Embodiment of virtual water of power generation in the electric power system in China

Applied Energy 151 (2015) 345–354

Contents lists available at ScienceDirect

Applied Energy journal homepage: www.elsevier.com/locate/apenergy

Embodiment of virtual water of power generation in the electric power system in China Xiaojie Zhu a, Ruipeng Guo a,⇑, Bin Chen b,c,⇑, Jing Zhang d, Tasawar Hayat c,e, Ahmed Alsaedi c a

College of Electrical Engineering, Zhejiang University, Hangzhou 310027, PR China State Key Laboratory of Water Environment Simulation, School of Environment, Beijing Normal University, Beijing 100875, PR China c NAAM Group, Faculty of Science, King Abdulaziz University, Jeddah, Saudi Arabia d Zhejiang Electric Power Corporation, Hangzhou 310007, PR China e Department of Mathematics, Quaid-i-Azam University, 45320 Islamabad, Pakistan b

h i g h l i g h t s  A virtual scarce water method is proposed to investigate the electricity power systems.  Virtual scarce water is transferred via power transmission system.  Virtual water flows from inland areas to coastal areas in the power system of China.

a r t i c l e

i n f o

Article history: Received 12 February 2015 Received in revised form 19 April 2015 Accepted 20 April 2015 Available online 15 May 2015 Keywords: Virtual scarce water Water intensity Electric transmission system China

a b s t r a c t Increasing severe water deficiency has led to an urgent need for better water resource management, especially in China. Electric power systems have been recognized as large water consumers; therefore, comprehensive analysis of their water use is needed. This study aims to analyze the flux and direction of virtual water and virtual scarce water within power system based on transmission–consumption water intensity (TCWI). A case study is then conducted to investigate China’s electric power system. The results show that including the water stress index (WSI) and virtual scarce flow concept largely influences the analysis of interregional virtual water flows. Regardless the WSIs, there are four regions exporting virtual water (northeast, north, northwest and central) and two regions exporting virtual scarce water (east and south). While considering the virtual scarce water, the central region becomes a big exporter with 144.12 GL of virtual scarce water outflow. In addition, the virtual water and virtual scarce water flux among these six regions reaches 726 GL and 163 GL, respectively. The electric transmission system transfers virtual scarce water from inland areas to coastal areas, which is roughly the opposite of the distribution of China’s water resources. The virtual water analysis incorporating the water scarcity not only largely increases the effectiveness of the results, but also provides more valuable and accurate information for water-efficient management and planning in electric power system. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction Water use grows rapidly with economic development, resulting in increasing water shortages. This is especially true for China, where the annual availability of renewable water resources per capita is only 25% of the world average [1]. Effective water resource

⇑ Corresponding authors at: 38 Zheda Street, Hangzhou 310027, PR China. Tel./fax: +86 571 87951542 (R. Guo), State Key Laboratory of Water Environment Simulation, School of Environment, Beijing Normal University, 19 Xinjiekouwai Street, Beijing 100875, PR China. Tel./fax: +86 10 58807368 (B. Chen). E-mail addresses: [email protected] (R. Guo), [email protected] (B. Chen). http://dx.doi.org/10.1016/j.apenergy.2015.04.082 0306-2619/Ó 2015 Elsevier Ltd. All rights reserved.

management is essential to ease China’s pressure on water resources [2]. Electric power systems are considered a major source of air pollutants, but their impact on water resources is often neglected [3]. With the biggest electric power system in the world, China’s electric power industry has consumed a great deal of water. Indeed, 79% of water withdrawal and 47% of water consumption of energy production in China in 2007 were associated with the electricity generation [4]. Furthermore, expected substantial increases in power generation capacity [5] will exacerbate the current water shortage. Thus, in-depth water use understanding within electric power systems is essential to improve water resource management in China.

346

X. Zhu et al. / Applied Energy 151 (2015) 345–354

Water use associated with different electricity generating technologies has received wide attention, particularly for the cooling systems of conventional plants that utilize heat to generate electricity [6–8]. Detailed descriptions and quantitative analyses of water use in various cooling systems show that water use varies widely among different systems [3,9]. Meanwhile, upstream water use plays a significant role in renewable plants without cooling systems, such as wind facilities [10]. Li et al. estimated the life cycle of wind power in China and found that indirect water use greatly affects the results of the water resource analysis [11]. Besides, water use for hydropower generation is very complicated since reservoirs usually have more than one function [12], such as water storage for agriculture or domestic use. Generally, only water evaporation is considered as the operational water consumption of hydropower plants, since it is hard to disaggregate the end uses of hydropower dam water into agriculture or domestic use [13]. In addition, local climatic differences have certain influences on water evaporation, which should be considered as a significant factor when water use associated with hydropower is investigated [14]. All these studies above focus on the power generation, while the power transmission is less investigated, which can be dealt with through the term ‘virtual water’. To achieve a more comprehensive water use analysis, Allan first proposed the term ‘virtual water’ in 1994, defining it as the water used to produce food crops that are traded internationally [15]. The concept was then further developed to include the volume of water required to produce a commodity or service [16,17]. Since then virtual water has been widely applied to water resource accounting on various scales, including cities [18–20], countries [21–23] and world [24,25]. All these analyses showed that virtual water accounting can bring comprehensive and accurate insights into water management issues. Recently, virtual water was also used to investigate the water issues within the specific engineering and technological systems or sectors, such as wastewater treatment system [26], building’s engineering system [27] and service sector [28]. Virtual water transfer concurrent with trade has shown a huge impact on water resource accounting [29–32]. Global virtual water trade network based on IO framework indicates that the number of trade connections and volume of virtual water trade have more than doubled in the past twenty years [33,34]. Incorporating water scarcity as a potential impact factor, Lenzen et al. [35] further provided new insight into the global virtual water trade. At national scale, the virtual water flows within the trade among UK and other three world regions have been investigated in [36,37]. The results of these estimates show that more than half of the UK’s consumption water footprint is from other countries through international trade, reflecting the importance of virtual water flows among different countries. Feng et al. [38] assessed virtual water flows in the Yellow River Basin (YRB) in China from a consumption perspective and found that activities outside the basin also influenced water resource use in the YRB. The provincial virtual water transfer in China was also evaluated by the multi-region input–output (MRIO) model, demonstrating that virtual water flows from northern areas to southern areas, and from underdeveloped areas to developed areas [39,40]. Similarly, virtual water can be used to analyze the water transfer associated with electric power systems, which reflects the life cycle assessment of goods and services for power generation in terms of water use [41]. Recently, life cycle water consumption by different types of power generation in China was analyzed by Feng et al. [42], where a hybrid method integrating the process-based life cycle analysis and the input–output life cycle analysis was used to calculate both the direct and indirect water consumption. Zhang et al. [4] used MRIO to analyze the regional life cycle water use associated with electric power industry. One

