reading summary

Paper instructions:
Skip the “Materials and Methods” section of Zhang and Anadon (2013) unless you are interested in reading the section. You are not required to show in-text citation of this assigned paper.

However, please have the reference section at the end of your document.

The term, “life cycle water use of energy production” refers to the direct use (i.e. by energy sectors) and indirect use (i.e. by other sectors).

Summarize the paper while answering the questions below:

Explain the linkage between energy and water resources.
Describe the difference between water withdrawal and consumption.
Which energy sectors (e.g. coal, crude oil, natural gas) show the highest water withdrawals, water consumption, and wastewater discharge in China (overall energy production NOT water use

per unit energy products)?
Which city has the highest water withdrawal for electricity productions? Why?
Based on the findings for the spatial distribution of water use and energy production, pick a region and make a recommendation(s) for the region, so the region can keep meeting the same

energy demand while reducing water withdrawal and/or consumption. Explain why you chose the recommendation(s).

Life Cycle Water Use of Energy Production and Its Environmental
Impacts in China
Chao Zhang*
,†, ‡,§
and Laura Diaz Anadon

Mossavar-Rahmani Center for Business and Government, Kennedy School of Government, Harvard University, 1350 Massachusetts
Avenue, Cambridge, Massachusetts 02138, United States

Belfer Center for Science and International A ffairs, Kennedy School of Government, Harvard University, 1350 Massachusetts
Avenue, Cambridge, Massachusetts 02138, United States
School of Economics and Management, Tongji University, 1239 Siping Road, Shanghai 200092, China
*S Supporting Information
ABSTRACT: The energy sector is a major user of fresh water resources in
China. We investigate the life cycle water withdrawals, consumptive water
use, and wastewater discharge of China ’ s energy sectors and their water-consumption-related environmental impacts, using a mixed-unit multire-gional input −output (MRIO) model and life cycle

impact assessment
method (LCIA) based o n t he Eco-in di cator 99 framework. En ergy
production is responsible for 61.4 billion m
water withdrawals, 10.8 billion
water consumption, and 5.0 billion m
wastewater discharges in China,
which are equivalent to 12.3%, 4.1% and 8.3% of the national totals,
respectively. The most important feature of the energy −water nexus in China
is the signifi cantly uneven spatial distribution of consumptive water use and
its correspon di ng environm en tal impacts caused by the geological
discrepancy among fossil fuel resources, fresh water resources, and energy
demand. More than half of energy-related water withdrawals occur in the east and south coastal regions. However, the arid north
and northwest regions have much larger water consumption than the water abundant south region, and bear almost all
environmental damages caused by consumptive water use.
Energy and water are two fundamental resources that support
most aspects of human well-being. It has been increasingly
recognized that energy and water sustainability are inextricably
The energy sector is the second largest water user
in the world in terms of withdrawals, following irrigation.
Almost every stage in the energy supply chain needs water in
various ways:
3, 4
e.g., water is used for drilling and fracturing in
oil and gas exploration;
large volumes of cooling and
processing water a re often n eeded in th ermal power
gen eration;
3,7− 9
and s om e f eedst ocks f o r biofuels need
considerably large amounts of water.
10 − 15
Besides its impact
in terms of the magnitude of water consumed and withdrawn,
energy production can also cause serious degradation to the
local aquatic environment. Fossil fuel extraction, especially coal
mining, can pollute aquifers and lead to ecosystem and health
A water-constrained future makes water an increasingly
important vulnerability in the energy sector s.
18 , 1 9
Wa t e r
shortages caused by an overexploitation of freshwater resources
and drought have already a ffected electric generation in some
parts of the world.
On the one hand, such an impact is likely to
occur with more frequency in the future due to climate
19, 20
On the other hand, driven by growing population
and economic growth, the energy sector will continue to
expand and this expansion will continue to increase the
pressures on fresh water demand in many nations.
20 − 22
example, Davies et al.
show that global consumptive water use
by electric power generation could increase by 4− 5 times from
2005 to 2095.
As a re fl ection of the growing concerns about the growing
pressures of energy production on water resources, there has
been an increasing number of studies aimed to systematically
quantifying the energy−water nexus in water scarce nations and
regions, such as the Middle East and North Africa;
and the United States.
27 −30
However, no compre-hensive study is currently available for China, even though
China’ s energy−water nexus presents some distinct features.
First, China is already su ffering from severe water shortages.
The average annual per capita renewable water resources in
China are only about one-third of the world’ s average.
annual per capita water resources i n t he 10 provincial
jurisdictions, where 34% of China ’ s population live, are well
below 1000 m
which is the widely regarded water scarcity
threshold (see Figure S1 in the Supporting Information (SI)).
Received: June 11, 2013
Revised: October 14, 2013
Accepted: October 14, 2013
Published: October 14, 2013
© 2013 American Chemical Society 14459 | Environ. Sci. Technol. 2013, 47, 14459 − 14467
Second, driven by China’ s unprecedented economic boom over
the past decade, total primary energy production has more than
doubled between 2000 and 2010.
The competition for limited
water resources among users in the energy industry, other
industries, and other users (e.g., irrigation and domestic use) is
intensifying. T hird, there exists a signifi cant geological
mismatch between the distribution of coal resources, the
dominant primary energy source in China, and that of water
resources. The three provinces with the largest coal outputs in
China, namely, Shanxi, Shaanxi, and Inner Mongolia, contribute
more than half of the total national coal output and 16% of
thermal power generation, but are only endowed with 3% of
national water resources.
Fourth, China is a major player in
the w orld energy marke t, w ith 17.3% of global energy
production and 17.5% of global consumption in 2010.
Most previous studies on investigating the energy−water
nexus have used technology-based bottom up accounting
methods to calculate water use (water withdrawals and/or
consumptive water use) in energy sectors. These studies
mu lt ip ly act iv it y le v el s o f d i ff erent e nergy p roduction
technologies with corresponding water intensity.
these studies calculate the direct water withdrawals and/or
consumption of the energy sector, indirect water use embodied
in the upstream supply chains of energy production are usually
not included. Furthermore, few studies accounting for the life-cycle water withdrawal, consumption, and discharge (mainly
using input − output analysis techniques) have assessed the
environmental impact of water use in the energy sector.
In this study, we focus on the life cycle water use of energy
production and its environmental impact in China. We quantify
life cycle freshwater withdrawals, consumptive water use, and
wastewater discharge of eight energy products (namely, coal,
crude oil, natural gas, petroleum products, coke, electricity,
heat, and gases) at the provincial level. The de fi nition of
consumptive water use adopted in this study is also consistent
with the widely accepted notion of blue water footprint.
cover the full supply chain of energy production, we base our
analysis on a mixed-unit multiregional input− output (MRIO)
model, in which data on energy consumption and water use in
physical units are merged with monetary input − output (IO)
tables. In the impact assessment step, the method and data
provided in Pfister and colleagues
is used to quantify damages
to human health, ecosystem quality, and resources caused by
life cycle water consumption of energy production. This
provides insights to researchers and policy makers interested in
(1) understanding the spatial characteristics of China’s energy−
water nexus, (2) mitigating environmental impacts related to
consumptive water use caused by energy production in hotspot
regions, and (3) identifying the water-related implications of
future energy plans and of severe droughts.
2.1. Mixed Unit MRIO Model. The multiregional input−
output analysis model (MRIO) is widely used to trace the
supply chain environment al impact s embodied in trade
activities from a consumption-based perspective,
37 −40
plenty of studies conducted at both global
41 −43
and national
44 − 47
The basic equation of a MRIO model containing R
regions can be expressed as follows:
=**=* − * *

