect of the outside air temperature on water recovery for different fan-power curves is
presented. The number in parenthesis for each fan-power curve represents the percentage on the
power requirement based on total power generation of a 500 MW power plant.
                                70           80           90
                                   Outside Air Temperature, °F
100
Appendix Figure 4.11 Effect of outside air temperature on FGD water recovery % for case
       7w-2.
   4.9.3.9  wESP Design Parameters

       The second part of the heat transfer optimization process involved the definition of ESP
design parameters directly related to the ESP dimensions, which also influences the heat transfer
phenomena. As indicated in the Methodology section, new cases for the ESP dimensions were
proposed and analyzed, including the original case presented in the previous section. The
dimensions analyzed along with the ESP parameters are presented in Appendix Table 4.12.
       The purpose of this part of the process was to define ESP dimensions that would result in
design parameters that best approach the typical ESP design values, or fall between the typical
ranges. This would guarantee high particulate collection efficiencies and a better operation. The
design parameters considered for the purpose of this analysis were the aspect ratio (AR), the
specific collection area (SCA) and the gas velocity. These parameters were determined for five
different cases including the original case. As observed in Appendix Table 4.12, one of the
problems with the original case is the low SCA and the high gas velocity achieved, which could
result in significantly low ESP performance. In light of this, the length, height and number of
ESPs were varied. The results indicate that when increasing the ESP's length and height, but
maintaining the AR within the typical range, a SCA within the typical range can be achieved.
Furthermore, the number of ESPs required to lower the gas velocity, and thus improve the
collection efficiency, has to be between 40 and 60 for the specific dimensions chosen.
                                           99

-------
     Appendix Table 4.12 Proposed cases for ESP dimensions and design parameters
               calculated
Case/Parameter
Flue gas rate (acfm)
ESP Length (ft)
ESP Height (ft)
ESP Channel Width
(ft)
No. ESP
AR
SCA
(ft2/1 000 acfm)
Gas Velocity (ft/s)
Original

20
20
1.5
25
1
20
21.8
A
9
50
40
1.3
60
1.25
245
5.2
B
80E+05
60
40
1.3
60
1.5
294
5.2
C

50
40
1.3
40
1.25
163
7.9
D

40
20
1.5
40
2
65
13.6
Typical Design Values
-
-
-
0.5-1.3
-
1-2
200 - 800
5-6lfAR=1.5
       It is important to note that the aspect ratio and the specific collection area play an
important role in the ESP collection efficiency and particle re-entrainment; however, the load of
particles in the flue gas entering the water recovery system is expected to be very low because
the flue gas had already gone through a particulate control system. Therefore, typical ESP design
values do not need to be strictly followed, but close approximations are desired. The gas
velocity, on the other hand, represents an important parameter for water vapor condensation,
where long residence times inside the ESP channels are required in order to reach a higher heat
transfer.
       Three cases A, B and C were simulated in order to find the water recovery efficiencies
for each case. The only difference between cases A and C is the number of ESP channels, 60 for
Case A and 40 for Case C. Case D was not considered for further analysis since the design
parameters calculated fall far off the typical ranges. For the new cases, ten different fin
configurations were proposed based on the results obtained in in the previous section. Simulation
results indicated that cases A-9 and C-9 exhibit the  highest water recovery while minimizing fin-
side fan power.  Also, the cost of the systems  in Cases A and C are expected to be lower when
compared to Case B since the ESPs length is smaller.
       The results from simulating Case A for each fin configuration are given in Appendix
Table 4.13, which shows the recovered FGD makeup water percentage, the total water per ESP
and the total fan power for a specific gas velocity for all the specific cases.  It can be observed
that the percentage of FGD makeup water recovered from the system  is high for all the
configurations proposed; thus, this is not a limiting  variable. On the other hand, the fin-side fan
power requirements vary from one configuration to the other. For instance, the total fan power
needed decreases when increasing the number of fins attached to the ESP walls. The total fan
power is directly related to the number of sections and number of ESP channels in the system. As
                                          100

-------
a reminder, a section corresponds to the channel on both exterior sides of the ESP where the
cooling air flows between two fins (see Appendix Figure 4.4). Therefore, decreasing the number
of fins decreases the number of sections, and thus, the total fan power required is reduced. The
same would be expected if the number of ESP channels is reduced, as in Case C. In summary,
the configurations of cases A6, A8, A9 and A10 exhibited the best results when considering
water recovery maximization and fan power minimization.
     Appendix Table 4.13 Results for Case A (length = 50 feet, height = 40 feet, no. ESPs
               60)
Case
A1
A2
A3
A4
A5
A6
A7
A8
A9
A10
FGD Make Up
(%)
266
217
267
264
268
265
262
261
244
258
Total Water per ESP
(gallon/hr)
911.64
744.31
915.87
905.34
917.86
909.87
897.47
896.15
837.49
884.78
Total fan Power
(Watts)
7.76E+07
9.02E+07
7.36E+07
1.05E+08
9.81 E+07
4.17E+07
4.48E+07
4.06E+07
1.75E+07
1.80E+07
       The simulations were also carried out for Case B and the results are presented in
Appendix Table 4.14. The behavior of the different configurations for Case B was similar to that
of the Case A configurations. It is observed that, even though the length of the ESPs is longer
than in Case A, no significantly higher water capture efficiencies are achieved when simulating
Case B. In addition, the fin-side fan power is not greatly reduced. Thus, Case B was not
considered for further analysis since its implementation supposes higher costs and footprint than
cases A and C, with no significant improvements in performance.
       In view of the results presented in Appendix Tables 4.13 and 4.14, two cases were chosen
for further analysis in this section. The two cases correspond to those where the highest water
recovery  is reached, while the least fin-side fan power consumption is required (Case A9 and
Case C9).
       Case A9
       The dimensions of the ESP  channel for the proposed case are:
       Length = 50 ft.
       Height = 40 ft.
                                          101

-------
       Width =1.3 ft.
       Number of ESP Channels = 60
     Appendix Table 4.14  Results for Case B (length = 60 feet, height = 40 feet, no. ESPs
               = 60)
Case
B1
B2
B3
B4
B5
B6
B7
B8
B9
B10
FGD Make Up (%)
268
259
268
268
269
267
267
263
247
261
Total Water per ESP (gallon/hr)
920.51
889.69
919.86
919.90
921.40
916.37
916.68
903.50
847.96
896.47
Total fan Power
(Watts)
7.28E+07
8.17E+07
7.01 E+07
9.81 E+07
9.31 E+07
3.97E+07
4.20E+07
3.89E+07
1.68E+07
1.72E+07
       The fin configuration with a better performance corresponded to 1000 fins for both
exterior walls of the ESP channel. The fin dimensions are:
       Fin length = 50 ft.
       Fin height = 40 ft.
       Fin depth = 1 ft.
       Fin thickness = Vi in.
       For Case A9, the water recovery as a function of fan power was estimated for different
cooling air temperatures, as shown in Appendix Figure 4.12. The water recovery is expressed in
gallons per minute or as a percentage of the water entering the system.
       It can be observed in Appendix Figure 4.12 that the system tends to reach optimized
water recovery values when the fan power is above 3 MW and the cooling air temperature is
equal to or below 60 °F. In those cases, the system reaches water recovery efficiencies higher
than 85 percent. Nonetheless, when the temperature is higher (80 °F), efficiencies achieved can
be as high as 70 percent or more.
       In Appendix Figure 4.13, the water recovery percentage is presented as a function of
cooling air temperature for different fin-side fan power. As expected, higher cooling air
temperature results in lower water capture, which reaffirms the results from Appendix Figure
4.12. The same results are shown in Appendix Figure 4.14, but the water recovery is expressed in
gallons per minute.
                                          102

-------
            1200
                                                     100
     2 X 106        3 X 106
Fin-side Fan Power, Watts
	40 °F
	60 °F
          80 *F
         100 °F
               -  40 °F
               -  60 °F
                  80 °F
                 100 8F
Appendix Figure 4.12  Water recovery vs. fin-side fan power for Case A9.

-------
  100
   80
   60
 o
 <_t
 01
 £40
   20
                             Fan Power = 6 MW
                                      Fan Power = 5 MW
Fan Power = 3 MW
Fan Power = 2 MW
Fan Power = 1 MW
Fan Power = 500 kW
    40        50         60         70         80        90        100
                           Outside Temperature, °F

Appendix Figure 4.13  Water recovery percentage vs. cooling air temperature for Case A9.
  100
   so
   60
 _
 S
   20
                             Fan Power = 6 MW
                                      Fan Power = 5 MW
Fan Power = 3 MW
Fan Power = 2 MW
Fan Power = 1 MW
Fan Power = 500 kW
    40
      50
60        70        80
   Outside Temperature, °F
90
100
Appendix Figure 4.14 Water recovery in gallons per min vs. cooling air temperature for Case
      A9.
                                      104

-------
       Case C9
       The dimensions of the ESP channel for the proposed case are:
       Length = 50 ft.
       Height = 40 ft.
       Width =1.3 ft.
       Number of ESP Channels = 60
       The fin configuration for this case is the same as Case A9. The results are first shown as
water recovery as a function of fan power for different cooling air temperatures (see Appendix
Figure  4.15). For this case, an optimized water recovery is not achieved; however, efficiencies as
high as 60 percent are obtained for cooling air temperatures below 80 °F. A 60 percent water
recovery represents around 700 gpm, which is about twice the amount of makeup water required
for the  FGD system of a typical 500 MW coal power plant. The water recovery plot against the
cooling air temperature for various fan power is shown in Appendix Figures 4.16 and 4.17,
expressed in terms of recovery percentage and gallons per minute, respectively.
       In conclusion, the results indicate that, for both cases, the water recovery percentage
increases when the cooling air temperature decreases. Nevertheless, high recovery percentages
are obtained when the outside temperature is around 80 °F. Furthermore, at the same
temperature, a large amount of water can be recovered using less than 1% of the plant's
generation power (considering a 500 MW power plant). As presented in Appendix Table 4.12,
the only difference between  cases A and C, regarding design, is the number of ESP channels
proposed for the system. This difference could have significant impact on the system's cost, but
not on  water recovery, since the amount of makeup water required for the FGD system is less.
       Also, optimization conditions are obtained when increasing the number of ESP channels
and decreasing the cooling air temperature, as can be observed in Appendix Figure 4.12. With 60
ESP channels and a cooling  air temperature of 40 °F, the maximum  achievable water recovery
percentage is around 95, with a fan power higher than 2.5 MW.
                                          105

-------
    1200
    1000
                                                              100
E
Q_
QD
OJ
o
Ol
       0
        0
      - 40 °F
      -- 60 °F
      — 80 °F
      	100 "F
1 X 106         2 X 106        3 X 106
          Fin-side Fan Power, Watts
4 X 106         5X106
            	40 °F
            	60 °F
            	  80 °F
            	100 °F
      Appendix Figure 4.15  Water recovery vs. fin-side fan power for Case C9.

-------
      1200
      1000
     I
     §600
      400
       200
                                         Fan Power = 6 MW
                                                   Fan Power = 5 MW
Fan Power = 3 MW
Fan Power = 2 MW
Fan Power = 1 MW
Fan Power = 500 kW
                  50         60         70        80        90       100
                                Outside Temperature, "F

Appendix Figure 4.16 Water recovery percentage vs. cooling air temperature for Case C9.
      1200
      1000
                                         Fan Power = 6 MW
                                                   Fan Power = 5 MW
            Fan Power = 3 MW
            Fan Power = 2 MW
            Fan Power = 1 MW
            Fan Power = 500 kW
         40        50         60         70         80         90
                                Outside Temperature, F

Appendix Figure 4.17  Water recovery vs. cooling air temperature for Case C9.
                                    107

-------
4.9.4  Appendix references:
Anderson, M., E. Lee, I. Tejedor, and T. Holm. 2003. "Photoactive removal of As(III) from
      water using novel active material." 226th American Chemical Society National Meeting,
      New York, N.Y. September 7-11, 2003. Oral presentation.

Bayless, D. J., M.K. Alam, R. Radcliff, and J. Caine. 2004. "Membrane-based wet electrostatic
      precipitation." Fuel Processing Technology, 85(6):781-798.

Cheng, L., J.-Y. Lee, T.C. Keener, and J. Yang. 2011. "Wastewater treatment for wet flue gas
      desulfurization systems in coal-fired power plants." EM 5: 16.

Daal, L., H. Kamphuis, A. Stam, T. Konings, M. Huibers, S. Rijen, and J. Ruijter. 2012.
      Evaluation of Different Water Vapor Capture Technologies and Energy Modeling Results
      for Membrane Technology. DNV-KEMA, Arnhem.

Energy Information Administration (EIA). 2014. http://www.eia.gov/electricity/data.cfm.

EPRI. Electric Power Research Institute. 2007a. "National Drinking Water Regulations." EPRI
      (1012549).

EPRI. Electric Power Research Institute. 2007b. "Treatment Technology Summary for Critical
      Pollutants of Concern in Power Plant Wastewaters" EPRI (1012549).

Folkedahl, B.C., G.F. Weber, andM.E. Collings. 2006. "Water Extraction from Coal-Fired
      Power Plant Flue Gas." Final Report from the University of North Dakota Energy &
      Environment Research Center to U.S. DOE/NETL. Report No. 2006-EERC-12-05.
      http://www.netl.doe.gov/File%20Library/Research/Coal/ewr/water/41907-Final.pdf

Heimbigner, B. 2007. "Treating FGD Wastewater: Phase 2 Clean Air Act Amendments make it
      Hot Topic." Industrial Water WorId 7(1).

Higgins, T.E., A.T. Sandy, and S.W.  Givens. 2009. "Flue gas desulfurization wastewater
      treatment primer." Power 153(3).

Holman, J. P. 1997. Heat Transfer. McGraw-Hill, New York.

Kabay, N., I. Yilmaz, S. Yamac, M. Yuksel, U. Yuksel, N. Yildirim, O. Aydogdu, T. Iwanaga,
      and K. Hirowatari. 2004. "Removal and recovery of boron from geothermal wastewater
      by selective ion-exchange Resins-II. Field Tests." Desalination, 167:427-438.

Khang, S. J., T.C. Keener, and J.Y. Lee. 2008. "Air-Cooled Condensing Wet ESP for Water
      Recovery and Removal of Residual Pollutants from Flue Gas." University of Cincinnati
      Invention Disclosure  109-005.

Levy, E., H. Bilirgen, and J. DuPont. 2011. "Recovery of Water from Boiler Flue Gas using
      Condensing Heat Exchangers." Final Project Report DOE/NETL Project DE-
      NT0005648.

Levy, E., H. Bilirgen, K. Jeong, M. Kessen, C. Samuelson, and C. Whitcombe. 2008.  Recovery
      of Waterfront Boiler Flue Gas.


                                         108

-------
Merrill, D.T., M.A. Manzione, D.S. Parker, JJ. Petersen, W. Chow, and A.O. Hobbs. 1987.
      "Field evaluation of Arsenic and Selenium removal by Iron coprecipitation."
      Environmental Progress, 6(2): 82-90.

Perry, R., D. Green, and J. Maloney. 1998. Heat transfer equipment. Perry's Chemical Engineers'
      Handbook (7th Edition), Section 11.

Sonstegard, J., and T. Pickett. 2009. ABMet® Biological Selenium Removal from FGD
      Wastewater. U.S. EPA Docket ID No. EPA-HQ-OW-2009-0819-1232.

U.S. EPA. 2013. "Critical Review of Treatment Technologies for Removal of Salts [Ca, HCO3,
      K, Mg, Na, SO4] from Water: Potential for Application to Coal Mining Impacted Water."
      National Risk Management Research Laboratory, Cincinnati, OH.

Wang, D., A. Bao, W. Kunc, and W. Liss. 2012. "Coal power plant flue gas waste heat and water
      recovery." Applied Energy, 91(l):341-348.
                                         109

-------
 5   Natural Gas Electric Power Generation and Water Usage
       Marissa S. Liang,' Timothy C. Keener,' Wei-Ning Wang," Pratim Biswas," Colin White,'"
       J. McDonaldiv

5.1  Introduction

       Natural gas is the second most common fossil fuel used for electric power generation,
accounting for 30% of total U.S. electricity production in 2012. Due to the relatively low natural
gas prices, relative abundance of natural gas and capital costs, a natural gas plant is a more
competitive choice for new generation capacity. According to EIA predictions in reference cases,
natural gas power plants will account for 63% of electricity capacity additions from 2012 to 2040
(EIA, 2012).
       Three aspects affect the use of natural gas in electrical generation: 1) the price of natural
gas compared with other fossil fuels, especially coal, 2) the ability to mine the significant
domestic reserves of natural gas, and 3) policies that reduce greenhouse gas (GHG) emissions.
An example of one such policy is the 2014 performance standard that was proposed to restrict to
1,000-1,100 pounds  of carbon dioxide per megawatt hour (MWh) for  newly constructed coal  and
natural gas-fired power plants to curb their GHG emissions.25
       Significant amounts of water are used in natural gas power plants, and other power plants
in general. They require large volumes of water for the generation of electricity, primarily to turn
turbines or for cooling in thermoelectric generation. Another use for water more specific to
natural gas is the consumption of large quantities during the gas extraction process. Both uses
greatly impact water resources, particularly freshwater, throughout the U.S.

5.2  Gas-fired Boiler Thermoelectric Plants

       According to the EIA database, about 8.2 trillion cubic feet of natural gas was consumed
as fuel  for conventional steam turbines for electricity generation in 2013. This amount doubled
the natural gas usage for electric power production since 1997. Electric power generated by
natural gas-fired boilers reached nearly 3,112 MMBtu in 2010.
       Natural gas-fired boiler steam turbines require cooling systems. Cooling system
technologies have been improving steadily and have significantly decreased water usage in
thermoelectric power plants over the past several years. Recirculating systems and recently
developed dry cooling systems have been widely used in  new thermoelectric power plants in
place of once-through cooling systems. In particular, recirculating cooling technologies have
been incorporated into about 200 newly constructed power plants built between 2000 and 2004,
1   University of Cincinnati, Department of Biomedical, Chemical, and Environmental
   Engineering
"  Washington University in St. Louis
'"  Pegasus Technical Services, Inc., Cincinnati, OH
1V  U.S. Environmental Protection Agency - National Risk Management Research Laboratory

25  See Chapter 2 of this report for further information regarding energy use trends.
                                          110

-------
including natural gas-fired boiler steam turbine thermoelectric plants. Water in circulating
cooling systems is kept in a closed-loop and reused; thus, water withdrawals are lower compared
to once-through processes. In dry cooling systems, air is used as a working fluid to lower the
cooling water temperature instead of water. Due to the significant amount of electricity required
for running cooling fans, dry cooling systems are more suitable for small scale plants and gas-
fired power plants.
       Among the 1,655 cooling systems in power plants that can each provide a combined net
summer capacity of 100MW or more in the U.S., 674 cooling systems are employed by natural
gas power plants. Of these, 422 (62.6%) are recirculating systems, while 197 (29.2%) are once-
through systems. The rest (51) are mostly dry-cooling systems (7.6%) and 1 hybrid system,
which can switch between wet and dry cooling (See Table 5.1)
     Table 5.1  System types by primary energy source in 2012
Primary energy
source
Coal
Natural gas
Nuclear
Other
Total
Once-
through
398
197
50
74
719
Recirculating
368
422
44
41
875
Dry cooling
4
51
0
1
56
Wet & dry
hybrid cooling
1
4
0
0
5
Total cooling
systems
771
674
94
116
1,655
       Note: Adopted from EIA (2012)

5.3  Gas Turbine and Boiler Cogeneration

       Gas turbine power plants primarily use natural gas as an energy source, though synthetic
gas from coal production can also be used within these plants. Gas turbine technology has
steadily advanced in recent decades, and represents a new trend for electricity generation. The
advantages of gas turbines include: 1) higher efficiency when used in a combine cycle
configuration, 2) flexibility in being turned on and off to meet electricity demand, and 3)
computer-based design and the development of advanced materials enable higher efficiency.
       The basic configuration of gas turbines mainly consists of a compressor (either a
centrifugal or axial), a combustion chamber, and a turbine integrated with an electrical generator.
Different from steam turbines, gas turbines use air instead of water as the working fluid. After
being accelerated by the compressor and slowed by a diffuser, fresh air flow is brought to a
higher pressure and mixed with natural gas. The mixed gas is then ignited to combustion. The
high-pressure and high-temperature gas produces shaft work output by expanding through the
turbine.
       The Brayton cycle is the ideal thermodynamic cycle typically used to represent gas
turbine operation. The Brayton cycle contains the following three thermodynamic processes:
1.   Isentropic compression: Air flow is drawn into the compressor and pressurized after
    acceleration through the compressor  and deceleration through the diffuser.
                                          Ill

-------
2.   Isobaric combustion: The mixture of compressed gas (air) and fuel (natural gas) is
    combusted under constant pressure to increase the specific volume.
3.   Isentropic expansion: The heated, pressurized, larger volume gases are expanded through a
    turbine.
       The exhaust products are then ejected at isobaric conditions from the turbine at the
original pressure. In practical  applications, energy losses occur during each process due to
friction and turbulence. In order to fully expel the exhaust gases, some pressure still remains in
the exhaust gases instead of returning it to the original pressure of the intake air.
       The efficiency of a gas turbine is limited by the temperatures and mechanical stress that
the engine materials can withstand. Due to the energy lost in the Brayton cycle and residual
energy in exhaust gas, the thermal efficiency is as low as 35-40% for a single-cycle gas  turbine
that produces 100 to 400 MW of electric power (Cengel et al., 2011; Ratliff et al., 2007)
       In electric power generation, a natural gas combined cycle (NGCC) configuration can
significantly increase thermal efficiency of gas turbines, up to 60% according to some reports
(Yuri et al., 2013; Hada et al., 2012; Uchida et al., 2012). A widely used configuration includes
the use of one or more natural gas turbine generators and a heat recovery steam generator
(HRSG), which is operated  using the Rankine cycle by extracting energy from  the hot exhaust
gases generated by the gas turbine. In this manner, the steam generator is used to recover a
significant fraction of waste heat from the gas turbine. The  steam turbine of the HRSG is
powered by high pressure steam and generates additional electricity. Low pressure steam exits to
a cooling tower and is condensed to warm water to recharge the HRSG (See Figure 5.1).
                                                       Hot Steam
             Cooling
             Tower
         LZ
IJ
 Makeup
 Water
  J
                          /"Steam ,
                         /CondenseX
                                                            Fuel for Optional
                                                              Duct Burner
                    Waste
                    waste
                    Water    MakeuDJ
                             Water
Figure 5.1  Schematic drawing of the gas turbine combined cycle power plant.
                                           112

-------
5.4  Water Usage

       As described earlier, producing energy from fossil fuels such as natural gas often
involves substantial amounts of water. Thermoelectric power plants generate power by boiling
water to produce steam that drives electricity-generating steam turbines. Furthermore, large
quantities of water are typically used to cool the steam to complete the power cycle.
Thermoelectric power generation consumed nearly 196 billion gallons per day, primarily for
cooling, in 2000. Approximately 70 percent of all thermoelectric water withdrawals are obtained
from limited supplies of freshwater, accounting for nearly 40 percent of all freshwater
withdrawals in the U.S. in 2000 (Dziegielewski, 2006). For natural gas electric power generation,
water is used during power generation, but it is also used to extract the natural gas (e.g.,
enhanced gas recovery, coalbed methane extraction, and shale gas extraction) from its
underground source.
5.4.1   Water usage for fuel extraction
       Natural gas in the U.S. has typically been extracted via drilling deep vertical wells that
require only relatively small amounts of water (UCS, 2013). However, on an overall basis, this
method generates significant amounts of "produced water" (~ 200 billion gallons/year) (U.S.
DOE, 2006; UCS, 2013). Produced water is generated from naturally occurring fluids in natural
gas-bearing formations (U.S. DOE, 2009). Several methods have been developed for disposing
of produced water, such as pumping it back into oil- or gas-producing wells to bolster
production, or injecting it deep into other formations away from groundwater resources (UCS,
2013).
       Over the past decade, the proportion of domestic natural gas production derived from
shale gas (unconventional gas reservoirs) has significantly increased, primarily due to
technological developments and innovations. As  a result, the price of the natural gas has steadily
decreased and shale gas has become a significant new source of natural gas in the U.S.
"Hydraulic fracturing" or "hydrofracking" has improved the economics  of accessing natural gas
from shale deposits. The hydraulic fracturing process is schematically shown in Figure 5.2.
Hydraulic fracturing of shale typically combines vertical drilling with horizontal drilling to
follow gas deposits (UCS, 2013). A fracturing fluid consisting of approximately 90% water, 9%
sand, and 1% chemical additives is injected into the gas deposits at high pressures and creates
fractures in the surrounding rock, which allows the natural gas to flow to the production well  and
through to the wellhead where it can be collected for distribution (UCS 2013). In 2012,  shale  gas
made up approximately 30% of total U.S. natural gas production, and  is  anticipated to grow to
almost 50% by 2040 (EIA, 2012).
       In spite of the economic advantages of increased shale gas extraction, there are several
potential environmental risks from increased use of hydraulic fracturing. For example,
groundwater could be contaminated with natural gas, volatile organic  compounds and/or the
chemicals used in the gas extraction process. A single hydraulic fracturing treatment has been
estimated to yield 15,000 gallons of chemical waste from the fracturing fluids if not properly
managed (Kenny et al., 2009). The quantity of produced water and sufficient treatment or reuse
of produced water may also be challenging. While the total amount of water required for
hydraulic fracturing is relatively small compared to the water withdrawn for thermoelectric
generation or for agriculture, the amount of water required for hydraulic fracturing may still be

                                          113

-------
 Roughly 200 tanker
 trucks deliver water lor
 the fracturing process.
 0 Pvet
 1.000
 2.000
 3rOOO
 4.000
 5,000

 6.000
 7,000
        A pumper truck injects a
        mix o( sand, water and
        chemicals into the well.
                      -r Tldh-T-i	J
                        frQ=MQO—-  PC?
Natural gas flows out of well
                                      i plant.
                            Water table
                         Well —
Hydraulic Fracturing
Hydraulic fracturing, or
"tracking." involves the injection
of more than a million gallons
of water, sand and chemicals
at high pressure down and
across into horizontally drilled
wells as far as 10,000 feet
below the surface. The
pressurized mixture causes the
rock layer, in this case the
Marcellus Shale, to crack.
These fissures are held open
by the sand particles so that
natural gas from the shale can
flow up the well.
                                           Well turns
                                           horizontal
                 Marcellus Shale
red water is stored in open
tn taken to a treatment
1 ' r^<
'OO ' ' OO' O

Storage Natural gas is piped
tanks to market.

~


1 	 ii 	 y —
        Natural gas
        flows from
        fissures
        into well
Mixture of
water, sand
and chemical
agents
                                                             The shale is fractured
                                                             by the pressure inside
                                                             the well.
                                                                                                       Graphic by Al Granberg
Figure 5.2  Diagram of hydraulic fracturing of natural gas. Adopted from Granberg,/ProPublica (2013).

-------
regionally or locally significant since hydraulic fracturing wells are often not collocated with
surface water sources (UCS 2013). Thus, local water sources could be adversely impacted
depending on the natural gas well  size, gas flow volume and duration, and number of other
natural gas wells in a particular area.

5.4.2   Water usage for electricity generation
       A majority of the natural gas-fired power plants in the U.S. are composed of natural gas
combined-cycle (NGCC) units (EIA, 2012). NGCC power plants require lower quantities of
water for cooling compared to conventional steam turbine technologies used in plants with
natural gas-fired boilers or in coal  or nuclear power plants (U.S. DOE, 2011). In a typical NGCC
power plant, dry cooling systems are used, which are more economical and smaller (about 30%)
than other thermoelectric options in a coal or nuclear power plant with the same electricity
output (GAO, 2009). About 8% of natural gas combined cycle plants in the U.S. use  dry cooling
technology, while 80% rely on recirculating systems. Fewer than 7% use once-through cooling
(Union of Concerned Scientists, 2013). Table 5.2 summarizes the water requirements for
different types of natural gas power plants (Macknick et al., 2012; Union of Concerned
Scientists, 2013).
     Table 5.2 Water requirements for natural gas power plant in gal MW"1 h"1 *


Natural Gas
Steam Turbine
Natural Gas
Combined
Cycle
Natural Gas
Combustion
Turbine
Once-Through
Withdrawal Consumption
10,000-60,000 95-291
7,500-20,000 20-100
0 0
Recirculating
Withdrawal Consumption
950-1,460 662-1,170
150-283 130-300
0 0
Dry-Cooling
Withdrawal Consumption
0-4 0-4
0-4 0-4
0 0
   Note:    * Union of Concerned Scientists, 2013: http://www.ucsusa.org/clean energv/our-energv-
            choices/energv-and-water-use/water-energY-electricitY-natural-gas.html.
       Based on water sources or water intake processes, once-through cooling systems and
recirculating cooling systems can be further grouped into the following subcategories (EIA,
2012):
          Once-through cooling systems:
            •   OC: Once through, with cooling pond(s) or canal(s)
            •   OF: Once through, freshwater
            •   OS: Once through, saline water

       -   Recirculating cooling systems:
            •   RC: Recirculating with cooling pond(s) or canal(s)
            •   RF: Recirculating with forced draft cooling tower(s)
                                           115

-------
            •   RI: Recirculating with induced draft cooling tower(s)
            •   RN: Recirculating with natural draft cooling tower(s)
       In general, water withdrawal for the thermoelectric power plants that use closed-cycle
cooling systems is much less than those equipped with once-through systems. In 2005,
recirculating cooling systems accounted for 8% of the total water withdrawal for thermoelectric
power plants; once-through systems made up for 92% (Kenny et al., 2009). According to the EIA
database  for the year 2010, cooling water withdrawal ranged from 0.30 to 8.05 m3 per 100MJ
generation for the plants having once-through cooling systems. This value ranged from 0.004 to
0.32 m3 per 100MJ electric generation for recirculating cooling systems, except those with
cooling ponds or canals (See Table 5.3)
  Table 5.3  Annual average withdrawal rate of cooling system for natural gas-fire steam turbine
            thermoelectric power plants
Cooling
type
oc
OF
OS
RC
RF
RI
RN
Annual average withdrawal rate
(m3/sec)
Max.
580.98
1465.49
325.89
859.61
65.60
17.22
37.58
Min.
22.75
29.08
182.33
2.60
2.20
0.45
19.02
Median
402.46
340.99
269.00
445.28
7.10
11.50
37.58
Cooling water withdrawal
per generation
(10-2m3/MJ)
0.30 - 7.50
0.35-8.05
1.50-7.11
0.05-9.40
0.04-0.32
0.004-0.12
0.20-0.21
Cooling unit,
n
4
50
6
13
28
24
3
5.5   Summary

       Natural gas is currently the second most-used fossil fuel for electric power generation.
Increased natural gas production through hydraulic fracturing, relatively lower prices of natural
gas, improved thermal efficiency of combined cycle gas turbine power plants, and reduced
carbon emissions make natural gas power plants an increasingly attractive option for new electric
generating capacity when compared with coal-fired power plants. EPA recently proposed new
GHG regulations for thermoelectric power plants, both new and old (see Chapter 2), which may
result in further increased usage of natural gas for electric power production. Moreover, existing
coal power plants are likely to stop increasing capacity. Thus, additional measures to reduce
carbon emissions will compete with expanding natural gas electric power generation.
       Water usage for natural gas power plants is an important factor to be considered in both
evaluating the current and future electric power plants. Reducing water usage is a key objective
to protect the environment and reduce costs. Natural gas prices have driven the recent expansion
of its use in power plants. However, there are concerns regarding water usage and contamination
from the process of acquiring it unless extraction processes and treatment of contaminated water
                                          116

-------
are managed in a sustainable manner. New technologies for fuel extraction and cooling are
essential to achieve the goal of minimal water use and the use of environmentally benign
technologies. Dry cooling systems in NGCC power plants are one such example, and should be
considered and used more often. NGCC power plants will result in a significant decrease in the
overall water footprint of electric generation from fossil fuels as they replace older coal-fired
power plants.
       Coupled with the industry's expected shift away from coal, water usage will further
decrease with new cooling technologies. The effects would be particularly significant in areas
with power plants nearing retirement. Specifically, replacement of these plants with competitive
and more efficient NGCC plants can reduce water consumption in water-stressed regions, or in
regions with shared aquifers. Both policy and market trends may accelerate the increased use of
natural gas within future electric power generation in the U.S., thus decreasing water
withdrawals and improving water quality.
5.6  References

Cengel, Y. A., M. A. Boles, and M. Kanoglu. 2011. Thermodynamics: An Engineering
       Approach. Vol. 5. McGraw-Hill, New York.

Dziegielewski, B., T. Bik, U. Alqalawi, S. Mubako, N. Eidem, S. Bloom. 2006 "Water Use
       Benchmarks for Thermoelectric Power Generation." Research Report of the Department
       of Geography and Environmental Resources Southern Illinois University Carbondale,
       Carbondale, IL 62901. http://web.extension.Illinois.edu/iwrc/pdf/238.pdf.

EIA. 2012. "Annual Energy Outlook 2013, Early Release Overview." 16 p.
       http://www.eia.gov/forecasts/aeo/er/pdf/0383er(2013).pdf. Accessed July 2014.

EIA. 2012. "AEO2013 early release overview. Table 1: Comparison of projections in the
       AEO2013 and AEO2012 reference cases, 2010-2040, Washington, DC."
       http://www.eia.gov/forecasts/aeo/er/pdf/0383er(2013).pdf. Accessed July 2014.

Government Accountability Office (GAO). 2009. "Energy-Water Nexus: Improvements to
       Federal Water Use Data Would Increase Understanding of Trends in Power Plant Water
       Use. Washington, DC."  http://www.gao.gov/products/GAO-10-23. Accessed July 2014.

Hada, S., M. Yuri, J. Masada, E. Ito, and K. Tsukagoshi. 2012. "Evolution and Future Trend of
       Large Frame Gas Turbines: A New 1600 Degree C, J Class Gas Turbine." ASME Turbo
       Expo 2012: Turbine Technical Conference and Exposition, American Society of
       Mechanical Engineers, 599-606.

Kenny, J.F., N.L. Barber, S.S. Hutson, K.S. Linsey, J.K. Lovelace, and M.A. Maupin. 2009.
       "Estimated use of water in the United States in 2005: U.S. Geological Survey Circular
       1344." 52 p. http://pubs.usgs.gov/circ/1344/. Accessed July 2014.
                                          117

-------
Macknick, J., R. Newmark, G. Heath, and K.C. Hallet. 2012.  "Operational water Consumption
       and Withdrawal Factors for Electricity Generating Technologies: A Review of the
       Existing Literature." Environmental Research Letters, 7:045802. 10pp.

ProPublica (Graphic by A. Granberg). 2013. "What Is Hydraulic Fracturing?".
       http://www.propublica.org/special/hydraulic-fracturing-national. Accessed July 2014.

Ratliff, P., P. Garbett, and W. Fischer. 2007. "The New Siemens Gas Turbine SGT5-8000H for
       More Customer Benefit." VGB Power Tech 87:128-132.

Uchida, S., A. Tsutsumi, Y. Takashima, and S.  Shiozaki.  2012. "Mitsubishi Latest Gas Turbine
       Combined Cycle Technology-World's First 1,600 C J-Series Gas Turbine."
       https://www.mhi-gl obal.com/company/technology/review/pdf/e491/e491018.pdf.
       Accessed July 2014.

Union of Concerned Scientists (UCS). 2013. "How it Works:  Water for Natural Gas."
       http://www.ucsusa.org/clean_energy/our-energy-choices/energy-and-water-use/water-
       energy-electricity-natural-gas.html. Accessed July 2014.

U.S. DOE. 2011. "Estimating Freshwater Needs to Meet  Future Thermoelectric Generation
       Requirements. Washington, DC." http ://netl .doe,  gov/research/energy-
       analvsis/publications/details?pub=371d7138-a050-46a8-a933-ea001c8f40ba. Accessed
       July 2014.

U.S. DOE. 2009. "Oil and Natural Gas Water Resources  Program. Washington, DC."
       http://www.springsgov.com/units/boardscomm/OilGas/Dept%20of%20Energy%20-%20
       Oil%20and%20Natural%20Gas%20Water%20Resources%20program.pdf. Accessed July
       2014.

U.S. DOE. 2006. "Energy Demands on Water Resources: Report to Congress on the
       Interdependency of Energy and Water." US Department of Energy, Washington, DC.
       http://www.docstoc.com/docs/49340912/Energy-Demands-on-Water-Resources_-Report-
       to-Congress—At-the. Accessed July 2014.

Yuri, M., J. Masada, K. Tsukagoshi, E. Ito,  and S. Hada.  2013. "Development of 1600 C-Class
       High-Efficiency Gas Turbine for Power Generation Applying J-Type Technology."
       Mitsubishi Heavy Industries Technical Review. 50(3): 1.
       http://www.mhi.co.jp/technologv/review/pdf/e503/e503001 .pdf. Accessed July 2014.
                                         118

-------
 6   Corn-Starch-Based Ethanol Production and Impacts on Water
     Resources
       Wei-Ning Wang1, Ying Li1, and Pratim Biswas1

6.1  Introduction

       Currently, ethanol constitutes approximately 93% and 91% of all biofuels produced 94%
of all biofuels produced in the U.S. in 2012 and 2013, respectively (EIA, 2014). Greater qualities
of ethanol are expected to be used as a transportation fuel in the future because of federal
policies. For example,  according to the Renewable Fuel Standard (RFS2), the volume of corn-
starch ethanol is capped at 13.8 billion gallons in 2013, but grows to 15  billion gallons by 2015
and is fixed thereafter (Schnepf and Yacobucci, 2013).
       At present, virtually all fuel ethanol in the U.S. is produced from fermentation of corn in
dry and wet milling plants, most of which are located  in U.S. Midwest states. In 2004, two-thirds
of total fuel ethanol was produced from dry mill plants, and the remaining one-third from  wet
mill plants (Wang,  2005). Besides corn-based and sugarcane-based varieties, ethanol can also be
produced from cellulosic biomass through fermentation of cellulose and semi-cellulose.
       Federal policies mandating increased use of ethanol and other biofuels in transportation
fuels may be hampered by the water issues faced in the corn-producing  areas of the U.S. unless
biofuel feedstock cultivation is transitioned to less water-intensive biomass crops, e.g.,
transitioning ethanol production from corn starch to a cellulosic feedstock such as switch grass27.
The majority of the corn grown in the U.S. is in Midwest states and a number of corn-producing
the  states that rely on groundwater for irrigation, such as Iowa and Nebraska, are in regions
where groundwater levels are falling. On a world-wide basis, the average consumptive water-
intensity of ethanol is 1826 gallons-FbO/gallon-ethanol, with 1260 gallons-FbO/gallon-ethanol
due to evapotranspiration and 566 gallons-FbO/gallon-ethanol due to irrigation. Average water
intensity of corn irrigation for ethanol production in the U.S. is lower on average than world-
wide figures at 113 gallons-FbO/gallon-ethanol and is highly variable, ranging from 15 to 934
gallons-FbO/gallon-ethanol (King and Webber, 2008), depending upon  regional differences in
irrigation levels.
       Previous studies and analyses have focused on overall net energy values (Farrell et al.,
2006; Graboski, 2002; Shapouri et al., 2004;  Wang, 2001) and net greenhouse gas (GHG)
emissions of corn-starch-based ethanol (EPA, 2010; Pimentel and Patzek, 2005; Shapouri et al.,
2004; Wang, 2001). However, less literature has been reported on the relationships between
energy input, water consumption, wastewater discharge, and CO2 emissions in the biorefmery
phase (i.e., conversion of corn to ethanol in ethanol plants). Pimentel and Patzek reported  that in
the  corn-to-ethanol process,  15 liters of water are mixed with each kg of corn, and to make 1 liter
of 99.5% ethanol, an input of 40 liters of water is needed when not including cultivation or other
needs (Pimentel and Patzek, 2005). Additionally, for each liter of ethanol produced,  about 13
1   Washington University in St. Louis

27  Cellulosic ethanol production is discussed separately within Chapter 7 of this report.
                                          119

-------
liters of very high-strength wastewater are produced. While it is not certain whether this data
took water recycling into consideration, other sources have reported much less water use when
water recycling is applied intensively in plant operations.
       The objective of the analysis contained within this chapter is to systematically evaluate
water use in the post-cultivation ethanol production processes by studying ethanol production
plants. The methodology applied is a mathematic modeling of the energy balance (including
thermal energy and electricity) and the mass balance (including corn input, water usage,
wastewater discharge, co-products and CO2 emissions) with respect to the various process
system boundaries. This analysis of corn ethanol production includes:
   a.   Construction of a mathematical model with an energy and mass balance for a typical
       corn-to-ethanol plant.
   b.  A summary of field visits to both pilot-scale and full-scale ethanol plants to gather first-
       hand ethanol production data.
   c.   A comparison of the different scales and successive generations of ethanol plants.
   d.  Recommendations for future research.
6.2  Energy and Mass Balance Model

6.2.1   Corn-to-ethanol production process overview
       Based on a literature review and various internet resources (Agricultural Marketing
Resource Center, 2001; ICM, 2007, 2014; National Corn-to-Ethanol Research Center, 2014;
McAlloon et al., 2000; Renewable Fuels Association, 2014; Seekingalpha, 2014), a system
diagram of the typical dry-mill corn-to-ethanol production process has been prepared (see Figure
6.1). The basic steps include milling, mashing, cooking, liquefaction, saccharification,
fermentation, distillation/dehydration, solids separation, evaporation and drying. If the ethanol
plant is taken as a single system, the overall inputs and outputs at the system boundary are shown
in Figure 6.2. The inputs are corn and water plus energy, while the outputs are ethanol, solids
(by-products), wastewater and CO2. In this chapter, a detailed analysis of energy and mass
balance calculations for each step in the process is presented.
                                           120

-------
                                Water
                                                                           To Atmosphere or
                                                                           Recovery Facilities
  Corn
       Grain Storage
Hammer
 Mill
Starch Slurry
  Tanks
   Fuel
  Fthanol
              Ethanol Storage       Q \       Molecular Sieve

                             Denaturant
                        Back to Slurry
                           Tank
                                                Liquid
                                                                                               co.
                                      Wastewater
                                       Discharge
                                   Centrifuge
                                  Grain Recovery
                                                                Wet Distillers
                                                                   Grain
                                                           Dried Distillers
                                                              Grain
                                                                            Grain Drying
Figure 6.1  System diagram of typical dry-mill corn-to-ethanol production process.
                                              Energy
Corn
Water



"* Corn to Ethanol Dry
Mill Process
1 1
Ethanol
Solids

                                          CO2   Wastewater
Figure 6.2  Inputs and outputs at the system boundary.
6.2.2   Mass balance model
        To resolve the material balance for the corn-to-ethanol plant, the unit operations of a
block flow diagram (BFD) must be defined (Mei, 2006). Figure 6.3 shows the BFD of a typical
ethanol plant with all the basic steps included.
                                                121

-------
( urn .
Wlfct
j:
i \v HI
_^ Millini! ^ Mushini:
A Mfil B
Irt »*irii prtr
!...* T
i.-i Diblillation

H«liil> I'D!
TruM
B«r
< iitumn i in
1 hin Miilnai
ViH C'lkr
ttmm
IMass ll:il;ilU'L-
c 1J 1 • Inkl
O - (lUtlft
OM,kiHft' <-»j Mi-MIH-MlH.
. . fc 11 .. ,, , .Ml- Mil •*• .MIZ + ...
' . Ma - Mol + Mo2 + „
CM
1 litiiL-l'iiJ Mjili \\ittr
H.J DM 1 Kil
,, , ^
KiTiiiL-ntation i>»! KM Scrubber — - — ^-CO^Oul
Icimnilri * WlttrnilH
(II- K"i
^ i.-i". ii. i ^»ilrwnt«r
lit,: T j.i T
tvaporalor Hui Jii Dryer tal
Njnip
ji:
Figure 6.3  Mass block flow diagram of ethanol production process.
       For each of the blocks, a material balance is written as:
       Material in = Material out
       or M. = M0 (/' = input; o = output)


                , = IX*
(6.2-1)

(6.2-2)


(6.2-1)
       where j represents the type of material inputs with a total of m inputs, and k represents
the type of material outputs with a total of n outputs.
       The mass balance calculation starts with corn inputs. Table 6.1 lists the composition of
corn (Mei, 2006). The starch (the actual material that makes ethanol) is then mixed with water
and turns into glucose followed by fermentation and the production of ethanol. The reaction
stoichiometry on a weight basis can be written as:
       Starch +  0.1111 Water -M. 1111 Glucose                                   (6.2-2)
       Glucose  -» 0.4589 Ethanol + 0.4641 CO2 + 0.05 other                        (6.2-3)
                                            122

-------
     Table 6.1  Corn composition
Component
Water
Starch
Protein
Oil
Other
Total
Mass Content %
15.0%
59.5%
7.7%
3.4%
14.5%
100%
       The yield of ethanol is a function of starch composition in corn, conversion efficiency of
starch to glucose and conversion efficiency of glucose to ethanol. Assuming a 100% efficiency
for both conversion processes, the typical yield is 2.5 to 2.85 gallons of ethanol per bushel of
corn (Mei, 2006). In this model, an average value of 2.7 gallons of ethanol per bushel of corn is
used as the corn-to-ethanol conversion rate. The CO2 emission profile can also be calculated
from equations 2-4 and 2-5 with a given mass of corn (or starch).
       Based on the equations 6.2-1 to 6.2-5, the process data and assumptions derived from
Mei (Mei, 2006), a spreadsheet-based mass balance model for the corn-to-ethanol production
process was established (Wang et al., 2014). Figure 6.4 shows an example worksheet of the mass
balance model.28 The model requires the amount of corn in kilograms as an input (cell B3 in
Figure 6.4). Based on the mass balance model illustrated in Figure 6.3, the model calculates the
mass of materials that go into and come out of each of the steps (column B for input and column
E for output in Figure 6.4) in the corn to ethanol production process (each box in Figure 6.3).
The final output from the model includes the amount of water used to produce ethanol (cell J6 in
Figure 6.4), and the amounts of ethanol (cell J8), carbon dioxide (cell J9), dry distiller grains
with solubles (cell J10) and wastewater (cell 111) produced. It demonstrates that with 1 kg input
of corn, 2.68 kg water is needed, and 0.32 kg ethanol and 0.33 kg dry distiller grains with
solubles (DDGS) can be produced with 0.31 kg CO2 emission and 2.72 kg wastewater discharge.
It should be noted that this model assumes that no water recycling technology is applied.
       28 An example of the model may be accessed via the following Internet URL:
http://www.aerosols.wustl.edu/education/energy/EthanolAudit/index.html
                                          123

-------

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
A
A
Inlet
Corn, Ai
B
Inlet
Corn powder,
BM
Water, Bi2
Inlet total, Bi
C
Inlet
Mash, CM
Steam, Ci2
Inlet total, Ci
D
Inlet
Liquefied
Mash, DM
Water, Di2
Inlet total, Di
K
Inlet
Fermenter
CO2, KM
Water, Ki2
Inlet total, Ki
Inlet
Fermented
Mash. Fi
Inlet total, Fi
G
Inlet
Solids, Gi
Inlet total, Gi
H
Inlet
This Stillage,
Hi
Inlet total, Hi
J
Inlet
Syrup, JM
Wet Cake, Ji2
Inlet total, Hi
B C
Milling
D
Outlet
Corn powder,
Ao

Mashing
Outlet
1 00 kg Mash, Bo
1.60 kg
2 60 kg Outlet total, Bo
Cooking/ Liquefaction
Outlet
Liquefied
2.60 kg Mash, Co1
Wastewater,
0.30 kg Co2
2.90 kg OuBet total. Co
Fermentation
Outlet
Fermented
2 67 kg Mash, Do1
Fermenter
0.61 kg CO2, Do2
3 28 kg OuBet total, Do
Scrubber
Outlet
CO2 Emitted.
0.31 kg Ko'
Wastewater,
0.17 kq Ko2
0.48 kg
Distillation
2.97 kg
2.97 kg
Solids Separation
2.44 kg
2.44 kg
Evaporator
1.85 kg
1.85 kg
Dryer
0.36 kg
0.59 kg
0 95 kg
Outlet total, Ko
Outlet
Ethanol, Fo1
Solids, Fo2
Wastewater,
Fo3
Outlet total, Fo
Outlet
This Stillage.
Go1
Wet Cake,
Go2
Outlet total, Go
Outlet
Syrup, Ho1
Wastewater,
Ho2
Outlet total, Ho
Outlet
DDGS, Jo1
Wastewater.
Jo2
Outlet total, Jo
E
1.00
F

kg

2 60
2.60
2.67
0.23
2.90
2.97
0.31
3.28
0 31
0.17
0.48
0 32
244
0.21
2.97
1 85
0.59
2.44
0.36
1.48
1.85
0.33
0.62
0.95

kg

G H I
^^

J
K
L
Input
Derived from Fan Mei. MS Thesis. 2006
Corn Feed
Water Usage
Total Mass In


Ethanol Out
kg CO2 Out
DDGS Out
kg
kg
kg


kg
kg
kg
Wastewai
Total Mas

er Out
s Out









kg

kg
kg


kg
kg
kg
kg


kg
kg
kg

kg
kg
kg



kg
kg


















kg





























1.00 kg
2.68 kg
3.68 kg


0.32 kg
0.31 kg
0.33 kg
2.72 kg
3.68 kg

































































































Figure 6.4 Microsoft Excel™ spreadsheet worksheet for the mass balance model. Adopted
      from Wang et al. (2014)
                                         124

-------
6.2.3   Energy balance model
       The energy needs for ethanol production is of great concern, and the availability of
economical and reliable energy sources is essential for stable operation of the facility. We have
performed a literature review on the total energy consumption of the corn-to-ethanol process, as
listed in Table 6.2. The reported energy consumption varies significantly from 40,850 to 75,118
Btu/gal, with an average of 53,750 Btu/gal. Pimentel et al. (2007) estimates are over 30,000
Btu/gal higher than Wang et al. estimates, and over 20,000 Btu/gal higher than the average value
of all  the studies (Wang, 2001). This is because of Pimentel's inclusion of energy expended on
capital equipment and energy for steel, cement and other materials used to construct the ethanol
plant  - components that were not included in most other studies. In this study, we used the
average value from literature, 53,750 Btu/gal, as the basis of our energy balance calculation.