benefit is that the impact of interregional electric power transmission is considered. This study was conducted to investigate the water issue of electric power systems, particular the electric transmission systems, with a new perspective on electricity flows rather than economic flows. Transmission–consumption water intensity (TCWI) in the electric power system was defined to transform electricity flows into virtual water flows. The electric network theory was then used to describe the migration of virtual scarce water within the electric power systems in the matrix form. Besides, the water stress index (WSI) concept was incorporated to calculate the virtual scarce water flows, thereby bringing a more comprehensive virtual water analysis. Finally, a case study to quantify virtual water and virtual scarce water flows concurrent with electricity flows in China’s electric power system was conducted. 2. Method and data An example of an electric power system with five regions and electricity flows is shown in Fig. 1. A region can be considered a component containing four branches and one node (Fig. 2a). There are local branches and interregional branches. The former contains a generator branch, which represents the local power plants, as well as a load branch that represents the local power consumers. The latter contains branches that connect other regions, which can be divided into line-in and line-out branches. Since water is required to produce electricity, the virtual water will move with electricity in the same direction. Consequently, the whole electric power system can be considered as a topological network composed of nodes and branches (Fig. 2b). It should be noted that in the analysis regions could be cities, provinces or countries, with more accurate estimations being obtained for smaller regions. 2.1. Method For convenience, the description of the method focuses on one region marked by j, the node in it is also marked by j. It is assumed that there are K generators and P loads in region j, and M lines are connected with node j. 2.1.1. Generator branch The electricity flux of the generator branch in region j can be calculated as the total amount of power generation flows injected into node j,

Region 1

Region 3

Region 2

Region 4

Electricity flow Power grid

Region 5

Load Generator Fig. 1. Schematic diagram of electric power system.

347

X. Zhu et al. / Applied Energy 151 (2015) 345–354

Node 2 Node 1

Generator

Line-in

Node 4

Node Load

Line-out

Node 3 (a) Schematic diagram of regional electric power system

Electricity flow

Node 5

Virtual water flow

(b) Schematic diagram of electric interregional virtual water flows Fig. 2. Regional electric power systems associated with virtual water flows.

K X Gk

where Gk represents the power generation of the kth power plant (k ¼ 1; 2; 3 . . . ; K). The water consumption of the kth power plant can be calculated as

It should be noted that, when considering power loss, the volume of virtual water flux of a line is unique, while electricity flux has two different values in each line. That is to say, the virtual water loss concurrent with the power loss in one line is allocated to its input node. The virtual water flux flowing into node j via the line-in branch can be represented as

wGk ¼ DGk  Gk

wnLinj ¼

enGj ¼

ð1Þ

k¼1

ð2Þ

where DGk represents the generation water intensity (GWI) (the water consumption per unit of electricity generated) of the kth power plant. Thus, the virtual water flux of the generator branch in region j can be represented as

wnGj ¼

K X wGk

In the same way, the virtual water flux out of node j through the line-out branch can be calculated as

X wLm

wnLoutj ¼

ð9Þ

2.1.3. Load branch Loads represent the electricity consumers, the sum of which represents the total amount of electricity consumption in a region. The electricity flux into the pth load (p ¼ 1; 2; 3 . . . ; P) is equal to its electricity consumption, which is represented by eLoadp . The electricity flux out of node j through the load branch can be represented as

enLoadj ¼

P X eLoadp

ð4Þ

The electricity flux into node j by the line-in branch can be represented as

X eLOm

ð8Þ

ð3Þ

2.1.2. Line branch The electricity flux of the mth line (m ¼ 1; 2; 3 . . . ; M) can be attained by electricity flow calculation, and a power loss DeLm will be generated when electric power is transferred through line m. Assuming eLIm and eLOm represent the electricity flux flowing in and out of the mth line, respectively, the relationship between them can be represented as

enLinj ¼

wLm

m!j

j!m

k¼1

eLIm ¼ eLOm þ DeLm

X

ð5Þ

ð10Þ

p¼1

The consumption water intensity of the pth load, DCp , is defined as the ratio of virtual water flux to the electricity flux into the pth load. In this way, the virtual water flux into the pth load can be calculated from

m!j

where m ! j indicates that electricity flow is from the m th line to node j. Similarly, the electricity flux flowing out of node j through the line-out branch can be calculated from

enLoutj ¼

X eLIm

ð6Þ

wLoadp ¼ DCp  eLoadp

Thus, the virtual water flux out of node j through the load branch, which also represents the total water consumption (including the virtual water flows in electric power system) in region j, can be represented as

j!m

The transmission water intensity of the mth line, DTm , is defined as the ratio of its virtual water flux to the electricity flux that flows into it. Thus, the virtual water flux of the mth line can be calculated from

wLm ¼ DTm  eLIm

ð7Þ

ð11Þ

wnLoadj ¼

P X wLoadp

ð12Þ

p¼1

2.1.4. Nodes Nodes are defined as parts of the system that do not generate or consume any electricity or water. In the other words, the flux into