ff W X(IA)Y1
where X* =[x
, ··· , x
, ··· , x
is the aggregated output vector
of all sectors in all regions;

11 12 1R
21 22 2R
R1 R2 RR
is the aggregated inter-regional direct requirement coe fficient
matrix; Y* =[Σ
+ ex
, Σ
+ ex
, ··· , Σ
+ ex
is the
aggregated fi nal use vector; y
is a column vector represents
goods or services produced in region r and consumed in region
s; ex
is a column vector represents exports from region r; and
f * =[f
, …, f
, …, f
] is the water use coeffi cient vector, referring
to water w it hdrawal, water con sumption, or wastewater
discharge intensities in different calculations. W is the total
water withdrawals, water consumption or wastewater discharges
driven by thefinal demand Y*.
Mixed-unit input− output models are an extension of
s tan da rd mo ne ta r y IO ana ly sis . Th ey in te gra t e mo ne tar y
input-output data and mass fl ows expressed in physical units
into a consistent framework and capture the characteristics of
bot h econ om ic tr an sact ions and m aterial fl ows in a n
This study constructed mixed-unit IO tables
for each province. Monetary input−output data of energy and
water sectors in the original IO tables are replaced by sector-wise energy consumption and water use data in physical units
extracted from various data sources. Then, provincial IO tables
are compiled into a mixed-unit MRIO table by using a gravity
to estimate inter-regional trade flows. To extract
supply chain water use by different energy products produced
in different regions, we use a deduction approach in the
calculation. More methodological details are presented in the
2.2. Data Sources. We put together a mixed-unit MRIO
table of China for 2007 (the latest year with data available),
including 30 provincial jurisdictions, eight energy products in
physical units, one water production and distribution sector in
physical units, and 22 nonenergy and nonwater sectors in
monetary units. Three groups of data are used in this research:
(1) provincial monetary IO tables; (2) provincial and sectoral
energy production, consumption, and trade data; and (3)
pro vinc ial a nd secto ral fresh wate r wit hdraw al, wate r con-sumption, and wastewater discharge data.
The monetary IO tables o f ea ch p rovinc e
are fi rst
transformed into a mixed-unit IO table with energy and
water data. Five energy sectors in the original monetary IO
tables are removed and replaced by eight energy products
expressed in physical units (Table 1 presents the mapping
between monetary sectors and physical accounts). Production,
consumption, and trade amounts of energy products are
sourced from the provincial energy balance tables
and the
e ne r gy c on su m pt io n s ur v e ys r e po r te d i n C hi na’ s s ec o nd
national economic census.
Water use data collection and estimation takes many steps.
Wat er wit hd raw al and wa ste wate r di sch arg e by ind ust rial
sectors are reported in the first national pollution source
census in China conducted by Ministry of Environmental
Protection (MEP) and National Bureau of Statistics (NBS).
This is the most comprehensive data set currently available that
provides sector-wise statistics of water withdrawal and waste-water discharge in each province. Information on irrigation
Environmental Science & Technology Article | Environ. Sci. Technol. 2013, 47, 14459 − 14467 14460
water withdrawal and water consumption in the agriculture
sector is extracted from the various provincial water resources
bulletins. Precipitation, or green water use, is not included,
since green water is usually not subjected to competitive
demand. The sector-wise water consumption is not reported in
the different industrial sectors. Thus, except for the “electricity ”
sector, we assume that the water consumption of the di fferent
industrial sectors equals their water withdrawal minus waste-water discharge. This means that the water withdrawals that do
not return back to the environment in the form of wastewater
(either treated or untreated) are “ consumed” through
evaporation, absorption by products, and/or other losses.
Consumptive water use by the “electricity ” sector in each
province needs to be estimated separately. Because of the lack
of information to provide reliable estimates at provincial level,
water consumption related to hydropower generation due to
ev aporation i s not includ ed in th is stud y. Therefore,
consumptive water use of “electricity ” sector in this study is
related to thermal power generation. Three types of cooling
technologies are used in China: i.e., once-through cooling,
closed-loop cooling, and air cooling, and they have very
di ff erent water consumption performance profi les. We fi rst
estimate the proportions of the three cooling technologies in
each province based on a literature survey and our own
calculations. Then, w e u se data on the average water
co ns u mp ti on f a c to r of th e di ff e re nt co ol in g te ch no lo gi es
reported in the China Electricity Council ’ s Thermal Power
Unit Benchmarking and Competition Data set
to calculate
total water consumption by thermal power generation in each
province. Details about t he steps o f estimating water
consumption data are provided in the SI. After compiling
provincial mixed-unit IO tables, we balance the provincial IO
tables at the national level and extend these single-province IO
tables to a MRIO table based on the method of estimating
inter-regional trade flows presented by Zhang and Qi.
3.1. Life Cycle Water Use Per Unit Energy Products.
National average values of life cycle water use for each of the
eight energy products by unit energy are presented in Table 2.
Electricity has the highest water intensity across all indicators.
Life cycle water withdrawals, water consumption, and waste-water discharge intensities of electricity are 5263, 234.2, and
656.7 m
/TJ, respectively, (or 18.9 m
, 0.84, and 2.36 m
MWh). Direct water use accounts for 82% of the electricity life
cycle water withdrawal, 63% of life cycle water consumption,
and 39% of life cycle wastewater discharge.
The life cycle water use of all other energy products is much
lower than that of electricity. Large proportions of their life
cycle water use are embodied in the upstream supply chains of
the other seven energy products. Coal is the dominant primary
energy source in China. Life cycle water withdrawals, water
consumption, and wastewater discharge intensity of coal are
106.4, 41.5, and 38.4 m
/TJ, respectively (or 2.22, 0.87, and 0.8
/tonne), in which 17%, 22%, and 74% are direct water use. It