      Table 6.2 Total energy consumption for corn-to-ethanol process
Literature
Pimentel and Patzek (2005)
Pimentel et al. (2007)
Lorenz and Morris (1995)
Wangetal. (1999)
Shapourietal. (2004)
Mei (2006)
Average Total Energy Demand
Ethanol Conversion Process (Btu/gal)
54684
75118
53956
40850
51779
46114
53750 (Btu/gal) or
15.0 MJ/L
       Generally, energy demand for an ethanol plant consists of thermal energy and electricity.
Thermal energy is used to produce steam, which can be used for cooking, liquefaction, ethanol
recovery and dehydration. Natural gas thermal energy is used for drying and stillage processing.
Electricity is used for grinding and running electric motors. Figure 6.5 shows the diagram of
energy flow through the corn-to-ethanol plant.
       A general energy balance equation for each individual block can be written as:
       Energy Input = Energy Output                                              (6.2-4)
Or
                                                                                 (6-2-5)
          input
          streams
                   output
                   streams
       where £} represents the total rate of energy transported by the/11 input or output stream of
a process, and Q and Ware defined as the rate of flow of heat and work into the process.
       The energy balance calculation procedure is adapted from Mei (Mei, 2006) and the
results are summarized in Table 6.3. For an ethanol conversion process, the majority of the
energy is used as thermal energy for cooking, liquefaction, distillation and drying. Electricity is
mainly used for milling, distillation and drying processes.
                                           125

-------
                                   Thermal Energy
                   Et A
Et C
      Corn->
Milling
A


Cooking/
C
w

Fermentation
D
fc.

Distillation
F
»,

Drying
J
                  Ee A
Ee C

                                            • Ethanol
Ee D
                                  Electrical Energy
Figure 6.5 Energy block flow diagram of ethanol production process. Labels: Et - thermal
       energy; Ee - electricity energy.
     Table 6.3  Energy flow in corn-to-ethanol process
Energy Flow
A -Milling
C- Cooking/Liquefaction
D - Fermentation
F - Distillation
J - Drying
Total
Thermal Energy (MJ/L)
0.21
2.81
-
4.76
6.22
14.0
Electricity Energy (MJ/L)
0.10
0.06
0.06
0.37
0.41
1.0
6.2.4  Flash-based interactive model
       Washington University in St. Louis developed an interactive model integrating mass and
energy balance at the ethanol plant system boundary using Adobe Flash™. Flash™ is a popular
multimedia software that can create animation and add interactivity to web pages. As shown in
Figure 6.6, the users of this Flash-based model have two options to start mass and energy balance
calculations by inputting either corn feed or ethanol plant capacity. For example, as shown in
Figure 6.6(a), if the user chooses "Corn Feed" as the input method, an input text box will show
up and allow the  user to type in the amount of corn that will be fed to the plant. Then by clicking
the "Run" button, the model will calculate and display the amounts of water and energy needed
for the process, the amounts of ethanol and DDGS that will be produced, and the amount of
wastewater and CO2 that will be generated and emitted, if no controls are installed. Similarly, as
shown in Figure 6.6(b), if the user chooses "Plant Capacity" as the input method, after typing in
the amount of ethanol that a plant is expected to produce, the model will calculate and display
the amount of corn, water and energy needed, as well  as the amount of co-products and
emissions that would be generated. The user friendly interface and the interactive feature make
                                          126

-------
this model a handy tool for researchers, plant managers, policy makers and the public to
understand the overall energy and environmental impact of the ethanol production process.
                             5£ Adobe Flash Player 9
                             File  View Control Help
                             Input either Corn Feed or Ethanol Plant Capacity
                                                    tons corn
                      -  n x
                        (a)
                                 Corn Feed
                                 Plant Capacity
1|
                                                 Energy
                                                   6030 MJ
                                     1  tons
                               Corn
                               Water
                                   2.68 tons
                  0.32 tons
                      Ethanol

                      DDGS
                  0.33 tons
                                               CO:  Wastewater
                                               0.31        2.72
                                              tons       tons
                            25 Adobe Flash Player 9
                            File  View Control  Help
                            Input either Corn Feed or Ethanol Plant Capacity
                             Q Corn Feed
                             Q Plant Capacity  |    1|    | tons ethanol

                                                 Energy
                                                 1884375 MJ
                                  3.125 tons
                              Corn

                              Water
                                  8.375 tons
                       (b)
                    1 tons
                      Ethanol
                      DDGS
                1.03125 tons
                                               CO2  Wastewater
                                            0.96875         8.5
                                              tons        tons
Figure 6.6  Flash-based interactive model on energy and mass balance - calculation based on
        (a) corn feed and (b) ethanol plant capacity. Adopted from Wang et al. (2014).
                                                  127

-------
6.3  Pilot Scale Results

6.3.1   Background
       Washington University in St. Louis studied the operations at the National Corn to Ethanol
Pilot Facility established at the Southern Illinois University at Edwardsville (SIUE) campus. The
purpose of this evaluation was to systematically identify the environmental and energy impacts
of the various processes in ethanol production. This included determination of the water
consumption, wastewater constituents, energy use, CO2 output and material input/output. This
information was used to understand key operations and their impacts on environmental
considerations.
6.3.2   Production process overview
       The National Corn to Ethanol Research Center (NCERC) is located on the campus of
SIUE. The NCERC is a not-for-profit research center focused on the validation of near-term
technologies for enhancing the economics and sustainability of renewable fuel production. The
facility has all of the unit operations and laboratory capabilities of a commercial facility but on a
much smaller scale. This is ideal for examining process parameters.
       The examination of production processes at NCERC focused on the dry grind ethanol
process. The pilot process is located in a 24,000 square foot complex. The scale of the operation
is approximately 17250th of a full-scale operation. The facility can process 100 to 400 bushels per
day of corn feed stock,  and is capable of running in a batch or continuous mode. The process
operations of the facility are similar to full-scale operations. Figure 6.7 provides an overview of
the process operations.  As can be seen in this diagram, the process operations include the same
unit operations as full-scale operations.
       The entire operation is equipped with online monitoring and controls to analyze the
process parameters and keep  historical records. Our review of this operation has shown that the
utility and environmental parameters are similar to those found in literature with the  exception of
the water balance. The NCERC has the option of directing its process wastewater to  an onsite
treatment system. For testing purposes, makeup water is routinely provided using city water
instead of recycled process water. This was advantageous since it allows the wastewater to be
sampled. Other process parameters where considered as the overall mass balance evaluation was
prepared.
6.3.3   Wastewater sampling and analysis
       Figure 6.7 shows an example of the process control system looking at the inputs and
outputs for the process water tank. This analysis only considered the dry-mill process. A careful
review of the plant identified the following major operations:
   •   Milling
   •   Mashing
   •   Cooking/Liquefaction
   •   Fermentation
   •   Distillation
   •   Solids Separation
                                           128

-------
        The Corn to Ethanol Process
        National Corn-to-Ethanol Research Center. SIUE
        © 2007 All High
                        -^-   Arnmonwi
                                                        quefaqtion

                                                               (Cooling]
                                                     Molecular
                                                       Sieve
                                                    IVY
          Dried Distillers
        Grain with Solubles
                         Drum Dryer
Evapol-tor
            (Thin
           Stillage)
^___Al	'       ( «•» Cpke)
                               (Whole
                               Stillaoe)
Figure 6.7 NCERC process flow diagram. Courtesy of SIUE.

       Additional operations included boiling, evaporation and drying. The pilot facility
recycled the condensate from the cooking/liquefaction and the distillation processes back into the
mashing tank. These are potential wastewater sources; however, the flow is minor, and good
design will normally route these streams as condensate water back into the mashing tank. Figure
6.8 shows the wastewater streams at this plant, and Figure 6.9 identifies the four locations chosen
for wastewater sampling. The locations sampled are:

    •   Location A - CO2 Scrubber
    •   Location B - Dryer
    •   Location C - Evaporator
    •   Location D - Thin Stillage
       It should be noted that the thin stillage discharge is from the centrifuge and continues
through the evaporator. This was a side stream sample taken to obtain a wastewater profile.  The
wastewater sample collection and analysis was conducted by American Bottoms Regional
Wastewater Treatment Facility (ABRWTF), which is a municipal wastewater treatment agency
located in Sauget, Illinois. They have  an extensive pretreatment program  due to the large amount
of industrial flow, and are fully trained and equipped to conduct wastewater sampling.
                                           129

-------
                          Process Water Tank
Figure 6.8 Process water tank flow scheme. Courtesy of SIUE.
 Corn
Water
i L-
A I Ao Bil
„ Milling ^ I\
^ Meal
fashing
B
B,,
Ma
t
Wastewrater
Fo3 J
Fol Distillation
F
Solids
From
Beer
Column
So
Sepa
<
Fo2
1
lids
*ation
Fi
Steam
I-
Cil
sh
Water
Dol
Fermented
Mash

Sample D
J
Got
Hi
Thin Stillage
Cooking/
Liquefaction
C
DI2 1
Co2





Col
Liquefied Mash
Oil
Fermentation
D

Wastewater
Ilo2|



Do2 Kil
^crmenter
Co2

Sample C

Evaporator
H


Hoi
Syrup

astewater
Water
1 Ki2
Scrubber
K

Ml

Go2
Wet Cake




Dryer
J
Ji2 ;
Kol
	 *- Wastewater
Ko2
Sample A

Sample B
Jol
	 »- DDGS
Figure 6.9 Wastewater sampling locations.
      The wastewater parameters tested for included the following:
                                     130

-------
       1.  Volatile organics
       2.  Semi-volatile acid and base/neutral
       3.  Metals (Cd, Cu, Zn, Mn, Pb and Fe)
       4.  Mercury
       5.  TKN-N and NH3-N
       6.  NO2-N, NO3-N and Ortho-P
       7.  Total Suspended Solids (TSS) and Total Dissolved Solids (TDS)
       8.  Chemical Oxygen Demand (COD)
       9.  Biological Oxygen Demand (BOD)
       The BOD/COD, the key indicator of wastewater strength, for all of these samples were
very high, as expected. The thin stillage sample indicates that it is not recommended to handle it
separately, and a process of subsequent evaporation and drying is necessary. Recycling of
scrubber water, evaporator water and dryer water back to the mashing operations appears to be a
viable option. The pilot plant does not recycle these flows in order to maintain consistent data in
their pilot test runs. A review of full-scale facilities shows that recycling of these flows is a
common practice. Treatment and discharge or discharge to a municipal sewer would be
considered costly to remove the high-strength organic waste or pay surcharges to the municipal
wastewater authority.

6.4  Full-Scale Plant Review

6.4.1   Overview
       Subsequent to the evaluation of the dry grind ethanol process described in Chapter 7, the
project team contacted and visited a full-scale ethanol plant to quantify specific inputs and
outputs achievable from modern ethanol  plants. With the explosive growth of ethanol facilities in
the U.S. since the passage of the Energy Policy Act of 2005, ethanol use has grown from 25
billion gallons in 2004 to over 100 billion gallons in 2010, rising to 120 billion gallons by 2014
(EIA, 2013). As more plants have been constructed, plant efficiency has significantly improved.
       This section summarizes full-scale data collected from operating ethanol plants.
Specifically, staff were interviewed at a new ethanol plant which launched operation in April,
2008. The staff reported current plant operating data, including corn to ethanol conversion rate,
total water use, total wastewater produced, thermal and electrical energy required and CO2
output.
6.4.2  Full-scale plant description and operation
       The new ethanol facility that was evaluated is located in  southern Illinois. Construction of
the facility began in October, 2006, and ethanol production began in April, 2008. The facility
uses the dry milling process to produce approximately 54 million gallons per year of ethanol, but
can be expanded up to a total plant production of approximately 108 million gallons per year.
The facility will use approximately 19 million bushels of corn annually. In addition to the
ethanol, plant staff expects the facility to produce 172,000 tons of dry distiller grain and 150,000
tons of CO2 annually.
       These additional plant outputs have considerable value. The dry distiller grains are now
sold as a replacement for corn in livestock feed. Given the higher costs of corn as the result of
the increase in demand from new ethanol plants, the dry distiller grains provide livestock owners
                                           131

-------
with a method to minimize their increased feed costs. In addition, the commercial-grade CO2
produced during the fermentation process can be captured, cooled and stored, and then sold into
the commodity market.
       Much of the improvement in plant performance is being driven by the high costs and
volatility of the primary plant inputs, along with increased industry output. The twin impacts of
increased production capacity and higher raw material costs contributed to a significant drop in
ethanol prices, suggesting that the building boom of 2005 and 2006 has ended (lessen, 2007;
Shirek, 2007). Corn prices have more than tripled this decade from around $2 per bushel to over
$7 per bushel, and the volatility of natural gas is well-known. Thus, there is considerable market
pressure on producers to operate the plants  as efficiently as possible.
       The project team's analysis of plant operations suggests that plant efficiencies are
improving. A summary of the plant's operating parameters as of August, 2008, after four months
of operation is presented on Table 6.4.
       The total energy use is overwhelmingly from thermal processes (i.e., steam), which is
used to heat the mixture. There is a modest amount of electricity used for pumping and air
handling, but this  accounts for less than 10  percent of the total plant energy inputs on a per
gallon of ethanol-produced basis. The wastewater generated by the facility is primarily cooling
tower and boiler blow-down. Water used for processing is recycled within  the process and
generally not discharged to the wastewater  stream. In fact, process wastewater (which includes
blow down) accounts for only about 25 percent of overall water use within this facility. The
remaining water use within the facility is used for cooling, and exits the plant as water vapor
from the cooling tower.
     Table 6.4  Summary of operating parameters from a Midwest, full-scale ethanol
                production facility
Corn to Ethanol Conversion Rate
Total Water Use*
Total Wastewater Produced
Total Energy
Distilled dry grains produced
Carbon dioxide
2.8
8.4
3.0
2.1
0.75
100,800
36,000
18
18
Gal ethanol per bushel of corn used
Gal water per bushel of corn used
Gal water per gal ethanol produced
Gal wastewater per bushel of corn used
Gal wastewater per gal ethanol produced
Btu per bushel of corn used
Btu per gal ethanol produced
Lbs per bushel of corn used
Lbs per bushel of corn used
 Note:  * Only accounts for water used by the facility to convert the corn into ethanol.
                                           132

-------
6.4.3   Assessment
       ICM, Inc. is a Wichita-based company specializing in the development and application of
ethanol processing technologies and has transitioned into a turnkey supplier to the ethanol
industry. The company's proprietary process accounted for 2.1 billion gallons per year of the
total 5.8 billion gallons per year of ethanol produced in the U.S. as of March, 2007. The project
team contacted ICM to ascertain the future for potential improvements to the operation of
ethanol plants in the U.S.
       The company currently offers a performance guarantee of 2.8 gallons of denatured
ethanol per bushel  of corn (U.S. No. 2 Yellow Dent). Further, they guarantee that plant natural
gas usage will not exceed 32,000 BTU per gallon of ethanol produced, and this value includes
any gas used to dry the distilled solids. Also, the guarantee includes a limit of 0.75 kW of
electricity per gallon of ethanol produced, which represents about a 15 percent improvement over
the performance reported by the operating facility in Illinois (see section 2).
       After corn,  energy is a facility's primary expense, so there are market forces at work to
minimize energy use. However, heat is essential to the fermentation and distillation processes,
and so a minimum  amount is needed. A typical, modern ethanol facility will use heat exchangers
to capture and reuse heat throughout the plant, as suggested in Figure 6.10.
Figure 6.10. Heat capture through heat exchangers greatly improves energy efficiency in the
       Illinois ethanol plant.

       Minimizing water use is also of great interest to ethanol producers. Technology is
available to build ethanol plants capable of achieving zero discharge, and such facilities make

                                           133

-------
permitting easier to complete, especially during expansions. However, in many locations zero
discharge is not necessary and a number of existing ethanol plants use local groundwater
sources, which are finite. In those cases, alternative water sources must be investigated,
particularly in water-stressed regions. These can include lower quality surface waters or gray
water.  Firms that specialize in water treatment for the ethanol industry suggest that a 30 percent
reduction in water use at ethanol plants is achievable. That level of reduction will be essential if
larger ethanol plants are built, and if the industry is to expand in any significant way outside of
the Midwest—particularly to the water-stressed regions of the Southwestern U.S.

6.5  Summary and Future Work

6.5.1   Summary
       In 2013, ethanol constituted approximately 91% of all biofuels produced in the U.S. A
majority of this ethanol is currently produced from corn starch.  However, a potential drawback
of producing  ethanol from corn starch is the amount of water used to produce feedstock from
corn cultivation. The irrigation of corn for ethanol production has a water-intensity that is an
order of magnitude higher than the average water intensity for producing ethanol from corn
starch, 113 gallons-FbO/gallon-ethanol vs. 13.4 gallons-FbO/gallon-ethanol (King and Webber,
2008) and thus needs to be taken into consideration separately from other stages of the
production process. Irrigation requirements vary widely by state and region. Adaptive planning
and future research investigating water use for corn ethanol production must take this regional
variation into account. The consumptive water losses by evapotranspiration from corn cultivation
for ethanol should also be carefully considered  in any future analyses of water use in the
production of ethanol. A potential alternative to corn ethanol is  to transition to less water-
intensive biomass crops for the feedstock such as alternative corn varieties that use less water for
irrigation, alternative sugar/starch crops or cellulosic feedstock  such as switch grass. Cellulosic
energy crops  on average have an irrigation water-intensity of approximately 25% of the
irrigation water-intensity of ethanol (Tidwell et al.,  2011).29
       The first phase of this work as presented in this report has resulted in a detailed review of
ethanol production from corn as a feedstock. Mathematical models that can be readily used by
design engineers and auditors have been created. Detailed review of a pilot-scale facility and a
full-scale plant were conducted. The key focus  areas in this chapter were the use of energy for
production of ethanol, the water usage (both quantity and quality) and the CO2 emissions. The
analysis was restricted to mass and energy balances around the plant.
       A theoretical model on mass balance and energy balance for the dry mill corn-to-ethanol
production process was  established in this work. The inputs of the model are corn  and water plus
energy, while the  outputs are ethanol, solids (by-products), wastewater and  CO2. The model was
presented in two ways: an Excel  Spreadsheet and a Flash-based interactive interface. The Excel-
based model gives details of the mass and energy balance calculations for each step in the
ethanol production process, while the Flash based model describes the overall inputs and outputs
at the system  boundary,  and has the option to start the mass and energy balance calculation by
inputting either corn feed or ethanol plant capacity. The user friendly interface and the
interactive feature make this model a handy tool for researchers, plant managers, policy makers
       29 See Chapter 7 for further discussion of cellulosic ethanol production.
                                           134

-------
and the public to understand the overall energy and environmental impact of the ethanol
production process.
       Assuming no water recycling in the process, the model demonstrates that with 1 kg input
of corn, 2.68 kg water is needed, and 0.32 kg ethanol and 0.33 kg DDGS can be produced with
0.31 kg CO2 emission and 2.72 kg wastewater discharge. Note that the theoretical model was
derived from Mei (Mei, 2006), which aimed to optimize the design of a 30 MMgpy corn-to-
ethanol facility. Hence, this model is appropriate to represent a full-scale facility, except it does
not account for wastewater reuse.
       Surveys of larger scale ethanol facilities were later conducted to provide reference data to
the theoretical model. We first studied the operations at the National Corn to Ethanol Pilot
Facility established at SIUE. This facility has all of the unit operations and laboratory
capabilities of a commercial facility but on a much smaller scale (approximately 17250th of a full-
scale facility). The operation process of the pilot-scale facility was similar to that described in the
theoretical model, and makeup water was routinely provided using city water instead of recycled
process water. Wastewater was sampled at four locations: CO2 scrubber, dryer, evaporator and
thin stillage. The analysis showed the BOD/COD for all these samples were considerably high,
as expected. The pilot plant did not provide any data on the amount of energy consumed and
wastewater discharged; however, recycling of scrubber water, evaporator water and dryer water
back to the mashing operations appears to be a viable option.
       Secondly, a full-scale new ethanol facility located in southern Illinois was evaluated. This
facility uses the dry milling process to produce approximately 54 MMgpy ethanol using
approximately 19 million bushels of corn annually. In addition to the ethanol, plant staff expects
the facility to produce 172,000 tons of dry distiller grain and 150,000 tons of CO2 annually. The
water used for processing in the facility is recycled within the process and generally not
discharged to the wastewater stream. The wastewater generated by the facility is primarily
cooling tower and boiler blow-down. They also reported a smaller value of energy consumption
compared with the theoretical model because they have used a series of heat exchangers for heat
capture, and thus enhanced plant energy efficiency.
       Minimizing water and energy use is of great interest to ethanol producers. Comparisons
of the results from the literature review and surveys of different scales of ethanol facilities
indicate the trend of improved energy efficiency and wastewater reuse rate for modern corn-to-
ethanol facilities. On the other hand, the high rate of wastewater reuse means that the discharge
could have high concentrations of heavy metals, BOD, TOC,  etc., which have to be removed.
The ethanol plants usually replace the entire stock of processing water after a certain amount of
time so that the contaminants would not build up to impose any potential hazardous effects. So
far, there is no standard operational reference or indicator to guide the practice of wastewater
reuse for ethanol plants. Analysis of wastewater streams from full-scale facilities would lead to a
better understanding of their potential reuse rate and treatment. In many locations,  however, zero
discharge is not necessary  and a number of existing ethanol plants use finite local groundwater
sources. In those cases, alternative water sources must be investigated. Currently, ethanol
processed from corn starch uses 2.7 - 40 gal water/gal ethanol, while ethanol processed from
alternative sources such as cellulose, switch grass or corn stover uses 9 - 15 gal water/gal
ethanol. This  level of water usage is still high compared to gasoline or diesel processed from
petroleum sources (1-2.5 gal water/gal fuel), and future research needs to focus on technologies

                                           135

-------
to reduce water use during the fuel processing stage.  The level of reduction in water use will be
essential if larger ethanol plants are built, and if the industry is to expand in any significant way
outside of the Midwest.
6.5.2  Future work
      Carbon dioxide emissions, and carbon balances in general, are going to be very important
aspects in the bio-fuels industry. Energy and water usage also remain important parameters. The
use of bio-refinery concepts and the coupling of other sources of CO2 offer significant potential
in the future. In addition, further investigation of water conservation, water reuse, zero-water-
discharge designs and use of waste-heat within ethanol production facilities will also need to be
further studied.

6.6   References

Agricultural Marketing Resource Center. 2001. "The U.S.  Dry-Mill Ethanol Industry."
      http://www.agmrc.org/media/cms/drvmill c40bbad756d35.pdf. Accessed July 8, 2014.

Energy Information Administration (EIA). 2014. Monthly Energy Review. U.S.  Energy
      Information Administration, July 2014. Office of Energy Statistics, U.S. Department of
      Energy, Washington, DC. DOE/EIA-0035 (2014/07).

Energy Information Administration (EIA). 2013. "Short-Term Energy Outlook, Number 21."
      http://www.eia.gov/todayinenergv/detail.cfm?id= 13891. Accessed on July 8, 2014.

Farrell, A.E., R.J. Plevin, B.T. Turner, A.D. Jones, M. O'Hare, and D.M. Kammen. 2006.
      "Responses to Comments on 'Ethanol Can Contribute to Energy and Environmental
      Goals'". Science, 312:1747-1748.

Graboski, M.S. 2002. "Fossil Energy Use in the Manufacture of Corn Ethanol. Report for the
      National Corn Growers Association."
      http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.170.7995&rep=repl&type=pdf

ICM,  Inc. 2014. http://www.icminc.com/innovation/ethanol/ethanol-production-process.html.
      Accessed June 3, 2014.

ICM Promotional Brochure. 2007. March, www.icminc.com.

Jessen, H. 2007. "Opportunities to Conserve Water." Ethanol Producer Magazine.
      http://ethanolproducer.com/articles/2734/opportunities-to-conserve-water/. Accessed
      June 3, 2014.

King,  C.W., andM.E. Webber. 2008. "Water intensity of transportation." Environmental
      Science and Technology, 42(21):7866-7872.

Lorenz,  D., and D. Morris. 1995. "How Much Energy Does It Take to Make a Gallon of
      Ethanol?"  August, 1995. Institute for Local Self-Reliance, Minneapolis, MN.
                                          136

-------
McAlloon, A., F. Taylor, and W. Yee. 2000. "Determining the Cost of Producing Ethanol from
       Corn Starch and Lignocellulosic Feedstocks." National Renewable Energy Laboratory
       Technical Report No. NREL/TP-580-28893.
       http://www.nrel.gov/docs/fy01osti/28893.pdf. Accessed August 2014.

Mei, F. 2006. "Mass and Energy Balance for a Corn-To-Ethanol Plant." M.S. Thesis,
       Washington University in St. Louis.
       http://crelonweb.eec.wustl.edu/theses/Masters/Fan%20Mei%20-Master%20thesis.pdf.
       Accessed June 3, 2014.

National Corn-to-Ethanol Research Center. 2014. "Advancing Biofuels Research."
       http://www.ethanolresearch.com/index.shtml. Accessed June 3, 2014.

Online Module for Corn to Ethanol Production developed by Washington University in St.
       Louis. http://www.aerosols.wustl.edu/education/energy/EthanolAudit/index.html.
       Accessed June 3, 2014.

Pimentel, D., and T.W. Patzek. 2005. "Ethanol production using corn, switch grass, and wood;
       Biodiesel production using soybean and sunflower." Natural Resources Research, 14:65-
       76.

Pimentel, D., T.W. Patzek, and G. Cecil. 2007. "Ethanol production: Energy, economic,  and
       environmental losses." Reviews of Environmental Contamination and Toxicology,
       189:25-41.

Renewable Fuels Association, http://www.ethanolrfa.org/pages/how-ethanol-is-made. Accessed
       July 8, 2014.

Schnepf, R., and B.D. Yacobucci. 2013. "Renewable Fuel Standard (RFS): Overview and Issues,
       CRS Report for Congress (No. R40155)." http://www.fas.org/sgp/crs/misc/R40155.pdf.
       Accessed June 3, 2014.

Seekingalpha.com. http://seekingalpha.eom/article/l 1431 -everything-vou-wanted-to-know-
       about-ethanol-production-but-were-afraid-to-ask-adm-hki-vse. Accessed July 8, 2014.

Shapouri, H., J. Duffield, and M. Wang. 2004. "The 2001 Net Energy Balance of Corn Ethanol."
       American Coalition for Ethanol.
       http://www.ethanol.org/pdf/contentmgmt/USDA energy balance 04.pdf. Accessed June
       3,2014.

Shirek, M. 2007. "Ethanol Experiences Growing Pains." Ethanol Producer Magazine.
       http://www.ethanolproducer.com/articles/3489/ethanol-experiences-growing-pains.
       Accessed June 3, 2014.

Tidwell, V.,  A.C.-T. Sun, and L. Malczynski. 2011. "Biofuel Impacts on Water."  No.
       SAND2011-0168.  Sandia National Laboratories.

U.S. Environmental Protection Agency. 2010. "EPA Lifecycle Analysis of Greenhouse Gas
       Emissions from Renewable Fuels. EPA-420-F-10-006"
       http://www.epa.gov/otaq/renewablefuels/420fl0006.pdf. Accessed July 11, 2014.
                                          137

-------
Wang, M., C. Saricks, and D. Santini. 1999. "Effects of Fuel Ethanol Use on Fuel-Cycle Energy
      and Greenhouse Gas Emissions." U.S. Department of Energy, Argonne National
      Laboratory, Center for Transportation Research, Argonne, IL.

Wang, M. 2001. "Development and Use of GREET 1.6 Fuel-Cycle Model for Transportation
      Fuels and Vehicle Technologies." Center for Transportation Research, Argonne National
      Laboratory, U. S. Department of Energy (ANL/ESD/TM-163).
      http://www.transportation.anl.gOv/pdfs/TA/l53.pdf. Accessed June 3, 2014.

Wang, M., 2005. In Updated Energy and Greenhouse Gas Emission Results of Fuel Ethanol, The
      15th International Symposium on Alcohol Fuels, September 26-28, 2005. San Diego, CA.
      http://www.transportation.anl.gov/pdfs/TA/375.pdf. Accessed June 3, 2014.

Wang, W.-N., D. Menon, Y. Li, J. Yang, and P. Biswas. 2014. "Web-based Educational Module
      Development for Water and Carbon Footprints Tracking."
      http://www.aerosols.wustl.edu/education/energy/EthanolAudit/index.html. Accessed
      August 2014.
                                         138

-------
 7   Water Usage within Lignocellulosic Biomass and Cellulosic
     Biofuels Production
       Bala P. Lingaraju1, Sowmya Karunakaran1, Vishnu Sriram1, and Joo-Youp Lee1
       In this chapter, the water requirements for biomass harvesting and cellulosic ethanol
production via a fermentation pathway were assessed on a volume-to-volume basis (i.e., gallons
of water consumption per gallon of ethanol produced) using data reported in the literature.31  The
water requirements were analyzed for a combination of three cellulosic feedstocks (i.e.,
hardwood, corn stover, and switch grass) and two pretreatment technologies (i.e., dilute acid and
ammonia fiber expansion) under different water network configurations with an overall goal of
zero wastewater discharge. The results indicate that the process water requirements are
significantly dependent on the selection of pretreatment processes and feedstocks, while the
effective cooling tower design and operation of a cooling tower offers an opportunity for saving
utility water (i.e., cooling water and steam).

7.1  Introduction

       The Energy Independence and Security Act (EISA) of 2007 calls for a fourfold increase
in the production of biofuels such as ethanol in the U.S. by 2022  (EISA, 2007). This will require
36 billion gallons per year of total renewable fuels by 2022, including the production of 15 and
21 billion gallons of conventional (i.e., crop-based ethanol such as from corn) and advanced
biofuels (i.e.,  1 billion gallons of biodiesel, 16 billion gallons of cellulosic ethanol and 4 billion
gallons from any other sources), respectively (EISA, 2007). Cellulosic ethanol is produced from
wood, grasses, or nearly any inedible part of plant material. The U.S. consume approximately
134.51 billion gallons of gasoline and 13.2 billion gallons of fuel ethanol in the 2012 calendar
year (U.S. EIA, 2014), of which approximately 1% is produced using cellulosic feedstock (U.S.
Bioenergy Statistics, 2014).
       There are currently -211 starch-based ethanol plants in the U.S. with a total name plate
capacity of 14875.4 MMgpy (RFA, 2014). Most of these ethanol plants use corn kernel as a
feedstock along with other starch-based feedstocks. Other feedstocks such as corn stover,
sorghum, sugarcane bagasse, beverage waste, wood waste and cheese whey are part of the small
percentages of feedstocks used in future cellulosic plants (Wallace et al., 2005). Corn stover is
the leaves and stalks of maize plants left after harvesting, and consists of stalk, leaves, and husk.
The current estimated operating capacity of the biorefmeries is 14,178.4 MMgpy (RFA,  2014).
There are also many plants under construction or expansion for the production of corn and
cellulosic ethanol with a potential capacity of 165 MMgpy (RFA, 2014). These plants are being
scaled up from pilot to commercial scale.
1   University of Cincinnati, Department of Biomedical, Chemical, and Environmental
   Engineering

31  The material in this chapter was originally published by the co-authors as Lingaraju et al.
   (2013) as part of work conducted for EPA under the Air, Climate and Energy Research
   Program.

-------
       Freshwater demand for energy production is a growing concern in the U.S. The increased
use of biofuels is expected to offer benefits such as the decreased dependence on foreign oil, but
may also present challenges such as the increased use of domestic agricultural resources, and
negative impacts on air quality and surface and ground water resources (Alvarez et al., 2010). A
National Research Council (NRC) report also indicates that the current estimates of consumptive
water use by ethanol biorefineries are 4 and 9.5 gallons of water per gallon of ethanol produced
from corn kernels and cellulosic feedstocks, respectively (Schnoor et al., 2008). Based on these
estimates, an additional 256 billion gallons per year of freshwater will  be required to meet the
annual production of 36 billion gallons of ethanol by 2022. However, these estimates do not take
the amount of water required for irrigation into consideration. As described in chapter 6, the
cultivation of corn has an average irrigation water intensity of approximately 113  gallons-
IHbO/gallon-ethanol and is highly variable, ranging from 15 to 934 gallons-FbO/gallon-ethanol
(King, 2008), depending upon regional differences in irrigation levels.  The intensity of
freshwater usage could be regionally problematic, representing an incremental withdrawal from
already marginally sustainable or unsustainable sources. For example,  current withdrawals in the
High Plains Aquifer (more than 1.5 billion gallons per day) are greater than the aquifer's
recharge rate, and the loss of this resource would be irreversible if the  current withdrawal rate is
not reduced (McMahon et al., 2007).
       Cellulosic ethanol has received growing attention in recent years as it does not compete
with food production,  and instead uses agricultural by-products and energy crops which can
grow even in arid regions (Dale, 2007; Chiu etal, 2009; Zink, 2007; Keeney and Muller, 2006).
It has been reported that the water consumption for irrigation significantly varies in terms of
region (Wu et al., 2009) for crop cultivation. However, agricultural residues in the early 2000s
comprised more than 70% of feedstock resources for cellulosic ethanol production, and the use
of agricultural residues does not require any additional water consumption for irrigation (Perlack
et al., 2005). The fraction of feedstock that is from agricultural residue is expected to decrease
given that EISA calls for the production of 16 billion gallons of ethanol from cellulosic feedstock
by 2022, and additional cellulosic material from biomass crops such as switch grass, short-
rotation woody crops (SRWC) and  other crops will be needed to meet  additional future cellulosic
ethanol demand. Switchgrass and SRWC are not considered  to be agricultural residue, and thus
could require significant amounts of water for irrigation. Tidwell et al. (2011) estimated that
approximately 4,000 and 2,800 MGD of water will be required to grow switch grass and SRWC
in 2030, respectively, to produce 80 billion gallons of cellulosic ethanol per year.32 The authors
estimated an average water intensity of 28 gal water/gal cellulosic ethanol produced for irrigation
(Tidwell et al., 2011),  approximately 25% of the average irrigation water intensity of corn
cultivated for ethanol production.
       In this study, a detailed water analysis for cellulosic ethanol production was focused only
on process and utility water requirements, and the amount of water required for irrigation was
not considered. Previous study results have been summarized in Table 7.1. There  are very few
studies available regarding the water quality and quantity requirements for cellulosic ethanol
production, and various units of measurement have been used to estimate water requirements to
meet respective objectives. In this study, the unit used is gallons of water per gallon of ethanol
32 Note that this is more double the entire volume of renewable fuels called for in 2022 under
   RFS2 and five times the RFS2 volumes for cellulosic ethanol production in 2022. See
   Chapter 2 for further discussion of RFS2.

                                           140

-------
produced (gal water/gal ethanol). This unit of measure was selected so that the water requirement
can be readily estimated by the production capacity. The primary purpose of this study is to
assess the water quantity and quality requirements for cellulosic ethanol production in terms of a
combination of different feedstocks, pretreatment technologies, and recycled water
configurations.

     Table 7.1  Literature review of water requirements for ethanol production
Water Requirement
Freshwater make-up*
Freshwater make-up*
Freshwater make-up*
Freshwater make-up*
Freshwater make-up*
Freshwater make-up*
Freshwater make-up*
Freshwater
Freshwater
Quantity
4.5
9
3.5-6
2.85
4.7
<3
6
1.5-4.3
50
3.4-4.6
Units of Water
Requirement
gal water / gal ethanol
gal water / gal ethanol
gal water / gal ethanol
gal water / gal ethanol
gal water / gal ethanol
gal water / gal ethanol
gal water / gal ethanol
gal water / gal ethanol
gal water / gal biofuel
gal water / gal ethanol
Feedstock
Corn Kernel
Cellulosic
Corn Kernel
Dried Distillers
Grain
Dried Distillers
Grain
Corn Kernel
Corn Stover
Switchgrass
Corn
Gasoline
Reference
(Schnooretal.,2008)
(Schnooretal.,2008)
(Keeney and Muller,
2006)
(Pfromm, 2008)
(Shapouri and
Gallagher, 2005)
(Swain, 2006)
(Montague, 2002;
Aden, 2007)
(Laser etal., 2009b)
(Hoekman, 2009)
(Wu etal., 2009)
   Note: * make-up is defined as net = inlet - outlet
         1 gallon = 3.79 liters
7.2   Cellulosic Ethanol Process

       A typical process configuration for cellulosic ethanol production is shown in Figure 7.1.
Major unit operations include pretreatment, detoxification, fermentation, distillation, solid
separation, drying, cooling, scrubbing, and wastewater treatment. In a typical dilute acid
pretreatment process, dilute sulfuric acid with a concentration of 2% by weight/weight is mixed
or contacted with biomass for cell wall rupture and to hydrolyze hemicellulose to other simple
monomeric sugars at temperatures of 160 to 220°C for periods ranging from minutes to seconds.
In the Ammonia Fiber Expansion (AFEX) pretreatment process, lignocellulosic biomass is
exposed to a dosage of liquid ammonia (1 to 2 kg ammonia/kg dry biomass) at elevated
temperature (132°C) and pressure (113 bar) followed by an instantaneous drop in pressure which
causes ammonia vaporization and explosive decompression of the biomass resulting in fiber
disruption (Mielenz and Mielenz, 2009). The biomass is transported by screw conveyors through
                                           141

-------
a pre-steamer, a pretreatment tank, a blowdown tank, and an ammonia/acid pretreatment tank to
form slurry. Solid-liquid separation is achieved by pressure filters.
       A detoxification process is employed to remove enzyme inhibitors, sugar degradation
products (e.g. furfural, 5-hydroxy methyl furfural) and chemicals that are toxic to
microorganisms in the fermentation tank. The extraction of by-products and acids is achieved by
non-dispersive membrane extraction and reactive membrane extraction; however, fouling caused
by biomass particulate matter is reported (Ramaswamy et al., 2013). Over-liming by adjusting
the pH to 10.0 by Ca(OH)2 is an alternative detoxification process. The hydrolyzate from the
detoxification tank is passed through saccharification and co-fermentation continuous stirred
tank reactors containing enzymes for digesting C5 and C6 sugars to produce ethanol. A typical
residence time is five days.
       A two-step distillation process hosts a beer-separation column and a rectification column
for repeated distillation to obtain pure ethanol.  A typical beer column having 22 trays takes feed
from fermenters to remove solids, lignin, insoluble proteins, and non-fermentable products as
bottoms slurry from the overhead ethanol product. The rectification column, having 25 to 30
trays, concentrates the ethanol vapors to produce 95% weight fuel grade ethanol (Summers,
2006). The wastewater is collected, reclaimed,  recycled, and pumped to the pretreatment reactor
and condenser in the distillation unit. The process and utility water requirements for the unit
operations are discussed in detail in Section 7.4.
       Feedstock for cellulosic ethanol production can be classified into the following three
categories:  forest residues, agricultural crop residues and perennial crops (Perlack et al., 2005).
In this report, three representative feedstocks (i.e., hardwood, corn stover and switch grass) and
two representative pretreatment technologies (i.e., dilute acid and Ammonia Fiber Expansion
[AFEX]) were selected based on the data available in the literature (Wooley et al, 1999;
Montague,  2002; Laser et al., 2009a). Detailed descriptions of the two pretreatment technologies
are given elsewhere (Mosier et al., 2005; Sun and Cheng, 2002).