348

X. Zhu et al. / Applied Energy 151 (2015) 345–354

node j should be equal to the flux out of node j, which can be represented as

einj ¼ enGj þ enLinj ¼ eoutj ¼ enLoadj þ enLoutj

ð13Þ

winj ¼ wnGj þ wLinj ¼ woutj ¼ wnLoadj þ wnLoutj

ð14Þ

where einj and winj represent the electricity flux and virtual water flux into node j, respectively; eoutj and woutj represent the electricity flux and virtual water flux out of node j, respectively. Nodes are characterized by the ratio of virtual water flux to electricity flux. Regarding node j, it can be represented as

Dnj ¼ winj =einj ¼ woutj =eoutj

ð15Þ

It should be noted that electricity becomes mixed when converging on one node. Thus, in region j the transmission water intensity of lines that belong to the line-out branch equals to the consumption water intensity of the loads

wnLoutj =enLoutj ¼ wnLoadj =enLoadj ¼ DTm ¼ DCp

ðm 2 ðj !ÞÞ

ð16Þ

where m 2 ðj !Þ represents the m th line belongs to the line-out branch of region j. That is to say, electricity in the m th line flows out of node j. From Eqs. (13)–(16), we obtain

Dnj ¼ woutj =eoutj ¼ wnLoutj =enLoutj ¼ wnLoadj =enLoadj ¼ DTm ¼ DCp

ðm 2 ðj !ÞÞ

ð17Þ

It can be seen that Dnj is equal to the transmission water intensity and the consumption water intensity as well, which is hereby termed as transmission–consumption water intensity (TCWI) of node j. 2.2. Extension Since there is usually more than one region in practice, the electric network theory is used to extend the method considering multiple regions. Eqs. (13), (14) and (17) were then rewritten in matrix form as Eqs. (18)–(20):

Ein ¼ EG þ Ain  ELin

ð18Þ

W in ¼ W G þ Ain  W L

ð19Þ

^Lin  BÞ1  W L ¼ ðE ^in Þ1  W in D ¼ ðE

ð20Þ

where (^) is used to represent the diagonalization of a vector; Ain is a modified incidence matrix of node-to-branch, which is defined as Ain ¼ fxij g, xij ¼ 1 when the jth branch connects the ith node and electricity flows into the ith node; otherwise, xij ¼ 0; B is a modified incidence matrix of branch-to-node, which is defined as B ¼ fyij g, yij ¼ 1 when the ith branch connects the jth node and electricity flows out form the ith node; otherwise, yij ¼ 0. Consequently, the nodal TCWI is

h i1 ^in Þ1  Ain  E ^Lin  B ^in Þ1  W G D ¼ I  ðE  ðE

ð21Þ

where I is the identity matrix. Then, the virtual water flow vector, W L , can be calculated by Eq. (20) as

^Lin  B WL ¼ D  E

ð22Þ

Water stress is commonly defined as the ratio of total annual freshwater withdrawals to total freshwater availability. The water stress index (WSI) concept defined and advanced by Pfister et al. is adopted in this paper to quantify the different impacts of the same amount of water consumption in water-rich and water-scarce

regions on local water resources and ecosystems [40,43] The threshold of WSI between moderate and severe water stress is set as 0.5 [43]. In this way, the virtual scarce water flow vector, W SL , can be calculated as [40]

^ L  B  WSI W SL ¼ W

ð23Þ

where WSI represents the vector of the regional water stress indexes. To obtain the regional total water consumption vector, W Load , Eq. (12) can be rewritten in matrix form as

W Load ¼ W G þ A  W L

ð24Þ

where A is an incidence matrix of node-to-branch, which is defined as A ¼ fxij g. When the jth branch connects the ith node, xij =1/-1 if electricity flows to/from the ith node, or xij ¼ 0 otherwise. Similarly, the regional total scarce water consumption vector, W SLoad , can be calculated as