is noteworthy that coal is the only energy product whose direct
wastewater discharge intensity exceeds its direct water with-drawal intensity. This is because large volumes of mine drainage
are generated and discharged during the coal mining process.
The water intensity of the “Gases” products is the smallest. This
is because a large proportion of gases are recovered as
byproducts in industrial production processes, such as coking,
and iron and steel making. Water use in the main production
processes is not allocated to the recovered gases, since their
economic values are rather small when compared with the main
products and the recovering of gases has a negligible impact on
the water use of the main production process.
Regional differences of water use intensity are signi ficant,
especially for water withdrawals. Taking electricity for example,
as illustrated in Figure 1, life cycle water withdrawal intensity
can be as high as 69.5 m
/MWh in Shanghai and as low as 1.2
/MWh in Qinghai. Such a large di fference is driven by a
multiplicity of factors, e.g., the mix of power generation
technologies, the adoption of diff erent cooling methods in
thermal power plants, and the structure of intermediate inputs
of the power sector. In Shanghai, all electricity is produced by
thermal power plants, but hydropower plays a dominant role in
Q in gh ai . S in c e in – st r ea mw a te r us e a nd ev ap or a ti o n f r om
reservoirs related to hydropower generation are not considered
in this study, the average water withdrawal and consumption
intensity of electricity production in Qinghai are both the
lowest. In terms of water consumption intensity, Beijing has the
highest value of 3.81 m
/MWh. Generally speaking, areas with
Table 1. Mapping the Relationship between Monetary
Sectors and Energy and Water Products
original monetary sectors
in the IO tables
corresponding physical accounts in
the mixed-unit IO tables
energy coal mining coal
crude oil and natural gas
crude oil
natural gas
oil refi ning and coking petroleum products
electric and heat
gas production and
water water production and
tap water
“Gases” includes coal gas and all kinds of recovered gases from
industrial processes, such as coke oven gas, blast furnace gas, and
converter gas.
Table 2. National Average Life Cycle Water Use Per Unit Energy Products (m
coal crude oil natural gas petroleum products coke electricity
heat gases
water withdrawal direct 18.4 37.2 13.6 87.2 36.1 4306.5 609.4 7.0
life-cycle 106.4 257.9 104.5 446.8 217.7 5263.0 901.3 42.4
water consumption direct 8.9 26.3 9.5 37.9 26.2 410.9 81.2 4.5
life-cycle 41.5 99.1 41.2 168.3 95.4 656.7 172.7 18.4
wastewater discharge direct 28.3 11.0 4.1 49.3 9.9 90.6 15.7 2.4
life-cycle 38.4 42.8 18.7 104.3 52.7 234.2 81.8 8.6
Including both coal- fired thermal power generation and noncoal power generation.
Environmental Science & Technology Article | Environ. Sci. Technol. 2013, 47, 14459 − 14467 14461
poor water availability and a higher proportion of thermal
power generation tend to have lower water withdrawals and
higher water consumption intensity of electricity. Places with
such characteristics are mainly located in northern China where
water resources are scarce, such as Beijing, Tianjin, Shandong,
Henan, and Shanxi province.
3.2. Sector Disaggregation of Total Life Cycle Water
Use Related to Energy Production.The energy sector is a
signi fi cant water user in China. Table 3 presents total life cycle
water use of energy production and its sectoral disaggregation.
Total life cycle water withdrawal, water consumption, and
wastewater discharge in 2007 amount to 61.4, 10.8, and 4.95
billion m
, respectively, which are equivalent to 12.3%, 4.1%,
and 8.3% of national totals, respectively. The proportion of
di rect wa ter u se by energy sectors is 87.3% for w ate r
withdrawal, 59.5% for water consumption, and 72.1% for
waste wate r discha rge. Ele ctricity ge n eration p lays a very
signi fi cant role in both in terms of water withdrawal and
water consumption, as it accounts for 79.3% of total withdrawal
and 47.0% of total consumption. For wastewater discharge, coal
production makes up the largest contribution, with 32% of total
wastewater−water discharge. Among other nonenergy sectors,
agriculture is the major source of embodied water use, with
9.4% of total life cycle water withdrawal and 34.7% of water
consumption. As expected given China ’ s heavy reliance on coal,
the energy− water nexus in China is dominated by coal-fi red
power generation.
Figure 1. Life cycle water withdrawal and water consumption per kWh
of electricity by region.
Table 3. Sectoral Disaggregation of Life Cycle Water Use of Energy Production in China
water withdrawal
wastewater discharge water consumption
sector million m
% million m
% million m
national total 504 208 59 545 266 195
total life cycle water use 61 356 100 4952 100 10 839 100
direct use by energy sectors 53 560 87.3 3568 72.1 6453 59.5
coal 1025.9 1.67 1584.4 32.00 495.5 0.85
crude oil 296.9 0.48 87.6 1.77 210.3 1.91
natural gas 39.3 0.06 11.8 0.24 27.5 0.25
petroleum products 1234.2 2.01 697.4 14.08 536.8 4.88
coke 344.0 0.56 94.6 1.91 249.4 2.27
electricity 48635.9 79.28 1023.6 20.67 4641.4 47.01
heat 1923.5 3.13 49.4 1.00 256.3 2.59
other gases 55.4 0.09 19.3 0.39 36.1 0.33
indirect use by other sectors 7796 12.7 1383 27.9 4386 40.5
agriculture 5788.8 9.44 0.0 0.00 3761.3 34.70
metal ores mining 106.5 0.17 42.2 0.85 64.3 0.59
nonmetal ores mining 19.6 0.03 4.8 0.10 15.4 0.14
food, beverage and tobacco 37.1 0.06 27.5 0.56 9.6 0.09
textile 38.8 0.06 31.3 0.63 7.6 0.07
wearing apparel and leather products 11.8 0.02 8.8 0.18 3.0 0.03
timber processing and furniture 4.6 0.01 2.6 0.05 2.0 0.02
paper products and education articles 223.1 0.36 179.8 3.63 43.3 0.40
chemical products 388.7 0.63 231.7 4.68 157.0 1.45
nonmetallic mineral products 25.9 0.04 7.1 0.14 18.8 0.17
ferrous and nonferrous metals 235.1 0.38 85.0 1.72 150.0 1.38
metal products 14.4 0.02 10.3 0.21 4.1 0.04
general and special purpose machineries 23.4 0.04 13.5 0.27 9.9 0.09
transport equipment 14.0 0.02 8.6 0.17 5.4 0.05
electrical equipment 8.6 0.01 5.0 0.10 3.6 0.03
communication and electronic equipment 4.6 0.01 3.3 0.07 1.3 0.01
measuring instruments and office supplies 2.3 0.00 1.3 0.03 1.0 0.01
other industrial products 4.1 0.01 2.7 0.06 1.3 0.01
construction 22.1 0.04 18.8 0.38 3.3 0.03
transportation, storage and postal services 128.4 0.21 109.2 2.20 19.3 0.18
wholesale, retail, lodging and catering 215.5 0.35 183.1 3.70 32.3 0.30