7.3   Biomass Harvesting and Biofuel Conversion

       The biomass harvesting process includes three operations: harvesting of biomass, raking
the crop residues into windrows, and baling of the windrows into square or cylindrical  shapes for
storage. Specialized equipment (i.e., combine)  is used for the single-pass, two-pass, or three-pass
method for biomass harvesting. In the single-pass method, the three operations are collectively
performed by a single piece of equipment. In the two-pass method, harvesting and windrowing is
performed by separate equipment. The three-pass method utilizes different equipment for each of
the three operations  (Ertl, 2013). The high moisture content of corn stover obtained after the
single pass harvesting method is reported to be -46% (Shinners et al., 2009).
                                           142

-------
Pro'*.



Cr.n




A..id


C on

LJ rwr — •*-
/*Otj rrrrr^fr


Rr"-y&.l



^.j.,l_, _,,,. ^^^^^j^.
V

J\|F — — 1^-


*,s, Wfltw 5.*



st^nsatf*

f *

PfWtfWi

J
13
i-J
*»
i^nSfSte "^
1 "•


1
U
y.l tt "Ki r "^.
1 "^
8 1


Ferin-pn
CclEuln^e F



itkm



'^lr,sr
,
|





3Cu,i«
1

icafwn



!•



s
*roduc.tnE>n







T



•*-



1


—




















j^V*^ _
"" '!' "1
1 :
il "5 [i
£ J-?
l' 1
_ _i !_„
• • »5;?» • • -


Cat's*



t 	 ftL^Myrl 	 A,


tfs.H





Wilier Split


r

Trcatifd Mht

A
iir
;,-fe
c

W**te water
TruaifwnT
4 1
5 "^
l.^"!!!.i4t<, ff g
4
1

DiM'INttoi
*
OenjfOFattei^,
Scrubbcr, Evapooitor

J




L 3Dcf|jndi Tis'^rt1}*; Hv^jp^r-isHiQn i^lnfliBfj^ l~c"^s
""¥;*£*: "Oft Watc-- i ,
! t t

Vt r^t-'llt &I I t^^-*~°f
. , MM
5
^j ^ *
5 ;3b g
~! «
'", f

Boiler . Burner — 1»- St->ii n
S^IWj,


mm Carr;«»d Q«'er
M(^«^ •iiiiiiiiiiiiiiiiiiiiiiiiiiiiin™^^^^^^^^ ja ,*

" 1
I MH*IJ-III- *>W;»*ir(f
kUian? J R


— . Waste Wpftr
Figure 7.1  Process schematic for cellulosic ethanol production

-------
       A dry or wet storage method is used to preserve the harvested biomass. Dry storage
involves field drying of the harvested biomass where the moisture content of freshly harvested
biomass is reduced to a desired level (e.g., <20%). The field drying process takes from several
days to weeks depending on ambient temperatures and rainfall during the harvest season. Wet
storage methods are proposed in regions where the humidity and rainfall precipitation does not
favor drying conditions. In the wet storage method, the freshly harvested non-dried biomass
containing >45% moisture content is stored in horizontal silos, airtight pits, or plastic wraps. The
wet storage method can be integrated with chemical or biological pretreatment methods because
of the growth of fermentative microorganisms under anaerobic conditions (Li etal., 2011). Two
schematic block diagrams of the biomass harvesting process and the equipment used are shown
in Figures 7.2 and 7.3, respectively.
       Pretreatment methods break down the lignin linings of the cell membrane in order to
facilitate easy access of cellulose to enzymes. Steam injection into the biomass cells in the
presence of ammonia or dilute acid leads to cell wall rupture. The biomass is washed with
freshwater and carried over to the fermentation chamber. The pretreatment methods used to
cause cell wall rupture are classified as 1) physical, 2) physicochemical and chemical, and 3)
biological pretreatment methods. Physical treatment includes milling, high pressure steaming,
and irradiation. Chemical and physiochemical methods include the use of explosion, gas, acid,
alkali, oxidizing agents, cellulose solvents, and lignin solvent extraction. Biological pretreatment
methods include the use of microorganisms such as fungi for causing cell wall rupture. Among
the aforementioned methods, physicochemical and chemical methods are efficient because of
delignification, complete hemicellulose hydrolysis and breakdown of cellulose. The ammonia
fiber expansion method (AFEX) takes physicochemical actions on the plant cells recovering 93%
of hemicellulose, and dilute acid treatment has chemical actions recovering 90% hemicellulose.
However, ammonia solvent evaporation and pH neutralization before enzymatic hydrolysis are
the major challenges faced when using chemical methods (Karimi, 2007). Depending upon a
combination of feedstocks used and the pretreatment type (AFEX or dilute acid), the quantity of
water required for the entire process varies (Zink, 2007).
       In the saccharification process, cellulose and hemicellulose are converted into simple C5
and C6 sugars by enzymatic activities. The freshwater containing bacterial broth is left for
several hours for the fermentation process to take place where the simple monomeric sugars are
converted into ethanol. The primary enzymes (cellulose and hemicellulase) and accessory
enzymes (endoglucanase, exoglucanase, glucocidase and glucosyl hydrolase) produced from
microbial activities of Saccharomyces cerevisiae, Clostridium thermocellum, Trichoderma
reesei, Escherichia coll, Zymomonas mobilis, Pachysolen tannophilus, C. shehatae, Pichia
stipitis, Candida brassicae andMucor indicus aid in the saccharification process (Sarkar et al.,
2012).
       After the conversion of sugars into ethanol, pure ethanol is separated from an ethanol-
water mixture by distillation. The current distillation process produces 95% weight ethanol
containing 5% water and by-products as impurities  (Summers, 2006). However,  pervaporation
using hydrophilic membranes for biofuel dehydration or organophilic membranes for biofuel
enrichment produces 99.9% (weight) fuel grade ethanol (Wang and Chung, 2012).  Steam is
required for preheating the ethanol-water mixture. The reboiler and condenser require steam and
cooling water, respectively, to recover products from the preheated mixture.
                                          144

-------
Figure 7.2 Unit operations involved in a biomass supply chain from field to biorefmery.
          1. Mowing
2. Raking
3. Baling

            4. Storage
Figure 7.3 Equipment used in the biomass harvesting process.
                                         145

-------
7.4  Water Usage and Wastewater Generation

       One of the constraints associated with cellulosic ethanol production is the water
requirement for the cultivation of energy crops. A major proportion of water usage for the
production is  to irrigate lignocellulosic biomass. It was estimated that in 2006, 5,616 million
gallons per day (MGD) was used for irrigating crops that were used to produce biofuel (primarily
from corn). The amount of water required for conversion to biofuels was relatively lower at 94
MGD.  It is predicted that the amount of irrigation water for feedstocks in the year 2030 would
increase to 11,458 MGD (4649, 4077 and 2822 MGD for corn, switch grass and SRWC,
respectively), while that for conversion would require 470 MGD (219 and 251 MGD for corn
and cellulosic ethanol, respectively; Tidwell et al., 2011). Water wastage as run-off during the
cultivation of plants is a significant issue. Irrigation water loss occurs due to run-off and
evaporation from the land.
       Water is required to wash off soil contaminants and other impurities from the biomass
matter  after harvest. Water usage is represented by the amount of water consumed to compensate
for the  loss of total moisture content  of the plant mass due to evaporation, evapotranspiration
(combination of transpiration and evaporation of water from leaves),  and wind loss.
Evapotranspiration water usage of different bioethanol crops in the U.S. is  estimated to be 500 to
4,000 gallons of water per gallon of ethanol produced (Dominguez-Faus et al., 2009). It was also
reported to take -2,500 gal water/gal biofuel, which includes 820 gal of irrigation water
(UNESCO, 2014). The moisture content (in percentage units) levels of the freshly harvested and
field-dried biomass (shown in Table  7.2a) should be regulated to avoid spoilage, mold formation
and for optimizing bioethanol yield.
       The irrigation water supply and the water loss conditions (evapotranspiration and wind
loss) determine the total plant moisture content during harvest. The total biomass moisture
content includes intrinsic and extrinsic moisture content. Intrinsic moisture content is the amount
of biomass moisture measured without the influence of the ambient weather conditions. Extrinsic
moisture content is the amount of moisture measured after exposing the biomass to the prevailing
weather conditions of the harvest season. Higher moisture content (>60%)  in the plant residues
depletes the net calorific value of cellulosic bioethanol produced in addition to potential
drawbacks such as biomass spoilage, mold formation in the stored bales, and dry  matter loss. To
avoid this, the storage bales should be kept at moisture content of less than or equal to 20% (Ertl,
2013).
       Field drying is an economical method to reduce the extrinsic moisture content of
feedstock. Moisture content is lost or gained depending upon the storage conditions. Freshly
harvested corn stover contains -47-66% moisture. After the baling process, the reduced moisture
content is -16% (Petrolia, 2008). Air-drying of the hardwood by exposing the stacked wood
boards to the  blowing winds reduces the moisture content to 17% (Roise et al., 2013). Freshly
harvested switch grass contains 43% moisture content and the  moisture level reduces to 10%
after a  week depending upon the weather conditions. In Wisconsin, post-harvest field drying
reduces the moisture level of switch grass from between 46 and 66% down to 20% (Sokhansanj
et al., 2009).
                                          146

-------
     Table 7.2a Literature review of moisture content during the biomass harvesting process
Feedstock
Corn stover
Corn stover
Freshly harvested
Switch grass
Switch grass after 3-7
days of harvest
Source
Moisture content
Moisture content
Moisture content
Moisture content
Quantity
47-66 %
30%
43%
10-17%
References
(Somervilleetal.,2010)
(Wuefa/., 2014)
(Sokhansanjetal., 2009)
(Sokhansanjetal., 2009)
     Table 7.2b Literature review of water consumption for different feedstocks
Feedstock
Hard wood
Switch grass
Corn
Source
Freshwater
Freshwater
Freshwater
Quantity
2.4 gal of water consumed /
gal of biofuel
1.9-9.8 gal of water
consumed / gal of biofuel
2.8-40 gal of water consumed
/ gal of biofuel
References
(Somervilleetal.,2010)

(Tidwelletal., 2011)
       The conversion of cellulosic biomass to biofuel requires freshwater. The biofuel
conversion process consists of the pretreatment of candidate feedstocks, fermentation, and
biofuel recovery. Water requirements are compared for a combination of different feedstocks and
pretreatment technologies as summarized in Table 7.3. The water used within the process was
classified into the following three categories: freshwater, recycled water, or carried-over water.
Freshwater is make-up water from an external source such as a river, reservoir, or well. Recycled
water refers  to either reclaimed or non-treated process water. Reclaimed water is the water after
wastewater treatment subject to aerobic and anaerobic treatment while non-treated water refers to
a direct evaporator condensate captured from the cooling process, which does not require further
treatment. Carried-over water is the process water carried over to the next unit operation,  such as
hydrolyzate  from a pretreatment reactor, the broth sent to  a fermentation reactor, or the ethanol-
water mixture sent to the distillation column.
       All of the four cases in Table 7.3 share a common process configuration with minor
variations in the distribution of fresh and recycled water among various unit operations. Table
7.3 summarizes the differences in process configurations and utilization of recycled water. A
zero wastewater discharge design concept was applied to all cases. The variations in different
water networks for these process configurations is discussed in more detail in section 7.4.1
                                           147

-------
     Table 7.3  Four cases analyzed for water quality and quantity requirements
Plant Aspect
Feedstock
Dry tonne/day
Washing
Pretreatment
Biological
Conversion
Distillation
Column Bottoms
Use of
condensate from
evaporator
Residue
Processing
Chilled Water
Case 1 (Wu et a/.,
2009)
Hardwood
2,000
Yes
Dilute Acid
Simultaneous
Saccharification and
Co-Fermentation
(SSCF)
Evaporative
concentration of
distillation column
bottom liquids
Condensate is used
as recycled water for
pretreatment without
further dilution
Residue is burned to
produce steam &
power in Rankine
Cycle
N/A
Case 2 (Montague, 2002)
Corn Stover
2,000
No
Dilute Acid
Separate Saccharification and
Co-Fermentation
Evaporative concentration of
distillation column bottom
liquids
Condensate is used as recycled
water after dilution with make-
up water
Residue is burned to produce
steam & power in Rankine
Cycle
N/A
Case 3 (Laser et a/.,
2009a)
Switchgrass
4,535
Yes
Dilute Acid
Simultaneous
Saccharification and
Co-Fermentation
(SSCF)
Evaporative
concentration of
distillation column
bottom liquids
Condensate is used
as recycled water for
pretreatment without
further dilution
Residue is burned to
produce steam &
power in Rankine
Cycle
N/A
Case 4 (Laser et
a/., 2009a)
Switchgrass
4,535
No
AFEX
Consolidated
Bioprocessing
Distillation column
bottom liquid is
sent to waste
treatment
Condensate is
eliminated
Residue processed
to produce fuel,
power and animal
feed
Chilled water is
added for ammonia
recovery
7.4.1   Water quantity
       A water quantity requirement analysis was based on an overall water balance for the
entire process, as well as for a water balance for each unit operation. Water input into the entire
ethanol plant shown in Figure 7.4 consists of 1) moisture in feedstock, 2) moisture in the air,
chemicals and nutrients, and 3) make-up water. The overall water output from the entire ethanol
plant consists of water losses including: 1) windage and evaporation loss, 2) venting to the
atmosphere, 3) moisture included in solid waste, and 4) other handling losses. A typical
cellulosic ethanol production process has the following unit operations on the process side:
pretreatment, fermentation, and product recovery. The utility side of the cellulosic ethanol plant
consists of a cooling tower, residue processing, steam generation, and wastewater treatment. A
water consumption rate was obtained by dividing a volumetric water flow rate in gallons per
hour by an ethanol production rate in gallons per hour.  The accuracy of the data is the same as
that of the source data used from the references. The purpose and water consumption of each unit
operation are given below.
                                           148

-------
                                 Vent / CO,
         Windage/ Evaporation
               Feedstock
                Moisture

              Chemicals,
             Nutrients, Air
              Well Water
  Ethanol Plant
(water streams inside the
plant include carried-over
  and recycled water)
-*• Ethanol products
                                           Land Fill

Figure 7.4  Overall water balance of cellulosic ethanol plant.

       Pretreatment: Pretreatment physicochemically changes the structure of cellulosic
biomass to make the enzymes easily accessible to the cellulose and hemicellulose fractions for
the conversion into fermentable C5 and C6 monomeric sugars during subsequent fermentation.
Major water requirements of this section consist of steam injection into the pretreatment reactor
as process water and dilution water for slurry and hydrolyzate carried over to the fermentation
unit in order to maintain a required solid-to-liquid ratio in the saccharification and fermentation
tank.
       Saccharification and Fermentation: In this unit operation, the cellulose and
hemicellulose fractions are converted into sugars by the action of cellulase followed by
fermentation by microorganisms. Water is required for the activity of microorganisms in the
broth.
       Product Recovery:  The product recovery section consists primarily of distillation
columns where the ethanol and water mixture is separated for ethanol production. The product
recovery section also receives the ethanol-water mixture from fermentation. Water is required to
preheat the feed to the distillation column and its reboiler.
       Cooling Tower:  Cooling water is required throughout the process as utility water. In the
National Renewable Energy Laboratory (NREL) process (Wallace et a/., 2005),  the following
process steps require cooling water: 1) cooling of pretreated hydrolyzate, 2) cooling of the flash
vent from the pretreatment unit and pneumapress air vent before being sent to wastewater
treatment, 3) temperature control for fermentation at 41°C, 4) cooling of rectification column
reflux, 5) cooling of the wastewater streams before entering anaerobic digestion, and 6) cooling
of the evaporator liquid into condensate for recycling water. Heated water is then cooled down
by free or forced convection with make-up water acting as a heat sink. In this process, a
significant  amount of water is evaporated from the cooling tower and is considered evaporation
and windage loss. A typical cooling tower operation can consume 75% of the water sent to the
cooling tower by evaporation, and recycle the remaining 25% as blowdown (Owens,  2007;
Wurtz, 2008). The make-up water requirement primarily results from this loss.
       Steam Generation:  Steam is required in the pretreatment section as process water and in
the product recovery section as utility water for the preheater and reboiler of the distillation
                                            149

-------
column. Freshwater is subjected to boiler feed water treatment before steam generation inside the
boiler, which requires the highest quality water.
       Four cases of cellulose ethanol production were analyzed for water consumption. In the
overall water balance shown in Table 7.4, freshwater make-up is the single-most important water
input into the system. Other minor water flow components include feedstock moisture and
moisture in chemicals and nutrients (i.e., 0-0.2 gal water/gal ethanol). Freshwater make-up
ranges from 9 to 15 gal water/gal ethanol. Relatively, cases 2, 3 and 4 require a greater amount
of freshwater than case 1 because of a smaller quantity of water recycled in their process
configurations as shown in Figures 7.5 through 7.8. The detailed results for individual unit
operations are given below.
     Table 7.4 Overall water balance*

Casel
(Dilute acid +
Hardwood)
Case 2
(Dilute acid +
Corn Stover)
Case 3
(Dilute acid +
Switchgrass)
Case 4
(AFEX +
Switchgrass)
Inlet (gallons of water per gallon of ethanol)
Feedstock Moisture
Air, Chemicals, Nutrients
Make-up water
Inlet Total
3
0
9
12
1
0
12
13
2
Negligible
15
17
1
Negligible
12
13
Outlet (gallons of water per gallon of ethanol)
Evaporation & windage Loss
Vent to Atmosphere
Moisture in Residues for
Landfill
Handling Loss
Outlet Total
7
2
2
1
12
7
2
3
1
13
7
3
5
2
17
10
2
1
0
13
Note: * This water balance does not include carried-over and recycled water used inside the process.
                                           150

-------
Feedstock
Moisture
{3 gal/gal)
Air,
Chemicals,
Nutrients —
(0 gal/gal)
Ethanot^—
Wj
Evaporator
Condensate
(3 gal/gal) RW
* Pretrc
s
^ Detoxi
so
,3
Fermei
ar
Celli
Prodi
•=1
Pro<
Recc
	 * 	
Steam
(3 galfgal)
atment L
nd
fication
_ -
< x
Q. y
3 ToWWT
3 (5 gal/gal)
s
itation
id L
jlase
iction
•o
1
a
Juct 1
>very |
	 r 	
Make-up water^ Fresh Water
(9 gal/gal) * Sp|it
Evaporator Compensate
Water for Pretreatment Treatfd Water
("1 gal/galfWrV
+{2 gal/gal) RW Mlx
MW (1 gal/gal) f~RW (2 ga
Water for Pretreatment *
(1 gai/gal) MW *
(4 gal/jal) RW
Cantrifuge L
(3 gal/gal)
I "
I
Water foe Cellylas® I
Production
(1 gal/gal) RW —
Pretreatment WW
(5 gal/gal) p
Rectification Column
Bottoms ^
(9 gai/gaij
1 *
ter for Vent Scrubber Scrubber Vent to
(1 gal/gal) Atmosphere
(2 nal/cialt
Recycle Water
jquid Mix and Split
RW" j
Waste Water
Treatment
•h
i
Water for Vent Scrubber
Boiler Water
""""*" (3gal?gai)MW
3 gal/gal MW
Cootingjower
•^ ^ Water . —
I/gal) I 5 gal/gal RVf
Water from WW
~7 gaPgaTfRw"
J"
Solic
	 ^-(
Centrifuge and
Evaporator
\ \

Evaporator Centrifuge Liquid
Condansate (3 gal/gai) RW
(Saal/aaltRW
|Jg |
(7 gal/gai)
\ Cooling
I Tower
|
/

Ty ^\
I
S ||i§ Boiler Water Ste
•= §.51 (3 gal/gal) MW (3 ga
i=|i p U
1 Boiler
Boiler Slowdown ~"^^
Is to Landfill t
J gal/gal)
Unil OperdtiOfl
am
/gal)

— — _ Reeve

Dver Water
up Water
ed Water
Water
Water G^Ucns oi W^lor p@r Gall-on of Ilhanpl
Process Water
Pretreatment
F«»««.,,i«
PfoducE

Cooling Tower
Steam
Generation
rzi,ao,sll,re
Total WaEer
C«rri«
-------
                   Feedstock
                    Moisture __
                    (1 gal/gal)


                  Air, Chemicals,
                    Nutrients 	
                    (0 gatfgal)
to
                     Ethanol-
                                   Process Water Section
             Steam
            (2 gal/gal)
               I
               *

         Pretreatment
               and
        Detoxification
Evaporator Condensate
    (1 gal/gal) RW   ,

Water for Pretreatment I
  '"(fi'gaUgalj MW
    (8 gal/gal) RW
                                                 To WWT
                                                (5 gal/gall
                                    Fermentation
                                         and
                                      Cellulase
                                     Production
           Product
          Recovery
                                                               Make-up Water
                                                                        ~*
                                                                               Fresh Water
                                     Evaporator and Windage Loss
                        Water for Vent Scrubber
                           (1 gal/gal) MW

                          Boiler Water
                         (2 galfgal) MW
            MW (9 gal/gal)
                                                                            Recyle Water Mix
                      Cooling Tower Water
                     I (8 gaVgal) = 3~
-------
                         Protest Wa!«r Section
      F&edstoek
       Moisture "
       (2 gaifgal)
     Air, Chemicals,
       Nutrients —
       (0 gat/gal)
Evaporator
Cortdensato Steam
H galtgal) RW <2 9*8*1!
1 1
— »*-
Pretreatment
and
_ .*. Detoxification




&
t
^

^--

^«
I I 	
< To WWT
| (5 gal/gal)
fe
Fermentation
and
Cellulase
Production
I
sS"
1
I
a
Product Recovery
t H



— •
                 Water for Vent Scrubber   Scrubber Vent to
                      (1 gal/gal)         A^m®"*™

              MW: M»N»-«i» Water; RW: Roeyelw! Water: WW: W
(UgaUgal)' 	
Evaporator Condensatt
(3 gal/gaiJRW *
'"(5*ga"9»l)MW
{2 sol/gal) RW
(J gal/gal) MW
(2 gal/gal) RW
Water tot C»slul»«
Production
ft ga&ga^ $)W
Pretreatment WW
(5 gal/gall
Vifwte wal»( ^


Plant Water
Distribution
(4 gsl/g*l) MW 4 (2g«trg*i)RW
Recycled Water
Distribution
Waste Water
Treatment
II
	 £ 	
Centrifuge and
Evaporator
I
Evaporator
Condensate
to Pretreatment
(T galfgal) RW
| (1 gal/gal) MW
! Evaporator and Wirsdage Loss
[ __ V 7
I |Make-upWater\Cooling|
N 	 '(2gal/gal)MW iTovrar I
|!6Balr'Bal)RW / \
,
~j 7 \
.|'i'i
P 0


^^ — ^^ — Recycled
^^ — - — - WasleWa


•r Wilt
Water
Water
tec
Gallons of Waief p
-------
                          Process Water Section
Evaporator and
Feedstock
Moisture •
(1 gal/gal)
Air, Chemicals,
Nutrients
(0 gallgal)

(1 gal/gal)
	 *~
Al
— +- Pretre
S3
^
Ferme
( ai
Cell
Prodi
£3
Se
1


Water for Vent
Scrubber
(1 gal/gal)
-EX
'atment
i
2.
1
f
ntation
id
ulase «*-
jction
6
a.
F


1
Scrubber Vent to
Atmosphere
<2 gal/gal)
(iz gaugai)
Plant Water
Distribution
Water for
•— t>i 1 gal/gal) MW (10 gal/gal)
i 	 Boiler Water __!„__.
?= (1 gal/gai) MW ^—^—y
Make-up Water \CoollngI
*~" (JgaWgaflMW nTow«r I
(9 aallaal) RW 1 ^^L,_,,_
|4 gal/gal) MW 4
(7 gallgal) RW "~> H
S3' S
II
fjl
2S
l!
1
Waste Wat<
Production
(1 gallgal) MW
i
JS
0.8
|f


— — — Recycled Water
— • — Waste Water
-1" Boili
t ^'rfl
Solid

Unit OpBfatiort
/ ^^
i " — '
- 'o
° 5 = (1 gallgal) MW (
f 1« I
£ o -» , J
sti K
•e* i ©
s h Boi
jr Slowdown |
ia'l/gal) WW
s to Landfill
gal/gal)


Stoam
1 flallgal)
er
»

Gallons of Water p*&r Gallon of E
Procft^S Wat^f
Pretroatm^nt
F^MMtHH.
Product
Rweovory
Total Water Input 13 1
Raw Material ^©fslufe ^ 1
Steam 1
Total Walsr Input 13
Camed Over W»l« 12 f
Total Water Input 14
Carried Ovei Water 13 f
UlMity Water
Cooling Tower
Steam
Total Water Input 14
Make*up Water 5
Total Watttf Input 2
R
Make-up Water 2

^ak©-up Wat
lecyc od Wai
M.iko-Lip Wat
(•eye ed Wai
M,ifce-up Wat
lecyc ed Wat

weycl^d Watf
^cyc sd Wate
Note: MW : Make-up Water; RW: Recycled Water; WW: Waste Water.
Figure 7.8 Process layout based on case 4 (switch grass + afex pretreatment).

-------
      Table 7.5  Water balance in individual unit operations

Biomass capacity (Dry tonnes per
hr)
Ethanol capacity (Gallons of
ethanol/hr)
Case 1 (Dilute
acid +
Hardwood)
83
6,181
Case 2 (Dilute
acid + Corn
stover)
83
8,219
Case 3 (Dilute
acid +
Switchgrass)
189
7,018
Case 4 (AFEX +
Switchgrass)
189
21,888
Process Water (gallons of water per gallon of ethanol)

Pretreatment




Fermentation



Product
Recovery



Total water
Raw materials
Reclaimed water
Make-up water
Process steam**
Total water
Carried over water
Reclaimed water
Make-up water
Total water
Carried over water
Reclaimed water
Make-up water
Input
17
3(18%)
9 (53%)
2(12%)
3(18%)
13
12 (92%)
1 (8%)
0 (0%)
13
12 (92%)
0 (0%)
1 (8%)
Output
17




13



13



Input
17
1 (6%)
9 (53 %)
5 (29 %)
2 (12%)
13
12(92%)
1 (8%)
0 (0%)
13
12(92%)
0 (0%)
1 (8%)
Output
17




13



13



Input
18
2 (12%)
8 (47%)
6 (29%)
2 (12%)
13
12(92%)
1 (8%)
1 (0%)
14
13(93%)
0 (0%)
1 (7%)
Output
18




13



14



Input
13
1 (8%)
7 (54%)
4 (30%)
1 (8%)
13
12 (92%)
0 (0%)
1 (8%)
14
13(93%)
1 (7%)
0 (0%)
Output
13




13



14



Utility Water (gallons of water per gallon of ethanol)
Cooling
Tower


Steam
Generation


Total water
Reclaimed water
Make-up water
Total water
Reclaimed water
Make-up water*
8
5 (65%)
3 (35%)
3
0 (0%)
3(100%)
1


3


8
5 (60%)
3 (40%)
3
0 (0%)
3(100%)
1


3


8
3 (35%)
5 (65%)
2
0 (0%)
2(100%)
1


2


14
9 (64%)
5 (36%)
2
0 (0%)
2 (100%)
4


2


Note:   Make-up water for steam generation* is used for process steam directly injected into the pretreatment
       reactor** and utility steam for the preheater and reboiler used for distillation. The utility steam used for
       distillation is not shown because its consumption is relatively small.
       Pretreatment and Detoxification: In the water requirement estimation, the dilute acid
pretreatment used for cases 1, 2 and 3 requires  17 to 18 gal water/gal ethanol irrespective of the
feedstock, while the AFEX pretreatment used for case 4 consumes 13 gal water/gal ethanol. The
                                              155

-------
process steam is directly injected into the pretreatment reactor, accounting for 2 to 3 gal
water/gal ethanol for cases 1, 2 and 3, and 1 gal water/gal ethanol for case 4. Dilute acid
pretreatment for switchgrass consumes the largest amount of water. On the other hand, a
combination of switchgrass and AFEX pretreatment in case 4 requires the lowest water
consumption of 13 gal water/gal ethanol due to a very high ethanol yield—almost three times as
much ethanol than case 3 (Laser et al., 2009a).
       Furthermore, operating conditions are different among the four cases. The typical
operating temperature and pressure of the dilute acid pretreatment are 12.1 atm and 190 °C,
while AFEX pretreatment is typically operated at 21 atm and 90 °C (Laser et a/., 2009a).
Although there is a stringent water quality requirement for slurry dilution in the pretreatment
reactor, the recycle water generated from the wastewater treatment unit meets the requirements
to be used as make-up water for slurry dilution (Merrick & Company, 1998). In Case 2, for
example, the condensate from the evaporator was directly used as process water without
wastewater treatment as shown in Figure 7.4. The AFEX pretreatment used  for Case 4 turned out
to consume the least amount of water for pretreatment. For cases 1 through 3, the output water
from the pretreatment  section is estimated to be -17 to 18 gal water/gal ethanol, and 12 gal
water/gal ethanol was  carried over to fermentation.  The difference between pretreatment and
fermentation, ~5 to 6 gal water/gal ethanol, is directed to wastewater treatment.
       Fermentation:  Water consumption in fermentation is minimum as the hydrolyzate carried
over from the pretreatment and detoxification units  is already diluted water. Approximately 1 gal
water/gal ethanol is added to the process for the growth of micro-organisms.
       Product Recovery: Fermented beer is distilled to separate ethanol from water in product
recovery operations. Cooling water is recirculated in the condenser of a distillation column
without any make-up water. A vent scrubber used to separate CO2 consumes ~1 gal water/gal
ethanol. Wastes are recovered from the distillation column bottom in order to fully utilize the
energy required to generate steam for this section (Merrick & Company, 1998). The waste heat is
used in evaporators and the distillation  reboiler for saving energy. This arrangement also helps
reduce the load on the wastewater treatment unit by concentrating the streams. The steam
requirement as utility water for the reboiler and preheater of the distillation feed stream is less
than 0.5 gal water/gal ethanol for all cases. This utility steam consumption is smaller than the
process steam consumption for the pretreatment section in all cases.
       Wastewater Treatment: The aerobic and anaerobic wastewater treatment unit originally
designed by Merrick and Company (Hill et al., 2006) was used in all four cases. High  chemical
oxygen demand (COD) streams were sent to  the wastewater treatment unit, and a 95% decrease
in the COD value  was assumed to be achieved in the reclaimed water streams. The input streams
of the wastewater  treatment unit come from all unit operations, including 1) Pretreatment: waste
from flash vents and an ion-exchange unit, and 2) Product recovery: part of the evaporator
condensate is reclaimed in the wastewater treatment unit and is routed back  to the process. Some
of the waste streams from distillation columns are sent to evaporators and centrifuges  located
after the distillation columns in series. The recycled water collected from centrifugation is also
re-used in the process without any wastewater treatment where water quality requirements are
not stringent (Montague, 2002). Such streams help reduce the load on the wastewater treatment.
                                          156

-------
7.4.2   Water quality
       Soluble solids, insoluble solids and chemical oxygen demand (COD) are the primary
indicators of water quality affecting the cellulosic ethanol production process. In this study, their
concentrations were calculated for the streams sent to wastewater treatment, reclaimed water and
recycled water without treatment. Then the COD of each stream was calculated by multiplying
the mass flow rate of each component by a COD factor reported in a previous NREL study
(Merrick & Company, 1998)
       Reclaimed water: Recycled water is a supplement to make-up water. High-strength
wastewater treatment processes consist of aerobic oxidation followed by anaerobic digestion
with a typical 95% COD reduction rate. A wastewater treatment feasibility study (Merrick &
Company, 1998) recommends on-site wastewater treatment instead of discharge to the publicly
owned treatment works (POTW) for cellulosic ethanol production. Consistent with this
recommendation, all of the cellulosic ethanol plants analyzed here were assumed to use on-site
wastewater treatment. The effluent from on-site wastewater treatment is recycled in the ethanol
production process, and this water stream is labeled as recycled water in the analysis. Distillation
columns are followed by centrifuges and evaporators to concentrate the waste streams from
product recovery, and the water separated from centrifuges and evaporators is recycled to the
distillation columns. The centrifuges and evaporators are used to remove insoluble solids and to
recycle water through a three-stage evaporation process (Merrick & Company, 1998). The
evaporator condensate is referred to as non-treated recycle water. In a recent NREL process
design (Montague, 2002), the size of the wastewater treatment  unit was reduced by routing the
water, containing a high level of suspended solids (SS), into the centrifuge-evaporator system.
Water containing a high level of dissolved solids (DS) typically has a high COD content, and is
routed to the wastewater treatment unit instead of the evaporator system (Wooley et al., 1999). In
this analysis, different water networks were applied to individual cases following the network
configurations used in the references (Wooley et al., 1999; Montague, 2002; Laser et a/., 2009a).
       The quality of water streams used within a cellulosic ethanol plant is summarized in
Table 7.6. The COD level of the influent wastewater streams is in the range of 15,000 to 50,000
mg COD/L. After wastewater treatment with a reduction of 95% or higher, the COD level
decreases to the range of 750 to 2,500 mg COD/L. A reduction in the COD level for different
cases  depends on the performance of aerobic and anaerobic processes used in wastewater
treatment, and the applied COD removal efficiencies used for the four analyzed cases are
summarized in Table 7.6.
       Cooling Tower: The primary water consumption in the cellulosic ethanol production
process stems from windage and evaporation loss. The water loss from the cooling tower takes
-75 to 80% of the make-up water supplied to the cooling tower. The windage and evaporation
loss for case 4 is largest because 14 gal water/gal ethanol are sent to the cooling tower, whereas
for cases 1, 2, and 3, only 8 gal  water/gal ethanol are sent to the cooling tower. The higher
cooling water requirement for case 4 results from the use of chilled water for ammonia recovery.
The loss can be  avoided with the promotion of advanced cooling technologies, and some of the
water saving options and alternatives under consideration and development are summarized in
Table 7.6. From a water requirement analysis of the current cellulosic ethanol plant designs and  a
literature survey of water requirements for thermoelectric power plants, it  is understood that a
major share of make-up water withdrawal is used as cooling water. Hence, the development of
advanced cooling technologies that can reduce cooling water consumption and utilize

                                          157

-------
unconventional water resources will be critical to reducing water requirements for energy
production (Rodgers and Castle, 2008; Feeley and Carney, 2005; Manan et al., 2004).
     Table 7.6 Water quality used for four cases

Biomass capacity (Dry tonne/hr)
Ethanol capacity (Gallons of
ethanol/hr)
Casel
(Dilute acid +
Hardwood)
83
6,181
Case 2
(Dilute acid +
Corn stover)
83
8,219
Case3
(Dilute acid +
Switchgrass)
189
7,018
Case 4
(AFEX +
Switchgrass)
189
21,888
Quality of Influent Wastewater to Wastewater Treatment
Total input mass flow rate
(kg/hr)*
COD (mg COD/L)*
COD (kg/hr)*
% Soluble Solids
% Insoluble Solids
179,283
29,164
5,511
<0.1
<0.1
97,265
14,575
1,446
<0.1
<0.1
187,603
49,283
4,833
<0.1
<1.5
1,323,100
22,333
29,917
<0.1
<1.5
Quality of Effluent Wastewater from Wastewater Treatment
Total output mass flow rate
(kg/hr)*
Aerobic & anaerobic removal
efficiencies (%)
COD (mg COD/L)*
COD (kg/hr)*
% Soluble Solids
% Insoluble Solids
173,154
92
2,237
388
<0.1
<0.1
95,805
90
1,315
126
<0.1
<0.1
181,822
96
1,747
318
<0.1
<0.1
1,225,845
97
746
920
<0.1
<1.5
 Notes:   N/A - Not Available;
         * - Calculation based on the values reported in the references (See Table 7.3 for references).

       Steam Generation: Steam is required for direct injection into the pretreatment reactor as
process water and for the preheater and reboiler in the product recovery section as utility water.
The steam requirement for the preheater and reboiler in the distillation column is less than 0.5
gal water/gal ethanol production, and is much less than that for the pretreatment section as shown
in Table 7.4. The water quality requirement for steam generation is very stringent. The scale
build-up, frequency of boiler blowdown, and use of boiler chemicals can be minimized by
reducing such impurities as colloids, silicates, zinc, and alumina (Wooley et al., 1999).
Allowable total suspended solids in the boiler feed water are less than 1% for all four cases
(Wooley et al., 1999; Montague, 2002). Because of these  stringent water quality requirements,
the wastewater treatment effluents shown in Table 7.6 are unlikely suited for stream generation.
Large potential exists in 1) recycling and reusing most of the wastewater, and 2) avoiding the
windage and evaporation water loss that can significantly reduce make-up water.
                                            158

-------
7.5  Technological Challenges and Opportunities for Water Reuse and
     Conservation

       The growing water footprint due to increased biofuels usage has led to increased focus on
water reuse and its conservation in the production of biofuels. In 2005, -3% of irrigation water
worldwide was used for the production of biofuels (UNESCO, 2014). It is predicted that the
proportion of irrigated water is expected to be greater than 8% of the total withdrawals by the
end of 2030 due to the increased production of biofuels (Ottman, 2008). In the production of
biofuels in a dry mill ethanol plant, 30% of the total water consumption goes into the
fermentation of corn, which results in the cooking and conversion of the biomass to dextrose.
The remaining 70% of the water is used for cooling tower and boiler operations (Minnesota
EQB, 2007). As discussed in the above section, cooling water consumption is very high. A major
loss of cooling water results from evaporation and drifts (Martin et al., 2010). Methods to reduce
the consumption of the cooling water are being investigated. High efficiency cooling systems are
a viable option in decreasing water consumption in the cooling tower. Air-cooling using forced
convection  fans has been proposed and such systems are being used instead of water for system
cooling (Aden, 2007). This could potentially decrease the amount of water evaporating from the
cooling tower. However,  the energy efficiency of dissipating heat is relatively low compared to
water-cooling. For cooling water recycling, the water quality should be high in order to avoid
scaling inside the cooling tower (Schnoor  et al., 2008). One method known as the HiCycler®
method, removes hardness of water and silica (Owens, 2007). This is a propriety method, which
helps to reduce costs in both cooling water make-up as well as cooling water blowdown
(CHEMICO, 2014). This process reduces blowdown by 95%; however, it does not help in
cutting the evaporative losses in a cooling tower as it focuses primarily on reducing blowdown
losses.
       Production of second-generation biofuels is increasing in part due to RFS2 requirements
(EPA, 2010; EPA, 2014). Second generation biofuels are fuels that can be manufactured from
various types of biomass. Examples of biomass that can be used include lignocellulosic biomass
and woody  crops such as switch grass and also include crops such as algae (IPIECA, 2012). With
second-generation biofuels, the use of membrane filtration is a feasible option that can reduce the
amounts of waste produced and water consumed  compared to the first generation fuels, such as
starch-based ethanol. The production processes related to the first generation biofuels did not
focus on increasing the process efficiency, but processes are now being developed to increase the
process efficiency in the production of second-generation biofuels (Chem.info, 2010; Koch,
2014).  The  production of a high concentration ethanol broth is being investigated to reduce the
energy consumption during the distillation process (Aden, 2007). A membrane technology,
known as pervaporation,  is also a potential method to increase the ethanol concentration and
thereby reduce water usage. Pervaporation is a method in which two or more miscible
components are made to permeate through the membrane. On the other side of the membrane, a
vacuum is applied. This evaporates the liquid that passes through the membrane (Ramaswamy et
al., 2013; Vane, 2004). The use of graphene oxide-based membranes for the recovery of water
for reuse has been investigated. These have been used in an application of pervaporation used to
separate water from ethanol (Tang et al., 2014). Highly selective membranes can save energy
when compared to the present distillation process. However, temperature-sensitive compounds
such as the  microorganisms, were not suitable for this process (Vane, 2004). The capital costs are
higher for heating the compounds, and formation of precipitates may also be a problem at


                                         159

-------
elevated temperatures. If solids are present, microfiltration or centrifugation can be used as an
option to remove them. In place of the distillation process, in situ biofuel separation within the
fermentation broth is under development with the help of membrane processes to improve
fermentation performance (Balan, 2014).
       Another opportunity for reusing water in the ethanol industry is the use of a novel
membrane solvent extraction technology (U.S. EPA, 2011).  This patented technology uses a
porous membrane that helps to separate ethanol from a fermentation broth using an extracting
solvent. The solvent forms an immiscible solution with the extracted ethanol, which can be
processed further. By separating ethanol from water, the cooling load decreases, which in turn
decreases the use of water. This process can effectively remove ethanol continuously during the
fermentation process, which in turn increases the fermentation process and also increases
throughput. This process is useful for second-generation ethanol production as it helps in
commercializing cellulosic biomass as a feedstock.  However, this technology has not been scaled
up to pilot-scale operation. The costs involved may also be a barrier when compared with current
separation techniques. Another technology, ultrafiltration, is used for clarifying the process
stream after conversion into sugars, which is used for the purification of biofuels. A major
problem encountered in this technology has been membrane clogging (Khanal, 2010).
       Reverse osmosis is used in downstream processing instead of an  evaporator for the
recovery of water (Cho et al., 2012). This process has reduced operating costs by almost 75%
(Jevons, 2011). A reverse osmosis process powered by photovoltaic renewable energy has also
been proposed. A disadvantage of this novel technique is that the pressure required to drive the
process is heavily dependent on climatic and weather conditions, which in turn, reduces the life
of the membrane.  Adequate pretreatment of water to remove minerals such as magnesium and
calcium ions can be used to prevent scaling in this process. However, this leads to an increase in
the operational and maintenance costs. The membrane technology's efficiency depends on the
selectivity of the membranes, scale-up of the processes to industrial scale, reliable and consistent
long-term performance, and the capital cost required to install these technologies. Another
technology proposed for the reuse of water is electrodialysis driven by solar energy (Khanal,
2010). This technology has advantages relative to reverse osmosis technology, as there is no
dependence on climatic and weather conditions. However, a major drawback of this technology
is that nonpolar compounds cannot be removed.
       The technologies to reduce the water consumption in the production of biofuels are
primarily focused on the use of membranes. The systems employing membranes are mainly
targeted towards reducing the energy consumption, which will in turn reduce water consumption
in biofuels production. However, there are some problems that must be resolved in order to use
these technologies at a commercial level. The selectivity of the membranes remains a significant
issue. The capital costs inherent with these processes also remain an issue, as they need to be
cost competitive with currently available processes.