d  W G þ A  W SL W SLoad ¼ WSI

ð25Þ

2.3. Data and scope China’s electric power system has been divided into six main regions as northeast, north, northwest, central, east and south [44]. When compared with the generation capacity and electricity consumption in other areas, those in Tibet and Taiwan are quite small [1], and their connections with other regions are very weak [44]; therefore, they were not included in this study. The major electricity flows among these six regions in 2010 are depicted in Fig. 3. The electricity flow volumes shown in the figure are the algebraic sums, as the interregional electricity flows changed their directions several times in 2010. It should be noted that the volumes reported here represent the electricity flux flowing into lines. Water resource data are also depicted to provide a background of China’s water resource distribution. It should be noted that Inner Mongolia is divided into two parts, i.e., east and west parts, the east part containing Chifeng, Hulunbeier, Xingan League and Tongliao belong to the northeast, while the rest cities belong to the north [45,46]. Accordingly, the water resources and population of Inner Mongolia are also divided into two parts. It is widely believed that there is a water shortage in the northwest owing to the extremely uneven distribution of water. However, water is actually quite abundant in some parts of this region, such as Qinghai, where there is much as 13,225 m3 of water per capita in 2010. Nevertheless, water is very scarce in other regions, such as Ningxia, where there is only 148.2 m3 per capita in 2010. In this study, we consider the northwest as a homogeneous entity, resulting in a value of 2500 m3 water per capita, which is relatively high in China. To make the results more accurate, in-station power consumptions for power generation are listed in Table 1. In addition, losses in power distribution are considered a part of the load, and the power loss rate on interregional connecting transmission lines is assumed to be 2% according to historical data. The electric power system was then modeled based on the proposed method (Fig. 4), in which, every region is composed of a power source unit that produces electric power, a load unit that represents regional terminal consumers and power loss during the power distribution, and an electric transmission system that delivers electric power from one region to another. Besides, in this study, we apply WSI calculation used by Feng et al. in [40] for these six main regions in China according to the methodology proposed by Pfister et al. [43]. In this case study, four types of electricity generation are considered, i.e., thermal, hydro, wind and nuclear. Other types of generation are omitted because of their small quantities (Table 2). The

349

X. Zhu et al. / Applied Energy 151 (2015) 345–354

Northeast

88.16 North

20.49 Northwest

26.85 123.87

Tibet

165.48

Central 401.14

East

234.24 South

Electricity flow (108 kWh)

Taiwan

Water resource (500m³/capita)

Fig. 3. Schematic diagram of electric power flows in China in 2010.

Table 1 In-station power consumption rate of China. Source: [44]. Item

Northeast

North

Northwest

East

Central

South

In-station consumption rate of thermal power (%) In-station consumption rate of hydropower (%)

7.21 0.85

7.28 0.56

5.89 0.7

5.26 0.31

5.96 0.28

6.37 0.42

basic GWIs are derived from [42] with an average value at the national scale considering all the life cycle water consumption. Since water consumptions in hydropower plants are largely dependent on local climatic differences, adjustments for GWI of hydropower are necessary when region scale is considered. The modified GWI of hydropower in the ith region, DHi can be calculated by solving Eq. (26) as follows:

! 8 6 X > < DHi =6 ¼ DHav erage > i¼1 : DHi =V i ¼ k

ð26Þ ði ¼ 1; 2; 3; 4; 5; 6Þ

where DHav erage is the national average GWI of hydropower; V i represents the average annual evaporation in the ith region, which can be derived from [47]; k is a constant, which is used to represent the direct proportion relationship. GWI of thermal power is also influenced by the regional difference due to various cooling technologies, e.g., more dry cooling technologies are used in inland areas in China [48]. Besides, in coastal areas, seawater is often used instead of freshwater in cooling systems [49,50], which should be excluded from this study [51]. The aforementioned issues are addressed by calculating the modified GWI of thermal power in the ith region, DTi through Eq. (27) as follows:

8 ! 6 X > > > RTi =6 ¼ ð1=2Þ  DTav erage > < i¼1

> RTi =DSi ¼ c > > > : DTi ¼ RTi þ ð1=2Þ  DTav erage

ði ¼ 1; 2; 3; 4; 5; 6Þ ði ¼ 1; 2; 3; 4; 5; 6Þ

ð27Þ

where RTi represents the direct GWI of thermal power in the ith region; DTav erage is the national average GWI of thermal power; DSi represents the GWI of thermal power in the ith region according to the data of sample units in [52,53] that contains detailed information about 456,300 MWe-Scale units and 323,600 MWe-Scale units; c is a constant, which is used to represent the direct proportion relationship. It should be noted that only water consumption during the operation of thermal power plants is considered [52,53], which is about half of life cycle water consumption of thermal power [42]. Thus, only half of GWI of thermal power is modified here, while the GWI associated with upstream remains unchanged. In China, nuclear power plants are located in coastal regions, and there are fifteen units in operation by 2012, thirteen of which use pressurized water reactor. Similar environment and technology lead to a similar water consumption of nuclear power in different regions. Additionally, water consumption of wind power is very small and has little relation to location. Accordingly, for nuclear and wind power, average value of GWIs at the country scale can be used to represent the regional values. The modified GWIs are listed in Table 3. The GWIs of thermal power in the south and east are much lower than those of other areas because many thermal plants use seawater as coolant in coastal areas. Except for the east and north, northwest has the lowest GWI of thermal power mainly due to the large application of dry cooling technology in thermal power plants there. Besides, the GWI of hydropower in the northwest is much higher than that in other areas, primarily because there is high wind and little

350

X. Zhu et al. / Applied Energy 151 (2015) 345–354

Northeast North Northwest Central East South

Transmission system Generator unit Load unit Electricity flow Fig. 4. Electric power network in China in 2010.

Table 2 Electric power generation data of China. Source: [48] Type

Northeast

North

Northwest

East

Central

South

Total

Thermal (108 kW h) Hydro (108 kW h) Nuclear (108 kW h) Wind (108 kW h) Others (108 kW h) Total (108 kW h)

2494 182 0 113 0 2789

10,396 70 0 266 0.04 10732.04

2769 822 0 44 0.03 3635.03

8646 729 414 42 0.46 9831.46

5127 3089 0 4 0.01 8220.01

4733 1961 334 17 0 7045

34,165 6853 748 486 0.54 42252.54

Table 3 Regional modified GWIs of China. Type

Northeast

North

Northwest

East

Central

South

Hydropower (L/kW h) Thermal (L/kW h) Nuclear (L/kW h) Wind (L/kW h)

14.82 3.83 3.10 0.56

18.69 3.53 3.10 0.56

25.92 3.46 3.10 0.56

15.53 2.60 3.10 0.56

14.87 4.08 3.10 0.56

16.96 2.18 3.10 0.56

vegetation, which results in about 50% more evaporation than other regions.