other services 478.6 0.78 406.9 8.22 71.8 0.66
Tap water withdrawals are included in the water withdrawals by each sector.
Environmental Science & Technology Article | Environ. Sci. Technol. 2013, 47, 14459 − 14467 14462
3.3. Spatial Distribution of Life Cycle Water Use
Related to Energy Production. In arid areas where water
resources are scarce, thermal power plants generally adopt
closed-loop cooling systems to avoid the need to withdraw
large amounts of water. Consequently, consumptive water use
has increased dramatically due to evaporation. This leads to
further depletion of the local water resources and intensi fi es
water shortages. The trade-off between water withdrawal and
water consumption is explicitly revealed in our calculation.
Subplot (a) and (b) in Figure 2 illustrate the provincial
breakdown of water withdrawal and water consumption related
to energy production in China.
Most energy-related water withdrawals occur in the eastern
and southern coastal regions of China. Water withdrawal in
Guangdong, Jiangsu, Zhejiang, and Shanghai are 10.6, 10.5, 7.0,
and 4.5 billion m
, respectively, which are equivalent to 23%,
24%, 30%, and 49% of their total local water withdrawal for all
sectors and account for 53% of the national total energy-related
water withdrawal. The Yangzi River Delta in the east coast and
the Pearl River Delta in the south coast are China ’ s major
manufacturing hubs and most populated areas. Electricity
production and consumption are high in these regions. In
addition, their better water resources availability, compared
with northern regions, results in a greater penetration of once-through cooling technology, thereby leading to large water
withdrawals in the four aforementioned provinces.
The spatial distribution of energy-related water consumption
di ff ers tremendously from that of water withdrawals. There is
generally an inverse pattern between the spatial distribution of
water consumption and freshwater resources in China. Thefi rst
fi ve provin ces with the largest energy -related water con-sumption are Shandong (995 million m
), Shanxi (879 million
), Henan (855 million m
), Hebei (718 million m
), and
Jiangsu (713 million m
), which together account for 38.4% of
the national total. Most areas of these provinces are located in
the Huang-Huai-Hai River Bain, where the most severe water
shortage problems in China occur.
The spatial distribution of wastewater discharge is quite
similar to that of water consumption. As shown in subplot (c)
in Figure 2, northern China also bears a large proportion of
energy-related wastewater discharges. Liaoning has the biggest
discharge amount, i.e., 526 million m
, which accounts for
10.6% of the nation total. Other major contributors include
Henan (422 million m
), Shandong (402 million m
), and
Hebei (334 million m
3.4. Environmental Impacts of Life Cycle Consump-tive Water Use of Energy Production. The environmental
impact needs to be understood to inform policy making on
water resources management.
The same volume of water
consumption may have more signi fi cant impacts on human
wellbeing and ecosystem health in water scarce regions than in
water-abundant regions. Considering China’ s large territory and
signi fi cant di fferences in spatial water resource distribution,
integrating the volumetric amount of water use by energy
production with regional specific environmental impacts is
critical to identifying the areas in China where the energy
infrastructure is posing or may pose the most di ffi cult water
resource challenges.
In this section, we use the impact assessment method
developed by Pfi ster and colleagues
to estimate the water-Figure 2. Spatial distribution of life cycle water withdrawals, water consumption, wastewater discharge, and environmental impacts related to
consumptive water use of energy production in China (values are proportional to the area of corresponding circles the scales are diff erent).
Environmental Science & Technology Article | Environ. Sci. Technol. 2013, 47, 14459 − 14467 14463
consumption-related environmental impacts induced by energy
production at the provincial level. This method, based on the
framework of Eco-indicator 99(EI99),
quantifi es potential
damages per cubic meter of water consumption on three
aspects: (1) human health measured as disability-adjusted life
years (DALY) caused by malnutrition due to water shortages
for irrigation; (2) ecosystem quality measured as water-shortage
related potentially disappeared fraction (PDF) of species in an
area over a period of time; and (3) resources represented by
energy requirements of desalination as a backup technology to
replace the depleted water resources. Damages per unit water
consumption, i.e., the characterization factors, and equivalent
EI99 scores of the three impact categories at the watershed
level provided in the literature
are aggregated at the provincial
level (see SI for details of data aggregation). Life cycle
freshwater consumption by energy production in each province
is then multiplied with corresponding characterization factors
and EI99 scores to calculate total damages.
National total damage to human health, ecosystem quality,
and resources related to life cycle water consumption by energy
production in China are 4027 DALY, 3.1 × 10
and 25.2 PJ, respectively. The aggregate impact translated into
EI99 points (pts) of the three categories are 104.7, 240.7, and
598.7 million pts, respectively. Potential damage to resources
(which accounts for 63.4% of the country’s total water-consumption-related damages of energy production) is larger
than damages to ecosystem quality (accounts for 25.5%) and
human health (accounts for 11.1%). A main reason is that
overwithdrawals of freshwater resources widely occur in many
watersheds in north China, where major energy production
bases are located. Huge amounts of energy are potentially
needed if the depleted water resources in those regions are to
be replaced by seawater desalination, which is regarded as a
backup technology for freshwater provision.
As shown in subplot (d) in Figure 2, the spatial distribution
of the environmental impacts caused by consumptive water use
is much more uneven than th e d is t r ib u t i on o f w at e r
consumption itself. Xinjiang, Hebei, Shanxi, Jiangsu, Shandong,
Henan, Gansu, and Inner Mongolia altogether bear 85.4% of
the total environmental impacts, while energy-related con-sumptive water use occurred in these provinces account for
51.4% of the national total. Compositions of impacts in these