7.6   Summary

       In this study, the process and utility water requirements for cellulosic ethanol production
has been assessed based on the published data in the literature. The study is not intended for a
life-cycle analysis of water usage from biomass production to waste disposal.  Instead it is
focused on a detailed engineering examination of water usage and the potential for water
                                          160

-------
conservation at a unit operation level for cellulosic ethanol production. Overall, 12 to 17 gallons
of water are consumed for each gallon of ethanol production as derived from a mass balance
analysis for four combined cases of feedstocks and pretreatment technologies. This is
approximately within the range of water-intensities reported for corn ethanol production within
Chapter 6 of this report. Among the four cases analyzed, the AFEX pretreatment technology
requires less process water (i.e., 13 gal water/gal ethanol) than the dilute acid pretreatment
technology (i.e., 17 to 18 gal water/gal ethanol). This is primarily due to the fact that a far greater
ethanol yield is obtained when AFEX pretreatment is used for herbaceous feedstocks  such as
switch grass. Hence, a process water requirement is primarily determined by the combination of
a feedstock and a pretreatment technology. The best pretreatment technology for a given
feedstock can be determined, and thus a process water requirement for a cellulosic feedstock can
also be determined.
       Utility water for cooling tower and steam generation accounts for -40 to 70% of the total
freshwater consumption in the process. The utility water requirements for the AFEX
pretreatment are higher (i.e., 14 compared to 8 gal water/gal ethanol for the dilute acid
pretreatment) due to the chilled water requirements for ammonia recovery. However,  the
consumption of total process and utility  water is not significantly different for the four analyzed
cases. The water networks for all cases were designed for zero wastewater discharge.
Nonetheless, the use of recycled water will primarily depend on process economics and will be a
function of the availability of freshwater make-up, and the capital and operating costs of an on-
site wastewater treatment system. Overall, alternative water saving options and advanced cooling
tower design are critically important to minimize freshwater consumption.

7.7  References

Aden, A. 2007. "Water usage for current and future ethanol production." Southwest Hydrology,
       6(5):22-23.

Alvarez, P.J., J.G. Burken, J.D. Coan, M.E.D. Oliveira, R.D. Faus, D.E.  Gomez, and A.M. Jaffe.
       2010. "Fundamentals of a sustainable biofuels policy."  James A. Baker III Institute for
       Public Policy and Rice University Department of Civil and Environmental Engineering.

Balan, V. 2014. "Current challenges in commercially producing biofuels from lignocellulosic
       biomass."  International Scholarly Research Notices Biotechnology, 2014:463074.

Chem.info. 2010. "Membrane Filtration Rules Biofuels."
       http://www.chem.info/articles/2010/ll/membrane-filtration-rules-biofuels. Accessed July
       2014.

CHEMICO International Inc. 2014. HiCycler®. http://www.chemico.com/HiCycler.htm.
       Accessed July 2014.

Chiu, Y.-W., B. Walseth, and S. Suh. 2009. "Water embodied in bioethanol in the United
       States." Environmental Science  and Technology, 43(8):2688-2692.
                                           161

-------
Cho, Y.H., H.D. Lee, and H.B. Park. 2012. "Integrated membrane processes for separation and
       purification of organic acid from a biomass fermentation process." Industrial &
       Engineering Chemistry Research, 51 (3 0): 10207-10219.

Dale, B. 2007. "Cellulosic Ethanol and Sustainability: There is no 'food vs. fuel' conflict."
       233rd ACS National Meeting, Chicago, IL.

Dominguez-Faus, R., S.E. Powers, J.G. Burken, and PJ. Alvarez. 2009. "The water footprint of
       biofuels: A drink or drive issue?" Environmental Science and Technology, 43(9):3005-
       3010.

EISA.  2007. "U.S. Public Law 110-140. Energy Independence & Security Act of 2007"  U.S.
       Senate Committee on Energy & Natural Resources. Washington, DC.
       http://www.gpo.gov/fdsvs/pkg/PLAW-110publl40/html/PLAW-110publl40.htm.

Ertl, D. 2013. "Sustainable Corn  Stover Harvest." Iowacorn.org,
       https://www.iowacorn.org/documents/filelibrary/research/research_reports/IowaCornRes
       earchBrochure  FINAL 31478BB786257.pdf.

Feeley, TJ. and B. Carney. 2005. "Innovative Approaches and Technologies for Improved
       Power Plant Water Management."  U.S. Department of Energy, Office of Fossil Energy,
       National Energy Technology Laboratory.

Hill, J., E. Nelson, D. Tilman, S.  Polasky,  and D. Tiffany. 2006. "Environmental, economic, and
       energetic costs and benefits of biodiesel and ethanol biofuels." Proceedings of the
       National Academy of Sciences of the United States of America, 103(30): 11206-11210.
       doi: 0604600103 [pii].

Hoekman, S.K. 2009. "Biofuels in the US-challenges and opportunities."  Renewable Energy,
       34(1): 14-22.

IPIECA. 2012. "The Biofuels and Water Nexus."
       http://www.ipieca.org/news/20120723/biofuels-and-water-nexus.

Jevons, K. 2011. "Membrane technology in production of biofuels." Ethanol Producer
       Magazine, March 2011. http://www.ethanolproducer.com/articles/7568/membrane-
       technology-in-production-of-biofuels

Karimi, T. A. 2007. "Enzyme based hydrolysis process for ethanol from lignocellulosic materials:
       A review." Bioresources, 2:707-738.

Keeney, D.,  and M. Muller. 2006. Water use by Ethanol Plants: Potential Challenges. Institute
       for Agriculture and Trade Policy, Minneapolis.

Khanal, S. K., ed.  2010. Bioenergy and biofuelfrom biowastes and biomass. ASCE Publications.

King, C.W. and M.E. Webber.  2008. "Water intensity of transportation." Environmental Science
       and Technology, 42(21):7866-7872.
                                          162

-------
Koch Membrane. 2014. "Membrane technology in the production of biofuels."
       http://www.kochmembrane.com/PDF s/KMS_Biofuel_Production_Applicati on.aspx
       Accessed May 3, 2014

Laser, M., H. Jin, K. Jayawardhana, and L.R. Lynd. 2009a. "Coproduction of ethanol and power
       from switch grass." Biofuels, Bioproducts and Biorefming, 3(2): 195-218.

Laser, M., E. Larson, B. Dale, M. Wang, N. Greene, and L.R. Lynd. 2009b. "Comparative
       analysis of efficiency, environmental impact, and process economics for mature biomass
       refining scenarios." Biofuels, Bioproducts and Biorefining, 3(2): 247-270.

Li, Y., J. Shi, and R. Redder. 2011. "Storing Lignocellulosic Biomass for Bio-Refining
       Industry."  The Ohio State University Extension.

Lingaraju, B.P, J.Y. Lee, and Y. J. Yang. 2013. "Process and utility water requirements for
       cellulosic ethanol production processes via fermentation pathway." Environmental
       Progress & Sustainable Energy, 32(2):396-405.

Manan, Z. A., Y.L.Tan, and D.C. Y Foo. 2004. "Targeting the minimum water flow rate using
       water cascade analysis technique." AIChE Journal, 50:3169-3183.

Martin, M., E. Ahmetovic, and I.E. Grossmann. 2010. "Optimization of water consumption in
       second generation bioethanol plants."  Industrial and Engineering Chemistry Research,
       50(7):3705-3721.

McMahon, P.B., J.K. Bohlke, and C.P. Carney. 2007. "Vertical Gradients in Water Chemistry
       and Age in the Northern High Plains Aquifer, Nebraska 2003." US Department of the
       Interior, US Geological Survey.

Mei, F. 2006. "Mass and Energy Balance for a Corn-To-Ethanol Plant." M.S. Thesis,
       Washington University in St. Louis.
       http://crelonweb.eec.wustl.edu/theses/Masters/Fan%20Mei%20-Master%20thesis.pdf.
       Accessed June 3, 2014.

Merrick & Company. 1998. "Wastewater Treatment Options for the Biomass-to-Ethanol
       Process. Report No AXE-8-18020-01." Final Report of Merrick & Company to National
       Renewable Energy Laboratory Merrick & Company:  Golden, CO.

Mielenz, J.R. and J.R. Mielenz. 2009. Biofuels: Methods and Protocols. Humana Press.

Minnesota Environmental  Quality Board. 2007. "Decrease Demand for Water used in Ethanol
       Production."
       http://www.eqb.state.mn.us/documents/forumUploads/61/61/Decreasedemandforwaterus
       edinethanolproduction.doc. Accessed April 28, 2007.

Montague, L. 2002. "Lignocellulosic Biomass to Ethanol Process Design and Economics
       Utilizing Co-Current Dilute Acid Prehydrolysis and Enzymatic Hydrolysis for Corn
       Stover." National Renewable Energy Laboratory Report NREL/TP-510-32438.
                                         163

-------
Mosier, N., C. Wyman, B. Dale, R. Blander, Y.Y. Lee, M. Holtzapple, and M. Ladisch. 2005.
       Features of promising technologies for pretreatment of lignocellulosic biomass.
       Bioresource Technology, 96(6):673-686.

Ottman, M. J. 2008. "Growing Crops for Biofuels: Implications for Water Resources."
       http://alfalfa.ucdavis.edu/+svmposium/proceedings/2008/08-259.pdf

Owens, S. R. 2007. "Reduce cooling tower water consumption by 20 percent." Ethanol
       Producers Magazine,  13(6):138-141.

Perlack, R.D., L.L. Wright, A.F. Turhollow, R.L. Graham, BJ. Stokes, and D.C. Erbach. 2005.
       "Biomass as Feedstock for a Bioenergy and Bioproducts Industry: The Technical
       Feasibility of a Billion-Ton Annual Supply." Technical Report. Oak Ridge National
       Laboratory, TN.

Petrolia, D.R. 2008. "The economics of harvesting and transporting corn stover for conversion to
       fuel ethanol: A case study for Minnesota." Biomass and Bioenergy,  32(7): 603-612.

Pfromm, P.H. 2008. "The minimum water consumption of ethanol production via biomass
       fermentation."  Open Chemical Engineering Journal 2.

Ramaswamy, S., H. Huang, and B.V. Ramarao. 2013. "Membrane Pervaporation." In
       Separation and Purification Technologies in Biorefineries, p. 259-299. John Wiley &
       Sons Ltd.

RFA. 2014. "U.S. Fuel Ethanol Industry Biorefineries and Production Capacity."
       http://www.ethanolrfa.org/bio-refmery-locations/. Accessed August  13, 2014.

Rodgers, J.H., and J.W. Castle. 2008. "An Innovative System for the Efficient and Effective
       Treatment of Non-traditional Waters for Reuse in Thermoelectric Power Generation."
       Final Technical Report to U.S. Department of Energy, DE-FG26-05NT42535.

Roise, J.P., G. Catts, D. Hazel, A. Hobbs, and C.  Hopkins. 2013. "Balancing biomass harvesting
       and drying tactics with delivered payment practice. Refining woody biomass feedstock
       logistics."  North Carolina State University.

Sarkar, N., S.K. Ghosh, S. Bannerjee, and K. Aikat. 2012. "Bioethanol production from
       agricultural wastes: An overview." Renewable Energy, 37(1): 19-27.

Schnoor, J.L., O.C. Doering, D. Entekhabi, E.A.  Hiler, T.L. Hullar, and D. Tilman. 2008. "Water
       Implications of Biofuels Production in the United States." National Academy of
       Sciences, Washington DC, USA.  http://www.Nap.edu/catalog/12039.Html.

Shapouri, H. and P. Gallagher. 2005. "USDAs 2002 Ethanol Cost-of-Production Survey."
       United States Department of Agriculture,  Office of the Chief Economist, Office of
       Energy Policy and New Uses.
                                          164

-------
Shinners, K.J., G.C. Boettcher, D.S. Hoffman, J.T. Munk, R.E. Muck, and P.J. Weimer. 2009.
       "Single-pass harvest of corn grain and stover: Performance of three harvester
       configurations. Transactions of the ASAE (American Society of Agricultural Engineers).,
Sokhansanj, S., S. Mani, A. Turhollow, A. Kumar, D. Bransby, L. Lynd, and M. Laser. 2009.
       "Large-scale production, harvest and logistics of switch grass (Panicum virgatum L.)-
       current technology and envisioning a mature technology." Biofuels, Bioproducts and
       Biorefining, 3 (2) : 1 24- 1 4 1 .

Somerville, C., H. Youngs, C. Taylor, S.C. Davis, and S.P. Long. 2010. "Feedstocks for
       lignocellulosic biofuels." Science (Washington) 329(5993):790-792.

Summers, D.R. 2006. "Rectifier Design for Fuel Ethanol Plants." Paper presented at AIChE
       Annual meeting, San Francisco, CA. November 14, 2006.

Sun, Y., and J. Cheng. 2002. "Hydrolysis of lignocellulosic materials for ethanol production: A
       review." Bioresource Technology, 83(1): 1-11.

Swain, B. 2006. "Water Conservation and Treatment Strategies for Ethanol Plants." Paper
       Presented at Governor's Ethanol Coalition, EPA Joint Meeting Kansas City, MO.

Tang, Y.P., D.R. Paul, and T.S. Chung. 2014. "Free-standing graphene oxide thin films
       assembled by a pressurized ultrafiltration method for  dehydration of ethanol." Journal of
       Membrane Science, 458:199-208.

Tidwell, V., A.C.-T. Sun, and L. Malczynski. 201 1. "Biofuel Impacts on Water." No.
       SAND2011-0168.  Sandia National Laboratories.

U.S. Bioenergy Statistics, U.S. Department of Agriculture Economic Research Service, last
       updated on April 30, 2014. http ://catalog. data. gov/dataset/u-s-bioenergy-statistics.
       Accessed August, 2014.

U.S. EIA. 2014. "Short Term Energy Outlook."  http://www.eia.gov/forecasts/steo/query/.
       Accessed June, 2014.

U.S. EPA. 2010. http://www.epa.gov/otaq/renewablefuels/420fl0007.pdf.

U.S. EPA. 2011. "Advanced Water Removal via Membrane Solvent Extraction."
       http://energy.gOv/sites/prod/files/2013/l l/f4/adv_water_removal_mse.pdf.

U.S. EPA. 2014. http ://www. epa. gov/otaq/fuels/renewablefuels/documents/420f 1 3 048 .pdf.

UNESCO. 2014. World Water Assessment Programme.

Vane, L.  2004. "Options for Combining Pervaporation Membrane Systems with Fermentors for
       Efficient Production of Alcohols from Biomass." Presented at AIChE Annual Meeting,
       November 2004. Topical Conference T4.
                                          165

-------
Wallace, R., K. Ibsen, A. McAloon, and W. Lee. 2005. "Feasibility Study for Co-Locating and
       Integrating Ethanol Production Plants from Corn Starch and Lignocellulosic Feedstocks."
       National Renewable Energy Laboratory, Golden, CO.

Wang, Y. and T.-S. Chung. 2012. "Membranes for biofuel separation." Asian Biotech (APBN)
       Magazine, 16:34-39.

Wooley, R., M. Ruth, J. Sheehan, K. Ibsen, H. Majdeski, and A. Galvez. 1999. "Lignocellulosic
       Biomass to Ethanol Process Design and Economics Utilizing Co-Current Dilute Acid
       Prehydrolysis and Enzymatic Hydrolysis Current and Futuristic Scenarios."  Technical
       Report. National Renewable Energy Laboratory, Golden, CO.

Wu, M., M. Mintz, M. Wang, and S. Arora. 2009. "Consumptive Water use in the Production of
       Bioethanol and Petroleum Gasoline."  Argonne National Laboratory Center for
       Transportation Research, Energy Systems Division.

Wu, M., Z. Zhang, and Y.-W. Chiu. 2014.  "Life-cycle water quantity and water quality
       implications of biofuels." Current Sustainable/Renewable Energy Reports, 1(1):3-10.

Wurtz, W. 2008. "Air Cooled Condensers Eliminate Plant Water Use."
       http://www.powermag.com/issues/features/1361.html. Accessed September 2008.

Zink, J. 2007. "Water-use Estimates for Ethanol Plant Go Up 60 Percent."  Tampa Bay, FL.
       http://www.sptimes.com/2007/03/09/Hillsborough/Water use  estimates fshtml.
                                          166

-------
 8   Biodiesel Production and Impacts on Water Resources
       Qingshi Tu ' and Mingming Lu1

8.1   Introduction

       As a renewable alternative to petroleum diesel, biodiesel has been widely used in the U.S.
and around the world. In the U.S., a record high of 1.8 billion gallons of biodiesel (including
1.34 billion gallons of biodiesel with the remainder being renewable diesel) was produced in
2013 (EIA, 2014), compared to 315 million gallons in 2010, 700 million gallons in 2008, and
just 25 million gallons in 2004 (Figure 8.1). A decline in biodiesel production from 2008 to 2010
was  due to the expiration of the blender's credit in 2008. Oil plant growth requires water
(irrigation or precipitation) and water is  needed to process oil seed into oil and the oil into
biodiesel. As an example, currently soybean remains the dominant biodiesel feedstock in the
U.S., which requires a large quantity of irrigation water consumption for growth. The goal of this
chapter is to investigate water consumption during biodiesel production.
                           U.S. Biodiesel Production
                           Source: National Biodiesel Board Annual Estimates
              1800
                                                                     2012   2013"
                                                * Year includes entire Biomass-Based Diesel category
Figure 8.1   Annual biodiesel production in U.S. (2003-2013). Source: National Biodiesel
       Board (http://www.biodiesel.org/production/production-statistics)
       This chapter will begin with a summary of the current status of the U.S. biodiesel
industry, followed by an estimate of water consumption for the processing of lipids into
biodiesel. Currently, soy oil is the primary source of lipids that are processed into biodiesel fuel
in the U.S.  The production of soy-based fatty-acid methyl ester (FAME) biodiesel (the biodiesel
   University of Cincinnati, University of Cincinnati Department of Biomedical, Chemical, and
   Environmental Engineering
                                           167

-------
process) are defined as consisting of the following three stages: soybean irrigation, soybean-to-
soybean oil processing, and biodiesel manufacturing. This analysis of water consumption from
the biodiesel process began with a survey of relevant literature. Water consumption was then
estimated by using characteristic allocation factors for each of these three stages. Both state-level
estimates and national averages of water consumption were determined within this analysis.
Results from this analysis were then compared with several relevant studies in detail. Water
consumption patterns of water-stressed areas were also summarized. The chapter concludes with
a discussion of future biodiesel trends, including the use of new feedstocks and the use of new
biodiesel production technologies since these changes can also affect water use. Water use from
algal biodiesel is also briefly summarized.

8.2    Water Consumption Estimates for the Soybean Oil  FAME Biodiesel
       Process in the U.S.

       A water consumption analysis for the processing of soybean oil  to biodiesel in the U.S.
using transesterification to FAMEs was  conducted. The analysis aimed  at evaluating the water
consumption in three stages of the process: soybean growth, soybean oil processing, and soybean
oil transesterification to FAME biodiesel.
8.2.1   Methodology

       Water consumption in the biodiesel process is estimated as the sum of irrigation water
use in the soybean growth stage (Wy), water use during soybean crushing and processing into
soybean oil (W^), and water use in biodiesel production (Wj). Water usage in fuel transportation
and distribution was not included due to a lack of data. The water consumption associated with
producing energy that is used to move and treat the water was also not included in the scope of
this study. Both Wy  and W2 focus on soybeans as a feedstock due to its  dominant market share in
biodiesel production in the U.S. and the availability of data. Wy, W2 and Wj are expressed in the
unit of "million gallons per year (MMgpy)." Ny, N2 and Nj are the normalized values for each
stage based on biodiesel produced in the unit of "gallons of water per gallons biodiesel
(gal/gal)," which is commonly used by other studies (Wu et al., 2009; Pate et al., 2008; Martin
and Grossmann, 2012).  The parameters for state-level water consumption are expressed as WT/,
W2/ and Wj/, withy representing each state. The overall total water consumption for the U.S.
(Wto?) is the sum of Wi, W2 and Wj, and corresponding normalized value N/0? is the sum of Ny,
N2 and Nj. As an example, details in estimating  of W#, W2/, Wj/, NT/, N2/ and Nj/ for Ohio are
provided in the Appendix to this chapter.

8.2.2   Data sources

   8.2.2.1  Soybean irrigation stage: irrigation water consumption (Wij, Nij)

       The "Farm and Ranch Irrigation Survey" (2008 is the most current) was used in this
analysis since U.S. Department of Agriculture (USDA) surveys are the  most comprehensive
surveys of irrigation water use for soybean agriculture in the U.S. (USB, 2010; USDA, 2008). In
estimating Wij, the following factors have been considered:  the fraction of soybeans processed
into biodiesel, the oil content of the soybeans, and the efficiency of the transesterification
reaction.
                                          168

-------
   8.2.2.2  Processing soybean into soybean oil stage: crushing, oil extraction, and crude soy
           oil refining (W2j, N2j)

       After harvesting, soybeans are transported to the refining plant for crushing, oil
extraction, and crude oil degumming. Water consumption in this stage is primarily related to
equipment operation, such as cooling tower make-up or water use in free fatty acid (FFA)
removal (USB, 2010; van Gerpen et al., 2004). The FFA is usually removed from soybean oil via
caustic refining (i.e., neutralizing the FFA with a caustic soda and using water to wash away the
soap formed). Other refining practices, such as bleaching and deodorizing are not as considered
at this stage since they are less typical. The consumptive water involved was summarized in an
aggregated form by the National Oilseed Processers Association (NOPA), which was the result
of a representative  survey among its member companies in 2008 (USB, 2010).

   8.2.2.3  Biodiesel production stage: crude biodiesel purification, cooling tower make-up
              j, N3j)
       In biodiesel manufacturing from soy oil, the following processes are found to be
associated with water use: biodiesel washing to remove residual glycerin and other impurities,
boiler make-up, and cooling tower make-up. The actual consumption can vary considerably
depending upon the system setup and the extent of heat economization used in the facility. Due
to the pretreatment requirements for the wash water prior to discharge, waterless
separation/purification or "dry-wash" technologies are increasingly practiced by biodiesel
producers to replace water washing. Even for water washing, the wash water is reused instead of
being discharged after one use. Boiler water make-up should be considered when distillation is
used to separate glycerin and other impurities from biodiesel, and the rates vary depending on the
distillation processes (vacuum or steam distillation) used in the facilities. The resultant boiler
water make-up from vacuum distillation can be much lower than for steam  distillation. Cooling
tower make-up water should be considered in Wj if the producer uses evaporative cooling towers
to condense process vapors (such as for methanol recovery) and cool liquid process streams. In
this analysis, these data were collected from actual biodiesel producers in addition to using data
from published literature (Tu et al., 2014).

  8. 2. 2. 4  States reporting zero water use

       Fifteen states (Alaska, Arizona, California,  Florida, Hawaii, Idaho, Montana, Nevada,
New Hampshire, New Mexico, Oregon, Pennsylvania, Rhode Island, Utah and Wyoming) are
not included in the calculation of WT/ and NT/, either due to negligible soybean growth or lack of
irrigation data from the USD A. Of these 15 states, the states of Florida, Montana and
Pennsylvania have sufficient soybean harvest data and accordingly are included in the
calculation of W2/ and N2/
       Four states (Colorado, Montana, Vermont and Wyoming) were not included in the
calculation of Wj/, total and normalized water consumption during the biodiesel manufacturing
stage, because there was no biodiesel production in those states prior to conclusion of this
analysis.
                                           169

-------
8.2.3  Results
   8.2.3.1  Water consumption in soybean growth stage (Wij, Nij)
       Appendix Figures 8.1 and 8.2 (see Chapter 8 Appendix) show the results of irrigation
water consumption (Wy/) and irrigation water intensity (Ny/) for soybean agriculture dedicated to
biodiesel production in 35 states in the U.S. While WT/ is a direct reflection of irrigation water
consumption, NT/ is an important measure of irrigation intensity regardless of the soybean growth
scale for a specific state. The irrigation  water use WT/ varies significantly from state to state, 0.00
to 15,953.00 MMgpy. The range of normalized irrigation intensity (Ny/) varies from 1058.20
gal/gal (Washington) to 0.00 gal/gal (states with minimal irrigation) with the weighted
nationwide average (Ny), median and weighted standard deviation being 61.78, 23.67 and 147.92
gal/gal, respectively.
       The states with negligible irrigation consumption (0.00 MMgpy) were Massachusetts,
Connecticut, Maine, Vermont, West Virginia and New York, primarily due to limited soybean
growth. In fact, Massachusetts, Connecticut, Maine, Vermont, West Virginia and New York rank
35th, 34th, 33rd, 31st, 29th and 23rd in terms of total amount of soybeans harvested. On the other
hand, the states with the highest irrigation water use, Arkansas (15,953.00 MMgpy), Nebraska
(9056.78 MMgpy), Mississippi (3714.78 MMgpy), Kansas (2514.84 MMgpy) and Missouri
(2456.94 MMgpy), are also major soybean producers, ranking 10th, 6th, 14th, 11th and 7th among
the 38 states that reported soybean harvests.
       The irrigation water intensity of 11 states (Washington, Arkansas, Colorado, Mississippi,
Nebraska, Texas, Delaware, Kansas, Louisiana, Georgia and Oklahoma) are above the national
average, with values of 1058.20, 674.30, 611.52, 285.90, 199.74, 190.08, 148.21, 127.09,
108.59, 106.75  and 88.13gal/gal, respectively. These 11 states represent 18.32% of the total
soybean harvest, and 36.10% of total biodiesel production capacity. As previously mentioned,
Arkansas, Mississippi and Nebraska are three states with both significant soybean growth and
significant irrigation water consumption. Although the states of Washington and Colorado have
very high irrigation water intensities, their total irrigation water consumption  (Wim, Wyco) is
lower than the national average due to much less soybean cultivation. On the other hand, Iowa,
Illinois, Minnesota, Indiana and Ohio account for 56.37% of U.S. soybean production, while
their irrigation water intensities are only 1.88, 4.01, 8.60, 7.85 and 0.71gal/gal, respectively. The
much lower NT/ values reported in states such as Illinois, Indiana, Iowa, Kentucky, Minnesota,
North Dakota, Ohio, South Dakota, Tennessee, etc., are due to much less irrigation used in
soybean production in these states.
       The vastly different irrigation ratios (irrigated acres vs. total acres) are the primary
contributing factors with respect to the wide range of state level irrigation water intensities. The
average irrigation ratio is 8.2% among these 35 states, with the range of 0% to 65.40% (with the
highest irrigation ratio for Arkansas), and 25 states have irrigation ratios below average. For the
top five soybean producing states, soybean irrigation ratios range from 0.02% to 1.72%. These
results indicate  that it may be advantageous to grow soybeans where less irrigation is needed
rather than in states that have high irrigation ratios. Due to the significant variation in irrigation
practices among U.S. states, a simple national average cannot accurately represent water use.
                                           170

-------
   8.2.3.2
Water consumption in the soybean processing and refining stage (Wy, N2j)
       A uniform value of the N2 (0.17 gal/gal) was used for calculation of water consumption,
and the range of W2/ varied from 0.003 to 112.00 MMgpy.

   8.2.3.3   Water consumption in the biodieselproduction stage (Wsj andNsj)

       Water consumption data for the biodiesel  washing process varies significantly among
different studies. The National Biodiesel Board (NBB) estimated that one pound of wash water
was needed for four pounds of biodiesel, which is equivalent to 0.22 gal/gal (Scott, 2010).  The
United Soybean Board (USB) conducted a life cycle assessment (LCA) for the soybean-to-
biodiesel process, where water used for biodiesel washing was reported as 0.26 gal/gal (USB,
2010). However, the water consumption reported from simulations was 0.03 (Haas et a/, 2005)
and 0.01 gal/gal (Zhang etal., 2003), respectively.
       The substantial difference in water use among data sources warranted data collection
from actual biodiesel manufacturers. In this analysis, inquiries were sent to 123 commercial
biodiesel producers listed by the NBB34. A total of 21 replies were received, among which six
reported water washing, 11 indicated dry purification  or "dry washing" and four considered this
information proprietary. The weighted average water consumption based on plant capacity was
0.12 gal/gal for biodiesel washing (Appendix Table 8.3). Therefore, water consumption in
biodiesel washing for this analysis was determined to be in a range from 0.12 to 0.26 gal/gal for
water washing processes and 0 gal/gal for dry washing.
       The dry wash method often consumes less water during distillation compared with water
washing. Water consumption for cooling tower make-up is presented in Appendix Table 8.4.
This again indicates the highly process-specific characteristics of actual biodiesel production
operations. The water consumption of cooling water make-up was averaged based on plant
capacity, and was also separately analyzed for dry washing and water washing.  Accordingly, the
cooling tower make-up for water washing was 0.275 gal/gal  and was 0.153 gal/gal for dry
washing. Only limited information was available  on boiler water make-up and the extent of dry
wash use among biodiesel producers, so these were assigned zero values.
       The water consumption rates in biodiesel  production (Nj) were summarized based on
three scenarios: water washing (upper range), water washing (lower range) and dry washing with
the corresponding values for Nj being 0.54, 0.4 and 0.15 gal/gal, respectively. On average, dry
washing consumes approximately one third of the total water consumed within the biodiesel
manufacturing process. A uniform Nj of 0.31 gal/gal was calculated by averaging the three
scenarios. Accordingly, the resultant water consumption in biodiesel production (Wj/) was
estimated based on Nj and biodiesel capacities in each state.
       In 2013, 46 out of the 50 states have biodiesel plants in operation (with no commercial-
scale biodiesel production in Colorado, Montana, Vermont and Wyoming). Texas has the highest
biodiesel production capacity in the U.S. (577 MMgpy), followed by Iowa, Missouri, Illinois and
Ohio. The average capacity is approximately 64 MMgpy per state, and detailed production
capacities in each state are summarized in the appendix (Appendix Table 8.1). The biodiesel
       34 The NBB maintains a list of biodiesel producers on the Internet at the following URL:
http://www.biodiesel.org/production/plants, accessed June 2014.

                                           171

-------
production capacities in the 46 biodiesel producing states varied from 0.25 MMgpy in Alaska to
577.25 MMgpy in Texas. Assuming that the purification and process water consumption rate Nj
(0.31 gal/gal) is uniformly applied to all biodiesel plants, the resultant Wj/ ranges from 0.08 to
178.47 MMgpy (see sample calculations in the Appendix to this chapter). As dry wash
technologies are increasingly practiced among the biodiesel industry, the water consumption for
this stage of biodiesel processing is expected to decrease with time.

   8.2.3.4  The total annual water consumption by states (Wtot, j, Ntot, j)

       The total quantity (Wtot) of consumptive water as the sum of water consumption from the
three stages is summarized for each state as the sum of water consumption in irrigation (Wy),
soybean-to-soybean oil processing (W2) and biodiesel production (based on capacity, Wj). The
fractions of water use at each stage are also estimated to better understand the relative
contribution. Figure 8.2 illustrates the total consumptive water (Wtotj) for the soybean-to-
biodiesel process in each  state. 49 states are included in this figure, with the exception of
Wyoming which had neither soybean growth/processing or biodiesel plants in 2007/08. Wtotj
ranges from 0.004 MMgpy to 16,016.11 MMgpy in these 49 states. Figure 8.3 shows the
normalized total water consumption for 49 biodiesel producing states. The range of Nto/j varies
from 0.17 gal/gal to 1058.68 gal/gal, with a national average of 62.26 gal/gal. On average,
irrigation represents 99.23% of the total water consumption, 0.27% for soybean crushing/refining
and 0.50% for biodiesel manufacturing. However, the fractions vary significantly among the
states.
       Water consumption for the ten states with the highest soybean harvest is listed in Table
8.1. These represent 83.31% of soybean harvest in the U.S. in 2008 (USDA, 2008). Most of
these major soybean-growing states are located in the Midwest region with the  exception of
Arkansas. The irrigation water intensities (Ny/) of these states are below national average with
the exception of Arkansas, Nebraska and Missouri. This again supports the fact that not all
soybeans in the U.S. are irrigated, and warrants further state level water consumption analysis.
       Table 8.2 lists water consumption for the ten states with the highest biodiesel capacities.
These ten states account for 66.6% of biodiesel production capacities in 2013 (Biodiesel
Magazine,  2013).  The states of Arkansas (#1 in irrigation water use), Mississippi (#3),  Missouri
(#5), Indiana (#8), Illinois (#9) and Iowa (#14) are both the highest in irrigation water use  (not
necessarily entirely soybean production) and biodiesel production. This may be an indication that
the soybeans produced have been consumed in close proximity, as biodiesel plants usually seek
nearby feedstock to reduce the cost of transport and storage. In contrast, water use from biodiesel
production accounts for a much larger fraction in the states of Washington and  Pennsylvania
when compared with soybean irrigation, as soybean growth in these states is relatively  low.
       It is noteworthy that in most of the states, Wij and Wj/ dominate the total water
consumption for biodiesel production except for Ohio, Illinois and Iowa where  W^/
consumptions account for 28.04%, 19.02% and 40.53%  of water use in their biodiesel processes,
respectively. This is due to high soybean harvest (5th, 2nd and 1st in the U.S., respectively), and
therefore a high percentage of water use in soybean oil processing.
                                           172

-------
Table 8.1 Total annual water consumption (Wtotj) in top 10 soybean harvesting states*
State
Iowa
Illinois
Minnesota
Indiana
Ohio
Nebraska
Missouri
South Dakota
North Dakota
Arkansas
Wtot,j(MMgpy)
400.577
484.8529
622.8166
488.7905
123.2441
9107.751
2556.946
289.2936
139.1979
16016.11
Wij/W»ot,j(%)
48.46
69.91
85.84
81.09
26.36
99.44
96.09
87.50
60.49
99.61
W3j/Wtotj(%)
23.50
11.07
3.28
7.65
33.11
0.02
2.22
0.75
19.55
0.23
  Note: * - Ranked by harvest.
Table 8.2 Total annual water consumption (Wtotj) in top 10 biodiesel producing states*
State
Texas
Iowa
Missouri
Illinois
Ohio
Indiana
Arkansas
Mississippi
Washington
Pennsylvania
Wtoy(MMgpy)
335.77
400.58
2,556.95
484.85
123.24
488.79
16,016.1
3,764.65
41.98
34.85
w,ywtoy(%)
45.58
48.46
96.09
69.91
26.36
81.09
99.61
98.68
16.75
0
WyWtoy(%)
53.16
23.50
2.22
11.07
33.11
7.65
0.23
0.95
83.23
98.70
  Note: * - Ranked by plant capacities
                                      173

-------
              Alaska
                                                                                     Total water consumption
                                                                                     (Million gallons per year)
                                 Hawaii
81-650


1153-16020

Not applicable
Figure 8.2 Total annual water consumption of the soybean-to-biodiesel process (Wtotj).
Consumption includes irrigation (USB, 2010; USDA, 2008), soybean processing into oil (USB, 2010) and biodieselproduction
from oil (Tuetal, 2014).

-------
^^-'-^*^ Alaska

Hawaii
                                                                                           Normalized Total Water Consumption
                                                                                           (Gallons H2O/Gallon Biodiesel)
                                                                                           NM
                                                                                              JO-20
                                                                                           H 21-79
                                                                                              [] 80-150
                                                                                           IB 150-300
                                                                                           m 300-1058
                                                                                              ^ Not apptecable
Figure 8.3 Total water consumption of the soybean-to-biodiesel process on per gallon biodiesel basis (Ntotj). Data from
       USB (2010), USD A (2008), and Tu et al. (2014).

-------
8.2.4  Discussion

   8.2.4.1  Regional impact analysis

       In this analysis, regional water use for biodiesel processing was also determined by
grouping the states into nine census regions (Doddler et al., 2011). Regional data averages out
differences among states as found in our study. Table 8.1 indicates that water use for the
biodiesel process mainly impacts the water resources in the Midwestern and Southern regions of
the U.S. Table 8.2 indicates that nine out of the ten top soybean growing states are in the East
North Central and West North Central regions, except for Arkansas in the West South Central
region. The states of South Dakota and North Dakota have high soybean harvests but are below
the national average in irrigation water use and total water use.
       In Table 8.3, water consumption of the biodiesel process for each region is  evaluated by
the fractional percentage of water consumed for irrigation (Pi) and biodiesel manufacturing (Ps)
for each region, and also the regional irrigation intensity (^tot-region).
     Table 8.3 Regional water consumption data

New England
Middle Atlantic
East North Central
West North Central
South Atlantic
East South Central
West South Central
Mountain
Pacific
Wtot(MMgpy)
4.22
60.95
1617.51
15654.20
981.66
3892.92
17114.78
42.16
71.78
Pi (%)
0.00%
23.73%
76.11%
96.43%
90.32%
97.35%
98.44%
51.50%
9.80%
Ps(%)
99.25%
73.10%
9.73%
1.30%
7.74%
1.94%
1.36%
48.40%
90.19%
Ntot-reg/on (gal/gal)
0.48
8.65
6.34
46.7
51.9
150
536
552
1059
       The West South Central (TX, OK, AR and LA) region accounts for 43.35% of the total
water consumption for biodiesel production and 5.13% of soybean harvest (the 3rd highest in the
U.S.). State-level data indicates that the high water consumption was primarily caused by
relatively high irrigation water intensities in AR, LA and OK, and large biodiesel production
capacity in Texas. The West North Central region had the second largest water consumption,
representing 39.65% of total water consumption, but with low total water intensity. The region
represents 53.25% of soybean growth in the U.S.  and 22.5% biodiesel capacity. New England
(six states) accounts for approximately 0.01% of the irrigation water consumption in the U.S.,
and this is predominantly from biodiesel manufacturing. Similar trends can be observed in the
Pacific and Middle Atlantic regions. The highest regional water intensity in the Pacific region is
due to the irrigation water intensity in Washington. Together, from Figures 8.3 and 8.4, it was
observed that the water use varies vastly within the West North Central. Agricultural water use in
Nebraska and Kansas is much larger than the national average, while such water use in other
                                           176

-------
states is less than the national average. This regional heterogeneity is also evident for the West
South Central, East South Central and South Atlantic Regions.
       Given the large variations among states within the same region, from a planning and
decision making standpoint, state level soybean water use data can be considerably more
accurate than regional data.

   8.2.4.2  Water-stressed areas

       For areas where the water supply is potentially constrained, the impact of biodiesel
production on water resources should also be analyzed with respect to future climate adaptation
considerations. A few studies have identified the water-stressed areas based on different criteria,
as summarized in Appendix Table 8.5. Accordingly, the following states are identified as water-
stressed in this study: Arizona, California, Colorado, Florida, Georgia (especially  southern
Georgia), New Mexico, Nevada and Texas (WRI, 2014). Accordingly, a summary of total annual
water consumption (Wtoy) is found in Table 8.4.
     Table 8.4  Total annual water consumption (Wtotj) for soybean biodiesel production in
                the states in the water-stressed areas
State
Arizona
California
Colorado
Florida
Georgia*
New Mexico
Nevada
Texas*
Wtoy(MMgpy)
14.84
24.26
21.75
12.13
225.24
0.46
0.31
335.77
W,yWtoy(%)
0
0
99.82
0
90.36
0
0
46.58
WyWtoy(%)
100
100
0
99.46
8.72
100
100
53.16
       Note:  * Southern Georgia (Yang et al., 2011) and two-thirds of Texas (WRI, 2014) have been
             reported as water stressed areas

       These states represent 1.6% of total water use, 0.46% of soybean harvest and 27.61% of
biodiesel production capacity in the U.S. The total water consumed for all stages of biodiesel
processing in these states is much lower than the national average of 1,152.05 MMgpy. For the
States of California, Arizona, Florida, New Mexico and Nevada, more than 99% of the water
consumption is in biodiesel production from oil due to very limited soybean growth in these
areas. Colorado and Georgia only account for 0.32% of the total soybean harvested and 0.63% of
total water consumption in the U.S., however, the irrigation intensity in Colorado is 611.52
gal/gal, the 3rd highest in the U.S., while Georgia ranks 10th with respect to irrigation water
intensity at 106.8 gal/gal. Texas accounts for 0.13% of total soybean growth and 19.71% of
biodiesel production capacity in the U.S., while its irrigation water intensity of 190.58 gal/gal is
the 6th highest.
                                           177

-------
       For companies located in water-stressed areas, such as Company 2 (Appendix Table 8.3)
in Texas and Company 4 in California, the adoption of water-saving technologies may be more
critical. If biodiesel production is to expand in these areas, water supply issues should be
considered within the decision making process.

  8.2.4.3  Comparison with existing studies

       Appendix Table 8.6 provides a detailed comparison of analysis contained within this
report compared with other similar studies by evaluating different parameters and assumptions
used.  In the soybean growth stage, all studies except King and Webber (2008) assumed a
complete consumption of irrigation water (i.e., none of the applied irrigation water was recycled
or reused).  Different irrigation ratios were used within these studies. The analysis within this
report is consistent with O'Connor (2010) in irrigation water intensity, and its water intensity in
biodiesel manufacturing (Nj) is within the same range as King and Webber (2008) and Harto et
al., (2010). All of these studies, and this analysis, found  that the water intensities within soybean
oil processing and biodiesel manufacturing were much smaller than agricultural irrigation water
intensity.
       O'Connor (2010) used an overall irrigation ratio of 8.2% by dividing irrigated acres of
soybean over total harvested acres. King and Webber (2008) separated the calculation by either
non-irrigation or 100% irrigation scenarios. Harto et a/.  (2010) averaged the irrigation ratios
from low, middle and high cost soybean farms, resulting in a national  average of 4%. Mulder et
al (2010) did not specify an irrigation ratio. Only King and Webber (2008) and the analysis
within this  report quantified water intensity during the soybean processing stage (N^). For the
biodiesel manufacturing stage (Nj), both King and Webber (2008) and Mulder et al. (2010) cited
data from Sheehan et al. (1998), and their values were 0.158 gal/gal and 3.63 gal/gal
respectively. Harto et al. (2010) used a value of 1 gal/gal from a 2006  U.S. DOE report (U.S.
DOE, 2006). The Nj used in this study is based on more current data from the biodiesel industry.
For the actual water use in "million gallons per year (MMgpy)," the values obtained in this study
are expected to be much lower than others since  only 17% of the soybean oil was processed into
biodiesel. Substantial data variation exists among individual states, which is an indication that
state level data is likely to be more accurate than both the national average and the regional data.

  8.2.4.4  Limitations of current study

       Although only parts of Texas (WRI, 2014) and southern Georgia (Yang et al., 2011) are
considered water-stressed areas instead of the entire state in each case, data is only available at
the whole state level, so the estimates in Table 8.3 are for entire states. In estimating W2 and Wj,
uniform allocation factors have been used, instead of using state-specific allocation factors due to
data limitations. For Wj, the 0.31 gal/gal may decrease as dry-wash methods are increasingly
adopted by biodiesel manufacturers.
       Indirect water consumption such as water consumed during fertilizer production and
water use for energy generation was not included in this analysis. Water loss factor was not
accounted for during the irrigation stage (i.e., irrigation water input during the irrigation was
assumed to be 100% consumptive).
       In real-life situations, some biodiesel producers import soybean oil from other states to
meet demand, especially those in the states where soybean growth is minimal (e.g., CA and

                                           178

-------
WA). The introduction of imported soybean oils may change the NT/ of the biodiesel produced in
a specific state, since usually there is a significant difference in the irrigation application between
states. Considering the fact that major soybean producing states such as IA and OH have more
than enough soybean oil available to meet the demand for in-state biodiesel production, it is
likely that the interstate export of soybean oil from these states may help to reduce the NT/ of
soybean biodiesel produced in other states. To quantify this phenomenon, robust data is needed
for the soybean oil trade across state boundaries for biodiesel production. Unfortunately, such
data is not readily available.