3. Results and discussion 3.1. Regional water consumption for different types of power generation The regional water consumption of different power generation types was presented in Table 4. Although the majority of national electric power is thermal power (Table 2), more than a half of national water consumption is caused by hydropower. This is especially obvious in the south, where hydropower only accounts for 27.8% of the power generation, but 75.57% of the water consumption. In addition, very little water is consumed by wind and nuclear power plants mainly due to the small generation capacity. However, it also can be seen that the power generation of wind power plants is equal to 1.42%, 7.09% and 64.97% of thermal power, hydropower and nuclear power plants, respectively, while the national water consumption of wind power plants is only 0.26%,

Table 4 Regional water consumption of different types of power in China. Region

Northeast North Northwest East Central South Total

Water consumption (108 L) Thermal

Hydro

Wind

Nuclear

Total

8863.32 34026.27 9016.43 21297.17 19671.44 9660.69 102535.33

2674.31 1300.97 21157.10 11286.27 45804.82 33118.87 115342.35

63.28 148.96 24.64 23.52 2.24 9.52 272.16

0.00 0.00 0.00 1283.40 0.00 1035.40 2318.80

11600.91 35476.21 30198.17 33890.37 65478.49 43824.48 220468.63

0.22% and 11.74% of them, indicating a significant water saving effect of wind power plants. Additionally, wind power is a clean energy that helps reduce carbon emissions [54]. Nevertheless, wind power is limited by technology and power security demand because the plants are weather-dependent [55,56]. More efforts should be made to increase the stability of wind power [57,58], which will lead to both reduced water consumption and carbon emissions associated with the electric power industry [59].

351

X. Zhu et al. / Applied Energy 151 (2015) 345–354 Table 5 Regional comparative efficiency of water use in power generation. Item 8

Regional power generation (10 kW h) Local water consumption (108 L) Regional water intensity (L/kW h) Comparative efficiency (%)

Northeast

North

Northwest

East

Central

South

2789 11600.91 4.16 79.47

10732 35476.21 3.31 100.00

3635 30198.17 8.31 39.79

9831 33890.37 3.45 95.89

8220 65478.49 7.97 41.50

7045 43824.48 6.22 53.14

Besides, it can be noted in Table 4 that local water consumption for local power generation is much different in various regions. Table 5 shows the regional comparative efficiency of water use in power generation. The comprehensive water intensity means the ratio of the local water consumption to the power generation. Comparative efficiency is derived by regarding the north as a benchmark since the north has the lowest comprehensive water intensity. The results shows that the northwest has the lowest comparative efficiency (39.79%), meanwhile the south and central region have comparative efficiencies about 50%. However, the north and east show relatively high comparative efficiencies above 95%. 3.2. Interregional virtual water and virtual scarce water flows concurrent with electricity flows The interregional virtual water and scarce water flows concurrent with electricity flows are depicted in Fig. 5. There are two virtual scarce water exporters (northeast and northwest), two virtual scarce water importers (east and south), and two virtual scarce water hubs (north and central). The northwest is the largest exporter with 55.51 GL of virtual scarce water, while the east is the largest importer with 75.40 GL of virtual scarce water. Besides, both the north and central regions have two electricity inflows and two electricity outflows. Electricity inflows bring in 27.99 GL of virtual scarce water concurrent with 10.865 TW h of electricity, and 47.47 GL of virtual scarce water concurrent with 15.072 TW h of electricity into the north and central regions. Meanwhile, the two electricity outflows remove 63.38 GL of virtual scarce water concurrent with 19.233 TW h of electricity, and 33.06 GL of virtual scarce water concurrent with 63.538 TW h of electricity from the north and central regions. From the Fig. 5, it also can be noted that the directions of virtual water flows and virtual scarce water flows are same, while their levels may be much different. It is especially true for the virtual water and virtual scarce water outflows of the central region. The virtual water flowing from the central region to the south reaches 186.4 GL, while the level of virtual scarce flow is less than one-tenth of that (12.2 GL). However, considering the north, the levels of its virtual water flows and virtual scarce water flows are very close. The virtual water flowing from the north to the central region is 8.9 GL, meanwhile the level of virtual scarce water flow is 8.8 GL. This is mainly due to the quite small level of WSI in the central region and the quite high level of WSI in the north region. The relatively abundant water resource should largely reduce the impact of the virtual water outflows, which is well expressed through the virtual scarce water flows. Besides, WSIs in different regions are also shown in Fig. 5. It can be seen that in the north and east, WSIs are relatively high (0.99 and 0.90 respectively), indicating an extremely severe water scarcity. However, in the south and central regions, the water resource is relatively abundant as their WSIs are below 0.1. Meanwhile, the northeast and northwest show a moderate water scarcity (below 0.5).The results also show that virtual scarce water flows in China’s electric power system are not necessarily from water-rich areas to water-scarce areas in 2010. As shown in Fig. 5, there are two virtual scarce water flows from the extremely water-scarce