hotspot regions are presented in Figure 3. The largest human
health impacts occur in Shandong, while the largest damage to
ecosy ste m qu ality and reso urce s bo th occ ur in Xi njia ng.
Howeve r, ind i re ct wate r consumption embodied in the
upstream supply chain of energy production is responsible for
most (76.2%) of the impacts in Xinjiang. In southern China,
water consumption-related environmental impacts are almost
negligible, compared with northern China. More details of the
water use in energy production and its corresponding impacts
by province are presented in the SI.
In t his s tu dy, life c ycle wate r us e and c o rre sp ondi ng
environmental impacts caused by water consumption in
China are investigated. Overall, China’ s energy sectors are
responsible for 12.3%, 4.1%, and 8.3% of the national total
water wit hdra wals, wa ter consu mptio n (exc lu ding th at of
hydropower) and wastewater discharge, respectively. Coal-fired power generation plays a dominate role in both energy-related water withdrawal s and water consumption. When
compared to other regions throughout the world, the share
of China’ s energy sectors in the total national water use is much
larger than that of other arid countries such as countries in the
Middle East,
but smaller than some water abundant countries
such as the United States, where nearly half of the total water
withdrawals are directly used by thermal power generation.
The most important feature of energy−water nexus in China is
the significantly uneven spatial distributions of water use and its
environmental impacts caused by the geological discrepancy
among fossil fuel resources, water resources, and economic
activities. While mo re th an half of energy -related wa ter
withdrawals occur in the eastern and southern coastal regions,
the arid northern region has much larger water consumption
than the water abundant southern region. The environmental
impacts related to the consumptive water use concentrate in
several hotspots provinces in northern China.
China is likely to continue to experience rapid expansion of
coal-based thermal power generation along with its rapid
economic development. According to the national 12th five-ye ar (2011 − 2015) development plan o f China ’ spower
total capacity of coal-based thermal power is expected
to reach 0.93 TW in 2015 and 1.17 TW in 2020, compared
with 0.71 TW in 2010, and accoun for 63.5% and 60.5% of the
t otal installe d cap acity, r e specti vely . Stim ulated by the
enormous impulse of developing local economy, local govern-ments in major energy production provinces have outlined even
m o r e a mb i ti o u s b l u e pr i n ts f o r t h ei r e ne r g y s e c t o r s . F or
ex amp l e, th e pl a nne d in cr em en ta l ca pac it ie s of coa l-fi red
electricity during the 12th fi ve-year in eight provinces in
north a nd northwest regions, i.e., Hebei, Shanxi, Inn er
Mongolia, Shandong, Henan, Shaanxi, Ningxia, and Xinjiang,
add up to 0.23 TW, which, surprisingly, exceeds China ’ s
national plan (see Figure S3 in the SI for planned incremental
ca pacit ies in th ese pro vinc es) . Coa l-base d th erma l pow er
capacity is planned to more than double in Shanxi, Shaanxi,
and Ningxia during 2010−2015. In Xinjiang, 4-fold increase
may be achieved in 2015 compared with 2010. Considering that
the existing environment impacts caused by energy production
are a lready signifi cant i n t hese regions, these e lec t ricity
generation plans involving signi fi cant e xpansions could
potentially contribute to more serious water-related environ-mental damages if the planned expansion does not rely on air-cooling technology.
Figure 3. Composition of environmental impact and share of direct
impact attributed to energy production in hotspot regions. Column
bars represent impact scores (left scale) and triangles represent share
of direct impact of energy production (right scale).
Environmental Science & Technology Article | Environ. Sci. Technol. 2013, 47, 14459 − 14467 14464
Relieving the pressure of growing energy production on
water resources in China is likely to require comprehensive
measures. Wide adoption of water-saving technologies in the
power sector in northern China, such as air cooling technology,
is of key importance.
Many advanced power technologies,
such a s ult r asupe r critical coal pow e r system s, l arge -sca le
circulating fluidized bed (CFB) technology, and integrated
gasi fi cation combined cycle ( IGC C) p lant s w ith h igher
conversion e fficiency and lower water intensities have begun
to be deployed since the late 2000s.
64 − 66
Compared with
traditional subcritical thermal power generation, ultrasupercrit-ical technology can reduce water consumption intensity by
more than 10% and IGCC by 40%.
Over the longer term, the challenge posed by the geographic
mismatch of water resources and energy production requires a
strategy of increased decoupling between the energy sector and
freshwater resource demand, especially in regions with high
environmental impacts. Options include (but are not limited
to ) th e in cr eas ed de pl oy me nt of wi nd po we r a nd so lar
photovoltaic power, which have started receiving government
support in the mid-2000s;
68 − 71
nontraditional water resources
utilization in the energy sector, e.g., using treated municipal
wastewater; and a larger role for the construction of coastal
power plants to allow the use of seawater for cooling.
Which of these options is preferable (or is even an option)
depends on the context, but the long-term plans for the energy
se c to r di sc u ss e d ab ov e do no t c on s id er wa te r r es ou r c es
limitations in areas where water is already scarce. As a top-down approach, this study still has some limitations in terms of
spatial resolution (for example, we were unable to identify
basin-level water use). If comprehensive plant-level information
of water use in energy production facilities in China becomes
available, more detailed analysis could be conducted. Combined
with an energy planning model, this type of data set could
provide important insights for energy and water planning.
In s um ma r y, we ha ve s ho wn th a t en e r gy is a ma j or
component of China’ s water scarcity challenge. Our spatially
disaggregated analysis allowed a better understanding of the
coupling characteristics between energy development and water
resources in China and identifying hotspot regions su ffering
most severe impacts, which provide important inputs to both
energy and water plans.