8.3  New Trends in Biodiesel Research and Development

8.3.1  Feedstock development
       Concerns have been raised about the first generation of biofuels, namely corn ethanol and
soybean biodiesel, regarding the diversion of agricultural resources away from food production
(Canali and Aragrande, 2010; Gasman and Liska, 2007; Tilman et al., 2009). Since feedstock
cost constitutes the highest fraction of the biodiesel production cost, newer and lower cost
feedstocks may provide attractive alternatives to the biodiesel industry while meeting national
policy goals for reducing dependency on fossil fuels. With more R&D on new feedstocks and/or
commercialization, the water use of these new feedstocks should also be studied in the future.
Based on various sources, the following feedstocks are discussed in this section: distiller grains
from corn ethanol, renewable diesel, algae and waste feedstocks.

   8.3.1.1 Distillers dried grains with solubles (DDGS) from ethanol plants

       Distillers dried grains with solubles (DDGS) is a by-product from the ethanol
manufacturing process. Currently, the primary outlet for DDGS is livestock feed (Klopfenstein et
al., 2007; Lumpkins et al., 2004; Shurson, 2002). However, the expansion of the bio-ethanol
industry has generated more than enough DDGSs, which is essential to make good use of in
order to keep the production cost low. Extraction of residual oil from DDGS for biodiesel
production has been proposed (Liu and Rosentrater, 2011). The potential feedstock supply from
recovered oil from ethanol plants is estimated to reach as high as 680 million gallons in 2022
(U.S. EPA, 2010). On average, residual oil takes up approximately 10 wt% of DDGS. The FFA
level in the extracted oil varies from case to case, with typical values of approximately 13 to 15%
(Haas, 2011). One potential concern is high wax concentration in the resulting biodiesel from the
recovered oil, since certain components of wax, such as steryl glucosides, are likely to precipitate
out when biodiesel is cooled, causing filter clogging. One way to solve this problem is to apply a
"winterization" processing step to fractionalize the wax components and remove them from the
biodiesel.

   8.3.1.2  Renewable diesel

       "Renewable diesel," also called "second generation biodiesel," refers to hydrocarbon
based diesel fuel refined from renewable sources, such as plant oil, fats or biomass instead of
from petroleum. Renewable diesel can come from three processes: hydro-treatment of biomass,
hydro-thermo processing of oils (vegetable oil and animal fat) and via the Fisher-Troche (F-T)
process. The key process is decarboxylation as opposed to the transesterification that is used to
produce FAME biodiesel (Knothe, 2010). Under the revised Renewable Fuel Standard (RFS2)

                                           179

-------
program, renewable diesel is credited at 1.7 RINs per gallon compared with 1.5 RINs per gallon
for biodiesel (primarily FAME). Industrial scale development investments have been made to
produce renewable diesel fuel. Neste Oil, a Finland-based oil refining company, has a production
capacity of up to 600 million gallons per year (MMgpy). In the U.S., Dynamic Fuels possesses a
renewable diesel plant with a production capacity of 75 MMgpy. Renewable diesel production
facilities under construction in the U.S. include a  137 MMgpy plant by Diamond Green Diesel
and an 85 MMgpy plant by Emerald Biofuels.
       At first, renewable light-distillate fuel oils such as renewable diesel fuel and renewable
jet fuel are of particular interest to the aviation industry and military  since in these two sectors,
vulnerability to the filter clogging and other cloud-point-temperature issues of biodiesel under
cold climate conditions limits its application. Renewable diesel and jet fuels may be particularly
suited for these applications since they can be processed into short chain paraffins that will not
cause the crystallization of fuel under cold weather conditions. But renewable diesel can have
wider usage, as it can essentially replace petro-diesel.
       Neste Oil has been in collaboration with Lufthansa for test flights of their renewable
aviation fuel (Reuters, 2012). In the U.S., UOP demonstrated the performance of their renewable
jet fuel in five aircraft from Gulfstream Aerospace as part of a trip from Gulfstream's
headquarters in Savannah, GA to Orlando, FL, where the National Business Aviation
Association (NB AA) convention was being held (PR Newswire). During an air show in
Maryland, two U.S. Air Force F-16 aircrafts also used UOP renewable jet fuel (UOP, 2011).  The
U.S. Navy has initiated several tests of algae-based renewable diesel in their vessels. Based on
initial success in small vessels, the Navy continued to test biodiesel with one of its destroyers.
The test is part of the program named the Navy's  Green Fleet Initiative, which aims to form the
capability of deploying a Navy fleet that is powered entirely by alternative fuels (Casey, 2011).
       It is still not possible to draw any definitive conclusions regarding water consumption
associated with renewable diesel production based on the current literature because only very
limited literature exists. As renewable diesel generally uses the same oil feedstocks as biodiesel,
the differences in water consumption reside within the renewable diesel manufacturing
processes. An initial life-cycle assessment published by Argonne National Lab (ANL) reported
that producing 84.19 Ib renewable diesel (out of 100 Ib feed oil) by UOP process generated 6.11
Ib wastewater and required cooling water input (partially consumed) of 1,356 Ib (ANL, 2008).

   8.3.1.3   Algae

       Research on algal biofuel began in the 1970s and has been a subject of considerable
research in recent years (NRC, 2012; U.S. EPA, 2010). The lipid content of most algae species
are higher than that of conventional oil crops (Mata et a/., 2010), which makes algae a promising
feedstock to produce significant amounts of biofuel  in the U.S. (Chisti, 2007). As compared with
conventional oil feedstocks, the cultivation of microalgae requires less land use in general, and
the use of algal oil is not competing with human consumption of edible oil (Singh et a/., 2011).
However, there are limited amount of algae biofuel companies that are able to grow algae and
produce biofuel in a commercial scale. It was estimated that 100 million gallons of algal
biodiesel would be produced in the U.S. in 2022 (U.S. EPA, 2010).
       U.S. DOE (2010) laid out an algal  biofuel technology roadmap, which listed areas of
research and development needed for future production of algal biofuels. In addition to the


                                           180

-------
technical and economic hurdles, water consumption is another significant obstacle in front of the
commercialization of algal biodiesel, considering the nature of algae growth.

   8.3.1.4  Water consumption in algal biodiesel production

       As part of the analysis of biodiesel production for this report, we collected feedback from
expert biodiesel industry sources (National Renewable Energy Laboratory, USD A, NBB and
biodiesel producers) regarding future trends of the biodiesel industry. Algae were recognized as
having potential to be a significant future biodiesel feedstock. A National Academies' report
(NRC, 2012) highlighted some concerns with respect to algal biofuel production, such as water
and nutrient demands, low energy output, leakage, etc.
       As described by DOE's algae technology roadmap (U.S. DOE, 2010) and within the
report by the National Academies (NRC, 2012), many technical issues still need to be solved,
such as improving algal strains for desired characteristics, advancing the materials and methods
for growing and processing algae into fuels, reducing energy requirements for multiple stages of
production, etc.
       Preliminary findings from our analysis of algal biodiesel production are presented in this
section. Similar to the analytical approach used within this report with respect to water use in the
soy biodiesel production process, water use in algae to biodiesel processing was also determined
primarily in the form of a summary of existing literature. The results showed that freshwater
consumption can be a significant concern if algae biodiesel is to undergo commercialization.
Switching to saline and/or wastewater use may be a potential solution.
       Water is used in the following algal biodiesel processes:  open pond cultivation,
harvesting and dewatering, algal oil extraction and biodiesel production via transesterification.
Water consumption during algae cultivation is mainly caused by the need to make up for the
evaporation losses. The evaporation loss in an open pond system is primarily affected by local
topography and by climate conditions. At the algae harvesting stage, dewatering is necessary to
reduce the carry-over water in the algae biomass down to a level at which it can be successfully
processed through oil extraction. In the harvesting and dewatering steps, water may be lost by
evaporation.
       After algae are harvested and dewatered,  extraction (e.g., by hexane) is performed to
separate the lipids from algae cells. The extraction process involves solvent recovery, which
requires make-up water for extraction facilities that include a boiler and cooling tower. The
water consumption during the biodiesel production stage includes crude biodiesel purification
(water washing) and process-related water consumption (e.g., make-up water for the cooling
tower).
       Table 8.5 summarizes the water consumption in these steps from several studies. These
studies were all focused on freshwater consumption. Harto et al. (2010) calculated the
evaporation losses that occurred in open pond systems based in the Southwest U.S. Water
consumption ranged from 32 gallons of water for 1 gallon of biodiesel to 656 gal/gal, depending
on the evaporation rates, with an average of 216 gal/gal. Wigmosta et al. (2011) calculated the
evaporative loss of freshwater in an open pond system using the Modular Aquatic Simulation
System 2-D (MASS2). MASS2 is a two-dimensional, depth-averaged hydrodynamic and
transport model developed at Pacific Northwest National Laboratories (PNNL, 2004). As a
result, the consumptive freshwater use due to evaporation was 1,421 gallons water per gallon

                                           181

-------
biodiesel. Yang et al. (2011) also investigated the freshwater consumption in an open pond from
algae cultivation to oil transesterification. A value of 591 kg water/kg biodiesel (i.e., 520 gal/gal)
was calculated when harvested water was recycled. Pate (2008) estimated a 1,000 gal/gal
evaporative water loss for a cooling photobioreactor based on a 50 MMgpy scale. Guieysse et al.
(2013) studied the variation of freshwater consumption by algae growth in five different climatic
regions and the results showed that the water footprint (WF) varied from 33.1 m3/GJ to 36.7
m3/GJ  (1,093 gal/gal to 1,212  gal/gal).
       The "Renewable Fuel  Standard Program (RFS2) Regulatory Impact Analysis" released
by U.S. EPA (2010) contains a LCA study on algae biodiesel conducted by the National
Renewable Energy Laboratory (NREL). The authors evaluated three scenarios for the algae
growth. The base case assumed the yield of algae to be 25 g/m2/day in an open pond and 63
g/m2/day in a photobioreactor (PER). The lipid concentration was assumed to be 25%. In the
aggressive case, the yield was 40 g/m2/day (op) and 100 g/m2/day (PER), and the lipid
concentration was 50%.  The third scenario, the max case, assumed the yield to be 60  g/m2/day
(open pond) and 150 g/m2/day (PER), with a lipid concentration of 60%. The result of net water
demand showed that water consumption in  an open pond was significantly higher than a PER,
and increasing yield and lipid  concentration was  an effective approach to reduce the total water
consumption for algae cultivation. Overall,  the water consumption of algae biodiesel production
is higher than that of biodiesel derived from soybean (62.26 gal/gal; this study) and waste
cooking oil (0.33 gal/gal; Martin and Grossmann, 2012), corn ethanol (10.1 gal/gal; Wu et al.,
2009) and cellulosic ethanol (1.9-9.8 gal/gal; Wu et al., 2009) reported in the literature. It is also
notable that the water consumption during the biodiesel production stage is significantly lower
than the cultivation, harvest and dewatering stages. This was also observed by Tu (2012) through
a water consumption analysis  for soybean biodiesel production in the U.S.

   8.3.1.5  Water saving methods

       In general, water savings can be achieved through using alternative water resources (e.g.,
saline water, municipal wastewater), improving reactor design and applying innovative
harvesting and extraction technologies.

•  Cultivation
       - Species for saline water
       Utilizing algae strains  that are tolerant to  saline water is expected to divert the algal
biofuel production towards reduced use and consumption of freshwater resources. Current efforts
are focused on isolating and genetically modifying marine and brackish algae species to
accommodate them for lipids production. Examples include a genetic manipulation of Dunaliella
tertiolecta, (Lane, 2012) and certain strains of Nannochloropsis sp. and Chlorella sp., which
possess both high lipid content and a preferred fatty acid profile that are crucial to successful
biodiesel production (Lim et al., 2012). Although growing salinity-tolerant algae for oil
production slows freshwater depletion, freshwater input is still necessary for preventing salt
build-up within ponds (Vasudevan et al., 2012). Murphy and Allen (2011) calculated the saline
and freshwater consumptions for massive cultivation of algae in open ponds. The results showed
a national average water consumption of 5.5 m3/m2 (4.4 m3/m2 saline and 1.1 m3/m2 freshwater)
per year. The utilization  of these saline strains of algae can be promising as it will reduce the use
of freshwater.
                                           182

-------
     Table 8.5  Literature summary of water consumption during different algal biodiesel
               production stages*
Study
Cultivation
Harvest &
Dewatering
Oil Extraction
Open Pond (OP)
Hartoetal. (2010)
Wigmosta et al. (2011)
Yang etal. (2011)
Guieysseetal. (2013)
RFS2RIA(2010)
165
1421
50
NA
514
1093-1212
974a, 383b, 31 7C
Biodiesel
Production
Total

1

6
NA
NA
216
1,421
520
1,093-1,212
974, 383, 317
Photobioreactor (PBR)
Hartoetal. (2010)
Pate (2008)
RFS2RIA(2010)
0
1000
17? 32b, 25C
40
NA
1
1
NA
41
1,001
72, 32, 25
  Note: * - All consumption in units of gallons of water per gallon of biodiesel.
        a Base case;b Aggressive case;c Max case
       - Immobilized microalgae culture
       Suspended algae growth is the most common practice under consideration for algal
biofuel production due to the improved economics of algae cultivation. However, separating the
suspended algae culture from water in the harvesting stage becomes energy and cost intensive
due to the difficulty in concentrating the suspended algae culture. Thus, immobilization of
microalgae has been proposed to ease separation (Hoffmann, 1998). The immobilized
microalgae cultures facilitate algal biomass recovery and hence reduce water losses during the
harvesting process. Mallick (2002) identified six processes for immobilizing algae strains:
covalent coupling, affinity immobilization, adsorption, confinement in liquid-liquid emulsion,
capture behind semipermeable membrane, and entrapment. Ozkan et al. (2012) reviewed existing
studies on immobilized algae production systems. The authors indicated that the lipid
productivity and growth rate of the immobilized growth  systems were not consistent when
compared to suspended algae cultivation systems.
       - Co-location with WWTP
       Using algae  as a treatment for wastewater has been previously investigated (Hoffmann,
1998). More recently, numerous studies have been published on using wastewater as growth
media for algae to avoid extensive freshwater consumption during the algae cultivation process.
One advantage of using wastewater is its higher nutrient content. Lundquist et al. (2010)
estimated that using wastewater could save approximately  10% of the operational cost of algal
biofuel production, mainly by reducing fertilizer use. Pittman et al. (2011) listed several factors
that determined the  success the of algae cultivation in wastewater, including operational
                                           183

-------
parameters (e.g. pH, temperature), nutrient level, toxin concentration and biotic factors. In
addition, algae species and starting density were also of key importance. The authors compared
several sources of wastewater and concluded that municipal sewage wastewater and agricultural
wastewater were preferred considering their high nutrient level, low toxin concentration and
wide availability. Industrial wastewater, on the other hand, typically contains lower nutrient
levels and higher levels of toxins and thus would not be as suitable for algae growth. Strum et al.
(2011) investigated algae growth in open ponds with wastewater from a local wastewater
treatment plant. The results showed that algal biomass productivity ranged from 0.78 to 15.9 g
dry weight per m2 per day over the course of the six month experiment. The analysis of the algal
oil profile and its resultant biodiesel indicated that algae produced under such a system could be
used as an alternative to oil seed crops for biodiesel production. The removal of 19% dissolved
nitrogen and 43% of dissolved phosphorus were an additional benefit to the environment realized
by this production method.
       Beal et al. (2012) performed research to assess the energy return on investment (EROI)
for integrating algae biodiesel production with wastewater treatment. The results indicated that
an additional  5.5 kJ of energy could be produced or saved for each liter of processed wastewater
through an integrated algae biodiesel production system, and the EROI was increased to 1.44 in
this case as compared with 0.42 for a separate, stand-alone biodiesel production system. Zhou et
al. (2011) isolated 60 algae-like microorganisms for algal biomass cultivation in concentrated
municipal wastewater (CMW). The CMW had high total suspended solid (TSS) concentration
and turbidity, allowing less light transmission for algae growth. 17 out of 60 strains survived in
the CMW, among which 60% were Chlorella species. The total algal oil concentration (wt%)
after a six-day cultivation period ranged from 17.41% to 33.53%.
       - Reactor design
       Considering the evaporation loss in open pond systems, the use of a photobioreactor
(PER) is an option to reduce the water consumption during algae cultivation. However, a crucial
obstacle associated with PER is difficulty in oxygen removal. Oxygen is the product of
photosynthesis from algae culturing and it starts to prohibit the photosynthesis when its
concentration is beyond air saturation (Carvalho et  al., 2006). Additionally, to overcome the
overheating issue in PER, cooling water spray may be necessary and the water consumption
associated can be huge (Pate, 2008).  Another option of reactor design is to utilize algal biofilms.
Algal biofilm formation initiates when algae cells attach to the growth surface. After initial
attachment, algae starts to secrete extracellular polymeric substances (EPS) to strengthen the
attachment. Finally,  the biofilm develops into a mature structure (Qureshi et al., 2005). Algal
biofilm is expected to provide a similar advantage of lower water loss as with immobilized algae
cultures since harvesting can be realized by scrubbing algae biomass from the growth surface.
Ozkan et al. (2012) introduced the application of a biofilm photoreactor for algae cultivation to
lower water and energy consumptions. The photoreactor was composed of a biofilm growth
surface, a nutrient medium recirculation system and an illumination system. The nutrient
medium was  delivered to the algae growth surface by dripping nozzles and at the end of the
surface the nutrient overflow would be collected and re-circulated to the dripping nozzles.
Harvesting was realized by mechanically scraping the thickened algae biomass with a squeegee.
The process provided an algae mass production rate of 0.71 g per m2 per day, and the total lipid
content was 26.8% by dry weight. More importantly, the reduction in water required was
approximately 45% in volume.
                                           184

-------
•  Harvesting and lipid extraction
       Unlike the conventional biodiesel process, feasible technology is still under investigation
for harvesting algae from water and extracting the algae oil for biodiesel production. Many
companies are developing various technologies, and details can be found in the Appendix to this
chapter.
       The following studies illustrate the potential to achieve a "water-efficient" harvesting or
extraction process. Wet lipid extraction methods eliminate algae harvesting and dewatering
steps, and thereby reduce energy consumption and water loss. The development of wet extraction
technology is deemed as a critical step for the algal biofuel industry to move towards
commercialization (Sills etal., 2013). Sathish and Sims (2012) employed in-situ hydrolysis of
lipids in algae, centrifuge separation and recovery of lipids by hexane for biodiesel production.
The process was able to extract 79% of the total transesterifiable lipids from a mixed-culture
algae with 84% moisture. Teixeira (2012) used a combination of dissolution and hydrolysis of
cell walls for algae extraction. It was found that at 100 to 140 °C,  most of the algal cell walls
were completely dissolved and lipid extraction  could be achieved  within 50 minutes. Cerff et al.
(2012) performed a screening study on magnetic separation of both fresh and marine algae
species. The results showed that over 95% separation  efficiencies  could be achieved, and
complete algae removal could be attained in approximately five minutes in a high gradient
magnetic filter system under continuous flow conditions. Levine et al. (2010) developed a two-
step, catalyst-free process that extracted algal oil and converted it  into biodiesel. Subcritical
water was applied to hydrolyze the intracellular lipids and to agglomerate the cells into solids
that retained fatty-acid rich lipids. In the second step, the solids underwent in-situ conversion by
ethanol under supercritical conditions.  Patil et al. (2011) studied the in-situ conversion of lipids
in wet algae biomass (with 90% water) into biodiesel in a one-step process. The lipids were
transesterified with supercritical methanol under 225°C  for 25 minutes with an optimum ratio for
wet algae and methanol (wt./vol.) of 1:9.

•  Resource demand
       As discussed previously, water will be a vital resource for  algal biodiesel production.  Site
selection will be crucial to the success  of an algal biodiesel plant and careful consideration must
be given to the availability of water resources (especially freshwater) and access to sunlight
radiation and carbon sources. Yang et al. (2011) discussed the impact of site selection on water
consumption for algae growth. Considering the tradeoff between solar irradiation and
evaporative water loss, their spatial analysis revealed that the water footprint of algae-derived
biodiesel decreased from north to south with  a roughly-defined boundary from the northern
border of California to the northern border of New York. On the northern side of this boundary,
the water footprint was estimated to be approximately 1,500 kg water/kg biodiesel, while in the
southern side the value became progressively smaller. On the other hand, the water footprint also
demonstrated a trend of decreasing from west to east.  Similarly, Wigmosta et al. (2011) pointed
out several preferred locations for microalgae growth  in terms of reducing water loss that
included the Gulf Coast region, most of the eastern seaboard and areas adjacent to the Great
Lakes. It was also noted that the optimization of site selection could lead to a reduction of water
consumption as large as 75%. Subhadra (2011) analyzed current water usage associated with
biofuel production in Southwestern U.S. The authors confirmed that this region was potentially
an optimum area for algal biodiesel production considering the availability of various resources.
But the authors also indicated that sustainable water management  policies would be imperative

                                           185

-------
to protect regional water resources and hydrologic patterns while developing algal biodiesel in
this region.

•  Data gaps
       - Data gaps with respect to studying the impact of algal biodiesel on freshwater
consumption
       As proposed by the NAP report (NRC, 2012), a more accurate model to estimate
evaporation losses from algal biodiesel production processes is needed. Guieysse et al. (2013)
compared nine evaporation models from existing literature and discovered significant uncertainty
regarding evaporation predictions. Currently, some studies use pan evaporation (Harto et al.
2010; Yang et al. 2011), which is not a very good surrogate for actual open-pond evaporation, to
approximate the evaporation behavior of algae raceways. Also, modeling of the water balance,
salt buildup and its management is of particular importance for future research. In addition, when
wastewater is used to minimize freshwater consumption, knowledge of water quality and its
influence on algae growth and an understanding of the resultant processing steps that may be
necessary (e.g., bioconcentration of heavy metals in algae may affect the further processing of
algae biomass) is another area possibly needing additional research. Information about water
resource distribution (both survey and ground water) needs updating so that analyses of
sustainable water withdraws can be performed with higher precision.
       There is potential for algae cultivation to have  negative influences on water quality from
the release of algae culture broth into surface water bodies and/or ground water aquifers. The
release could occur as a result of overflow caused by extreme weather events, improper
operation or noncompliant discharges,  or due to defects in the design, construction and
maintenance of open ponds. Considering the nutrients in the algae growth broth, the release may
lead to eutrophication in surface water bodies. In addition, many  algae species can
bioconcentrate heavy metals if flue gases are used as the CO2 source for algae growth (O'Dowd
et al., 2006). If wastewater is used, waterborne toxicants contained in the released algae culture
broth may pose a risk to both surface and ground water sources, and threaten drinking water
quality.
       In addition to the algae growth  stage, the harvesting and extraction stage may also result
in negative impacts on water quality through incidental release or spillage of the culture broth.
Wastewater generated during the algal  biodiesel production stage, on the other hand, is expected
to have a similar impact on water quality as transesterification of other lipid feedstocks.
       - Data gaps regarding the impact of algal biodiesel on the water quality of surface and
ground water resources:
       To assess the impact of algae biodiesel production on the  quality of surface and ground
water resources, an initial step would be to develop a representative database of factors such as
the typical nutrient load and the quality of the non-freshwater used for algae growth.
Additionally, background information about the region of interest would be necessary for impact
analysis. Examples include the current loading and concentrations of nitrogen  and phosphorus-
containing compounds, herbicide, heavy metals and salinity in surface and ground water sources.
                                           186

-------
   8.3.1.6  Drought-tolerant plants

       New and improved crop species, such as the ones that can use less water, marginal lands,
etc., may also help reduce the water intensity for biodiesel production. In order to reduce water
consumption, several drought-resistant feedstocks have been proposed, including jatropha,
camelina, and pennycress.
       Jatropha is a genus of a long history and belongs to the Euphorbiaceae family. Jatropha is
known for its high adaptability to a wide range of soil quality and irrigation/precipitation
conditions (Makkar and Becker, 2009). It has been reported  by several studies to be more
capable of withstanding drought stress than other lipid crops (Niu et al., 2012; Rao et al., 2012;
Ye et al., 2009). According to Gardner et al. (2012), the oil productivity of jatropha is
approximately 194 gallons per acre, more than four times as much as that of soybeans. One
downside of jatropha oil is that it usually contains high FFA levels, which require acid
esterification as a pretreatment (Tiwari et al., 2007). Several pilot studies have been performed to
investigate  the feasibility of jatropha cultivation and the outcomes were not as promising as
expected. It was found that if jatropha is cultivated in the marginal areas, the yield is
compromised (Louma, 2009).  However, growth of jatropha is not likely to be feasible in regions
above the 30-degree latitude in North America (U.S. EPA, 2010).
       Camelina belongs to the Brassicaceae family. Similar to jatropha, camelina can survive
in drought and semiarid conditions, and requires less fertilizer and pesticide application (Moser,
2010).  Camelina has an oil productivity of approximately 60 gallons per acre, slightly higher
than that of soybeans (Gardener et al., 2012). In addition to the oil, camelina meal can be used as
animal feed (Ehrensing and Guy, 2008). Unlike jatropha oil, camelina oil generally contains a
very low level  of FFA (Sanford et al., 2009).
       Pennycress is also a member of the Brassicaceae family.  Pennycress is able to tolerate
low temperatures, minimal fertilizer inputs and limited water supply. Moreover, its low
temperature tolerance makes it possible to grow pennycress  during winter when arable lands are
typically fallow. Analysis shows that pennycress contains around 30 wt% oil content and the
field yield is expected to be approximately 100 gallons per acre (USDA-ARS, 2006; Voegele,
2010).
       Other waste oil containing materials can also become biodiesel feedstocks, especially at
the community scale, such as the fats, oils and greases (FOG) from food services and the sewer
system. An EPA study indicated that FOG in the sewer system is the number one cause of sewer
overflows and  it is mainly landfilled (U.S. EPA, 2007). The  advantages include low costs and
cost savings from landfill tipping  fees. The difficulties usually lie in the quantity, waste
collection, and, most challenging of all, oil extraction from these feedstocks.

8.3.2   Biodiesel and bio-based dieselfuel technology development
       Along with the rapid expansion of the biodiesel industry in the U.S., there has also been
active research and development in search of new feedstocks, technology innovation and further
process improvement.
                                           187

-------
   8.3.2.1  Biodiesel dry wash

       As a purification process step, biodiesel water washing has several advantages. It is easy
to operate, has a relatively low cost, and it is effective in removing glycerin and alkaline
catalysts. Its disadvantages include emulsion formation, the need for post-washing heating (to
remove moisture content), a long process time, water consumption and wastewater generation.
Emulsion not only prolongs the purification time but also leads to the loss of biodiesel yield.
More importantly, wet washing consumes considerable amounts of water and generates
wastewater that usually requires pretreatment or discharge permitting. With increased concerns
regarding biodiesel sustainability and regional water stress, dry wash technologies have been
increasingly adopted by biodiesel producers, especially for newly constructed biodiesel
processing plants. Major dry wash technologies include using adsorbents and ion-exchange
resins. A typical adsorbent used by the biodiesel industry is a synthetic magnesium silicate. This
adsorbent is able to adsorb polar compounds such as methanol, glycerin, and FFAs (Bryan,
2005). Use of ion-exchange resins, such as Purolite® PD 206, the Na+ and K+ from the sodium
hydroxide and/or potassium hydroxide catalysts, is exchanged  with H+ from a reactive group on
the surface of the resin beads to break down soaps (Biodiesel TechNotes, 2010).
       The advantages of dry washing processes include reduced purification time, waterless
processing, and increased final biodiesel recovery (Sims, 2011). The disadvantages of dry
washing methods include sorbent cost, increased metal concentrations in biodiesel (such as Na,
K and even Ca) and waste disposal issues as most of the sorbents are not regenerated.
       For adsorbents such as magnesium silicates, one concern is the increased purification cost
since it cannot be regenerated. Also, it has been found that some fine silicate particles can remain
in the purified biodiesel, potentially causing engine deposit or wear issues. For ion-exchange
resins, the elevated acid value after purification is an inherent problem, so the FFA level should
be relatively low before ion-exchange resin purification can be applied (Smith, 2012).
       Some biodiesel producers are developing sorbents (proprietary research and
development) from waste materials, such as saw dust, wood chips and waste coffee grounds.
These materials resemble activated carbon and can be burned as boiler fuel (i.e., for steam
generation or evaporative distillation) after use.

   8.3.2.2  Heterogeneous catalysis

       Homogeneous catalytic transesterification is the standard technology in the biodiesel
industry, with alkaline catalysts such as NaOCFb, NaOH and KOH. The advantages are the use
of low-cost catalysts, high reaction rates and high yield reactions.  With increased biodiesel
production and more stringent environmental regulations, the disadvantages are that the catalysts
are non-recyclable, they are hard to separate from biodiesel, and often require neutralization
during biodiesel processing. Therefore, heterogeneous catalysis using solid catalysts is under
active research.
       A variety of materials have been tested as catalysts: (1) alkaline earth metal oxides and
derivatives such as MgO, CaO,  SrO, etc., (2) carbohydrates such as sucrose and cellulose, (3)
alkaline inserted complexes and zeolites, (4) transition metal oxides and derivatives, (5) mixed
metal oxides and derivatives, (6) hydrotalcite metal oxides, (7) cation-exchange resin, (8)
sulfated oxides, and (9) biocatalysts (Singh Chouhan and Sarma, 2011).
                                           188

-------
       Heterogeneous catalysis of FAME eliminates the use of potentially hazardous acid and
alkali materials, which also reduces the efforts for subsequent purification. The heterogeneous
process is usually developed to be more effective with low quality feedstocks that contain larger
amounts of free fatty acids (FFA). For example, the current process of high FFA feedstocks uses
FbSC^esterification and excess methanol. Additional energy and material are required for
excessive methanol recovery and neutralization of FbSO/t. The heterogeneous process is
developed to combine the esterification and transesterification in a single step and thereby has
the potential to reduce methanol use. Additionally, heterogeneous catalysts are reusable.
Currently the downsides of the heterogeneous process includes high energy input (e.g., high
temperature and pressure). Only a few biodiesel producers use heterogeneous catalysis since it
requires significantly different process equipment from most existing biodiesel production
operations. McNeff et al. (2008) developed a heterogeneous catalysis process, called Mcgyan®,
which has been applied by the Ever Cat Fuels™, LLC biodiesel manufacturer (3 M-gallons per
year capacity). To date, very few applications have been found in the biodiesel industry, and
some resistance in adopting has been observed.

8.4   Summary

       The current biodiesel industry still relies heavily on oil crops like soybeans.  The water
consumption associated with biodiesel production from  oil crops mainly stems from irrigation,
with minor consumptions from oil processing and biodiesel manufacturing stages. The results
suggest that on average irrigation accounts for 61.78 gallons (gal) of water for a gallon of
soybean biodiesel, while soybean processing (0.17 gal/gal) and biodiesel production (0.31
gal/gal) stages consume much less. The total  water consumption intensity for the  entire biodiesel
production process was found in this analysis to be 62.26 gal/gal, which is considerably lower
than reported in previous literature. However, water consumption from the three stages varies
significantly from state to state, which warrants state-level water consumption analyses in order
to improve decision making with respect to water resource management.
       The need to develop lower cost feedstocks will remain critical for the biodiesel industry.
The need to understand their impact on water resources is also critical. Distiller grains are
expected to become a major biodiesel feedstock (U.S. EPA, 2010). Renewable diesel has the
potential to become a significantly important technology since it is a direct replacement for
diesel and jet fuel petroleum distillate fuels. If large-scale production of renewable diesel is to
happen in the future,  its impact on water  resources should also be and further analysis should be
conducted to characterize water used in renewable diesel fuel manufacturing and  its upstream
feedstock production.
       Production of biodiesel from algae has the potential to reduce competition for agricultural
resources between biofuel production and food crop production, and to reduce the cost of lipids
used for transesterification to FAME. Important advances have been made in developing more
sustainable algal lipid production processes through the use of wastewater and brackish water for
algal growth. With technology improvements, low quality lipid feedstocks, especially feedstocks
from waste such as the trap grease from restaurants, sewer pipeline, and other lipid containing
wastes, will be increasingly used for biodiesel production. These serve the dual benefits of waste
reduction and renewable fuel  production. Limitations of these feedstocks include their limited
quantity (i.e., they will contribute to only a small  fraction of the biodiesel supply) and that they
will require pretreatment to extract the oil fractions useable for FAME production. A better

                                           189

-------
understanding of the life cycle water needs for waste feedstocks will be necessary. Similarly, a
comprehensive evaluation of the sustainability matrix of any crop-based feedstocks would also
be necessary before large scale use of any of the new crop-based feedstocks for FAME biodiesel
or renewable diesel fuel production.

8.5  References

Argonne National Laboratory (ANL). 2008. "Life-Cycle Assessment of Energy and Greenhouse
       Gas Effects of Soybean-Derived Biodiesel and Renewable Fuels." Report Number
       ANL/ESD/08-2. http://www.transportation.anl.gov/pdfs/AF/467.pdf. Accessed June
       2014.

Beal, C.M., A.S. Stillwell, C.W. King, S.M. Cohen, H. Berberoglu, R.P. Bhattarai, R.L.
       Connelly, M.E. Webber, and R.E. Hebner. 2012. "Energy return on investment for algal
       biofuel production coupled with wastewater treatment."  Water Environment Research,
       84(9):692-710.

Biodiesel Magazine. "Biodiesel Plant List."  http://www.biodieselmagazine.com/plant-list.jsp.
       Accessed March 2013.

Bryan, T. 2005. "Adsorbing it all." Biodiesel Magazine, March 2005.
       http://www.biodieselmagazine.com/articles/239/adsorbing-it-all. Accessed August 2014.

Canali, M., and M. Aragrande. 2010. "Biofuels versus Food Competition for Agricultural
       Resources: Impacts on the EU Farming Systems." In The Economic Impact of Public
       Support to Agriculture. Springer, New York. pp. 191-209.

Carvalho, A.P., L.A. Meireles, and F.X. Malcata. 2006. "Microalgal reactors: A review of
       enclosed system designs and performances." Biotechnology Progress, 22:1490-1506.

Gasman, G.K., and J.A. Liska. 2007. "Food and fuel for all: realistic or foolish?" Biofuels,
       Bioproducts and Biorefining, 1:18-23.

Casey, T. 2011. "U.S. Navy conducts its largest algae biofuel test ever."
       http://cleantechnica.com/2011/ll/29/u-s-navy-conducts-its-largest-algae-biofuel-test-
       ever/. Accessed December 2012.

Cerff, M., M. Morweiser, R. Dillschneider, A. Michel, K. Menzel, and C. Posten. 2012.
       "Harvesting freshwater and marine algae by magnetic separation: Screening of separation
       parameters and high gradient magnetic filtration." Bioresource Technology, 118:289-
       295.

Chisti, Y. 2007. "Biodiesel from microalgae." Biotechnology Advances, 25(3):294-306.

Dodder, R., T. Felgenhauer, W. Yelverton, and C. King. 2011. "Water and greenhouse gas
       tradeoffs associated with a transition to a low carbon transportation system." In ASME
       2011 International Mechanical Engineering Congress and Exposition. American Society
       of Mechanical Engineers. 531-547.
                                          190

-------
Ehrensing, D.T., and S.O. Guy. 2008. http://extension.oregonstate.edu/catalog/pdf/etn/etn8953-
       e.pdf. Accessed August 2012

EIA. 2014. "Monthly Energy Review. U.S. Energy Information Administration."  Office of
       Energy Statistics, U.S. Department of Energy, Washington, DC. DOE/EIA-0035
       (2014/07).

Gardner, J.C., N.W.T.  Quinn, J. Van Gerpen, and J. Simonpietri. Chapter 8: Oil seed and algal
       oils as biofuelfeedstocks.
       http://www.swcs.org/documents/resources/Chapter  8 Gardner  Oilseed B191CC8A1C
       9DB.pdf. Accessed August 2012.

Guieysse, B., Q. Bechet, and A. Shilton. 2013. "Variability and uncertainty in water demand and
       water footprint assessments of fresh algae cultivation based on case studies from five
       climatic regions." Bioresource Technology, 128:317-323.

Harto, C., R. Meyers, and E. William. 2010.  "Life cycle water use of low-carbon transport
       fuels."  Energy Policy, 38:4933-4944.

Haas, MJ. 2011. "Extraction and use of DDGS lipids for biodiesel production."  In Distillers
       grains: production, properties, and utilization. CRC Press, Baca Raton, FL. pp 487-502.

Haas, M.J., AJ. McAloon, W.C. Yee, and T.A. Foglia. 2005. "A process model to estimate
       biodiesel production costs."  Bioresource Technology, 97:671-678.

Hoffmann, J.P. 1998. "Wastewater treatment with suspended and non-suspended algae."
       Journal of Phycology, 34:757-763.

King, C.W.,  andM.E. Webber. 2008. "Water intensity of transportation."  Environmental
       Science and Technology, 42(21):7866-7872.

Klopfenstein, T.J., G.E. Erickson, and V.R. Bremer. 2007. "Board-invited review: Use of
       distillers byproducts in the beef cattle feeding  industry." Journal of Animal Science,
       86(5):1223-1231.

Knothe, G. 2010. "Biodiesel and renewable diesel: A  comparison." Progress in Energy and
       Combustion Science, 36:364-373.

Lane, J. 2012. "Saltwater algae just fine for biobased fuels, products: New research."
       http://www.biofuelsdigest.com/biobased/2012/ll/27/saltwater-algae-just-ftne-for-
       biobased-fuels-products-new-research/. Accessed September2013.

Levine, R.B., T. Pinnarat, and P.E. Savage. 2010. "Biodiesel production from wet algal biomass
       through in situ lipid hydrolysis and supercritical transesterification." Energy Fuels,
       24:5235-5243.
                                          191

-------
Lim, D.K.Y., S. Garg, M. Timmins, E.S.B. Zhang, S.R. Thomas-Hall, H. Schuhmann, Y. Li, and
       P.M. Schenk. 2012. "Isolation and evaluation of oil-producing microalgae from
       subtropical coastal and brackish waters."
       http://www.plosone.org/article/fetchObj ect. action?uri=info%3 Adoi%2F 10.1371 %2Fj our
       nal.pone.004075l&representation=PDF. Accessed September 2013.

Liu, K., K.A. Rosentrater. 2011. Distillers grains: production, properties, and utilization. CRC
       Press, Baca Raton, FL.

Louma, J.R. 2009. "Hailed as a Miracle Biofuel, Jatropha Falls Short of Hype."  Yale
       Environment 360, May 2009.
       http://e360.yale.edu/feature/hailed as a miracle biofueljatropha falls  short of hype/2
       1477. Accessed August 2009.

Lumpkins, B.S., A.B. Bata, and N.M. Dale. 2004. "Evaluation of distillers dried grains with
       soluble as a feed ingredient for broilers." Poultry Science, 83:1891-1896

Lundquist, T.J., 1C. Woertz, N.W.T. Quinn, and J.R. Benemann. 2010. "A realistic technology
       and engineering assessment of algae biofuel production." http://www.ascension-
       publishing.com/BIZ/Algae-EBI.pdf Accessed January 2012.

Makkar, H.P.S., and K. Becker. 2009. "Jatropha curcas, a promising crop for the generation of
       biodiesel and value-added coproducts." European Journal of Lipid Science Technology,
       111:773-787.

Mallick, N. 2002. "Biotechnological potential of immobilized algae for wastewater N, P and
       metal removal: A review."  BioMetals, 15:377-390.

Martin, M., and I.E. Grossmann.  2012. "Simultaneous optimization and heat integration for
       biodiesel production from cooking oil and algae." Industrial and Engineering Chemistry
       Research, 51(23):7998-8014.

Mata, T.M., A.A. Martins, andN.S. Caetano. 2010. "Microalgae for biodiesel production and
       other applications: A review." Renewable and Sustainable Energy Reviews, 14(1):217-
       232.

McNeff, C.V., L.C. McNeff, B. Yan, D.T. Nowlan, M. Rasmussen,  A.E. Gyberg, B.J. Krohn,
       R.L.  Fedie, and T.R. Hoye.  2008. "A continuous catalytic system for biodiesel
       production."  Applied Catalysis A: General, 343:39-48.

Moser, B.R. 2010. "Camelina (Camelina sative L.) "Oil as a biofuels feedstock: golden
       opportunity or false hope?" Lipid Technology, 22(12):270-273.

Mulder, K., N. Hagens, and B. Fisher. 2010. "Burning water: a comparative analysis of the
       energy return on water invested." AMBIO, 39:30-39.

Murphy C.F., and D.T. Allen. 2011. "Energy-water nexus for mass cultivation of algae."
       Environmental Science and Technology, 46:5861-5868.
                                          192

-------
National Research Council of the National Academies (NRC). 2012. Sustainable development of
       algal biofuels in the United States. The National Academies Press, Washington D.C.
       www.nap.edu.

Niu, G., D. Rodriguez, M. Mendoza, J. Jifon, and G. Ganjegunte. 2012. "Response ofJatropha
       curcas to salt and drought stresses." InternationalJournal of Agronomy. 2012:632026.
       doi: 10.1155/2012/632026

O'Conner T. 2010. National Biodiesel Board, Personal communication.

O'Dowd, W.J., H.W. Pennline, M.C. Freeman, EJ. Granite, R.A. Hargis, CJ. Lacher, and A.
       Karash. 2006. "A technique to control mercury from flue gas: The thief process." Fuel
       Processing Technology, 87(12): 1071-1084.

Ozkan A, K. Kinney, L. Katz, and H. Berberoglu. 2012. "Reduction of water and energy
       requirement of algae cultivation using an algae biofilm photobioreactor." Bioresource
       Technology,  114:542-548.

Pate, R. 2008. "Biofuels and the Energy-Water Nexus." Sandia National Laboratories.
       http://www.sandia.gov/energy-water/docs/EW_RoadmapSum_Solar8-06.pdf. Accessed
       July, 2012.

Patil, P.D., V.G. Gude, A. Mannarswamy, S. Deng, P. Cooke, S. Munson-McGee, I. Rhodes, P.
       Lammers, and N. Nirmalakhandan. 2011. "Optimization of direct conversion of wet
       a/gae to biodiesel under supercritical methanol conditions." Bioresource Technology,
       102:118-122.

Pittman, J.K., A.P. Dean, and O. Osundeko. 2011. "The potential of sustainable algal biofuel
       production using wastewater resources."  Bioresource Technology, 102(1): 17-25.

Qureshi N, B. Annous, T. Ezeji, P. Karcher, and I. Maddox. 2005. "Biofilm reactors for
       industrial bioconversion processes: employing potential of enhanced reaction rates."
       Microbial Cell Factories, 4:24.

PNNL. 2004. "MASS2 Modular Aquatic Simulation System in Two Dimensions-Theory and
       Numerical Methods."  Pacific Northwest National Laboratory, Richland, WA.
       http://www.pnl. gov/main/publications/external/technical_reports/pnnl-14820_1 .pdf.
       Accessed May 2014.

Rao, A.V.R.K., S.P.  Wani, P.  Singh, K. Srinivas, and C.S. Rao. 2012. "Water requirement and
       use by Jatropha curcas in a semi-arid tropical location." Biomass andBioenergy,
       39:175-181.

Reuters. 2012. "Neste Oil's NExBL renewable aviation fuel used for the first time on an
       intercontinental Lufthansa flight."
       http://www.reuters.com/article/2012/01/12/idUSl 1909 l+12-Jan-2012+HUG20120112.
       Accessed December 2012.
                                          193

-------
Sanford, S.D., J.M. White, P.S. Shah, C. Wee, M.A. Valverde, and G.R. Meier. 2009.
       "Feedstock and biodiesel characteristics report."
       http ://www.regfuel. com/sites/default/files/pdf/Feedstock%20and%20Biodiesel%20Chara
       cteristics%20Report.pdf. Accessed Aug. 2012.