north region to the relatively moist east and central regions. Besides, with the smallest WSI, south still imports 12.19 GL of virtual scarce water from the central region. Interregional virtual scarce water flows in China’s electric power system are not consistent with expectations based on water resource endowments. The spatial discrepancy of virtual scarce water flows and regional water resource endowments may be a result from the big domestic mismatch of electricity demand and supply [60]. Major primary energy sources for power generation in China are located far away from the power consumption centers. For example, coal is mainly located in the inland northwest, while most electricity generated by it is consumed in the eastern coast. Although much coal has been transported by train or by ship to the generators nearby the power consumption centers [61], a large amount of electricity is still generated near the coal production areas thereby leading to an enormous demand for power transmission, which is also the main reason for the ‘‘West–East Power Transmission Project’’ [62]. In 2008, more than 46% of the nation’s total coal output is from the three coal-rich inland regions, namely Inner Mongolia, Shanxi and Shaanxi. However, nearly 34% of total electric power is consumed in the four economically fast developing coastal regions, namely Shandong, Zhejiang, Guangdong and Henan [63]. Therefore, although inland areas are generally in a water-scarce situation, they have to consume more water to produce more electricity for other regions, forming the virtual scarce water flows from inland areas to coastal areas. Regarding the south, more than 727.4 TW h of electricity is needed while only 704.5 TW h of electricity is generated there in 2010. This big gap between power generation and power demand drives the huge virtual scarce water flux concurrent with electricity flux from the central region to south region. Besides, geographical location is also an important factor influencing the virtual scarce water flows. Considering the north, its position between northwest (or northeast) and east should be the main reason. Since there is no direct power transmission line between the east and northwest (or northeast), flow has to go through the north to transfer electricity from the northwest (or northeast) to the east. Generally speaking, the electric transmission system transfers virtual scarce water from inland areas to coastal areas, which is roughly the opposite of the distribution of China’s water resources. 3.3. Regional water and scarce water consumption The regional total water and scarce water consumption are shown in Table 6, as well as the regional virtual water and virtual scarce water influxes. It can be seen that there are four regions (northeast, north, northwest and central) exporting virtual water and two regions (east and south) importing virtual water. An interesting observation here is that the central region as a virtual water hub has exported much more virtual water than the other two virtual water exporters (northeast and northwest). Three Gorges Dam may be the main reason, which in 2010 exported 30.651 TW h and 13.696 TW h of electricity to the east and south, respectively. However, regarding the regional virtual scarce water influx, the largest virtual water exporter, the central region, becomes a virtual scarce water importer by importing 14.41 GL of virtual scarce water,

352

X. Zhu et al. / Applied Energy 151 (2015) 345–354

Fig. 5. Interregional virtual water and virtual scarce flows concurrent with electricity flows in China in 2010.

Table 6 Regional water and scarce water consumption of electric power system in China. Item

Northeast

North

Northwest

East

Central

South

Total water consumption (108 L) Total scarce water consumption (108 L) Virtual water influx (108 L) Virtual scarce water influx (108 L)

11234.21 3401.86 366.7 111.04

35374.08 34830.39 102.13 353.93

28998.88 10779.47 1199.29 555.07

37632.8 31704.57 3742.43 754.04

61539.92 4424.85 3938.57 144.12

45688.75 2769.77 1864.27 121.88

showing the significant influence of WSIs. Regardless of WSIs, the central region as a virtual water exporter may help other regions alleviate the water shortage. However, incorporating the WSIs, the central region becomes a virtual scarce water importer that would further exacerbate water shortage in other regions. It can be seen that a much different perspective can be obtained through incorporating the WSI and virtual scarce water flow concept. Besides, it can be noted that the central region has the largest total water consumption (6153 GL), while considering the WSIs, its total scarce water consumption is only 442 GL and at the same time the north becomes the largest consumer of scarce water (3483 GL).

3.4. Limitations In this paper, the electricity flow between any two regions has only one direction at the time interval of one year. At shorter time interval (e.g., seasons, months, days), however, the electricity flows will be much more complex, which may have more diverse virtual water and virtual scarce flow directions. Also, the regional generation water intensity is derived through modifying the national average value, which is somewhat rough. For example, only regional average annual evaporation is utilized to modify the generation water intensity of hydropower, while the reservoir areas are not