*S Supporting Information
Additional information as noted in the text. This material is
available free of charge via the Internet at

Corresponding Author
*(C.Z.) Phone: +86 21 65984236; fax: +86 21 65986304; e-mail:,
The authors declare no competing fi nancial interest.

This work was conducted while Chao Zhang was a joint
Giorgio Ruff olo fellow in the Sustainability Science Program in
the Mossavar-Rahmani Center for Business and Government
and Energy Technology Innovation Policy Fellow at the Belfer
Center for Science and International Affairs, both at the
Harvard University Kennedy School of Government. Support
from Italy’ s Ministry for Environment, Land and Sea, the
Science, Technology and Public Policy program (STPP), and
feedb ack from members of both g roups a re gratefully

(1) Schnoor, J. L. Water energy nexus. Environ. Sci. Technol. 2011 , 45 ,
(2) Hightower, M.; Pierce, S. A. The energy challenge. Nature 2008 ,
452, 285 − 286.
(3) Mielke, E.; Anadon, L. D.; Narayanamurti, V. Water Consumption
of Energy Resource Extraction, Processing, and Conversion; Energy
Technology Innovation Policy Discussion Paper 2010-15; Belfer
Center for Science and International Affairs, Harvard Kennedy School:
Cambridge, MA, 2010; files/ETIP-DP-2010-15-final-4.pdf.
(4) Energy Demands on Water Resources: Report to Congress on the
Interdependency of Energy and Water ; U.S. Department of Energy:
Was hington, DC, 2006; nergy-water/docs/121-RptToCongress-EWwEIAcomments-FINAL.pdf.
(5) Lifecycle Analysis of Water Use and Intensity of Noble Energy Oil
and Gas Recovery in the Wattenberg Field of Northern Colorado ; Noble
Energy, Inc., Colorado State University: Boulder, CO, 2012; cewc.
c o lo s ta te . ed u/ w p -c o nt e nt / up l oa d s /2 0 12 /0 4 /W at er – I nt en si ty – R ep o rt -Final.pdf.
(6) Jordaan, S.; Anadon, L.; Mielke, E.; Schrag, D. Regional water-and land-use implications of reducing oil imports with natural gas,
shale oil, or biofuels in the United States. Environ. Sci. Technol. 2013 ,
DOI: 10.1021/es404130v.
(7) Sovacool, B. K.; Sovacool, K. E. Identifying future electricity-water tradeoffs in the United States. Energy Policy 2009 , 37 , 2763 −
(8) Macknick, J.; Newmark, R.; Heath, G.; Hallett, K. A Review of
Operational Water Consumption and Withdrawal Factors for Electricity
Generating Technologies; Technical Report NREL/TP-6A20-50900;
U.S. National Renewable Energy Laboratory: Golden, CO, 2011;
(9) DiPietro, P.; Gerdes, K.; Nichols, C. Water Requirements for
Existing and Emerging Thermoelectric Plant Technologies; Technical
Report DOE/NETL-402/080108; U.S. National Energy Technology
Laboratory: Mor gantown, WV, 2008; www.netl.doe .gov/ energy-analyses/refshelf/PubDetails.aspx?Action=View&PubId=137.
(10) Dominguez-Faus, R.; Powers, S. E.; Burken, J. G.; Alvarez, P. J.
The water footprint of Biofuels: A drink or drive issue? Environ. Sci.
Technol.2009 , 43 , 3005 − 3010.
(11) Gerbens-Leenes, W.; Hoekstra, A. Y.; van der Meer, T. H. The
water footprint of bioenergy.Proc. Natl. Acad. Sci., U. S. A.2009 , 106
(25), 10219 − 10223.
(12) Gheewala, S. H.; Berndes, G.; Jewitt, G. The bioenergy and
water nexus. Biofuel Bioprod. Bioresour. 2011 , 5 (4), 353−360.
(13) Gerbens-Leenes, P. W.; Hoekstra, A. Y.; Meer, T. v. d. The
water footprint of energy from biomass: A quantitative assessment and
consequences of an increasing share of bio-energy in energy supply.
Ecol. Econ. 2009 , 68 (4), 1052 −1060.
(14) Mishra, G. S.; Yeh, S. Life cycle water consumption and
withdrawal requirements of ethanol from corn grain and residues.
Environ. Sci. Technol. 2011 , 45 , 4563 − 4569.
(15) Murphy, C. F.; Allen, D. T. Energy-water nexus for mass
cultivation of algae. Environ. Sci. Technol. 2011 , 45 , 5861 −5868.
(16) Thirsty Coal: A Water Crisis Exacerbated by China’s New Mega
Coal Power Bases; Green Peace: Beijing, 2012;
eas tas ia/Global/east as ia/publ icati on s/repor ts/climate-energ y/2 012/
(17) Mishra, V. K.; Upadhyaya, A. R.; Pandey, S. K.; Tripathi, B. D.
Heavy metal pollution induced due to coal mining effluent on
surrounding aquatic ecosystem and its management through naturally
occurring aquatic macrophytes. Bioresour. Technol. 2008 , 99 , 930 −936.
Environmental Science & Technology Article | Environ. Sci. Technol. 2013, 47, 14459 − 14467 14465
(18) Schaeffer, R.; Szklo, A. S.; Lucena, A. F. P. d.; Borba, B. S. M. C.;
Nogueira, L. P. P.; Fleming, F. P.; Troccoli, A.; Harrison, M.;
Boulahya, M. S. Energy sector vulnerability to climate change: A
review. Energy 2012 , 38 ,1 −12.