Sathish, A., and R.C. Sims. 2012. "Biodiesel from mixed culture algae via a wet lipid extraction
       procedure." Bioresource Technology, 118:643-647.

Scott, D. 2010. National Biodiesel Board, Personal Communication.

Sheehan, J., V. Camobreco, J. Duffield, M. Graboski, and H. Shapouri. 1998. "Life cycle
       inventory of biodiesel and petroleum diesel for use in an urban bus." NREL/SR-580-
       24089; National Renewable Energy Laboratory: Golden, CO.
       http://www.nrel.gov/docs/legosti/fy98/24089.pdf. Accessed March 2013.

Shurson, J. 2002. "The value and use of distiller's dried grains with solubles (DDGS) in swine
       diets." http://www.ddgs.umn.edu/articles-swine/2002-
       Shurson-%20Carolina%20nutr%20conf.pdf. Accessed June 2011.

Sills, D.L., V. Paramita, MJ. Franke, M.C. Johnson, T.M. Akabas, C.H. Greene, and J.W.
       Tester. 2013. "Quantitative uncertainty analysis of life cycle assessment for algal biofuel
       production." Environmental Science and Technology, 47:687-694.

Sims, B. 2011. "Finding the right purification approach."
       http://www.biodieselmagazine.com/articles/7661/fmding-the-right-purification-approach.
       Accessed March 2012

Singh, A.,  S.N. Poonam, and J.D. Murphy. "Renewable fuels from algae: an answer to debatable
       land based fuels." Bioresource Technology, 102(1): 10-16.

Singh, Chouhan, A.P., and A.K. Sarma. 2011. "Modern heterogeneous catalysts for biodiesel
       production: A comprehensive review." Renewable and Sustainable Energy Reviews,
       15(9):4378-4399.

Smith, W.  2011. Pacific Biodiesel Inc. Personal communication.

Smith, W.  2012. "Biodiesel purification: Finding the right fit."
       http://www.biodieselmagazine.com/articles/8462/biodiesel-purification-fmding-the-right-
       fjt. Accessed September 2013.

Strum, B.S.M., E. Peltier, V. Smith, and F. de Noyelles. 2011. "Control  of microalgal biomass
       and lipid production in municipal wastewater-fed bioreactors." Environmental Progress
       and Sustainable Energy, 31(1):10-16.

Subhadra,  E.G.  2011. "Water management policies for the algal biofuel  sector in the
       Southwestern United States" Applied Energy, 88:3492-3498.

Teixeira, R.E. 2012. "Energy-efficient extraction of fuel and chemical feedstocks from algae."
       Green Chemistry, 14:419-427.
                                          194

-------
Tilman, D., R. Socolow, J.A. Foley, J. Hill, E. Larson, L. Lynd, S. Pacala etal. 2009.
       "Beneficial biofuels—the food, energy, and environment trilemma."  Science,
       325(5938):270.

Tiwari, A.K., A. Kumar, and H. Raheman. 2007. "Biodiesel production from jatropha oil
       (Jatropha curcas) with high free fatty acids: an optimized process." Biomass and
       Bioenergy, 31:569-575.

Tu, Q. 2012. "Assessment of Selected Sustainability Aspects of Biodiesel Production: Water and
       Waste Conservation."  University of Cincinnati Master's Thesis.

Tu. Q., M. Lu and Y. J. Yang. 2014. "Water Consumption Estimates of Biodiesel Process in the
       U.S." Publication pending in ES&T.

United Soybean Board (USB). 2010. "Life cycle impact of soybean production and soy industrial
       products." http://www.sovbiobased.org/wp-content/uploads/2010/02/Sov-Life-Cycle-
       Profile Report.pdf Accessed March 2013.

United States Department  of Agriculture (USDA). 2008. "Farm and Ranch Irrigation Survey."
       United States Department of Agriculture, Washington, DC.
       http://www.agcensus.usda.gov/Publications/2007/Online Highlights/Farm and Ranch I
       rrigation Survey/index.asp. Accessed March 2013.

USDA-ARS. 2006. "Midwestern weed may inspire newfound respect."
       http://www.ars.usda.gov/is/pr/2006/061101 .htm. Accessed August 2012.

UOP. 2011. "Honeywell Green Jet Fuel™ powers U.S. Air Force Thunderbirds Demonstration."
       http://www.uop.com/honeywell-green-jet-fuel-powers-air-force-thunderbirds-
       demonstration/. Accessed December 2012.

U.S. Department of Energy (DOE). 2006. "Energy demands on water resources: Report to
       Congress on the interdependency of energy and water."  U.S. Department of Energy,
       California. http://www.sandia.gov/energy-water/docs/121-RptToCongress-
       EWwEIAcomments-FINAL.pdf Accessed May  2013).

U.S. Department of Energy (DOE). 2010. "National algal biofuels technology roadmap."
       http://wwwl.eere.energy.gov/bioenergy/pdfs/algal_biofuels_roadmap.pdf Accessed
       September 2013.

U.S. Environmental Protection Agency (EPA). 2010. "Renewable Fuel Standard Program
       (RFS2) Regulatory Impact Analysis."  Document Number EPA-420-R-10-006.
       http://www.epa.gov/otaq/renewablefuels/420rl0006.pdf Accessed June 2014

U.S. Environmental Protection Agency (EPA). 2006.  "Controlling Fats, Oils and Grease
       Discharges from Food Service Establishments."  EPA-833-F-07-007. Office of Water,
       National Pretreatment Program (40 CFR 403).

van Gerpen, J., B. Shanks, R. Pruszko, D. Clements, and G. Knothe. 2004. "Biodiesel Production
       Technology." NREL/SR-510-36244. National Renewable Energy Laboratory. Golden,
       CO. http://www.nrel.gov/docs/fy04osti/36244.pdf Accessed on March 2013.

                                         195

-------
Vasudevan, V., R.W. Stratton, M.N. Pearlson, G.R. Jersey, A.G. Beyene, J.C. Weissman, M.
       Rubino, and J.I. Hileman. 2012. "Environmental performance of algal biofuel technology
       options."  Environmental Science and Technology, 46(4):2451-2459.

Voegele, E. 2010. "Field pennycress shows feedstock potential." BiodieselMagazine, October
       Issue.

Wigmostra, M.S., A.M. Coleman, R.J. Skaggs, M.H. Huesemann, and L.J. Lane. 2011. "National
       microalgae biofuel production potential and resource demand." Water Resources
       Research, 47:WOOH04.

World Resources Institute (WRI). 2014. "Water stress magnifies drought's negative impacts
       throughout the United States."  http://www.wri.org/blog/2014/04/water-stress-magnifies-
       drought%E2%80%99s-negative-impacts-throughout-united-states. Accessed June 03,
       2014.

Wu, M., M. Mintz, M. Wang, and S. Arora. 2009. "Water consumption in the production of
       ethanol and petroleum gasoline." Environmental Management, 44(5):981-997.

Yang, J., M. Xu, X. Zhang, Q. Hu, M.  Sommerfeld, and Y. Chen. 2011. "Life-cycle analysis on
       biodiesel production from microalgae: Water footprint and nutrient balance."
       Bioresource Technology, 102:159-165.

Ye, M., C. Li, G. Francis, and H.P.S. Makkar. 2009. "Current situation and prospects of
       Jatropha curcas as a multipurpose tree in China."  Agroforest Systems, 76:487-497.

Zhang, Y., M.A. Dube, D.D. McLean,  and M. Kates. 2003. "Biodiesel production from waste
       cooking oil: 1. process design and technological assessment." Bioresource Technology,
       89:1-16.

Zhou, W., Y. Li, M. Min, B. Hu, P. Chen, and R. Ruan. 2011. "Local bioprospecting for high-
       lipid producing microalgal strains to be grown on concentrated municipal wastewater for
       biofuel production." Bioresource Technology, 102:6909-6919.
                                          196

-------
8.6  Appendix

8. 6.1   Sample calculation for irrigation water consumption and normalized water intensity
      for Ohio

   8. 6. 1. 1  Irrigation water consumption for Ohio (Wion)

       An example of calculation procedures for Wi and Ni for Ohio is provided below.
Following the same principle, results can be calculated for the state-level irrigation water
consumption for soybean growth by using state-specific data from USDA reports (USD A, 2007,
2008). According to Table 28 in 2008 Farm and Ranch Irrigation Survey (USDA, 2008), the total
irrigated area for soybean through primary water distribution methods is 1,056 acres in Ohio and
the average acre-feet applied per acre is 3.2. Therefore, the total volume of irrigation water for
soybean in 2007 is:
       VT = 1056 x  3.2 = 3379.2 acre-feet
       Since one acre-foot equals 325,851 gallons:
       VTj =  3379.2 x 325851 = 1.10 x 109gallons of water


       Mass-based allocation:

       According to the assumptions above, 19.5% (Fsoy) of soybean was oil, about 17% (Fuse)
of the soybean oil  in 2007 was used for biodiesel production (Centrec Consulting Group, 2013)
and 89% (FBioD) oil was eventually converted into biodiesel (USB, 2013).
       So for the calculation
            = VTj x Fsoy x Fuse x FBioD

            _ 1.10 x 109  x  19.5% x 17% x 89%
            ~             1000000
            = 12.25 million gallons of water (MMgpy)
   8.6.1.2  Normalized irrigation water intensity for Ohio (Nion)

       Assume one bushel soybean weighs about 60 pounds and the density of soybean
biodiesel is 7.4 Ib/gallon. According to Table 28 in 2008 Farm and Ranch Irrigation Survey
(USDA, 2008), total harvested soybeans in bushel are 191,559,567, the normalized irrigation
water consumption per bushel of soybean is:


                       109
       K!  = 1.10 x
                   191559567
          = 5.75 gallons of water/bushel soybean
                                          197

-------
Appendix Table 8.1 Biodiesel plant capacity in each state
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Plant Capacity
(MMgpy)
13.3
0.25
48
119
78.465
0
3.8
5
39
63.5
4.5
5.5
173.6
121
304.5
3.9
63
19.2
1.5
7.5
0.5
49.75
66
115.5
183.5
0
State
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia




Plant Capacity
(MMgpy)
5
1
5.75
90
1.5
25.25
21.6
88
132
36
17.94
111.26
2
90.3
7
53
577.25
10
0
16
113
3




  So, the allocated water consumption for biodiesel from one bushel is:
  RZ = 5.75 x 19.5% x 17% x 89% = 0.17 gallons water I bushel soybean
  Finally, the irrigation water consumption based on every single gallon of biodiesel is:
                                      198

-------
       NIOH =0.71 gallons water I gallon soybean biodiesel

   8.6.1.3  Allocation factor for soybean growth stage

       FI = Fsoy x Fuse * FBioD  = 19.5% x 17% x 89% = 0.03
8.6.2   Calculation of normalized water consumption and sample calculation for water
       consumption in the soybean crushing and processing stage for Ohio

   8.6.2.1  Normalized water consumption (N2) in the soybean crushing and processing stage

       According to the life cycle report by United Soybean Board (2010), the water
consumption during soybean processing and refining stage is: 1,167 and 65.9 kg/1,000 kg
soybean oil for the two steps. Below is the conversion of water consumption occurred in this
stage into normalized value based on one gallon of biodiesel.
       _ (1,164 + 65.9)/c# H20        1 m3       900  kg soybean oil
    N2= 1,000kg soybean oil * 1,000 kg H20 *T^?        x Fuse x FBioD
                  = 0.17 gal/gal
       As stated in the main text, N2 is assumed to be uniformly applicable to all the states in
this study.

   8.6.2.2  Water consumption during soybean crushing and processing for Ohio (W2on)

       Also for the total water consumption in this stage (W2), the calculation is performed
based on the same allocation principles.  Below is the sample calculation for the State of Ohio.
       From Table 28 in 2008 Farm and Ranch Irrigation Survey (USDA, 2008), the harvested
soybean in 2007 is 191,559,567 bushels, which translates into 5.2 X 109 kg. By applying the
consumption factor of 1,229.9 kg water /1000 kg oil (Appendix Table 8.2), the total water
consumption before allocation is 6.4 x 109 kg. Following the same allocation procedure, the
total water consumption during soybean crushing and processing stage for Ohio is 49.95 MMgy.
                       60 Ib soybean         0.454 kg     1     (1,164 + 65.9)fcg H20
 W20H = 191,559,567 x - - - x Fsov x - - x - x - -
   20H                    bushel       soy       Ib      1,000    1,000 kg soybean oil
                              1 gal H20        1 MMgy
              X
                                            1.000.000 gal/yr
   8.6.2.3  Allocation factor for soybean crushing & processing stage

                        F2 = Fuse x FBioD = 17% x 89% = 0.15
                                         199

-------
8. 6. 3   Sample calculation for normalized and total water consumption for biodiesel
       manufacturing in Ohio

   8. 6. 3. 1  Normalized water consumption (Ns) in the biodiesel manufacturing stage

       Three scenarios are proposed in this study to account for water consumption from
different purification methods (water/day wash) and process  operations (cooling tower make-up).
Assuming water wash  and dry wash both account for 50% of current biodiesel purification
technology, an averaged value from the data representing different scenarios is obtained through
the following equation:
           rl  r,
     N3 = j- x [(Water Washupper + Cooling Towerwater) + (Water Washlower

                                        M                     )   !
                   + Cooling Towerwater)\  +  Cooling Towerdryt x — =  0.31 gal /gal

       Where: Water Washupper and Water Washiower are the washing water consumptions
(gal/gal) from upper and lower scenarios; Cooling Towerwater and Cooling  Towerdry are the
volumes of cooling tower make-up water (gal/gal) for water wash and dry wash scenarios.

   8. 6. 3.2   Total water consumption (Wson) in biodiesel manufacturing stage for Ohio

       Wj is calculated by following equation:
       W3 = N3 x Total Biodiesel Plant Capacity

       For a specific state, such as Ohio, the product of Nj (0.31) and total biodiesel plant
capacity (132 MMgpy) yield a WSOH of 40.92 MMgy.
8. 6.4   Water-saving technologies developed in the algae industry

•  Reactor design
       A commercial application of the biofilm reactor is the "Algal Turf Scrubber" by
Hydromentia (http ://www.hydromentia. com/). A plastic mesh is used as the growth substrate for
filamentous algae and the mature biofilm is scraped by a scrubber to harvest algae biomass.
•  Harvesting and lipid extraction
       Algaeventure Systems, Inc. (http ://algaevs. com/) received $5.9 million from U.S.
Department of Energy's Advanced Research Projects Agency-Energy (ARPA-E) and developed
the solid liquid separation (SLS) system for the harvesting of algae, which significantly lowers
the cost of separation compared with centrifuge separation systems. Evodos (www.evodos.eu)
developed a low-cost, high-performance centrifuge system that could reach over 95% of the
separation efficiency and minimize the extracellular water attached to the algae biomass paste.
Aurora Algae Inc. (http ://www.aurorainc. com/) developed a unique process that removed protein
and carbohydrates from the algae growing media and left behind the water/lipid mixture that
could be easily separated. AER Sustainable Energy (www.aer-bio.com) developed an enzymatic
hydrolysis process that lysed cell walls of algae. United Technologies Inc. (www. uniteltech . com)
developed a hydrolysis technology that focused on producing fatty acids from algae biomass to
avoid the water removal and lipid extraction steps. Diversified Technologies Inc.
                                          200

-------
(www.divtecs. com) utilized pulsed electric field technology to rupture the cell wall of the algae
biomass to release the lipids and hence facilitate the oil extraction process. Phycal's
(www.phycal.com) proprietary non-destructive oil extraction technology continuously recovers
oil from algal cells. The process mixes the algae culture with a lipid extraction solvent and uses
sonication for separation. OriginOil (www.originoil.com) developed an electro-assisted
harvesting device that coagulates and floats the algae cultures for skimming collection. New Oil
Resources (www.newoilresources.com) employed a one-step process that converts biomass into
fuels. The technology uses water  at high temperature and pressure to de-polymerize the biomass
into smaller compounds and allow recovery.
     Appendix Table 8.2 Water consumption data in biodiesel wash collected from
               biodiesel manufacturers
Data Sources
Company 1
Company 2
Company 3
Company 4
Company 5
Company 6
Feedstock
Multi-feedstock
Animal Fats
Waste Cooking Oil
Multi-feedstock
Multi-feedstock
Waste Cooking Oil
Animal Fats
Washing Water
(gal/gal)
0.1
0.0125-0.015
0.84
0.25-0.375
0.09-0.1
0.06
Production Capacity
(Million Gallons per year)
3
1.25
4.5
12
180
1.5
       r Company names omitted at their requests.
       The variation in the data reported by these companies can be attributed to a number of
factors, including water reuse practices, washing water properties (e.g., acidic/warm), plant size,
as well as water availability and pricing.
     Appendix Table 8.3 Water consumption in cooling tower make-up
Data Sources
Company 7
Company 8
Feedstock
Virgin oil
Virgin oil
Waste cooking oil
Waste cooking oil
Multi-feedstock
Purification method
Dry wash (silicate)
Water wash with recycle
Dry wash (silicate)
Water wash with recycle
Dry wash (silicate)
gal water/ gal biodiesel
0.12-0.15
0.19-0.21
0.27-0.3
0.33-0.36
0.03-0.05
       Data from Company 7 indicates that using low quality feedstock corresponds to higher
make-up water, which may be due to the need to recover an excessive amount of methanol (e.g.,
for the esterification reaction). Company 8 has much smaller make-up water consumption, which
                                          201

-------
is achieved by integrating other cooling approaches such as an air chiller. The biodiesel
production capacities of these two companies are 29 MMgpy and 14 MMgpy respectively.
     Appendix Table 8.4 Summary of water-stressed states from literature
Studies
EPRI report 1
Hurdetal.2
Scown et al. 3
Yang4
Criteria
Water Supply Sustainability Index
Level of development, natural variability,
dryness ratio, groundwater depletion,
industrial water use flexibility and
institutional flexibility
Palmer Drought Index
Available precipitation
Water stressed areas
AL, AZ, CA, FL, GA, ID, LA, NM,
NV, TX, WA
AZ, CA, CO, KS, NM, NV, TX, UT
Southwestern US
AZ, CA, CO, FL, GA, NV
    Note:
EPRI, 2003;    2 Hurd et al., 1999;3 Scown et al., 2011;4 Yang, 2010
     Appendix Table 8.5 Key assumptions and parameters of the studies

O'Connor1
King and
Webber2
Harto et al.3
Mulder et a/.4
This study
Parameters and Assumptions
Irrigation water
loss %
% soybean
irrigated
Energy-water in
irrigation
Fertilizer water
use
100%
8.2%
No
No
79.7%
either 100% or
0%
0.1 58 gal/gal
No
100%
4%
No
1 1 gal/gal
100%
NA
No
No
100%
Overall:
8.2%
No
No
Parameters and Assumptions
Normalized
irrigation
consumption (Nt)
Normalized
consumption
during soybean
crushing &
processing (N2)
Soybean oil-to-
biodiesel (Ns)
Nfof (gal/gal)
79 gal/gal
NA
NA
79
200 gal/gal
0.009 gal/gal
0.1 58 gal/gal
200.32
11 9.5 gal/gal
No
0.5gal/gal
131
71 6.35 gal/gal
No
3.63gal/gal
719.98
61. 78 gal/gal
0.1 7 gal/gal
0.31 gal/gal
62.26
    Note: 10'Connor, 2010; 2King and Webber, 2008; 3Harto et al., 2010; "Mulder et al., 2010.
                                            202

-------
100000


 10000
 1000


 ; 100 •


  10
    Ulli
     AR NE MS KS MO LA MN IN  IL NC Ml SD GA IA TX Wl DE MO OK ND SC KY VA OH TN NJ CO  AL WA CT ME MA VT WV NY
                                           State

Appendix Figure 8.1 Irrigation water use at state level for soybean growth (Wl, million
       gallons per year). Note that 35 out of 50 states have data.
10000 i

 1000 •

  00
  u

   1

  01
Illlll,
    WA Aft CO MS NE TX 06 KS LA GA OK MO MO NC NJ SC Ml AL VA Wl MN SO IN ICY TN IL NO  IA OH CT ME MA NY VT WV
                                           State
Appendix Figure 8.2 The irrigation water intensity for soybean growth by state (Nl, gallon
      water per gallon biodiesel).
8.6.5  Appendix 8 References
Centrec Consulting Group, LLC. 2013. "Economic impacts of biodiesel production on the
      soybean sector, revisited." http://www.biodiesel.org/reports/2010120l_gen-423.pdf
      Accessed January 2013.

Electric Power Research Institute (EPRI). 2003. "A survey of water use and sustainability in the
      united states with a focus on power generation." Electric Power Research Institute:
      California.
      http://my.epri.com/portal/server.pt?space=CommunityPage&cached=true&parentname=
      ObjMgr&parentid=2&control=SetCommunitv&CommunityID=404&RaiseDocID=0000
      00000001005474&RaiseDocType=Abstract id. Accessed March 2013.

Harto, C., R. Meyers, and E. Williams. 2010. "Life cycle water use of low-carbon transport
      fuels." Energy Policy, 38(9):4933-4944.
                                         203

-------
Kurd, B., N. Leary, R. Jones, and J. Smith. 1999. "Relative regional vulnerability of water
       resources to climate change."  Journal of the American Water Resources Association,
       35(6):1399-1409.

King, C.W., andM.E. Webber. 2008. "Water intensity of transportation." Environmental
       Science and Technology,  42(21):7866-7872.

Mulder, K., N. Hagens, and B. Fisher. 2010. "Burning water: A comparative analysis of the
       energy return on water invested." Ambio, 39(l):30-39.

O'Conner, T. 2010. National Biodiesel Board, Personal communication.

Scown, C.D., A. Horvath, and I.E. McKone. 2011. "Water footprint of US transportation fuels."
       Environmental Science and Technology, 45(7):2541-2553.

United Soybean Board (USB). 2013. "Life cycle impact of soybean production and soy industrial
       products." http://www.sovbiobased.org/wp-content/uploads/2010/02/Sov-Life-Cycle-
       Profile Report.pdf. Accessed March 2013.

United States Department of Agriculture (USDA).  2007. "Census of Agriculture."  United States
       Department of Agriculture. Washington, DC.

United States Department of Agriculture (USDA). 2008. "Farm and Ranch Irrigation Survey."
       United States Department of Agriculture. Washington, DC.

Yang, YJ. 2010. "Topographic factors in precipitation dynamics affecting water resource
       engineering in the contiguous U.S.: Notes from hydroclimatic studies." World
       Environmental and Water Resources Congress: Challenges of Change. Providence, RI.
                                          204

-------
 9   Impacts of Electric Generation on Water Resources: Regional
     Assessment and Adaptation
       Patcha Huntrai and Timothy C. Keener1

9.1  Introduction

       Thermoelectric power plants in the southwest U.S. are most likely to face challenges
regarding water withdrawal and consumption due to the arid climate. The impacts of
thermoelectric water withdrawals are exacerbated by regional population migration. The
southwest and southeast regions of the U.S. were selected for a sustainable future projection case
study within this report.  Southwestern states in particular (e.g., Arizona, California, New
Mexico, Nevada, Texas and Utah) face severe water-stressed conditions (Yates et al. 2013a;
Yates et al. 2013b). Consumptive water (i.e., water that is effectively removed from the system
and not available for other uses) is an important consideration for water-scarce regions, and is
particularly relevant in future energy resource development (NREL, 2012). Physical and
regulatory compliance limitations associated with high water withdrawals can lead to water-
related power plant curtailments and shut downs even in water-rich regions,  such as the 2007
southeast U.S. drought (NETL, 2009).
       As discussed in further detail within Chapters 2 and 4, the electric power sector is
responsible for a significant fraction of total water withdrawals in the U.S. For example,
thermoelectric power plant operations are estimated to be responsible for 36 to 41 percent of
total freshwater withdrawals (Meldrum, 2013b). Regulatory changes, policy changes, and shifts
in the dispatch of electricity by plant type are expected to significantly impact management of
local, regional, and national water resources (NREL, 2012).

9.2   Regional Integrated Water-Energy Resource Management

       There is geographical variation in the relative importance of water withdrawals and
consumption due to regional water resource availability, environmental considerations and  water
allocation requirements (NREL, 2012). Due to the uncertainty  of climate change and the
increasing demand  for water to cool thermoelectric generation  capacity, there will be substantial
competition for water in many regions. Water-stressed regions are of particular concern as water
withdrawal rates are greater than 60 percent of mean annual runoff (NREL, 2012; Raskin et al.,
1997; Waggoner et al., 1990). Evaluating the water usage of all electricity-generating
technologies will provide critical information necessary for decision makers to develop strategies
to relieve critical stresses from water resources.
       The most common forms of electric generation in the U.S. such as coal, natural gas, and
nuclear power are highly water-consumptive. In contrast, the least water-consumptive forms of
generation are those of emerging technologies such as wind and solar photovoltaic technologies.
       Figure 9.1 shows the water intensity of electricity generation (represented in gallons per
megawatt). These water intensities demonstrate the consumptive demand for cooling systems,
   University of Cincinnati, Department of Biomedical, Chemical, and Environmental
   Engineering
                                         205

-------
which vary considerably depending on the fuel source and the type of cooling technology used.
Researchers have found that many low-carbon-intensity, renewable sources of electric generation
such as wind, solar photovoltaic (PV), geothermal energy, and some types of concentrating solar
power (CSP), consume far less amounts of water when compared to non-renewable sources such
as fossil-fuel thermoelectric generation (WRA, 2008).
      Photovoltaic and wind generation require far less water per unit energy produced than do
conventional thermoelectric generation. In water-deficient regions, such generation sources with
low water requirements may be in increasing demand. Another example of renewable electric
generation is a wet cooling CSP plant that uses water to condense steam downstream of a steam
turbine. Even though wet cooling CSP  designs consume a great amount of water compared to dry
cooling CSP, the wet cooling CSP may be required in arid climates because a dry solar CSP
plant may overheat due to an insufficient temperature differential in the heat exchanger.
Overheating of a CSP plant reduces efficiency, thus decreasing electricity output (WRA, 2008).
          2000

          1800

          16OO

          MOO

          I2OO

          1OOO

          eoo

          6OO

          4OO

          2OO
                             WATER INTENSITY OF ELECTRICITY GENERATION
                       EMERGING
                     TECHNOLOGIES
  CONVENTIONAL
   GENERATION
Ill     ,       ll
                                            RENEWABLE*
                                                     f ^ //  f f  f
                                                     *? A1  •?  •*  *V  rf  iV
                *3>
   -? ^ ^      c  ^
  rar*/         £

                                                ,•?' #'
                                f //     / /
              g  & o
             J* (^
             -^  cr
Figure 9.1 Typical rate of water consumption for electricity generation (consumptive use).
      Error bars are included for technologies for which multiple data points were available.
      Adopted from Western Resource Advocates (2008).
       The life cycle water use research by Mel drum et al. (2013 a) estimates water withdrawal
and water consumption for selected electricity-generating technologies. These water use
                                         206

-------
estimates were classified as either "water withdrawal" or "water consumption." Water
withdrawal is defined as the water diverted from a water source for use, and water consumption
is defined as the portion of withdrawn water that is not returned to the immediate water resource
after use (Meldrum et al. 2013b). Water use factors were developed for each of the three main
life cycle stages: the fuel cycle, the power plant life cycle, and power plant operation. These
three stages can be seen in Tables 9.1 to 9.3. The fuel cycle includes fuel extraction, fuel
processing, and fuel transportation. The power plant life cycle includes component
manufacturing, power plant construction and power plant decommissioning. The life cycle water
usage factors are primarily influenced by the cooling water demands for thermoelectric
generation. Of the renewable energy sources, two generation technologies (PV and wind) used
the least amount of water across the life cycle. These two technologies only require water for
occasional cleaning purposes (see Table 9.2 and Table 9.3). The life cycle water use for the
different technologies varies based on the  spatial and temporal requirements associated with the
fuel cycle and the power plant life cycle. Because of this variability, the water consumption and
water withdrawal for electricity generation should be considered separately across different life
cycle stages (Meldrum et al. 2013b). Thus, the following stated water usage will be considered
solely in the context of power plant operation, and more specifically, the cooling water process.
     Table 9.1  Range of fuel cycle water consumption and withdrawal estimates for
               selected generation technologies and sub-categories

Generation Technology
Coal
Coal
Natural Gas
Natural Gas
Nuclear
Nuclear
Sub-Category
Surface Mining
Underground Mining
Conventional Gas
Shale Gas
Centrifugal
Enrichment
Diffusion Enrichment
Consumption (gal/MWh)
Min Median Max
6.1 21.9 58.1
16.9 55.5 229.9
1.1 4.0 25.9
2.9 16.1 208.7
12.9 55.5 290.6
42.3 87.2 317.0
Withdrawal (gal/MWh)
Min Median Max
6.1 21.9 60.8
16.9 58.1 229.9
4.0 5.0 34.3
5.0 16.9 219.3
12.9 55.5 290.6
60.8 140.0 422.7
  Note: Adopted from Meldrum et al. (2013a,b).
                                           207

-------
     Table 9.2 Range of power plant equipment water consumption and withdrawal
               estimates for selected generation technologies and sub-categories

Generation Technology
Coal
Natural Gas
Nuclear
CSP
Geothermal
PV (C-Si)
PV(Other)
Wind
Consumption (gal/MWh)
Min
<0.5
<0.5
<0.5
79.3
2.1
10.0
5.0
<0.5
Median
1.1
1.1
<0.5
158.5
2.1
81.9
6.1
1.1
Max
25.1
1.1
<0.5
169.1
2.1
208.7
6.9
9.0
Withdrawal (gal/MWh)
Min
<0.5
<0.5
<0.5
97.8
<0.5
1.1
<0.5
13.00
Median
1.1
<0.5
<0.5
161.2
2.9
95.1
18.0
25.9
Max
11.9
1.1
<0.5
169.1
10.0
1611.6
1400.3
81.9
   Note:  Adopted from Macknick et al. (2012).
     Table 9.3 Power plant operations cycle water consumption and withdrawal estimates
               for selected generation technologies and sub-categories

Generation Technology
Coal: Pulverized Coal (sub-
critical)
Coal: Integrated Gasification
Combined Cycle
Natural Gas: Combined Cycle
Natural Gas: Combustion
Turbine
Nuclear
CSP: Trough
CSP: Power Tower
Geothermal
PV
Wind: On-shore
Sub-Category
Cooling Tower
Cooling Tower
Cooling Tower
No Cooling
Cooling Tower
Cooling Tower
Cooling Tower
Binary,
Dry Cooling
Flat Panel
n/a
Consumption (gal/MWh)
Min Median Max
200.8 528.4 1294.6
34.3 317.0 449.1
47.6 208.7 290.6
50.2 50.2 343.5
581.2 713.3 898.3
554.8 898.3 1902.2
739.8 819.0 871.9
264.2 290.6 634.1
1.1 6.1 25.9
1.1 1.1 2.1
Withdrawal (gal/MWh)
Min Median Max
449.1 660.5 1188.9
161.2 396.3 605.0
150.6 251.0 766.2
422.7 422.7 422.7
792.6 1109.6 2589.2
871.9 951.1 1109.6
739.8 739.8 739.8
264.2 290.6 634.1
1.1 6.1 25.9
1.1 1.1 1.1
   Note: Adopted from Meldrum et al. (2013a,b) and Macknick et al. (2012).

       The location of a power plant also has a major impact on the technology and processes
that are used for generating electricity and, therefore, has a direct impact on water consumption
                                           208

-------
     Table 9.4 National electricity water crisis areas
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
County
Mecklenburg
Lake
Will
Queens
Cobb
Dallas
Coweta
Denver
Montgomery
St. Charles
Washington
Bexar
Calvert
Harris
Tarrant
Multnomah
Contra Costa
Fort Bend
Wake
Suffolk
Clark
Montgomery
Total
Electricity in
2025 in (MW)
17,950
12,987
27,399
11,613
3,480
6,170
6,180
4,503
3,776
3,350
3,203
9,222
12,938
4,462
2,704
5,402
4,759
19,656
5,967
5,062
20,148
2,871
Population Growth
1995 to 2025 (per
sq.mile)
1,528
1,064
806
8,056
2,049
1,437
510
1925
757
533
632
555
533
1,179
1,170
548
678
851
1,266
1,184
642
647
Summer Water
Deficit in 2025
(inches)
28.72
18.1
16.67
12.68
9.34
6.6
5.56
4.98
4.45
4.33
4.2
2.98
2.92
2.4
2.34
2.24
1.99
1.88
1.65
1.65
1.52
1.52
Metropolitan Area
Charlotte, NC
Chicago, IL
Chicago, IL
New York, NY
Atlanta, GA
Dallas, TX
Atlanta, GA
Denver, CO
Washington, DC and
Baltimore, MD
St. Louis, MO
St. Paul, MN
San Antonio, TX
Washington, DC and
Baltimore, MD
Houston, TX
Dallas, TX
Portland, OR
San Francisco, CA
Houston, TX
Raleigh, NC
Boston, MA
Las Vegas, NV
Houston, TX
    Note: Adapted directly from Sovacool and Sovacool (2009).
and withdrawal. Sovacool and Sovacool (2009) identified 22 National Electricity-Water Crisis
Areas (shown in Table 9.4), and showed that there is likely to be a trade-off between the water
needed to satisfy demands for drinking, agriculture and other uses in these areas and the water
needed for new thermoelectric generation capacity. These metropolitan areas have a combined
population growth of at least 500 people per square mile, a demand of at least 2,700 MW of
electric capacity, and a projected summer water deficit of at least 1.52 inches by 2025 (See Table
9.4) (Sovacool and Sovacool, 2009). Their findings showed that ten of the National Electricity-
Water Crisis Areas—Atlanta, Charlotte, Chicago, Denver, Houston, Las Vegas, New York, San
                                           209

-------
Francisco, St. Louis, and Washington, DC—plan to add a collective 149,892 MW of
thermoelectric capacity by 2025 with the potential use of 29.41 trillion gallons of water per year
for thermoelectric cooling (nearly 81 billion gallons of water per day) (Sovacool and Sovacool,
2009). A case study was conducted to follow the water and energy management of one of these
ten National Electricity-Water Crisis Areas, Las Vegas.

9.2.1   Case study: Las Vegas
       The Las Vegas metropolitan area is one of the ten national electricity-water crisis areas
depicted in Table 9.4. Growing populations, including tourists and business visitors, have
increased demand for electricity and water supplies. Water resources in the Las Vegas
metropolitan area have been a concern due to the ongoing drought conditions in the Colorado
River Basin. The basin supplies water to Lake Mead, which accounts for 90% of Southern
Nevada's water. Managing demand and meeting long-term needs are urgently required along
with extensive planning to make sure that water consumption in the area falls within the amount
of water available for use. The following subsections describe ongoing problems, existing water
conservation efforts and long-term water resources management options for Las Vegas.

   9.2.1.1  Drought in Lake Mead

       Lake Mead is the reservoir for the southwest states' share of Colorado River water
resources. It provides water for 20 million people in southern Nevada, southern California, and
Arizona. Since 2000, Lake Mead's elevation at the Hoover Dam has declined more than 80 feet
(see Figure 9.2).
                            Lake Mead Elevation during 1950-2013
   1250
   1200
   1150
 £ 1100
   1050
   1000
Figure 9.2 Lake Mead elevation from 1950 to 2012.

                                          210

-------
       If Lake Mead's water level drops below 1,075 feet, the U.S. Department of the Interior
will declare a water shortage on the river (SNWA, 2012). Figure 9.2 shows monthly records of
elevation at Lake Mead from 1950 to 2013. The lowest elevation of 1,082 feet occurred in
November 2010, during which time the Department of the Interior was about to declare a
shortage on the river. These elevation drops lead to a reduction in Nevada and Arizona's
available Colorado River water allocation (SNWA, 2012).
                         Renewable^              .Hydroelectric
                           Energy
                            12%
Figure 9.3 Electricity generation by source type for Nevada as of February 2014. Data from
       EIA(2014c).
   9.2.1.2  Water resource management for electricity generation in Las Vegas area areas

       Most of Nevada's electricity is generated thermoelectrically, a process that relies on a
significant amount of water resources due to the water required for steam cooling and other
processes. Of Nevada's total electricity production, 53 percent is from natural gas and 27 percent
is from coal. Moreover, from Nevada's energy portfolio in 2014, hydroelectric power accounts
for 8 percent of total electricity generation (see Figure 9.3). Renewable energy (excluding
hydroelectric) accounts for 12 percent of the electricity production in Nevada. Of the total
renewable energy in Nevada, 82 percent came from geothermal energy and solar energy (EIA,
2014c).
       During the past decade, the nation battled drought and adapted to dramatic changes in
weather patterns. These warm, arid conditions are expected to continue through the next decade
(SNWA, 2012). To cope with these changes, the Southwest Nevada Water Authority (SNWA)
employs a multi-faceted approach to  add flexibility to the state's water resources. Their primary

                                          211

-------
approaches to water resource management included implementing a water conservation program,
treating and reusing wastewater and extending water supplies, as described below.

•      Water conservation programs in Las Vegas
       To promote the efficient use of water and reduce water waste, the SNWA has created
conservation programs in the Las Vegas area. Conservation is considered to be a valuable water
resource that reduces overall demand and extends supplies. The SNWA utilizes several
conservation tools to reduce water usage that include a combination of regulation, water pricing,
incentives and education.
       Established in 1991, the conservation program in southern Nevada aims to reduce water
use for both indoor and outdoor consumption. The primary conservation programs are  focused
on regulating outdoor water uses, which make up the majority of consumptive water demand in
southern Nevada.
       In 1991, local government agencies adopted watering restrictions that prohibited watering
during the hottest times of the day in the warmer months (SNWA,  2009). In 2003, the  SNWA
member agencies adopted more policies that include additional restrictions on landscape
watering, vehicle washing, lawn installation, mist systems and golf course water use during
declared droughts (SNWA, 2009).
       Landscape water, for example, is prohibited from 11 a.m. to 7 p.m. in the summer, and
limited to one day a week in winter and three days a week in spring and fall (SNWA, 2009). For
residential vehicle washing, a positive shutoff nozzle is required, and commercial vehicle
washing is prohibited unless water is captured to be treated and reused. Moreover, for lawn
installation, turf is prohibited in the new residential  front yards and commercial development.
The use of commercial mist systems is limited to the summer months, and fountains and
ornamental water features are prohibited except as allowed by jurisdiction policy. Golf courses
are subject to mandatory water budgets of 6.3 acre feet of water (SNWA, 2009).
       Water rates are one of the most effective conservation tools. The SNWA member
agencies have adopted conservation-oriented water rates that support the community's
conservation goals. The SNWA regularly assesses the rate to ensure that it corresponds with
inflation and maintains their effectiveness in encouraging conservation (SNWA, 2009).
       The SNWA also has a number of "water smart" incentive programs that invite the
community to participate in the conservation effort. These water smart programs include a water-
smart landscape rebate program, efficient landscape irrigation equipment, water-efficient
technologies, water-smart car washes, a pool cover rebate program, water smart contractor
program, water smart homes and water upon request.
       To educate communities about the importance of conservation and how they can
conserve water most effectively, the SNWA also created public education programs such as the
water conservation coalition, water smart innovations, a conservation helpline, publication and
media, demonstration gardens and FkO university (SNWA, 2009). Details regarding both the
SNWA incentive programs and education programs are summarized within the Appendix to this
chapter.
       SNWA's conservation efforts have reduced water consumption by roughly 21 billion
gallons annually between 2002 and 2008 (SNWA, 2009). As a result, conservation remains an

                                          212

-------
important element in planning and balancing the resource and infrastructure needs in southern
Nevada.
•      Wastewater reuse and recycle
       Wastewater is reclaimed, treated and used as a resource in southern Nevada (SNWA,
2012). When southern Nevada directly reuses or returns flows to Lake Mead, additional
reclaimed Colorado River water for irrigation or cooling are credited. These credits are referred
to as "Return-flow credits." Reclaimed water accounts for roughly 40 percent of domestic water,
making it the second-largest water resource (SNWA, 2012). Southern Nevada reclaims
wastewater through return-flow credits or direct reuse. Approximately 200,000 acre-feet or
62,700 million gallons of urban flow wastewater and runoff are returned to the  Colorado River
each year for return-flow credits. Direct  reuse accounts for about 17,000 acre-feet per year or
5,542 million gallons per year (SNWA, 2012).

•      Extending water resources
       Intentional Created Surplus (ICS) are credits that accumulate when water agencies
conserve or introduce additional water into the Colorado River. These credits can be earned
through Tributary Conservation, Importation, System Efficiency and Extraordinary ICS.
       The SNWA has created approximately 124,000 acre-feet (40,400 million gallons) of
Tributary Conservation ICS from conveying Muddy and Virgin rivers water rights to Lake Mead
for Colorado River credit (SNWA, 2012). For the Imported ICS, the SNWA transports
groundwater from Coyote Spring  Valley to Lake Mead by constructing facilities for conveying
groundwater to Lake Mead, which totaled 4,000 acre-feet of groundwater or 1,300 million
gallons in 2012 (SNWA, 2012). An "Extraordinary Conservation ICS" is the credit that is
applied when both the "Imported  ICS" and "Tributary Conservation ICS" are not used during the
year and thus converted to credit for the  following year.
       Southern Nevada has an additional 40,000 acre-feet or 13,000 million gallons of
Colorado River water available for consumptive use each year, which was created by the Warren
H. Brock Reservoir near Gordon Wells,  California to capture unused Colorado  River water that
eventually passes into Mexico (SNWA,  2012).
       SNWA's system Efficiency ICS  also received water credit from the Yuma Desalting
Plant (YDP) that conserved 30,000 acre-feet or 9,700 million gallons of irrigation return flow
water, which was then returned to the Colorado River to adjunct water delivery obligations to
Mexico (SNWA, 2012).
       Moreover, Arizona and California struck an agreement with Nevada that allows the
SNWA to bank an additional 200,000 to 400,000 acre-feet of Colorado River water between
2012 and 2016 in their resources.
       Regarding infrastructure adaptation for water resource management, the SNWA is
constructing a third intake tunnel  for treatment and distribution, located approximately 350 feet
beneath the surface of Lake Mead, to keep Lake Mead above shortage levels during critical
drought conditions. The project is expected to be completed in 2014 (SNWA, 2012).
                                          213

-------
9.3  Holistic Water Resource Adaptation

9.3.1   Future energy production scenarios
       In 2011, nearly 90 percent of electricity in the U.S. came from thermoelectric (coal,
natural gas and nuclear) power plants (Clemmer etal. 2013), yet this percentage is projected to
fall to 83 percent by 2040 (Figure 9.4) (EIA, 2014a). Note that the most recent EIA projections
due not yet include projected impacts due to proposed GHG regulations.36
       Growth in renewable generation is supported by many state requirements and by
greenhouse gas (GHG) emission regulations. The proportion of U.S. electricity generation
coming from renewable fuels (including conventional hydropower) is expected to increase from
12 percent in 2012 to 16 percent in 2040 (EIA, 2014a).
               1   3
               c
               o
                             History
2012
Projections
                                                           Natural gas
                                                              'u clear
                            35%


                            16%

                            16%
                    0
                                                  Oil and other liquids
                            132%

                            1%
                          2040
                     1990      2000     2010     2020     2030
Figure 9 4  Electricity generation by fuel, 1990-2040 (trillion kilowatt-hours). From EIA
       (2014a).
       To analyze the impact of different fuel sources on water withdrawals and water
consumption for future electricity generation, the base and future scenarios were modeled using
the Regional Energy Deployment System (ReEDS), an electricity model developed by the
National Renewable Energy Laboratory (NREL) (Clemmer et al., 2013). The electricity
portfolios within the model take into account existing state and federal regulations, and the
relative economics of different electricity generating-technologies (Clemmer et al., 2013).
ReEDS is a linear optimization program that analyzes the future impacts of different electricity-
generating technologies on water withdrawals and consumption, along with carbon emissions,
electricity and natural gas prices in the U.S. (Clemmer et al., 2013). The major future electricity
generation technologies within the model include supercritical coal and integrated gasification
       36 See Chapter 2 for further details regarding pending GHG standards for new and
existing electric generating units.
                                          214

-------
combined cycle coal (IGCC), natural gas combined cycle (NGCC), natural gas combustion
turbines (steam/Rankine cycle), fossil fuels with carbon capture and storage (CCS), nuclear,
hydropower, wind, solar photovoltaic (PV), concentrating solar power (CSP), geothermal power
and biopower with storage (Clemmer et al., 2013). When developing future scenarios, the cost
and performance estimates for different electricity generating-technologies need to be taken into
consideration. Key estimates for the ReEDS model are shown in Table 9.4. The U.S. Energy
Information Administration (EIA) estimates that electricity demand in the U.S. will grow 0.8
percent per year between 2010 and 2050 (EIA, 201 la).
     Table 9.5  Electricity modeling scenarios
Scenario
1) Reference Case
2) Carbon budget, no technology
targets
3) Carbon budget and
higher nuclear and
coal with carbon capture and
storage (CCS)
4) Carbon budget and
higher energy efficiency
and renewable energy
Key Assumptions and targets
Existing state and federal regulations
Electricity sector contribution to a 170-Gt C02 eq US carbon
budget through 2050
29% nuclear generation by 2035 and 36% by 2050
15% coal with CCS generation by 2035 and 30% by 2050
20% reduction in electricity use by 2035 and 35% by 2050
50% renewable generation by 2035 and 80% by 2050
     Note: Adapted from Clemmer et al. (2013)
       Because the electricity generation portfolio varies greatly in different regions of the U.S.,
the ReEDS model also determines the geographic distribution of the technologies at the regional
level based on relative economics, resource potential and electricity demand (Clemmer et al.,
2013). For water-stressed regions such as the southwest and southeast, where water is limited for
withdrawal and consumption, future energy production was used for the projected electricity
generation scenarios (Table 9.4).