X. Zhu et al. / Applied Energy 151 (2015) 345–354

considered. More detailed data and model associated with electric power plants can improve the accuracy of current analysis. 4. Conclusions In this study, the water issue of China’s electric power system, particular the electric transmission system, is investigated by the proposed method. The results reveal that a large volume of virtual scarce water (16.3 GL) is transferred in China’s electric power system. It is found that virtual scarce water is transferred from inland areas to coastal areas, which is roughly the opposite of the distribution of China’s water resources. It is expected that electric power systems will comprise an increasing amount of renewable energy [64,65]. In 2014, the Asia Pacific Economic Cooperation (APEC) introduced a plan to double the share of renewables in the APEC energy mix by 2030. However, resources are always located far from load centers in China, including renewable energy resources. For example, hydropower resources are mainly in the south, while wind and solar power resources are primarily in the north, northwest, and northeast, and electricity consumption is concentrated in developed areas, such as the east and central regions. Additionally, unlike fossil fuels, renewable resources are largely dependent on the terrain or climatic conditions and therefore cannot be transported. The electric power transmission is inevitable when these renewable energy resources are utilized. Accordingly, interregional electric power transmission will grow as renewable energy increases, leading to more interregional virtual water and virtual scarce water flux in China. This study provides a feasible way to analyze the interregional virtual water and virtual scarce water flows in electric power systems, which may facilitate a better coordination of power supply and water resource management in China. Acknowledgements This work was supported by funds from Creative Research Groups of the National Natural Science Foundation of China (Grant No. 51121003), Major Research Plan of the National Natural Science Foundation of China (Grant No. 91325302), National Natural Science Foundation of China (Grant No. 41271543), and Specialized Research Fund for the Doctoral Program of Higher Education of China (Grant No. 20130003110027). References [1] NBS – National Bureau of Statistics of China. China Statistical Yearbook 2011. China Statistics Press, Beijing; 2011 [in Chinese]. [2] Jiang Y. China’s water scarcity. J Environ Manage 2009;90:3185–96. [3] Baum E, Chaisson J, Evans L, Lewis J, Marshall D, Thompson J. The last straw: water use by power plants in the arid west. Land Water Fund Rockies 2003. [4] Zhang C, Anadon LD. Life cycle water use of energy production and its environmental impacts in China. Environ Sci Technol 2013;47:14459–67. [5] International Energy Agency (IEA). World Energy Outlook 2009. Paris, France; 2009. [6] Macknick J, Newmark R, Heath G, Hallett KC. Operational water consumption and withdrawal factors for electricity generating technologies: a review of existing literature. Environ Res Lett 2012;7:045802. [7] Salazar JM, Diwekar U, Constantinescu E, Zavala VM. Stochastic optimization approach to water management in cooling-constrained power plants. Appl Energy 2013;112:12–22. [8] Tidwell VC, Macknick J, Zemlick K, Sanchez J, Woldeyesus T. Transitioning to zero freshwater withdrawal in the U.S. for thermoelectric generation. Appl Energy 2014;131:508–16. [9] Dziegielewski B, Bik T, Alqalawi U, Mubako S, Eidem N, Bloom S. Water use benchmarks for thermoelectric power generation. Southern Illinois University Carbondale; 2006. [10] Fthenakis V, Kim HC. Life-cycle uses of water in US electricity generation. Renew Sustainable Energy Rev 2010;14:2039–48. [11] Li X, Feng K, Siu YL, Hubacek K. Energy-water nexus of wind power in China: the balancing act between CO2 emissions and water consumption. Energy Policy 2012;45:440–8.

353

[12] Siddiqi A, Anadon LD. The water-energy nexus in Middle East and North Africa. Energy Policy 2011;39:4529–40. [13] Torcellini P, Long N, Judkoff R. Consumptive water use for US power production. Tech. Rep. NREL/TP-550-33905, Golden, CO; 2003. [14] Herath I, Deurer M, Horne D, Singh R, Clothier B. The water footprint of hydroelectricity: a methodological comparison from a case study in New Zealand. J Clean Prod 2011;19:1582–9. [15] Al Alawi J, Abdulrazzak M. Water in the Arabian Peninsula: problems and perspectives. Water in the Arab World: Perspectives and Progress, Division of Applied Sciences, Harvard University, Cambridge, Mass; 1994. [16] Allan JA. Water stress and global mitigation: water food and trade. Arid Lands Newslett 1999:45. [17] Hoekstra A, Weitsel A, Starink M. Perspectives on water: an integrated modelbased exploration of the future. Utrecht (the Netherlands): International Books; 1998. [18] Chen GQ, Li JS. Virtual water assessment for Macao, China: highlighting the role of external trade. J Clean Prod 2015;93:308–17. [19] Zhou SY, Chen H, Li SC. Resources use and greenhouse gas emissions in urban economy: ecological input–output modeling for Beijing 2002. Commun Nonlin Sci Numer Simul 2010;15:3201–31. [20] Han M, Guo S, Chen H, Ji X, Li J. Local-scale systems input-output analysis of embodied water for the Beijing economy in 2007. Front Earth Sci 2014;8:414–26. [21] Chen GQ, Chen ZM. Carbon emissions and resources use by Chinese economy 2007: a 135-sector inventory and input–output embodiment. Commun Nonlin Sci Numer Simul 2010;15:3647–732. [22] Chen ZM, Chen GQ, Zhou JB, Jiang MM, Chen B. Ecological input–output modeling for embodied resources and emissions in Chinese economy 2005. Commun Nonlin Sci Numer Simul 2010;15:1942–65. [23] Zhao X, Chen B, Yang ZF. National water footprint in an input–output framework—a case study of China 2002. Ecol Model 2009;220:245–53. [24] Chen Z-M, Chen GQ. Virtual water accounting for the globalized world economy: national water footprint and international virtual water trade. Ecol Ind 2013;28:142–9. [25] Hoekstra AY, Mekonnen MM. The water footprint of humanity. Proc Nat Acad Sci 2012;109:3232–7. [26] Shao L, Chen GQ. Water footprint assessment for wastewater treatment: method, indicator, and application. Env Sci Technol 2013;47:7787–94. [27] Meng J, Chen GQ, Shao L, Li JS, Tang HS, Hayat T, et al. Virtual water accounting for building: case study for E-town, Beijing. J Clean Prod 2014;68:7–15. [28] Li JS, Chen GQ. Water footprint assessment for service sector: a case study of gaming industry in water scarce Macao. Ecol Ind 2014;47:164–70. [29] Feng K, Chapagain A, Suh S, Pfister S, Hubacek K. Comparison of bottom-up and top-down approaches to calculating the water footprints of nations. Econ Syst Res 2011;23:371–85. [30] Guan D, Hubacek K. Assessment of regional trade and virtual water flows in China. Ecol Econ 2007;61:159–70. [31] Zhao X, Yang H, Yang Z, Chen B, Qin Y. Applying the input-output method to account for water footprint and virtual water trade in the Haihe River Basin in China. Environ Sci Technol 2010;44:9150–6. [32] Mubako S, Lahiri S, Lant C. Input–output analysis of virtual water transfers: case study of California and Illinois. Ecol Econ 2013;93:230–8. [33] Chen Z, Chen G, Xia X, Xu S. Global network of embodied water flow by systems input-output simulation. Front Earth Sci 2012;6:331–44. [34] Dalin C, Konar M, Hanasaki N, Rinaldo A, Rodriguez-Iturbe I. Evolution of the global virtual water trade network. Proc Nat Acad Sci 2012;109: 5989–94. [35] Lenzen M, Moran D, Bhaduri A, Kanemoto K, Bekchanov M, Geschke A, et al. International trade of scarce water. Ecol Econ 2013;94:78–85. [36] Feng K, Hubacek K, Minx J, Siu YL, Chapagain A, Yu Y, et al. Spatially explicit analysis of water footprints in the UK. Water 2010;3:47–63. [37] Yu Y, Hubacek K, Feng K, Guan D. Assessing regional and global water footprints for the UK. Ecol Econ 2010;69:1140–7. [38] Feng K, Siu YL, Guan D, Hubacek K. Assessing regional virtual water flows and water footprints in the Yellow River Basin, China: a consumption based approach. Appl Geogr 2012;32:691–701. [39] Zhang C, Anadon LD. A multi-regional input–output analysis of domestic virtual water trade and provincial water footprint in China. Ecol Econ 2014;100:159–72. [40] Feng K, Hubacek K, Pfister S, Yu Y, Sun L. Virtual scarce water in China. Environ Sci Technol 2014;48:7704–13. [41] Lenzen M. Understanding virtual water flows: a multiregion input-output case study of Victoria. Water Resour Res 2009;45:W09416. [42] Feng K, Hubacek K, Siu YL, Li X. The energy and water nexus in Chinese electricity production: a hybrid life cycle analysis. Renew Sustainable Energy Rev 2014;39:342–55. [43] Pfister S, Koehler A, Hellweg S. Assessing the environmental impacts of freshwater consumption in LCA. Environ Sci Technol 2009;43:4098–104. [44] CEP-Editorial Board of the China Electric Power Yearbook. China electric power yearbook 2011. Beijing: China Electric Power Press; 2011 [in Chinese]. [45] Inner Mongolia Water Resources Bulletin 2010. Department IMWR; 2011 [in Chinese]. [46] Inner Mongolia Statistical Yearbook 2011. Bureau IMAR, China Statistics Press, Beijing; 2011 [in Chinese]. [47] Ground meteorological data. China Meteorological Administration(CMA); 2011 [in Chinese].