(19) Vliet, M. T. H. v.; Yearsley, J. R.; Ludwig, F.; Vögele, S.;
Lettenmaier, D. P.; Kabat, P. Vulnerability of US and European
electricity supply to climate change. Nat. Climate Change2012 , 2 ,
(20) Chandel, M. K.; Pratson, L. F.; Jackson, R. B. The potential
impacts of climate-change policy on freshwater use in thermoelectric
power generation. Energy Policy 2011 , 39 , 6234 − 6242.
(21) Estimating Freshwater Needs to Meet Future Thermoelectric
Generation Requirements; Technical Report DOE/NETL-400/2008/
1339; U.S. National Energy Technology Laboratory: Morgantown,
WV, 2008; lpower/ewr/pubs/
(22) Carrillo, A. M. R.; Frei, C. Water: A key resource in energy
production. Energy Policy 2009 , 37 , 4303 −4312.
(23) Davies, E. G. R.; Kyle, P.; Edmonds, J. A. An integrated
assessment of global and regional water demands for electricity
generation to 2095. Adv. Water Resour.2013 , 52 , 296 −313.
(24) Siddiqi, A.; Anadon, L. D. The water-energy nexus in Middle
East and North Africa. Energy Policy 2011 , 39 , 4529 −4540.
(25) Scott, C. A. The water-energy-climate nexus: Resources and
policy outlook for aquifers in Mexico. Water Resour. Res. 2011 , 48 (6);
doi: 10.1029/2011WR010805.
(26) Hardy, L.; Garrido, A.; Juana, L. Evaluation of Spain ’s water-energy nexus. Water Resour. Dev.2012 , 28 , 151 − 170.
(27) Klein, G.; Krebs, M.; Hall, V.; O’Brien, T.; Blevins, B. B.
California ’s Water-Energy Relationship ; CEC-700-2005-011-SF; Cal-ifornia Energy Commission: Sacramento, CA, 2005;
gov/2005publications/CEC-700 -2005-011/CEC-700-2005-011-SF.
(28) Scott, C. A.; Pierce, S. A.; Pasqualetti, M. J.; Jones, A. L.; Montz,
B. E.; Hoover, J. H. Policy and institutional dimensions of the water-energy nexus. Energy Policy 2011 , 39 , 6622 −6630.
(29) Stillwell, A. S.; King, C. W.; Webber, M. E.; Duncan, I. J.;
Hardberger, A. The Energy-Water Nexus in Texas.Ecol. Soc. 2011 , 16
(1), 2.
(30) The Water-Energy Nexus in the American West ; Kenney, D. S.;
Wilkinson, R., Eds.; Edward Elgar Publishing, Inc.: Northampton,
(31) World Development Indicators 2012; World Bank: Washington,
DC, 2012; files/wdi-2012-ebook.pdf.
(32) NBSC (National Bureau of Statistics of China). MEP (Ministry
of Environmental Protection) China Statistical Yearbook on Environment
2010 ; China Statistics Press: Beijing, 2011.
(33) NBSC. China Energy Statistical Yearbook 2010 ; China Statistics
Press: Beijing, 2011.
(34) Key World Energy Statistics 2012; International Energy Agency:
Paris, 2012; ions/fre epublicat ions/publi cation/
(35) Hoekstra, A. Y.; Chapagain, A. K.; Aldaya, M. M.; Mekonnen,
M. M. The Water Footprint Assessment Manual: Setting the Global
Standard; Earthscan: London, 2011.
(36) Pfister, S.; Koehler, A.; Hellweg, S. Assessing the environmental
impacts of freshwater consumption in LCA. Environ. Sci. Technol.
2009 , 43 , 4098 −4104.
(37) Wiedmann, T.; Lenzen, M.; Turner, K.; Barrett, J. Examining
the global environmental impact of regional consumption activities -Pa rt 2: Rev ie w of in pu t- ou tpu t mo de ls for th e as se ssm en t of
environmental impacts embodied in trade. Ecol. Econ. 2007 , 61 ,15−
(38) Wiedmann, T. A review of recent multi-region input− output
models used for consumption-based emission and resource account-ing. Ecol. Econ. 2009 , 69 , 211 − 222.
(39) Wiedmann, T.; Wilting, H. C.; Lenzen, M.; Lutter, S.; Palm, V.
Quo Vadis MRIO? Methodological, data and institutional require-ments for multi-region input−output analysis.Ecol. Econ.2011 , 70 ,
1937 −1945.
(40) Ewing, B. R.; Hawkins, T. R.; Wiedmann, T. O.; Galli, A.; Ercin,
A. E.; Weinzettel, J.; Steen-Olsen, K. Integrating ecological and water
footprint accounting in a multi-regional input− output framework. Ecol.
Indic. 2012 , 23 ,1 −8.
(41) Peters, G. P.; Minx, J. C.; Weber, C. L.; Edenhofer, O. Growth
in emission transfers via international trade from 1990 to 2008. Proc.
Natl. Acad. Sci., U. S. A.2011 , 108, 8903 −8908.
(42) Lenzen, M.; Kanemoto, K.; Moran, D.; Geschke, A. Mapping
the structure of the world economy. Environ. Sci. Technol. 2012 , 46 ,
8374 −8381.
(43) Davis, S. J.; Peters, G. P.; Caldeira, K. The supply chain of CO
emissions. Proc. Natl. Acad. Sci., U. S. A.2011 , 108 (45), 18554 −
(44) Wiedmanna, T.; Wood, R.; Minx, J. C.; Lenzen, M.; Guan, D.;
Harris, R. A carbon footprint time series of the UK − results from a
multi-region input-output model. Econ. Syst. Res. 2010 , 22 ,19− 42.
(45) Liang, Q.-M.; Fan, Y.; Wei, Y.-M. Multi-regional input−output
model for regional energy requirements and CO