   9.3.1.1  National electricity generation

       Electricity generation from coal-fired power plants is expected to steadily decrease by 37
percent between 2010 and 2050 (Figure 9.5, EIA 201 la). This decline is initially due to
announced coal plant retirements (included in the model), resulting primarily from low natural
gas prices, the implementation of EPA regulations and state requirements for energy efficiency
and renewable energy, and the higher cost of new coal plants compared with natural gas
(Clemmer et al., 2013). Nuclear generation is also projected to decline due to the estimated 60-
year lifetime for existing nuclear plants and the relatively high cost of building new plants.
Renewable energy generation is expected to more than triple by 2030, due to state renewable
electricity standards and federal tax credits. A more than six-fold increase is expected by 2050
due to projected cost reductions that make some technologies more economically competitive
(Clemmer et al., 2013). Of the renewable energy technologies, wind and solar PV are expected to


                                           215

-------
make the biggest contributions, providing 55 percent of the total generation in the U.S. by 2050
(Figure 9.5 scenario 4).
                 Scenario 1
Scenario 2
      6,000
    "S" 5,000
                                          6,000
            2010  2020   2030  2040   2050
                                                2010   2020   2030   2040  2050
                 Scenario 3
Scenario 4
      6,000
                                          6,000
Efficiency
Wind-Offshore
Wind-Onshore
PV
CSP
Geotnermal
Hydro
Biopower
                                                                            ^•i nuclear
                                                                            »»Coal-IGCC-CCS
                                                                            —
                                                                            --- Bus-bar Demand
            2010  2020   2030  2040   2050
                                                2010  2020   2030   2040   2050
Figure 9.5  National electricity generation by scenario. Adopted from Clemmer et al. (2013).
       Scenario 1, reference case; Scenario 2, carbon budget, no technology target; Scenario
       3, carbon budget with coal with CCS and nuclear targets; Scenario 4, carbon budget
       with efficiency and renewable energy targets.
   9.3.7.2  Southwest

       The southwest region analyzed by Clemmer et al. included California, Nevada, Arizona,
Utah, New Mexico, Colorado and Wyoming. The southwest currently relies on natural gas (36
percent), coal (33 percent), nuclear energy (14 percent), hydroelectric power (10 percent) and
other renewable energy sources (7 percent) (EIA, 201 la). In the reference case (Figure 9.6), non-
hydro renewable generation increases from 7 percent to more than 38 percent, while natural gas
generation grows to 39 percent (Clemmer et al., 2013). Renewable generation increases
appreciably especially in scenario 4 that shows an increase to over 95 percent by 2050. With
scenario 4, geothermal generation utilizing enhanced geothermal systems technology (EGS)
                                            216

-------
provides a much larger share of renewable generation (36%) as it becomes more economically
viable (Clemmer et al., 2013; EERE, 2007).
                Scenario 1
      Scenario 2
      900
           2010   2020  2030   2040

                Scenario 3
2010  2020   2030   2040  2050

      Scenario 4
Wind-Offshore
Wind-Onshore
PV
CSP
Geothemial
Hydro
Biopower
                                                                           ^  •
                                                                           --- Bus-bar Demand
           2010   2020  2030   2040  2050
                                              2010  2020   2030  2040   2050
Figure 9.6  Electricity generation in the southwest by scenario. Adapted from Clemmer et al.
       (2013). Scenario 1, reference case; Scenario 2, carbon budget, no technology target;
       Scenario 3, carbon budget with coal with CCS and nuclear targets; Scenario 4, carbon
       budget with efficiency and renewable energy targets.
   9.3.1.3  Southeast

       The southeast region analyzed by Clemmer et al. included Mississippi, Alabama,
Tennessee, Georgia, South Carolina, North Carolina and Virginia. In 2010, the southeast relied
heavily on coal (47 percent), nuclear (27 percent) and natural gas (17 percent) (EIA, 201 la).
Under the reference case, gas generation is projected to provide 75 percent of the region's total
generation by 2050, as it replaces retiring coal and nuclear plants (Clemmer et al., 2013). Figure
9.7 also shows that the southeast has an increase of renewable energy generation due to the
region's relatively high intensity of solar PV, biomass and offshore wind resources. Scenario 4
shows an increase in solar PV to 28 percent, biomass to 26 percent and offshore wind to 15
percent by 2050.
                                            217

-------
                Scenario 1
     Scenario 2
                 2020  2030   2040

                Scenario 3
                                 2050
2010   2020   2030

     Scenario 4
                                                                 2040
i  i Efficiency
  Wind-Offshore
^•Wind-Onshore
  PV
  CSP
  Geothermal
  Hydm
^H Biopoxver
SSNGas-CC-CCS
^:,-
•M Nuclear
«*»Coal-IGCC-CCS
^
	Bus-bar Demand
           2010   2020  2030   2040   2050
                                                2010   2020   2030   2040   2050
Figure 9.7  Electricity generation in the southeast by scenario. Adapted from Clemmer et al.
       (2013). Scenario 1, reference case; Scenario 2, carbon budget, no technology target;
       Scenario 3, carbon budget with coal with CCS and nuclear targets; Scenario 4, carbon
       budget with efficiency and renewable energy targets.

9.3.2  Water Demand Distribution and Water availability
       According to the US Geological Survey, in 2005, power plants accounted for 41  percent
of total freshwater withdrawals (USGS, 2005a). In 2005, thermoelectric power plant cooling
processes accounted for 49 percent of all water withdrawals (freshwater and saline water) in the
U.S., compared to 31 percent for agricultural uses and 11 percent for public supply (Kenny et al.
2009). Water withdrawals from thermoelectric power maintained a steady trend from 1985 to
2005 (Figure 9.8). Focusing on water consumption (consumptive use), the U.S. electric sector
constituted about 3 percent of the national total in 1995, compared to more than 75 percent for
the agricultural sector and 12 percent for public supply (Solley et al., 1998).
                                            218

-------
               300
  250
ro
•o
S 200
Q.


t150

                50
       ,1
                                                                       300
                                                                       250
                                                                          1
                                                                           E
                                                                       150
100 =
   £
50

0
                    1950 1955 1960 1965  1970 1975 1980 1985 1990 1995 2000 2005
                           Thermoelectric power withdrawals    Population
Figure 9.8 Thermoelectric-power water withdrawals 2005. Adopted from USGS (2005a).

       The most important determinant for water demand variation is the choice of cooling
technology (Macknick et al., 2011). As mentioned in Chapter 4, conventional thermoelectric
generation uses two types of cooling systems: recirculation systems with cooling towers and
once-through systems without cooling towers. Both systems use "wet cooling" of process steam
via heat exchangers and can utilize either saline water or freshwater for cooling (NREL, 2012).
Cooling systems can thus be classified under four general types based upon the source of cooling
water and the type of cooling system utilized: once-through cooling using freshwater (OTF),
once-through cooling using saline water (OTS), re-circulation cooling using freshwater (CCF)
and re-circulation cooling using saline water (CCS) (NREL, 2012). The  five states with the
highest water withdrawals for each of the four general system types are shown in Tables 9.5 and
9.6.
     Table 9.6  The top 5 states with the highest thermoelectric-power water withdrawals for
               once-through cooling type in 2005
Withdrawals for once-through cooling in million gallons per day
Groundwater
State
HI
OH
IA
AZ
IN
Fresh
25.3
16.7
13.6
5.13
4.27
%
28.0%
18.5%
15.1%
5.7%
4.7%
State
HI
FL



Saline
1450
3.26



%
99.8%
0.2%



Surface water
State
IL
Ml
TN
OH
TX
Fresh
11800
9140
8750
8550
8180
%
9.3%
7.2%
6.9%
6.7%
6.4%
State
CA
FL
MA
NJ
NY
Saline
12600
11300
5940
5190
4880
%
22.4%
20.1%
10.6%
9.2%
8.7%
   Note: Adopted from USGS (2005).
                                          219

-------
     Table 9.7 The top 5 states with the highest thermoelectric-power water withdrawals for
               recirculation cooling type in 2005
Withdrawals for recirculation cooling in million gallons per day
Groundwater
State
LA
TX
AZ
MS
MO
Fresh
97.4
55.8
45.3
35.4
17.7
%
23.2%
13.3%
10.8%
8.4%
4.2%
State
UT
NJ
AL
AK
AZ
Saline
4.18
0.01
0
0
0
%
99.8%
0.2%
0.0%
0.0%
0.0%
Surface water
State
KY
PA
LA
NC
TX
Fresh
2670
2320
1670
1660
1450
%
17.3%
15.1%
10.8%
10.8%
9.4%
State
NJ
FL
TX
MA
DL
Saline
273
140
35.2
14.8
8.4
%
57.9%
29.7%
7.5%
3.1%
1.8%
   Note: Adopted from USGS (2005)

       In 2008, about 43 percent of thermoelectric generators in the U.S. used once-through
cooling, 56 percent used recirculating and 1 percent used dry-cooling. Including renewable
energy generations, 30 percent of the total electricity generation in the U.S. involved once-
through cooling, 45 percent recirculating cooling, and 2 percent dry-cooling (UCS, 2013). In
2010, referring to Chapter 4-Figure 4.2, the total electricity generation increased to 58.2 percent
for recirculating systems and reduced to 41.6 percent for once-through cooling systems.
Generally, less water is required for withdrawal when cooling water is recycled through cooling
towers or ponds compared to once-through cooling. Consumptive water loss (i.e., loss through
evaporation) for recycled cooling systems is approximately 60 percent of the total water
withdrawal for that method, but accounts for only two percent of total water withdrawal for
once-through cooling (Solley et al., 1998).
       Sources of energy and cooling technologies used to generate electricity determine the
quantities of water consumption (Macknick et al., 2011). The water demand projection for power
plant cooling can be estimated using electricity generation (by source of energy and cooling
technology type) and average water use (withdrawals or consumptions) in gallons per unit of
electricity (Kilowatt-hour) for each type, as shown in Figures 9.9 and 9.10. The data in Figures
9.9 and 9.10 were calibrated with the ReEDS estimated dispatch of a historical year (2006) to
estimate regional variations and calculate consumption values. From this calibration, the water
withdrawal and consumptive factors for each cooling technology were found to vary widely
between regions (NREL, 2012).
       As reported in Chapter 4.4.2, the volume of water withdrawn in coal plants  operating
once-through systems in 2010 was, on average,  30 times greater than those operating
recirculating systems. The comparison was in close agreement with that found in the database in
2006 shown in Figures 9.9 and 9.10.
                                          220

-------
^ O)



?!
*o
(0 r-
K O)



1 £
                                            withdrawal rates for tecirculating-. dry-

                                            cooling, and non-thermal systems are

                                            shown on the right axis
                                                                              I
                                                                              —
                                                                              7)
                                                                              D)
                                                                              1
                                                                              M

                                                                              e
                                        Cooling System
Figure 9.9  Overview of water withdrawal factors by technology. Adopted from NREL

       (2012).
                 1,200
                                          Cooling System



Figure 9.10 Overview of water consumption factors by technology. Adopted from NREL

       (2012).
                                            221

-------
   9.3.2.1  Water demand projection for future electric power generation scenarios

       Changes in future electricity generation in the U.S. will have important implications for
future water use. Changing water availability due to competing demands and climate variability
has been taken into account in water demand projections for electricity generation (Clemmer et
al., 2013).
       The ReEDS model calculates results of future sources of electricity generation with low-
carbon emissions. The energy portfolio results will include both thermoelectric and renewable
energy according to the scenarios (Table 9.4). The future electricity generation model also
incorporated changes in energy costs, technologies, policies and regulations that impact future
electricity generation planning in the U.S. (Clemmer et al., 2013).
       The choice of models and model assumptions are important. The cost and performance
assumptions for different electricity-generating technologies used in the reference scenarios were
retrieved from EIA's Annual Energy Outlook 2011 (Clemmer et al., 2013; EIA, 201 la). The
Assumption for Annual Energy Outlook (AEO) 2011 was based on laws and regulations in effect
before October 31, 2010, and was used in the National Energy Modeling System (NEMS) to
generate the projections in the AEO 2011 (EIA, 201 Ib).
       The ReEDS model can produce results for 134 power control authorities (PCAs) (shaded
areas) for all electricity-generating technologies, and for 356 wind and concentrating solar power
(CSP) resource regions (see Figure 9.11). ReEDS has a higher spatial resolution than the NEMS
model. Regions can be aggregated up to the state level, the regional transmission organization
(RTO), the North American Electricity Reliability Council (NERC) regional level (darker black
lines), or for the three major electricity interconnections (red lines). Thus, ReEDS offers greater
resolution for analyzing water impacts at any relevant geographic scales (Clemmer et al.,  2013).
       Clemmer et al., 2013 applied the ReEDS model to determine which types of electricity
generating-technologies at the national and regional levels would meet the carbon budget and
technology target. These selections are then used to calculate the impacts on national water
withdrawals and consumption from the electricity sector. The model also takes into account the
impact on electricity and natural gas prices. However, according to Clemmer et al., 2013, the
produced results are specifically for national, southwest and  southeast regions, and did not
mention the electric dispatch between interconnected regions.
       From the reference scenario's results in the ReEDS model (Clemmer et al., 2013),
national water consumption (consumptive use captures evaporative losses from the cooling
process) increases slightly (0.6 percent) by 2030 as increased electricity demand is met primarily
with natural gas combined cycle plants with no substantial change in coal and nuclear
generation.
       Water consumption in 2050 of the base  scenario, is 34.2 percent lower than the 2010
levels, as coal and nuclear generation is substantially reduced and replaced with natural gas and
renewable generation (Clemmer et al., 2013). Scenario 4 (see Table 9.4) in the ReEDS model
predicts a substantial reduction in water consumption in 2050 due to a reduction in electricity
demand and increased penetration of renewable technologies, which decreased to 85.2 percent
from 2010.
                                          222

-------
                  Interconnect
            n  RT°
                  PCA
                  Wind/CSP Region
Figure 9.11  ReEDS modeling regions including the Interconnect, Regional Transmission
       Organization (RTO), Power Control Authorities (PCA), and Wind/Concentrating Solar
       Power (CSP) Region. Adopted from Clemmer et al. (2013).
9.3.3   Adaptation of electricity generation water use

   9.3.3.1  Water conservation

Water conservation through cooling technologies
       Most thermoelectric plants use water in steam cooling systems and the water loss from
evaporation accounts for most of the water consumption at the majority of thermoelectric plants
in the U.S. (Ciferno et al., 2010). As mentioned in Chapter 4, the two main methods of cooling
are once-through systems and wet-recirculating (closed loop) systems. New power plants use
once-through cooling because  of the significant amount of water withdrawals and the disruptions
to the local ecosystems (EIA 2014b) and to comply with EPA ecological flow and discharge
temperature regulations.
       In 2010, the U.S.  Geological Survey (USGS) initiated a study to estimate water
consumption by thermoelectric power plants as part of the USGS National Water Use
Information Program and the agency's broader mission to provide scientific information to
manage U.S. water resources (Diehl et al., 2013). Macknick, Newmark and Hallett from the
National Energy Technology Laboratory also have developed water consumption and water
withdrawal factors for the operational cycle across and within fuel technologies and for differing
cooling technologies. They found that concentrating solar power technologies and coal facilities
with carbon capture and sequestration capabilities have the highest water consumption when
                                          223

-------
using recirculating cooling systems. Non-thermal renewables such as photovoltaic and wind have
the lowest water consumption in cubic meter per megawatt hour (Macknick et al., 2012). Details
of operational water consumption and withdrawal factors are depicted in Figures 9.9, 9.10, and
9.12, and in Table 9.3. As Figure 9.12 shows, the cooling system employed has a greater impact
on water consumption than the particular technology used for generating electricity (Macknick et
al., 2012).
       Dry cooling is another option for reducing water-consuming cooling technology. Water
usage in cooling systems can be reduced by two to ten times (see Figure 9.12) when utilizing dry
cooling technology instead of a re-circulating cooling system. Thus, the choice of cooling system
plays a significant role in the planning and development of any future thermoelectric generation
capacity (Macknick et al., 2011).

Water conservation during high demand periods using hybrid technologies
       Dry cooling systems, known as air cooled condensers, are cooling systems that condense
steam and transfer the waste heat to the atmosphere without the consumption of water. Dry
cooling systems use air instead of water to cool the steam exiting a steam turbine, which can
decrease total water consumption by more than 90 percent (UCS, 2013). All of the heat rejected
from the steam is absorbed in the form of sensible heat gain in the ambient air (Stultz and Kitto,
1992). The drawback of the dry cooling system is high energy consumption that might increase
costs or decrease efficiency (Macknick et al., 2012). Dry cooling systems consume a large
amount of energy because the systems require air to be passed over the steam by one or more
large fans. This requires a significant amount of electricity (EIA, 2014b). Thus, the dry cooling
systems are less suited for large plants that use a substantial amount of steam such as those
powered by coal or nuclear energy (EIA, 2014b).
       By combining wet and dry cooling  systems, the hybrid-cooling system has potential for
water conservation and energy-efficient production. Hybrid cooling systems use water for
cooling during summer months and air for cooling during cooler months, or operate in unison,
which increases overall cooling efficiency (WRA, 2008). Hybrid technologies are more likely to
be used with certain generation technologies, such as Concentrating Solar Power (CSP) trough,
CSP tower and Geothermal Binary (Figure 9.12). The water  consumption reduction for hybrid
cooling systems from re-circulation cooling accounts for 63 percent and 78 percent for CSP
trough and CSP tower, respectively (Figure 9.12).

   9.3.3.2   Water reuse and replacement

Saline water use for cooling systems
       Of the available cooling technologies (Table 9.7), recirculation cooling systems with
saline water and once-through cooling systems with saline water are alternatives that should be
considered for freshwater replacement in coastal regions. The top five states that withdraw
surface saline water for the once-through cooling processes are California,  Florida, Maryland,
New Jersey and New York (Table 9.5) (USGS, 2005). For re-circulation cooling processes, the
top five states that use saline water are New Jersey, Florida, Texas, Maryland and Delaware
(Table 9.8) (USGS, 2005).
                                          224

-------
                                   Recirculating
                                     Cooling
Once-through
  Cooling
 Pond
Cooling
 Dry          Hybrid  No Cooling
Cooling        Cooling   Required
to
to
fc-
•

CSP
Tow
CSP
nrough G
few
>0
O
4

Bo-
pi«f
Skn
iP B
• .T-
Bo
J
O
$
Usual
Gas
?- Svorr
•
«as NXUr
Con
c>

t
$
>
NSIWd
Gas
Comored
Cycle
«CCS
tfG& 0
ved Gen
ft

»*
';
&

C . 1
Sub-
DM
Co
Sun
m crc
ere
Cod
iGCC
21
er- &
a su,
:••
w<
<
1

Coil
Sut-
cilea
wCCS
ri &
Ml- IG
01 vvl
CS
I
f






o o^
lo!o$^
0
Geo-
f'fisi
1
^ »
cc SH
cs
Gas
!.l'i ' r| . .!
•a 0/cte
3ni Nai
a
SB
Cc
Site
r,l
:,
mi
a
•a
c
ai

O
1
Cod
ftjdw
a p
KT- S
aJ


*
•I
4


<

,s

0
>


-------
       In 2005, thermoelectric-power withdrawals made up an estimated 56,700 million gallons
per day of saline water from surface water and 1,450 million gallons per day of saline water from
groundwater (USGS, 2005b). The total  saline withdrawals from both surface and groundwater
accounted for 29 percent of the total water withdrawals for thermoelectric-power in the U.S.
(USGS, 2005a).
       The use of high salinity makeup water for a cooling tower typically imposes a loss of
operating efficiency of four to eight percent. Due to the requirement of corrosion-resistant
construction materials, the cost in construction can increase 35 to 50 percent compared to
freshwater towers of comparable cooling capability (Maulbetsch and DiFiippo, 2010).
     Table 9.8  Cooling-system types used to classify plant by cooling system technology*
EIA Cooling Type
Dry(air) cooling system
Hybrid cooling systems
Once-through cooling systems
Recirculating with cooling
systems
Cooling System Type
Air cooling systems
Hybrid: recirculation cooling pond(s) or canal(s) with dry cooling
Hybrid: recirculating with forced draft cooling tower(s) with dry cooling
Hybrid: recirculating with induced draft cooling tower(s) with dry cooling
Once through with cooling pond(s) or canal(s)
Once through, freshwater
Once through, saline water
Recirculating with cooling pond(s) or canal(s)
Recirculating with forced draft cooling tower(s)
Recirculating with induced draft cooling tower(s)
Recirculating with natural draft cooling tower(s)
   Note: Adopted from (EIA, 2010).
Treated wastewater for cooling systems
       As the availability of freshwater for cooling processes in thermoelectric power
production becomes increasingly limited, alternative sources of water for power plant cooling are
of interest for both existing and future power plants (Vidic et al., 2009). Reclamation of
wastewater for cooling can use millions of cubic meters of freshwater especially in semi-arid and
arid areas that require extensive wastewater reuse. Reused wastewater for power plants, if
properly planned, may provide an economically efficient solution for water shortage situations.
       According to recorded inventories of U.S. power plants, a total of 38 power plants across
15 states use reclaimed water in the cooling water system (Vidic and Dzombak 2009).  The power
plants that use treated municipal wastewater are located in Arizona, California, Colorado,
Florida, Iowa, Massachusetts, Maryland, Minnesota, New Jersey, Nevada, Oklahoma,  Oregon,
Rhode Island, Texas and Virginia (Vidic and Dzombak 2009). In 2012,  amongst 5,400 power
                                          226

-------
plants, 60 plants used reclaimed water for cooling systems, which were found mostly in
California, Florida, Texas and Arizona (Cooper, 2012; Veil, 2007).
       Among all possible sources of impaired water that could potentially be used in power
production, secondary treated municipal wastewater is the most common and widespread source
in the U.S. Therefore, particular attention is given to a comprehensive analysis of the quantities,
availability and proximity of this impaired water for use in existing and future power plants
(Vidic and Dzombak 2009).
       Vidic and Dzombak (2009) studied the use of wastewater in recirculating cooling water
systems at thermoelectric power plants. Their evaluation included an assessment of water
availability based on proximity and relevant regulations, as well as the feasibility of managing
cooling water quality with traditional chemical management schemes. Their feasibility study
includes chemical treatment to prevent corrosion,  scaling and biofouling.
       Their assessment in 2007 revealed that 81  percent of power plants proposed for
construction would have a sufficient cooling water supply from one to two publicly owned
treatment works (POTW) within a  10-mile radius, while 97 percent of the proposed power plants
would be able to meet their cooling water needs with one to two POTWs within 25 miles.
Moreover, 75 percent of the existing thermoelectric power plants in 2007 would have a sufficient
cooling water supply from one to two POTWs within a 25 mile radius.
       While there are no federal regulations specifically related to impaired water reuse, a
number of states have introduced regulations. Veil (2007) summarized that only nine states had
regulations or requirements for industrial water reuse activities. These states are California,
Florida, Hawaii, New Jersey, North Carolina, Oregon, Texas, Utah and Washington.

9.4   Summary of U.S. GAO Report on Climate Change Energy Infrastructure
      Risks and Adaptation

       This section summarizes major findings related to the water-energy nexus that are
contained within the 2014 U.S. Government Accounting Office (GAO) report entitled "Report to
Congressional Requesters - Climate Change Energy Infrastructure Risks and Adaptation Efforts"
(GAO, 2014). According to the U.S. GAO, the energy sector's demand for water will
increasingly compete with rising demands from the agricultural and industrial sectors, among
others (GAO, 2014). EPA found that water from snowpack declined for most of the western
states from 1950 to 2000, with losses at some sites exceeding 75 percent (EPA, 2010). Annual
stream flows are expected to decrease in the summer for most regions and drought conditions
have become more common and widespread over the past 40 years in the Southwest, southern
Great Plains and Southeast (GAO, 2014). Moreover, groundwater resources are being depleted in
multiple regions (USGCRP, 2013; USGS, 2013). Research by the Electric Power Research
Institute (EPRI) indicates that approximately 25 percent of existing electric generation in the
U.S. is located in counties projected to be at high or moderate water supply sustainability risk in
2030 (EPRI, 2011). According to GAO (2014), the USGCRP's 2009 studies suggest that every
one percent decrease in precipitation results in a two to three percent drop in stream flow. In the
Colorado Basin, such a drop decreases hydropower generation by three percent.
       Hydroelectric generation is a major source of electricity in some regions of the U.S.,
particularly the northwest, and is highly sensitive to changes in precipitation and river discharge
                                          227

-------
(GAO, 2014). Rising temperatures can reduce the amount of water available for hydroelectric
generation due to increased evaporation (USGCRP, 2009; GAO, 2104). Increased evaporation
rates or snowpack changes can affect both the volume and timing of water available for
hydroelectric generation (GAO, 2014). Water is also required for coal and uranium mining, the
extraction and refining of petroleum and natural gas, and for biofuel energy crop production
(GAO, 2014).
       To develop adaptation strategies for infrastructures potentially impacted by changes in
precipitation patterns and drought, GAO cited specific examples of technological options to
improve climate resilience, including: enhancing restoration technologies and practices to
maintain or expand regional wetlands and other environmental buffer zones, increasing the
efficiency of electric generation through integration of technologies with higher thermal
efficiencies than conventional coal-fired boilers, and improving water reservoir management and
turbine efficiency for more efficient hydroelectric generation (GAO, 2014; U.S. DOE, 2013).
GAO (2014) also summarized the federal government's role in energy infrastructure as it relates
to water resources management (See Table 9.9).
       Because electricity generation infrastructures are vulnerable to severe weather that  can
interrupt operations, the ability to adapt to water supply changes is especially necessary for
plants that rely on water resources (GAO, 2014). This adaptive capacity is built through water
and natural resource governance that invest in infrastructures to provide increased water storage,
such as dams and reservoirs, desalination plants, wastewater recycling facilities, groundwater
wells and urban storm water drainage systems (Smith and Barchiesi, 2013).
       U.S. DOE Office of Fossil Energy and the National Energy Technology Laboratory
(NETL) are developing advanced water management technologies applicable to fossil fuel  and
other power plants  in three specific areas (GAO, 2014; Ciferno et al., 2010):
1.  Nontraditional  sources of process and cooling water to demonstrate the effectiveness of
   utilizing lower  quality water for power plant needs.
2.  Innovative research to explore advanced technologies for the recovery and use of water from
   power plants.
3.  Advanced cooling technology research that examines wet, dry and hybrid cooling
   technologies.
       These initiatives can help advance the adaptive efforts that private companies are making
to incorporate less water-intensive technology (GAO, 2014).
Infrastructure Adaptation in Las  Vegas
       To demonstrate water shortage adaption, a power plant in Las Vegas and  its application
of dry cooling technology provides a valuable example. The plant's dry-cooled technology at
Silverhawk Power  Station, located 35 miles north of Las Vegas, supports the water agency's
conservation efforts by using 90 percent less water than a typical water-cooled plant. The facility
also incorporates strict emission limits and the Best Available Control Technology for air
quality. As a result, Silverhawk meets stringent air quality requirements, and will increase the
availability of electric power to southern Nevada (SNWA, 2014).
       Solar Power application is another example of infrastructure adaptation. The Southern
Nevada Water Authority (SNWA) has incorporated various photovoltaic (PV) technologies into
its water system  operations. Solar panels provide covered parking at both the River Mountains
                                          228

-------
Water Treatment Facility and the Alfred Merritt Smith Water Treatment Facility, which
produces a total of 308 kilowatts of clean energy (SNWA, 2014).


      Table 9.9 Summaries of selected federal roles in energy infrastructure for water
                  management
    Summaries of Selected
           Federal
                        Roles in Energy Infrastructure
 Agency
 Key activities related to
 energy infrastructure
Federal Energy Regulatory Commission (FERC)
Hydropower
• Issues licenses for the construction of new hydropower projects and for the
continuance of existing projects (relicensing)
• Oversees ongoing project operations, including dam safety inspections and
environmental monitoring
 Agency
 Key activities related to
 energy infrastructure
Environmental Protection Agency (EPA)
• Regulates waste discharges into U.S. waters for discharge and treatment wastewater
from power plants, petroleum refineries, and oil and gas extraction facilities
• Regulates cooling water intake structures for power plant cooling systems
• Prevents contamination of underground drinking water resources from underground
wells associated with natural gas and oil production
Source: GAP, 2014	
 Agency
 Key activities related to
 energy infrastructure
U.S. Department of the Interior Bureau of Reclamation (Reclamation)
Hydropower
• Assist in meeting the increasing water demands of the West while protecting the
environment and the public's investment in these structures.
• Emphasis on fulfilling water delivery obligations, water conservation, water recycling
and reuse, and develop partnerships with customers, states, and Native American
Tribes, and find ways to bring together the variety of interests to address the competing
needs for limited water resources.
Source: Bureau of Reclamation,  2013	
 Agency
 Key activities related to
 energy infrastructure
Bureau of Land Management (BLM)
• Provide sites for new modern transmission facilities needed to deliver clean power to
consumers
• Review and approve permits and licenses from companies to explore, develop, and
produce both renewable and non-renewable energy on Federal lands.
Source:  BLM, 2014	
 Agency
 Key activities related to
 energy infrastructure
U.S. Army Corp of Engineers (USAGE)
• Expand the usage of alternative fuels, such as biofuels, in vehicles and vessels
• Complete energy and water audits
• Implement energy and water conservation measures identified by the audits
• Develop balanced and informed assessments of the safety of dams and evaluate,
prioritize and justify dam
Source: USAGE, 2014	
    Note:  Adapted from Appendix II of GAO (2014).
                                                 229

-------
9.5   Summary

       Due to prolonged droughts and population growth, there are concerns over water supplies
to sustain agricultural requirements, municipal water use and energy production technologies that
rely on water cooling and hydropower. Many demands and various consumers competing for
water result in frequent over-allocation of water resources.
       To support the sustainability of water supplies, water conservation programs together
with operationalized water activities, such as recycling and water reuse including water
infrastructure adaptation and water policies, were initiated. For example, the Southwest Nevada
Water Authority employs a multi-faceted approach to reduce the drought exposure and create
diverse and flexible water resources (SNWA, 2012). Their primary approaches to water resource
management include implementing a water conservation program, treating and reusing
wastewater, and extending water supplies.
       Over the next several decades, increasing energy demand is expected to aggravate water
shortage issues, especially the availability of freshwater for thermoelectric cooling (U.S. DOE,
2008). Moreover, while renewable energy sources reduce our reliance on coal-fired power plants
that emit climate-changing greenhouse gases (GHG),  some heavily rely on water sources.
Hydropower production, for example, is sensitive to total runoff and reservoir levels.
       Droughts have forced utilities in the Southwest to consider the use of new cooling
technologies and sources of water to cool thermoelectric plants (Walton, 2010). Several projects
are exploring the potential use of non-traditional sources of process and cooling water. The use
of wastewater for thermoelectric cooling is increasing in the Southwest although the supply of
urban wastewater may also be decreasing in some locations due to conservation policies
(Walton, 2010). Municipal wastewater still appears to be the impaired source of water most
likely to be locally  available in sufficient and reliable quantities to provide cooling for
thermoelectric generation (U.S. DOE, 2010). Using reclaimed water for cooling systems has
several advantages, as it helps to save potable water and provides a reliable supply.
       Available water savings in thermoelectric plants can be achieved via air cooling, through
the use of non-traditional or impaired water sources, by recycling of plant wastewater and by
increasing plant thermal efficiency (WNA, 2014). Advanced cooling technologies can also
provide alternative  approaches to reduce water consumption. For example, as discussed in
Chapter 4, plants can implement condensing modules within the cooling towers or apply
filtration methods to prevent scaling and increase the cycles of concentration. Other methods to
reduce water consumption  in cooling systems include the use of an air-cooled condensers (ACC),
as well as the use of ice thermal storage within coal-fired cooling systems, which cools the
intake-air to gas turbines. Furthermore, technologies that can recover usable water from
alternative sources, such as water from the flue gas emitted by coal-fired power plants, can also
help reduce water consumption. Discussed in Chapter 4, these technologies include liquid
sorption technologies, flue gas sorption membranes and condensing heat exchangers.
       Another approach for reducing freshwater consumption in coal-fired power plants, as
mentioned in Chapter 4, is  to enhance the fuel and improve plant efficiency. IGCC plants need
approximately one-third of the engineered cooling when compared with conventional coal
thermoelectric plants (WNA, 2014).

                                          230

-------
       Using dry cooling systems or hybrid systems is a possibility for water-scarce regions
when integrated with energy sources such as solar power and geothermal energy. Hybrid cooling
systems are more applicable for cooling when utilizing an air-cooled condenser (ACC) for heat
load rejection during summer months, which increases overall cooling efficiency.
       Facilities can also consider saline water as a potential water source. Desalination is
gaining attention in solar power plants in the southwestern area. Most desalting facilities  are
found in California and coastal regions. There is a possibility for a dispatch being shifted towards
coastal areas and regional interconnects due to the abundance of salt water sources. In the Las
Vegas case, the SNWA struck an exchange agreement with California to invest in desalination
facilities in California in exchange for the use of a portion of California's Colorado River
apportionment.
       Water usage is influenced mostly by the cooling water demands for thermoelectric
generation. The generation technologies that use non-renewable energy sources of coal, natural
gas and nuclear power tend to require substantially more water withdrawal and water
consumption than technologies that use renewable energy sources. Of the renewable energy
sources, two generation technologies (photovoltaic and wind) used the least amount of water for
cooling.
       While a variety of sources have been used to supplement or replace freshwater for
cooling systems, there are limitations for those  applications. For example, a dry cooling system
can decrease a power plant's efficiency. Moreover, the low quality of impaired water can cause
the cooling system to be less effective, and a power plant's  distance from a wastewater source
can make using impaired water infeasible.
       Local efficiency can greatly improve if power sources and water resources are in  a
proximity to the electric power plants, or if they are regionally interconnected. Availability of
these resources may prove to be a significant factor of new  power plant capacity to be built at or
near existing facilities.
       In conclusion, water climate, and energy issues are closely interrelated and cannot be
addressed in isolation (Vidic et al., 2009). As both population and energy demand continue to
increase in  the U.S., freshwater scarcity will become a critically important issue. Both impaired
and saline waters have potential to serve as alternative water sources to help meet future
thermoelectric cooling demand (Vidic et al., 2009).  There is already some experience with the
use of impaired water for thermoelectric cooling.  Examples include the use of treated municipal
wastewater and the use of seawater in coastal. There will be an increasingly urgent need to find
alternative water resources to replace freshwater demand for thermoelectric cooling purposes,
particularly in  water-stressed regions of the U.S.
       Additionally, more research is needed to assess the regional and local water impacts of
different types of electricity generation, and to analyze the water impacts of electricity-sector
choices. More  studies of viable energy resources and the impacts of geographical limitations may
be useful in adapting the use of water locally in the generation sector. Finally, a review of the
limitations  of state and federal regulations on impaired water use and a series of feasibility
studies of technologies to facilitate the use of impaired water are recommended.
                                           231

-------
9.6  References

Bureau of Land Management (BLM). 2014. "Energy." U.S. Department of Interior.
      http://www.blm.gov/wo/st/en/prog/energy.html. Accessed August 2014.

Bureau of Reclamation. 2013. "Mission."  U.S. Department of Interior.
      http://www.usbr.gov/main/about

Ciferno, J.P., R.K. Munson, J.T. Murphy. 2010. "Department of Energy/National Energy
      Technology Laboratory's Water-Energy Interface Research Program: December 2010
      Update."
      http://www.alrc.doe.gov/technologies/coalpower/ewr/water/pdfs/Water Program Overvi
      ew.pdf Accessed August 2014

Clemmer, S., J. Rogers, S. Sattler, J. Macknick, and T. Mai. 2013. "Modeling low-carbon US
      electricity futures to explore impacts on national and regional water use." Environmental
      Research Letters, 8 (1):015004. doi: 10.1088/1748-9326/8/1/015004.

Cooper, I, 2012. "Municipal wastewater for cooling towers opportunities and obstacles."
      Golden Associates.

Diehl, T.H., M.A. Harris, J.C. Murphy, S.S. Hutson, and D.E. Ladd. 2013. "Methods for
      estimating water consumption for thermoelectric power plants in the United States." U.S.
      Geological Survey Scientific Investigations Report 2013-5188, 78 p.,
      http://dx.doi.org/10.3133/sir20135188

EIA. 2010. "Form EIA-860 database, Annual  electric generator report."
      http://www.eia.gov/electricity/data/eia860/index.html. Accessed August 2014.

EIA. 201 la. "Annual Energy Outlook 2011."
      http://www.eia.gov/forecasts/archive/aeol l/pdf/0383(201 D.pdf. Accessed August 2014.

EIA. 201 Ib. "Form EIA-0383 database, Assumptions to the Annual Energy Outlook 2011."
      http://pbadupws.nrc.gov/docs/ML1209/ML12090A838.pdf

EIA. 2014a. "Form EIA-0383 database, Annual Energy Outlook 2014"
      http://www.eia.gov/forecasts/aeo/er/pdf/03 83 er%282014%29.pdf

EIA. 2014b. "Today in Energy: Many newer power plants have cooling system that reuse water."
      http://www.eia.gov/todayinenergy/detail. cfm?id= 14971

EIA. 2014c. "Nevada: State Profile and Energy Estimates newer power plants have cooling
      system that reuse water."  http://www.eia.gov/state/data.cfm?sid=NV#ReservesSupply

EERE. 2007. "Enhanced Geothermal Systems."  http://energy.gov/eere/geothermal/enhanced-
      geothermal-systems-0

EPA. 2010. "Climate Change Indicators in the United States."  EPA 430-R-10-007. Washington,
      DC. EPA.

                                         232

-------
EPRI. 2011. "Water Use for Electricity Generation and Other Sectors: Recent Changes (1985-
       2005) and Future Projections (2005-2030)." Palo Alto, CA. Electric Power Research
       Institute.
       http://www.epri.com/abstracts/Pages/ProductAbstract.aspx?ProductId=00000000000102
       3676.

Energy Technology Laboratory's Power Plant-Water R&D Program. U.S. Department of
       Energy, National Energy Technology Laboratory, Pittsburgh, PA.

Kenny, J.F., N.L. Barber, S.S. Hutson, K.S. Linsey, J.K. Lovelace, and M.A.  Maupin. 2009.
       "Estimated Use of Water in the United States in 2005." Circular 1344. Reston, VA. U.S.
       Geological Survey, http://pubs.usgs.gov/circ/1344/.

Macknick, I, R. Newmark, G Heath, and K.C. Hallet. 2011. "A Review of Operational Water
       Consumption and Withdrawal Factors for Electricity Generating Technologies".
       NREL/TP-6A20-50900. Golden, CO. National Renewable Energy Laboratory.

Macknick, J., R. Newmark, G. Heath, and K.C. Hallett. 2012. "Operational water consumption
       and withdrawal factors for electricity generating technologies: a review of existing
       literature." Environmental Research Letters, 7(4):045802.

Maulbetsch, J.S., and M.N. DiFilippo. 2010. "Performance, Cost, and Environmental Effects of
       Saltwater Cooling Towers." PIER Final Project Report, California Energy Commission.
       http://www.energv.ca.gov/2008publications/CEC-500-2008-043/CEC-500-2008-
       043.PDF

Meldrum, J.R., S. Nettles-Anderson, G. Heath, and J. Macknick. 2013a. "Life cycle water use for
       electricity generation: a review and harmonization of literature estimates."
       Environmental Research Letters, 8:015031.

Meldrum, J. R., I.E. Macknick,  G.A. Heath, and S.L. Nettles-Anderson. 2013b. "Life Cycle
       Water Use for Electricity Generation: Implications of the Distribution of Collected
       Estimates."  Paper presented at the ASME 2013 Power Conference, Boston, MA.

NETL. 2009. "Impact of Drought  on US Steam Electric Power Plant Cooling Water Intakes and
       Related Water Resource Management Issues."  Washington, DC. National Energy
       Technology Laboratory US Department of Energy) DOE/NETL-2009/1364.
       http://www.netl.doe.gov/File%20Library/Research/Coal/ewr/water/fmal-drought-
       impacts.pdf.

NREL. 2012. Renewable Electricity Futures Study. Edited by M.M. Hand, S. Baldwin, E.
       DeMeo, J.M. Reilly, T. Mai, D. Arent, G. Porro, M. Meshek, and D. Sandor, D.
       NREL/TP-6A20-52409. Golden, CO. National Renewable Energy Laboratory.
       http://www.nrel.gov/analysis/re_futures/.

Raskin, P., P.H. Gleick, P. Kirshen, R.G.Jr. Pontius, and K. Strzepek. 1997. "Water Futures:
       Assessment of Long-Range Patterns and Prospects. Comprehensive Assessment of the
       Freshwater Resources of the World."  Stockholm Environment Institute. Document
       prepared for UN Commission for Sustainable Development, 5th Session.
                                          233

-------
Smith, D. M., and S. Barchiesi. 2013. "Environment as Infrastructure: Resilience to Climate
      Change Impacts on Water through Investments in Nature."  Gland, Switzerland: Union
      for Conservation of Nature (IUCN).
      http://cmsdata.iucn.org/downloads/iucn  environment as infrastructure l.pdf.

Solley, W. B., R.R. Pierce, and H.A. Perlman. 1998. Estimated use of water in the United States
      in 1995. Edited by U.S. Department of Interior, U.S. Geological Survey.
      http://pubs.usgs.gov/circ/1998/1200/report.pdf

Southern Nevada Water Authority (SNWA). 2009. "SNWA Water Resource Plan."
      http://www.snwa.com/assets/pdf/wr_plan.pdf.

Southern Nevada Water Authority (SNWA). 2010. "Water Use Fact."
      http ://www. snwa.com/consv/goal s_facts .html.

Southern Nevada Water Authority (SNWA). 2012. "SNWA Annual Report 2012"
      https://www.snwa.com/assets/pdf/about  reports annual.pdf.

Southern Nevada Water Authority (SNWA). 2014. "Sustainable Projects."
      http://www.snwa.com/env/sustain.html.