354

X. Zhu et al. / Applied Energy 151 (2015) 345–354

[48] Statistical Data Collection of Electric Power Industry of 2010. Department of statistics and information, China Electricity Council; 2011 [in Chinese]. [49] Macknick J, Sattler S, Averyt K, Clemmer S, Rogers J. The water implications of generating electricity: water use across the United States based on different electricity pathways through 2050. Environ Res Lett 2012;7:045803. [50] Yu F, Chen J, Sun F, Zeng S, Wang C. Trend of technology innovation in China’s coal-fired electricity industry under resource and environmental constraints. Energy Policy 2011;39:1586–99. [51] Cai B, Zhang B, Bi J, Zhang W. Energy’s Thirst for Water in China. Environ Sci Technol 2014;48:11760–8. [52] National 300 MWe-scale thermal power unit benchmarking and competition dataset. China Electricity Council: Beijing; 2012 [in Chinese]. [53] National 600 MWe-scale thermal power unit benchmarking and competition dataset. Electricity Council: Beijing; 2012 [in Chinese]. [54] Desideri U, Yan J. Clean energy technologies and systems for a sustainable world. Appl Energy 2012;97:1–4. [55] Lin J, Sun Y-z, Cheng L, Gao W-z. Assessment of the power reduction of wind farms under extreme wind condition by a high resolution simulation model. Appl Energy 2012;96:21–32. [56] Hagspiel S, Papaemannouil A, Schmid M, Andersson G. Copula-based modeling of stochastic wind power in Europe and implications for the Swiss power grid. Appl Energy 2012;96:33–44.

[57] Morales JM, Mínguez R, Conejo AJ. A methodology to generate statistically dependent wind speed scenarios. Appl Energy 2010;87:843–55. [58] Wang J, Botterud A, Bessa R, Keko H, Carvalho L, Issicaba D, et al. Wind power forecasting uncertainty and unit commitment. Appl Energy 2011;88:4014–23. [59] Nanduri V, Saavedra-Antolínez I. A competitive Markov decision process model for the energy–water–climate change nexus. Appl Energy 2013;111:186–98. [60] Ma H, Oxley L, Gibson J, Li W. A survey of China’s renewable energy economy. Renew Sustainable Energy Rev 2010;14:438–45. [61] Mou D, Li Z. A spatial analysis of China’s coal flow. Energy Policy 2012;48:358–68. [62] Ming Z, Honglin L, Mingjuan M, Na L, Song X, Liang W, et al. Review on transaction status and relevant policies of southern route in China’s West-East Power Transmission. Renew Energy 2013;60:454–61. [63] Wang Q, Chen Y. Status and outlook of China’s free-carbon electricity. Renew Sustain Energy Rev 2010;14:1014–25. [64] Liu W, Lund H, Mathiesen BV, Zhang X. Potential of renewable energy systems in China. Appl Energy 2011;88:518–25. [65] Sayigh A. Renewable energy—the way forward. Appl Energy 1999;64:15–30.