emissions in China.
Energy Policy 2007 , 35 , 1685 −1700.
(46) Zhang, Z.; Yang, H.; Shi, M. Analyses of water footprint of
Beijing in an interregional input−output framework.Ecol. Econ. 2011 ,
70 , 2494 − 2502.
(47) Feng, K.; Siu, Y. L.; Guan, D.; Hubacek, K. Assessing regional
virtual water flows and water footprints in the Yellow River Basin,
China: A consumption based approach.Appl. Geogr. 2011 , 32 , 691 −
(48) Weisz, H.; Duchin, F. Physical and monetary input−output
analysis: What makes the difference? Ecol. Econ. 2006 , 57 , 534 − 541.
(49) Hawkins, T.; Hendrickson, C.; Higgins, C.; Matthews, H. S. A
mixed-unit input-output model for environmental life-cycle assessment
and material flow analysis. Environ. Sci. Technol. 2007 , 41 , 1024 − 1031.
(50) Batten, D. F.; Boyce, D. E. Spatial interaction, transportation,
and interregional commodity flow models. In Handbook of Regional
and Urban Economics , Vol. 1; Nijkamp, P., Ed.; North Holland:
Amsterdam, 1987; pp 357− 406.
(51) Erlander, S.; Stewart, N. F. The Gravity Model in Transportation
Analysis: Theory and Extensions ; VSP: Utrecht, The Netherlands, 1990.
(52) China Regional Input-Output Tables; China Statistics Press:
Beijing, 2011. (In Chinese)
(53) China Energy Statistical Yearbook 2007 ; China Statistics Press:
Beijing, 2008. (In Chinese)
(54) China Economic Census Yearbook; China Statistics Press: Beijing,
2010. (In Chinese)
(55) Dataset of China Pollution Source Census; China Environmental
Science Press: Beijng, 2011. (In Chinese)
(56) National 600 MWe-Scale Thermal Power Unit Benchmarking and
Competition Dataset ; China Electricity Council: Beijing, 2010. (In
(57) National 300 MWe-Scale Thermal Power Unit Benchmarking and
Competition Dataset ; China Electricity Council: Beijing, 2010. (In
(58) Zhang, Y.; Qi, S. China Multi-Regional Input-Output Models:
2002 and 2007 ; China Statistics Press: Beijing, 2012.
(59) Ridoutt, B. G.; Huang, J. Environmental relevancethe key to
understanding water footprints. Proc. Natl. Acad. Sci., U. S. A. 2012 ,
109 (22), E1424.
(60) Goedkoop, M.; Spriensma, R. The Eco-Indicator 99: A Damage
Oriented Method for Life Cycle Impact Assessment, Methodology Report;
e Consultants.: Amersfoort, the Netherlands, 2000.
(61) Kenny, J. F.; Barber, N. L.; Hutson, S. S.; Linsey, K. S.; Lovelace,
J. K.; Maupin, M. A. Estimated Use of Water in the United States in
2005 ; U.S. Geological Survey Circular 1344; U.S. Department of the
Interior: Washington, DC, 2009;
(62) Research Report on the 12th Year (2011− 2015) Plan of Electric
Sector; China Electricity Council: Beijing, 2012. (In Chinese)
Environmental Science & Technology Article | Environ. Sci. Technol. 2013, 47, 14459 − 14467 14466
(63) Wang, P. Innovation and development of direct air cooling
technology for thermal power generati on in China. Electrical
Equipment 2008 , 3, 270 −272 (In Chinese).
(64) Zhao, L.; Gallagher, K. S. Research, development, demon-stration, and early deployment policies for advanced-coal technology
in China. Energy Policy 2007 , 35 , 6467 − 6477.
(65) Liu, H.; Ni, W.; Li, Z.; Ma, L. Strategic thinking on IGCC
development in China. Energy Policy 2008 , 36 ,1 −11.
(66) Chen, W.; Xu, R. Clean coal technology development in China.
Energy Policy 2010 , 38 , 2123 − 2130.
(67) DiPietro, P.; Gerdes, K.; Nichols, C. Water Requirements for
Existing and Emerging Thermoelectric Plant Technologies; Technical
Report DOE/NETL-402/080108; U.S. National Energy Technology
Laboratory: M organtown, WV, 2008;
(68) Wang, Z.; Qin, H.; Lewis, J. I. China ’s wind power industry:
Policy support, technological achievements, and emerging challenges.
Energy Policy 2012 , 51 ,80− 88.
(69) Twelfth Five Year (2011−2015) Plan of Energy Development ;
China State Council: Beijing, 2013. (In Chinese)
(70) Grau, T.; Huo, Molin; Neuhoff, K. Survey of photovoltaic
industry and policy in Germany and China. Energy Policy 2012 , 51 ,
20 − 37.
(71) Hua, Z.; Wang, J.; Byrne, J.; Kurdgelashvili, L. Review of wind
power tariff policies in China. Energy Policy 2013 , 53 ,41−50.
(72) Li, H.; Chien, S.-H.; Hsieh, M.-K.; Dzombak, D. A.; Vidic, R. D.
Escalating water demand for energy production and the potential for
use of treated municipal wastewater. Environ. Sci. Technol. 2011 , 45 ,
4195 − 4200.
(73) Shen, M.; Han, M. Environmental pollution and water resources
utilization in coal-fired power industry. China Environ. Prot. Ind.2011 ,
1,43− 48 (In Chinese).
Environmental Science & Technology Article | Environ. Sci. Technol. 2013, 47, 14459 − 14467 14467