Sovacool, B.K., and K.E. Sovacool. 2009. "Preventing national electricity-water crisis areas in
      the US." Technical Report, Energy Governance Program at the Centre on Asia and
      Globalization, Lee Kuan Yew School of Public Policy at the National University of
      Singapore.

Stultz, S.C., and J.B. Kitto. 1992. Steam, its generation and use (40th ed.). Barberton, Ohio,
      Babcock & Wilcox, 131p.

Union of Concerned Scientists (UCS). 2013. "Energy and Water Uses."
      http://www.ucsusa.org/clean  energy/our-energv-choices/energy-and-water-use/water-
      energy-electricitv-cooling-power-plant.html.

U.S. Army Corps of Engineers (USAGE). 2014. "Mission Overview: Dam Safety Program."
      http://www.usace.army.mil/Missions.aspx

U.S. Climate Change Science Program (USGCRP). 2009. Second National Climate Assessment,
      Global Climate Change Impacts in the United States.
      http://www.globalchange.gov/browse/reports/global-climate-change-impacts-united-
      states.

U.S. Climate Change Science Program (USGCRP). 2013. "Draft Third National Climate
      Assessment, Climate Change Impacts in the United States." Chapter 4, Energy Supply
      and Use.

U.S. Department of Energy (U.S. DOE). 2008. Estimating Freshwater Needs to Meet Future
      Thermoelectric Generation Requirement. Report No. DOE/NETL-400/2008/1339.
                                         234

-------
U.S. Department of Energy (U.S. DOE). 2010. Department of Energy/National Energy
       Technology Laboratory's Water-Energy Interface Research Program: December 2010
       Update.

U.S. Department of Energy (U.S. DOE). 2013. "U.S. Energy Sector Vulnerabilities to Climate
       Change and Extreme Weather. Report # DOE/PI/0013.
       http ://energy. gov/sites/prod/files/2013/07/f2/20130716-
       Energv%20Sector%20Vulnerabilities%20Report.pdf.

U.S. Geological Survey. 2005a. "Summary of Estimated of Water Use in the United State in
       2005." http://water.usgs.gov/edu/wupt.html.

U.S. Geological Survey. 2005b. "Thermoelectric-power water use."
       http ://water.usgs. gov/edu/wupt.html.

U.S. Geological Survey. 2013. "Groundwater Depletion in the United States (1900-2008)"
       Scientific Investigation Report 2013-5079. Reston, VA: USGS.

U.S. Government Accountability Office (GAO). 2014. "Climate Change: Energy Infrastructure
       Risks and Adaptation Efforts." GAO-14-74. http://www.gao.gov/assets/670/660558.pdf.

Veil, J.A. 2007. "Use of Reclaimed Water for Power Plant Cooling." ANL/EVS/R-07/3, U.S.

Vidic, R.D.,  D.A. Dzombak, Ming-Kai Hsieh, H. Li, Shih-Hsiang Chien, Y. Feng, I. Chowdhury,
       and J. D. Monnell. 2009. Reuse of Treated Internal or External Wastewaters in the
       Cooling Systems of Coal-Based Thermoelectric Power Plants. Final Technical Report
       Submitted to U.S. Department of Energy National Energy Technology Laboratory.

Walton, B. 2010. "Water Scarcity Prompts Different Plans to Reckon With U.S. Energy Choke
       Point." Circle  of Blue, August 17, 2010.
       http://www.circleofblue.org/waternews/2010/world/water-scarcity-prompts-different-
       plans-to-reckon-with-energy-choke-point-in-the-u-s/.

Waggoner, P.E. 1990. Climate Change and U.S. Water Resources. John Wiley and Sons, New
       York.

Walton, B. 2010. "Water Scarcity Prompts Different Plans to Reckon With U.S. Energy Choke
       Point." Circle  of Blue, August 17, 2010.
       http://www.circleofblue.org/waternews/2010/world/water-scarcity-prompts-different-
       plans-to-reckon-with-energy-choke-point-in-the-u-s/.

Western Resource Advocates (WRA). 2008. "A sustainable paths: Meeting Nevada's Water and
       Energy Demands." http://www.westernresourceadvocates.org/water/NVenergy-
       waterreport.pdf.

World Nuclear Association (NWA). 2014. "Cooling Power Plants". WNA Information Library
       (Updated September 2014). http://www.world-nuclear.org/info/current-and-future-
       generation/cooling-power-plants/.
                                          235

-------
Yates, D., K. Averyt, F. Flores-Lopez, J. Meldrum, S. Sattler, J. Sieber, and C. Young. 2013a.
       "A water resources model to explore the implications of energy alternatives in the
       southwestern US." Environmental Research Letters, 8(4):045004. doi: 10.1088/1748-
       9326/8/4/045004.

Yates, D., J. Meldrum, and K. Averyt. 2013b. "The influence of future electricity mix
       alternatives on southwestern US water resources."  Environmental Research Letters,
       8(4):045005. doi: 10.1088/1748-9326/8/4/045005.
                                          236

-------
9.7  Appendix

      Appendix Table 9.1 Detailed summary of Southern Nevada Water Authority's
                 incentive programs
Water Smart
Landscape Rebate
Program
This program provides incentives for residential and commercial property owners to
upgrade lawn to water efficient landscaping.
The current program rebate amount is $1.50 for the first 5,000 square feet of lawn removed
and $1 for additional lawn removed up to $300,000.
 Efficient Landscape
 Irrigation Equipment
This program pays up to half the cost of replacing inefficient irrigation controllers with smart
controllers that can interrupt irrigation whenever the valley receives significant rainfall.
These controllers are capable of reducing water use by 15 to 30 percent.
Water Efficient
Technologies
Business customers who choose from proven conservation technologies that conserves at
least 500,000 gallons of water per year, qualify for a rebate of up to $150,000 per property.
Water Smart Car
Wash
The water smart car wash program is a public-private partnership that encourages residents
to use commercial car wash facilities, which recover all of their wastewater for treatment
and reuse, instead of washing their vehicles at home.
 Pool Cover Rebate
 Program
The SNWA pool cover rebate program pays up to half the cost of a swimming pool cover.
Typical use of a cover is estimated to save 13,000 gallons annually on an average-size
pool.
Water Smart
Contractor Program
Landscape contractors who participate in the program need to ensure that their project
meets specific criteria to conserve water. To obtain status as water smart contractor,
licensed landscape contractors must attend SNWA water efficiency training and pass a
proficiency exam.
Water Smart Home
The water smart home program certifies new homes as water smart, ensuring that
homeowners are purchasing a home that can save as much as 75,000 gallons of water per
year.
Water Upon Request
The SNWA and several local partners teamed up with local restaurants, which agree to
serve water only when patrons request it.
    Note: Adopted from SNWA (2009).
                                                237

-------
     Appendix Table 9.2 Detailed summary of Southern Nevada Water Authority's
                 education programs
Water
Conservation
Coalition
This coalition is a public-private partnership formed by community leaders to help increase
water-efficient practices through initiatives such as speakers bureau, Business-to-Business
Challenge and various public projects.
Water Smart
Innovations
In 2008 the SNWA with the U.S. EPA's WaterSense program hosted the inaugural
WaterSmart Innovations Conference & Expo to share information about conservation
programs and water-efficient technologies.
Conservation
Helpline
The Conservation Helpline is an information line that customers can call to obtain
conservation information or report water waste.
Publications and
Media
The SNWA regularly executes a comprehensive campaign of television, print and radio ads
that educates the community on the need for water conservation and offers help through
the SNWA website and Conservation Helpline. The SNWA also produces and distributes
publications to help customers conserve water.
Demonstration
Gardens
The SNWA promotes visits to the Springs Preserve, a 180-acre facility that offers hundreds
of examples of water-efficient landscaping, as well as classes by master gardeners and
horticulturists. The SNWA also funded conservation grants of up to $5,000 to develop
demonstration projects for their own campuses.
    University
The SNWA has partnered with the Springs Preserve to develop a comprehensive education
program known as H20 University for teachers in the Clark County School District.
Note: Adopted from SNWA (2009).
                                                238

-------
 10 Conclusions and Recommendations
       Timothy C. Keener1, Marissa S. Liang1, and Joseph McDonald11

10.1 Conclusions

       The use of hydrocarbons and other fossil fuels (natural gas, refined petroleum products,
coal, etc.) for electricity generation and transportation, respectively, are the largest sources of
energy demand and GHG emissions in the U.S. Cooling systems in thermoelectric generating
facilities and biofuel crop cultivation consume significant quantities of water, approximately 3-
4% for each of the two sectors. Each step in the production of fuels for the energy and
transportation industries, and the use of these fuels for producing electricity involves the
withdrawal, and sometimes consumption, of substantial amounts of water. For example, large
amounts of water are required for drilling, extraction, and conversion of petroleum into products
such as gasoline and diesel, which are the primary sources of fuel for the transportation industry.
Other sources of fossil fuels such  as shale or tar sands involve the use of extraction techniques
which often require large amounts of energy and water. The U.S. Department of Energy
estimates that approximately 70 - 260 million gallons of water per day is used for coal mining,
including the  water used for washing coal and cooling drilling equipment (U.S. DOE, 2006).
After washing, most coal is transported to power plants by truck, rail  or barges. In a  few cases,
however, finely ground coal is transported via pipelines as a slurry, which involves the use of
hundreds of gallons of water per megawatt-hour (MWh) of electricity produced (Meldrum,
2013).
       The majority of electricity in the U.S. is generated by combusting fossil fuels in a boiler
to produce steam, and using the kinetic energy of the steam to generate electricity using steam
turbines. In addition to the water required to produce steam, other uses of water in power plants
include:

•   Surface water withdrawals to condense steam after it passes through the steam turbine, with
    significant water consumption due to evaporation in  cooling towers;
•   Water required for scrubbing flue gases  to meet Clean Air Act regulations;
•   Water required to dispose of fly ash;
•   Water lost during desalination.
       The amount of water used in electric power plants has been found to be largely dependent
on the type of fuel used in the plants. In 2012, coal was used to generate 37% of the  total
electricity, followed by natural gas, nuclear  power, and renewable forms energy (hydroelectric,
wind, solar, geothermal) at 30%, 19%, and 12%, respectively. Overall, this report has found a
general trend  of decreasing water-intensity as carbon-intensity of electric generation is decreased
or as plant efficiency is increased.
       Over reliance on non-renewable fuels and a push for greater energy independence led to
the introduction of the Renewable Fuel Standards (RFS) under the 2005 Energy Policy Act,
which was later amended in the Energy Independence and Security Act (EISA). EISA calls for a
1   University of Cincinnati, Department of Biomedical, Chemical, and Environmental
   Engineering
11  U.S. Environmental Protection Agency - National Risk Management Research Laboratory
                                          239

-------
reduction in annual petroleum consumption by at least 20%, and an increase in the use of
alternate fuels/biofuels such as ethanol and biodiesel by 10% by 2015, and a four-fold increase
by 2022. There has been a steady increase in the demand for energy in the last century, and the
demand is expected to increase, albeit at a slower rate, in the next few decades. In addition, the
water consumption for producing the biofuels is significantly higher than for producing
petroleum fuels, and this is dominated by the irrigation requirements of biomass cultivation.
Increased use of biofuels may result in increased use of domestic agricultural,  surface, and
groundwater resources unless the water-intensity of biomass cultivation can be reduced.
Transitioning from corn cultivation to the use of agricultural residues and cellulosic energy crops
provides opportunities to reduce the water-intensity of both biomass cultivation and biofuel
production in the U.S.
       When taking into account the most recent AEO 2014 forecast trends of electrical demand,
sources of electricity, consumption of biofuels and estimations of the  expected rate of water
consumption for these sectors in the U.S., the demand of water for energy is expected to increase
over the next several decades unless new technologies that reduce water usage are  implemented
throughout the energy sector. Regulations that are expected to result in a reduction of the carbon
intensity of electric power generation (e.g., the recently proposed EPA GHG regulations
impacting new and existing thermoelectric generation)  and the eventual transition of ethanol
biofuels from cultivation of corn to cellulose (e.g.,  regulated volumes within the EPA RFS2
program) have significant potential to offset the impact of water usage from the energy sector.
       The availability of water for power generation and biomass production for biofuels will
also be affected by extraneous factors such as climate change, population growth and
redistribution, domestic consumption, land use and regulatory criteria such as minimum
ecological stream flow requirements, the Clean Air Act (CAA), and the Clean Water Act
(CWA). All of these factors directly or indirectly impact water quality and quantity in water
sources that will be used for cooling purposes or crop irrigation, thus  influencing the ways
energy production will adapt in the future. Climate changes are known to cause precipitation
variations in intensity, frequency, seasonality, and quantity, leading to variations in surface water
flow and groundwater levels, which in turn affect energy production processes. Climate change
can also change the nature of precipitation (snowfall to rainfall) and the melting of snow pack -
increasing spring runoff while reducing summer streamflow. Large-scale climatic model system
simulations indicate that elevated GHG from power plant emissions and  emissions from the
transportation sector will further intensify atmospheric-oceanic interactions, thus producing
greater impacts on climatic systems. Model simulations project 10% or more decreases in
precipitation in southwest U.S.; changes in precipitation from snow to rain, resulting in a
decrease in snow pack in the Rockies and Pacific northwest, and hence, stream flow; decreased
precipitation and increased drought in southeast U.S.; and increased precipitation and flooding
along the Great Lakes and northeast U.S. that may  be responsible for  water-borne pathogens and
nutrient runoff.
       Although there is uncertainty among different climatic models that consider different
predictive variables, some general trends tend to emerge. Limited water availability in areas most
affected by climate change, especially in areas with decreasing precipitation and increasing
drought, is expected to increase water competition  for power generation and biomass cultivation.
This is especially pertinent for future energy production planning as the potential impacts of
changing climate, population growth, and changing water quality on water usage, competition,
                                           240

-------
mandated minimum ecological stream flows, water discharge criteria, and consumption in power
plants needs to be considered in determining plant parameters. These parameters include the type
and source of fuel, type of cooling to be used, and the treatment technologies that will be used to
meet regulatory criteria. The effects of climate change on biomass cultivation are somewhat
different from power generation in that biomass cultivation is dependent on land in addition to
water-related factors mentioned above and the impact of regional differences on the irrigation
needed for biomass cultivation results in much greater regional variation in water-intensity.
       The electric generation and transportation sectors are the largest sources of emissions of
CO2, NOx,  SOx and paniculate matter in the U.S.  These two sectors are also the largest
contributors of GHG and short-term climate forcing emissions and thus are the primary
anthropogenic contributors to global warming and climate change. Irrigation and thermoelectric
power generation (especially the amount of water required to condense steam) are the largest
users of water when considering water usage in areas affected by climate change. Various direct
and indirect steps can be taken to minimize water use in thermoelectric power plants, depending
on whether the main purpose is to directly reduce water consumption or if the purpose is to
indirectly contribute to more sustainable water usage using advanced cooling technologies, water
reuse and recovery, use of non-traditional sources  for process and cooling water, and use of
advanced wastewater treatment technologies. The  use of recirculating systems (cooling towers)
with thermoelectric generation is increasing due to regulations impacting discharge temperature
and minimum ecological water flow. However, although once-through cooling systems tend to
withdraw more water than recirculation systems with cooling towers, recirculating systems
consume more water than once-through systems due to evaporative losses. This is expected to
lead to a trend of increased water consumption as newer plants are brought on line unless
consumptive losses are reduced. For example, a typical conventional coal-fired power plant
withdraws between 20,000 - 50,000 gallons of water/MWh and 500 -  1,200 gallons of
water/MWh for once-through and recirculating cooling systems, respectively. The estimates for
water consumed were between 100 - 317 gal/MWh and 480 - 1,100 gal/MWh for the two
systems, respectively. An alternative is to use air-cooled systems instead of water-cooled
systems, albeit at a loss of energy efficiency. Air-cooled systems are rarely used for coal-fired
thermoelectric generation but are used with increasing frequency for generation from natural-gas.
Other opportunities to improve energy efficiency and reduce fuel and water usage include the use
of supercritical and ultra-supercritical steam turbine plants, the use of supercritical oxy-coal
systems with staged combustion, the use of high efficiency air-cooled heat exchangers, using
condensing heat exchangers that condense water vapor in the flue gas for water makeup or reuse,
using treated effluent from wastewater treatment plants, reducing wastewater discharge and
using improved methods of flue gas scrubbing and flue gas desulfurization.
       Water use for CO2 capture and sequestration from future coal-fired power plants was
modeled as part of this report.  Methods to capture  or eliminate CO2 from the flue gas from coal-
fired thermoelectric  plants include post-combustion separation; pre-combustion separation,  in
which the fuel is decarbonized prior to combustion; oxy-fuel combustion, which uses pure
oxygen for combustion, and which has the added advantage of reducing NOx in flue gas; and
integrated gasification combined cycle (IGCC), wherein coal is converted at high temperature to
form hydrogen/CO syngas, which after purification can be burned cleanly in a hybrid gas turbine
with exhaust heat from the turbine used to provide heat energy for a steam turbine cycle. The
captured CO2 is either sequestered underneath the  ground or in oceans,  or the gas is compressed,

                                          241

-------
dried, purified and stored for industrial use. Potential knowledge gaps with respect to CO2
capture and sequestration include estimates of regional storage capacity for underground
sequestration, potential leakage rates, cost data, and remediation options among others. However,
there appear to be no insurmountable technical barriers to use geological or oceanic storage as an
effective sequestration option for CO2, thus leading to significant reductions in emissions of
greenhouse gases into the atmosphere. The use of IGCC with CO2 capture and sequestration has
potential to significantly reduce water-intensity on a consumptive basis relative to coal-fired
electric generation in part due to reduced steam cooling requirements and the potential to use
dry- or hybrid wet-and-dry steam cooling.
       The trend towards increased electric generating capacity from Natural-gas fired power
has advantages that include reduced fuel costs, reduced greenhouse gas emissions relative to
other fossil fuels, reduced criteria air pollution and air toxic emissions and the greater
accessibility and abundance of natural gas in the U.S. through the use of newer extraction
technologies such as hydraulic fracturing. Hydraulic fracturing involves drilling both vertically
and horizontally to follow a geologic formation with trapped gas, injecting a mixture of water
and chemicals at high pressure to fracture rock in the formation and allowing the gas to flow
from the fractured rock to a wellhead where it is collected and processed for future use. The use
of natural gas in electric power generation is expected to increase from 30% of total electricity
production in the U.S. to approximately 63%  in 2040. In addition to conventional, steam turbine-
based natural gas power plants that are very similar to conventional coal power plants, natural
gas can also be used to generate electricity via gas turbines as part of a natural gas turbine/steam
turbine combined cycle (also known as natural gas combined cycle or NGCC). Similar to
coal/IGCC systems, NGCC systems derive half or more of their generating capacity from the gas
turbine and thus the steam cooling requirements for the steam turbine used in these systems are
greatly reduced when compared to more conventional boiler/steam turbine systems. The reduced
cooling requirements of the steam turbine stage result in reduced evaporative losses in the case
of systems using cooling towers. The reduced steam cooling capacity can also enable the use of a
dry cooling system or a hybrid wet-and-dry cooling system for steam cooling instead of cooling
towers. The use of NGCC power plants can greatly help in reducing water consumption in water-
stressed areas, thus leading to decreasing water withdrawals and improved water quality. Most
new generating capacity from natural gas is expected to be via NGCC plants in order to meet
expected GHG regulations. Other advantages of NGCC systems include much greater flexibility
in powering the system on or off to meet electricity demand and the higher thermal efficiency of
the Brayton/Rankine  combined cycle relative to the Rankine cycle typical  of conventional
boiler/steam turbine plant designs.
       In the past, small amounts of water had been used to extract natural gas from deep
vertical wells. A single hydraulically fractured well can produce large volumes of fracture and
formation water. The quantity and quality of this produced water varies across geologic
formations. Much of the produced water is disposed of via injection wells, but increasing
quantities of produced water are being treated for reuse. Other potential disadvantages of the
hydraulic fracturing process include the potential to contaminate groundwater and chemical
waste that has to be treated and disposed.
       The amount of water withdrawn and consumed during electricity generation using coal
(with and without integrated gasification combined cycle (IGCC) and CO2 capture) and natural
gas (gas-fired steam boiler/Rankine cycle and NGCC) is listed in Table 10.1. For comparison
                                          242

-------
purposes, the amount of water withdrawn and consumed in nuclear power plants (the third most
common form of thermoelectric generation) and concentrated solar power (an emerging
thermoelectric technology) is also listed in the table. Note that the comparisons in Table 10.1 are
only for recirculating cooling systems (e.g., cooling towers). The values listed in the table cover
a wide range, and include both actual and modeled electricity generated from different types of
coal (bituminous, sub-bituminous, lignite, etc.), different ambient air and water temperatures,
and different water quality among other factors. Each of these variables affects water withdrawal
and water consumption rates to a different extent,  and the effect of each variable needs to be
considered during power plant design. For comparison purposes, water withdrawn and consumed
in thermoelectric plants with once-through cooling systems are also summarized in Table  10.2.
Note that water withdrawals for legacy plants with once-through cooling systems are one to two
orders of magnitude higher than for newer plants with recirculating cooling systems. Both the
IGCC and NGCC generation have significantly reduced water intensity on both a withdrawal and
consumptive basis than other major thermoelectric generation systems and have a potential for
further reductions in water-intensity via the use of dry cooling systems.
     Table 10.1 Water-intensity on a withdrawal and consumptive basis for thermoelectric
               generation using different sources of energy and using recirculating cooling
               systems
System
Coal
IGCC
NG (Rankine/steam-
turbine)
NGCC
Nuclear
Concentrated Solar
Power
Water withdrawals
gal/MWh*
500-1200
161-605
950-1460
150-283
793-2589
740-1110
10-4m3/MJ**
5.3-12.6
1.7-6.4
10.0-15.4
1.6-3.0
8.3-27.2
7.8-11.7
Water consumption
gal/MWh
201-1189
34-449
662-1170
130-300
581-898
555-1902
10-4m3/MJ
5.0-11.6
0.4-4.7
7.0-12.3
1.4-3.2
6.1-9.4
5.8-20.0
Chapter
4&9
9
5
5
9
9
   Note: * English units of gallons per megawatt-hour; ** SI units of cubic meter per mega-joule.

       Petroleum-derived fossil fuels represented over 95% of all transportation energy
consumed in the U.S. in 2012. Gasoline (63%), diesel fuel (22%) and jet fuel (15%) are the most
widely used types of petroleum-based transportation fuels. Biofuels contributed only 4.5% of the
total energy consumed for transportation in 2012, the majority of which (4.1% of the total
transportation energy) is ethanol blended into gasoline. A dependence on non-renewable fuels,
concern about global warming, and a push for greater energy independence led to the
introduction of the Renewable Fuel Standards (RFS), and later the Energy Independence and
Security Act (EISA), which called for a reduction in annual petroleum consumption and an
increase in the use of alternate fuels/biofuels such as ethanol and biodiesel. In 2012, ethanol
constituted 94% of all biofuel produced in the U.S. on a volume basis, and was produced
primarily from corn (the other sources being sugar cane and cellulose). Irrigation is the most
                                          243

-------
significant source of water uptake and consumption for corn ethanol production and varies
significantly from state to state. A mass- and energy- balance based model of ethanol refining
presented in Chapter 6 estimated that 0.32 kg of ethanol, 0.33 kg of dry distiller grain (can be
used as replacement for corn in livestock feed), and 2.72 kg of wastewater is produced per kg of
corn and 2.68 kg of water raw material. The water used for cooking processes (-75%) exits the
plant as water vapor but the water used for other processes (-25%) is usually  recycled during
ethanol production in modern facilities. Technology is available to build ethanol production
plants that are capable of achieving zero water discharge, if necessary. Using  lower quality
surface or gray waters for ethanol production processes is also possible, which could play an
important role in reducing the impact of ethanol processing in water-stressed regions. Research
and development should continue in the areas of zero water discharge designs, the use of lower
quality process water and the increased use of waste heat within ethanol production processes.
                                  i withdrawal and consumptive basis for thermoelectric
Table 10.2 Water-intensity on a withdrawal and consumptive basis for thermoelectric
          generation using different sources of energy and using once-through cooling
          systems*
System
Coal (once-through)
Natural gas (Rankine,
once-through)
NGCC (once-through)
Nuclear (once-through)
Water withdrawals
gal/MWh*
20,000-50,000
10,000-60,000
7,500-20,000
25,000-60,000
10-4m3/MJ»
210.3-525.7
105.0-631.0
78.9-210.0
262.8-630.8
Water consumption
gal/MWh
100-317
95-291
20-100
100-400
10-4m3/MJ
1.1-3.3
1.0-3.1
0.2-1.1
1.1-4.2
Chapter
4&9
5
5&9
9
  Note:  * - Once-through cooling systems represent legacy power plant designs. Future plant construction will use
        either recirculating, dry or hybrid wet-dry cooling systems.
       # - gallons per megawatt-hour; ## - cubic meter per mega-joule.

       In addition to corn, other raw materials that can be used to produce ethanol include
sugarcane, sorghum, beverage waste, cheese whey, cellulose and hemi-cellulose, corn stover,
hardwood and switch grass. One potential drawback of using cellulose as a source of ethanol
instead of corn is an increase in the amount of water required during processing to produce an
equivalent amount of ethanol from the two sources. Cellulosic ethanol uses agricultural by-
products and energy crops that can grow even in arid regions (Dale, 2007; Chiu et a/., 2009;
Zink, 2007, Keeney, 2006), so a significant advantage of cellulosic ethanol is reduced water
uptake and consumption due to a reduced need for irrigation. Use of agricultural by-products and
cultivation of cellulosic energy crops can also potentially reduce competition between ethanol
and food production.
       There are a number of cooling steps within cellulosic ethanol production that result in
water consumption through evaporation losses that need to be made up with fresh, recycled
                                           244

-------
treated, or recycled carried-over water. It is possible to save water during ethanol production by
recycling and reusing most of the wastewater as well as by taking steps to avoid evaporation
losses. Additional technologies that can be used to minimize water usage from ethanol
production include using pervaporation, which uses membrane separation to separate ethanol
from water instead of steam distillation, membrane solvent extraction, which uses porous
membranes to separate ethanol from the fermentation broth using an extracting solvent.
Thermophilic yeasts can also be used to minimize the amount of cooling required for feed going
into the fermentation unit and thus reduce cooling water usage.
       Biodiesel, which is currently made primarily from transesterified soy oil in the U.S., is
the second most common biofuel used in the U.S. In 2013, 46 out of the 50 states had biodiesel
plants in operation. Irrigation is the most significant source of water uptake and consumption for
biodiesel production and varies significantly from state to state. The water consumption study of
biodiesel production detailed in Chapter 8  suggests that, on average, irrigation accounts for 61.78
gallons (gal) of water for a gallon of soybean biodiesel while soybean processing (0.17 gal/gal)
and biodiesel production (0.31 gal/gal) stages consume much less. The total water consumption
intensity for biodiesel processing was found to be 62.26 gal/gal, which is much lower than values
reported in other existing literature. Chapter 8 of this report also investigated water consumption
in potentially water-stressed areas. One recommendation for future work will be to characterize
the inter-state trade of biofuel feedstocks and its impact on regional and  state-level water
resources since soy that is processed into biodiesel is a fungible commodity that can be
transported for processing into biodiesel in locations far removed from where soy agriculture is
occurring. To achieve this, robust data on soybean and soy oil trade across state boundaries is
needed to fully account for the impact of irrigation water consumption on or regional biodiesel
production.
       Distiller grains  are expected to become a major feedstock for post-2022 biodiesel
production. In addition, algae is also considered a potential feedstock for biodiesel production.
Water that can be used  in algal biodiesel processes include open pond cultivation, harvesting and
dewatering, algal oil extraction, and producing biodiesel via transesterification. While very large
evaporative water losses occur during algal cultivation and in the harvesting and dewatering
steps, dewatering is necessary to prevent carry-over of the algal biomass to the oil extraction step
as it may prevent separation of lipids from algal cells. The extraction process involves solvent
recovery, which requires make-up water for cooling towers and a boiler. Water consumption
during the biodiesel production stage for algal biodiesel remains the same as that for soybeans.
Overall, water consumption for the production of biodiesel from algae is much higher than the
consumption for production of biofuels from soybeans, waste  cooking oil, corn ethanol and
cellulosic ethanol.  Switching to saline and/or wastewater reuse may be necessary to alleviate
concerns regarding the  large quantities of freshwater use in algal biodiesel production. With
technology improvements, low quality feedstocks, especially feedstocks from waste such as the
trap grease from restaurants and sewer pipeline and other oil containing  wastes, can be
increasingly used for biodiesel production. Limitations of these feedstocks include the limited
quantity, i.e.,  they will  contribute to only a small fraction  of the biodiesel supply, and they will
require pretreatment to extract the oil fractions. A better understanding of the life cycle water
needs for waste feedstocks will be necessary. The development of new processes to refine
biological oils and fats  into "renewable diesel fuel" has the potential to become a technology
with high impact since  it is a direct replacement for diesel and jet distillate fuels and thus

                                           245

-------
represents a potentially higher market volume than for biodiesel blends with petroleum diesel
fuel.
       The amount of water consumed for producing ethanol and biodiesel from different fuel
sources is listed in Table 10.3. For comparison purposes, the amount of water consumed to
produce gasoline and diesel from petroleum is also listed.
     Table 10.3 Water intensities on a consumptive basis for producing different types of
                biofuels
Fuel
Gasoline
Ethanol (corn)
Ethanol (cellulose/switch grass +
SRWC/corn stover)
Diesel Fuel Oil
Biodiesel (soy, current hydroxide TE)
Biodiesel (waste oil, acid-ester)
Biodiesel (algal, hydroxide TE)
Water used for Fuel
Processing
1-2.5
2.7-40(13.4)1
12-172
1-2.5
0.3-0.5
0.3
1
Water used for Crop Irrigation or
Petroleum Extraction
0
15-934(113)1
27 - 691 (28)3
0
1-1059 (62)4
0
40-1421 (554)4
  Note: Water intensity in unit of gal hhO/gal fuel or m3 hbO/m3 fuel
       1 Approximate average value based on King and Webber, 2008
       2 Chapter 7 of this report
       3 Approximate average value based on Tidewell et al. 2011  projections for 2030
       4 Approximate average value based on literature discussed  in Chapter 8 of this report.

       As shown in table 10.3, ethanol processed from corn starch uses 2.7 - 40 gal water/gal
ethanol, while ethanol processed from alternative sources such as cellulose, switch grass or corn
stover uses 12-17 gal water/gal ethanol. The water-intensity for cellulosic ethanol processing
falls approximately within the range of water-intensity for corn ethanol processing. This level of
water usage is still high compared to gasoline or diesel processed from petroleum sources (1 -
2.5 gal water/gal fuel), and future research needs to focus on technologies to reduce water use
during the fuel processing stage. In addition, as mentioned earlier, the amount of water required
for irrigation purposes is often more than 2 orders of magnitude higher than the amount of water
required to convert biomass to transportation fuel. Average water-intensities are shown in Table
10.3 for comparison purposes, but water-intensities for irrigation vary considerably on a regional
basis. State and regional adaptive planning or the development of future water use models should
carefully consider regional differences in irrigation water-intensity for biofuels. The water-
intensities shown in Table 10.3 also do not take into consideration water lost by
evapotranspiration. Development of state or even local evapotranspiration water intensities for
biomass that are specific to U.S. cultivation should be a topic for future research.
       Development of Federal policies that increase the use of ethanol, biodiesel and other
biofuels in transportation fuels may be hampered by the water issues faced by communities in
key agricultural regions of the U.S. unless biofuel feedstock cultivation is transitioned to less
                                            246

-------
water-intensive biomass crops, e.g., transitioning ethanol production from corn starch to a
cellulosic energy crops. For example, the cultivation of corn requires approximately 113 gallon-
FbO/gallon-ethanol on average compared to approximately 13 gallon-FbO/gallon-ethanol
required to convert corn to ethanol. Given that the majority of the corn grown in the U.S. is in the
Midwest, and many areas rely mostly on groundwater for irrigation in states such as Iowa and
Nebraska where groundwater levels are falling, it might not be sustainable for these states to
maintain corn production to meet future ethanol demand. A potential alternative is to transition
to less water-intensive biomass crops for the feedstock such as alternative corn varieties that use
less water for irrigation, alternative sugar/starch crops or cellulosic feedstock such as switch
grass. By 2022, EISA calls for the production of 16 billion gallons of ethanol from cellulosic
feedstock while the production of corn-based ethanol is capped at 1 billion gallons, additional
cellulosic material such as switch grass and short-rotation woody crops (SRWC) will need to be
grown to meet the additional cellulosic ethanol demand. Switchgrass and SRWC are not
considered to be agricultural residue,  and thus would also require water for irrigation in some
regions  of the U.S. In 2006, 5,616 MOD of water was used for irrigating crops  that were used to
produce biofuel (primarily from corn). The amount of water required for conversion of the corn
to biofuels was relatively lower at 94  MGD. It is predicted that the amount of irrigation water for
feedstocks to produce approximately  15 billion gallons of corn-based ethanol and 80 billion
gallons  of cellulosic ethanol per year in the year 2030 would increase to 11,458 MGD of water
(4649, 4077 and 2822 MGD for corn, switch grass and SRWC, respectively), while water used
for conversion processes would require 470 MGD (219 and 251 MGD for corn and cellulosic
ethanol, respectively). It is estimated that the irrigation water intensity to produce 80 billion
gallons  of cellulosic ethanol from switch grass and SRWC in 2030 would be approximately 28
gal water/gal cellulosic ethanol produced, which is still considerably lower than the irrigation
water intensity to produce corn-based ethanol. Thus, shifting production from corn-based ethanol
to cellulosic ethanol is expected to lead to significant reductions in the water intensity for
irrigation of biomass crops. The irrigation water-intensity for cellulosic ethanol should also be
considered a conservative estimate. The Tidwell et al. (2011) analysis of cellulosic ethanol water
use was based upon a scenario with cellulosic ethanol production in 2030 that is more than
double the entire RFS2 2022 volume for all renewable fuels, five-times the RFS2 volumes for
cellulosic ethanol production, and is substantially higher than EIA projections.  Such a high
volume of production would result in considerable cellulosic biomass cultivation in regions that
would require a relatively high degree of irrigation and thus a scenario using a more realistic
scenario of approximately 16-20 billion gallons of cellulosic ethanol production may result in a
reduction of irrigation water-intensity for cellulosic and should be the subject of future research
in this area.
       Due to prolonged droughts and population growth, there are concerns over water supplies
to sustain food production, municipal water use, and energy production technologies that rely on
water cooling and hydropower. Many demands and various consumers competing for water
result in frequent over-allocation of water resources. To support the sustainability of water
supplies, water conservation programs together with operationalized water activities, such as
recycling and water reuse including water infrastructure adaptation and water policies, have been
initiated. For example, the Southwest Nevada Water Authority (SNWA) employs a multi-faceted
approach to reduce the drought exposure and create diverse and flexible water resources
(SNWA, 2012). Their primary approaches to water resource management are implementing a
water conservation program, treating  and reusing wastewater, and extending water supplies.
                                           247

-------
Droughts have forced utilities in the Southwest to consider new cooling technologies and sources
of water to cool thermoelectric plants (Walton, 2010). Several projects are exploring the
potential use of non-traditional sources of process and cooling water. The use of wastewater for
thermoelectric cooling is increasing in the Southwest although the supply of urban wastewater
may be decreasing in some places due to conservation policies (Walton, 2010). Municipal
wastewater still appears to be available in sufficient and reliable quantities to provide cooling for
thermoelectric generation (U.S. DOE, 2010). Using reclaimed water for cooling systems has
several advantages, as it helps to save potable water and provides a reliable supply. Water
savings in thermoelectric plants can be achieved via air cooling, use of non-traditional water
sources, recycling plant freshwater, and by increasing plant thermal efficiency (WNA, 2014).
Other methods to reduce water consumption in cooling systems include the use of ice thermal
storage within coal-fired cooling systems, which cools the intake-air to gas turbines, as well as
the use of an air-cooled condensers (ACC). Furthermore, technologies such as Liquid Sorption
Technologies, Flue Gas Sorption Membranes, and Condensing Heat Exchangers that can recover
usable water from alternative sources, such as water from the flue gas emitted by coal-fired
power plants, can also help reduce water consumption. Using dry cooling systems or hybrid
cooling systems is possible in water-scarce regions where dry cooling may be feasible for certain
plant configurations such as IGCC, NGCC, concentrated solar and geothermal generation. While
a variety of technologies have been used to supplement or replace freshwater for cooling
systems, there are some of limitations. For example, a dry cooling system can result in decreased
power plant efficiency. Moreover, low quality impaired water can cause cooling system
inefficiencies and a power plant site's proximity to wastewater sources may not be feasible.
       Increasing energy demand in the coming decades is expected to aggravate competition
for water and especially the availability of water used for electricity generation (U.S. DOE,
2008). Moreover, while renewable energy sources reduce the carbon-intensity of electric
generation, some low-carbon intensity systems still heavily rely on water sources. Hydropower
production, for example, is known to be sensitive to total runoff and to reservoir levels. IGCC
plants need approximately one-third as much engineered cooling when compared with
conventional coal thermoelectric plants (WNA,  2014). Concentrated solar  generation systems
have water requirements that are similar to other thermoelectric generation systems.  Of the
renewable forms of electric generation, two technologies, photovoltaic and wind, use the least
amount of water for cooling.
       Facilities can also consider saline water as a potential water source for cooling water and
other uses. There is also potential for electricity dispatch be shifted via regional  interconnects
towards coastal areas due to their abundance of salt water sources for cooling water.

10.2 Recommendations and Future Work

       Water, climate and energy issues are intricately and closely interrelated and cannot be
adequately addressed in isolation. Increasing population and increased energy demand will result
in increased competition between sectors for surface water. It is both inevitable and urgent to
find alternative water resources to help offset increasing freshwater demand for thermoelectric
cooling. Impaired waters and saline waters are potential alternative water sources that can help
meet cooling needs of thermoelectric generation. There is already some experience in the U.S.
                                          248

-------
with the use of impaired waters for thermoelectric cooling, including the use of treated municipal
wastewater and seawater.
       One issue that has not been addressed in this report is the amount of energy required to
transport and treat source water for power plant use as well as the energy required to treat
wastewater from power plants. In addition, ways to conserve and reuse water in power plants
should receive additional attention. A second issue that has not been addressed within this report
is the use of waste biomass from  secondary wastewater treatment processes to produce electricity
(e.g., using microbial fuel cells), biofuels or methane gas, which can be used as a fuel source in
wastewater treatment plants. In addition, the amount of water and energy required to produce
fuel (coal, gas, oil, biofuels, etc.) in a form suitable for energy production nor the water and
energy requirements for nuclear thermoelectric power generation have been fully addressed in
this report and are topics for further investigation. In the case of biofuels, there are large data
gaps in terms of missing irrigation water usage data, especially for certain types of feedstocks
such as switch grass, algae, hard wood, etc., and for data that shows regional or state differences
in the U.S. Since the amount of water required for irrigation far exceeds the amount of water
required for biomass conversion to biofuels, future research should focus increasing on the size
and scope of water monitoring networks for biomass cultivation, and improving modeling tools
to estimate irrigation water usage.
       Water usage data is readily  available for coal-fired and natural gas-fired power plants, but
there is an imbalance in data availability for other energy sources such as nuclear, solar, wind,
hydroelectric and biomass. Life-cycle water use for generating electricity from renewable, low-
carbon-intensity energy sources such as hydroelectric, solar, wind, biomass and geothermal
systems should be studied in greater detail. In addition, since nuclear energy contributes a
significant fraction of the electricity produced in the U.S., future studies should also conduct a
detailed analysis of energy consumption, water withdrawal, and water consumption for nuclear
power in conjunction with future analyses for coal and natural gas electric generation.
       Climate change and regulations such as RFS, RFS2 and EISA mandate reductions in the
use of petroleum-based fuels, and increases in alternative fuels such as the biofuels ethanol and
biodiesel, which can be produced from multiple sources. Biofuels also have significant water
requirements, especially at the cultivation stage. Some of the same water-conserving techniques
used by the energy industry such as recycling process water or using treated municipal
wastewater can also be used in the biofuel industry. Additionally, more research is needed to
assess the regional and local water impacts of different types of electricity generation, and to
analyze the water impacts of electricity-sector choices. More studies of viable energy resources
and the impacts of geographical limitations may be useful in adapting the use of water locally in
the energy generation sector. Finally, a review of the limitations of state and federal regulations
on impaired water use, and a series of feasibilities study of technologies to facilitate the use of
impaired water are recommended. Because of the linkages between water, energy and climate
change, water-intensity on both a withdrawal and consumptive basis should also be integrated
into regulatory models and analyses that characterize energy use and GHG emissions. Specific
examples include detailed electric dispatch models such as the Integrated Planning Model (U.S.
EPA, 2014) and detailed transportation fuel life cycle assessments such as the Greenhouse
Gases,  Regulated Emissions, and Energy Use in Transportation Model (Argonne  National
Laboratory, 2013).

                                           249

-------
10.3 References

Argonne National Laboratory. 2013. "GREET Model."  https://greet.es.anl.gov/. Accessed
       August, 2014.

Chiu, Y.W., Walseth, B., & Suh, S. 2009. "Water embodied in bioethanol in the United States."
       Environmental Science and Technology, 43:2688-2692.

Dale, B.E. 2007. "Cellulosic ethanol and sustainability: There is no 'food vs. fuel' conflict."
       Paper presented at the 233rd American Chemical Society (ACS) National Meeting and
       Exposition, Chicago, IL USA, March 25-29, 2007.

Keeney, D. and M. Muller 2006. Water use by ethanol plants: Potential challenges. Institute for
       Agriculture and Trade Policy, Minneapolis, MN USA.
       http://www.iatp.org/files/258 2  89449.pdf. Accessed August, 2014.

King, C.W., andM.E. Webber. 2008. "Water intensity of transportation." Environmental
       Science and Technology, 42 (21):7866-7872.

Meldrum, J., S. Nettles-Anderson, G. Heath, and J. Macknick. 2013. "Life cycle water use for
       electricity generation: a review and harmonization of literature estimates." Environmental
       Research Letter, 8 (1) 015031. doi:10.1088/1748-9326/8/l/015031

Southern Nevada Water Authority (SNWA). 2012. "SNWA Annual Report 2012"
       https://www.snwa.com/assets/pdf/about_reports_annual.pdf

Tidwell, V., A.C.-T. Sun, and L. Malczynski. 2011. "Biofuel Impacts on Water." No.
       SAND2011-0168. Sandia National Laboratories.

U.S. EPA. 2014. "EPA Power Sector Modeling."  http://www.epa.gov/powersectormodeling/.
       Accessed August, 2014.

U.S. Department of Energy (U.S. DOE). 2006. "Energy Demands on Water Resources: Report to
       Congress on the Interdependency of Energy and Water." Washington, DC.
       http ://www. sandia. gov/energy-water/docs/121 -RptToCongress-EWwEIAcomments-
       FINAL.pdf.

U.S. Department of Energy (U.S. DOE). 2008. "Estimating Freshwater Needs to Meet Future
       Thermoelectric Generation Requirement." Report No. DOE/NETL-400/2008/1339, U.S.
       Department of Energy, National Energy Technology Laboratory, Pittsburgh, PA.

U.S. Department of Energy (U.S. DOE). 2010. "Department of Energy/National Energy
       Technology Laboratory's Water-Energy Interface Research Program: December 2010
       Update."

World Nuclear Association (NWA).  2014. "Cooling Power Plants". WNA Information Library
       (Updated September 2014). http://www.world-nuclear.org/info/current-and-future-
       generation/cooling-power-plants/.
                                         250

-------
Zink, J. 2007. "Water-use estimates for ethanol plant go up 60 percent." Petersburg Times,
       Tampa Bay, FL. March 9, 2007.
                                          251

-------