Format And Type Fonts

CCHHEEMMIICCAALL EENNGGIINNEEEERRIINNGG TTRRAANNSSAACCTTIIOONNSS

VOL. 45, 2015

A publication of

The Italian Association

of Chemical Engineering

www.aidic.it/cet
Guest Editors: Petar Sabev Varbanov, Jiří Jaromír Klemeš, Sharifah Rafidah Wan Alwi, Jun Yow Yong, Xia Liu

ISBN 978-88-95608-36-5; ISSN 2283-9216 DOI: 10.3303/CET1545074

Please cite this article as: Guo Z., Liu P., Ma L., Li Z., 2015, A grid-structure based multi-region optimisation model for the

development of power generation sector in china, Chemical Engineering Transactions, 45, 439-444

DOI:10.3303/CET1545074

439

A Grid-Structure Based Multi-Region Optimisation Model for

the Development of Power Generation Sector in China

Zheng Guo, Pei Liu*, Linwei Ma, Zheng Li

State Key Lab of Power Systems, Department of Thermal Engineering, Tsinghua University, Beijing,100084, China

liu_pei@tsinghua.edu.cn

Power demand in China has increased over years and is expected to further expand in the future. The

disparity of resources distribution and load dispatch for regions requires regional analyses rather than

taking China as a single entity. Meanwhile subsidy policies for clean energy and power prices will largely

affect the development of the power sector in a long-term. In this paper, a multi-region model based on

grid structures is built aiming for maximising profits gained by China’s power generation sector from 2013

to 2050. A case study is performed demonstrating deployment trends for technologies and the utility of

transmission capacity. The results indicate different developing strategies for regions and the importance

of energy subsidies. In addition, power transmission utility is analysed and discussed.

1. Introduction

China has experienced a rapid economic growth as well as sharp increase in power demand, from 1,347.3

TWh in 2000 to 4,976.8 TWh in 2012 (NBSC, 2014). Electricity demand was expected to further increase

in the future (Hu et al., 2009) and pathways for the development of power capacity are worth studying.

Many studies focus on analysing China’s power sector by taking the whole country as a single entity. Cai

et al. (2007) analysed the development of China’s power sector in various policy scenarios by the Long-

range Energy Alternatives Planning System (LEAP) and gave CO2 reduction potential. Zhu and Fan (2010)

adapted portfolio theory to optimise China’s power sector in three scenarios. Zhang et al. (2012)

developed a model optimising costs for China’s power sector while considering carbon markets and CCS

technologies. However, these studies do not distinguish regional natural resources and power demand

differences.

Gnansounou and Dong (2004) considered power transmission between two provinces in China and

integrate their electricity markets. Wang and Nakata (2009) divided China into coastal and inland areas

and discussed the development of clean coal technologies concerning different environmental policies.

Notably, these simplified divisions cannot reflect the actual situation of China’s power sector. Cheng et al.

(2015) divided China into 10 parts based on physical grid structures and minimised total costs with the

consideration of electricity transmission. Nevertheless, this work does not consider transmission capacity

and cross-region transmission. Though studies for India (Parikh and Chattopadhyay, 1996) and Greece

(Koltsaklis et al., 2014) took regional variations and electricity transmission into account, electricity markets

and policies in these countries are much different from China. Thus, this paper established a multi-region

model, considering actual grid structures and capacity, resource distribution and energy subsidy policies,

in order to give an insight into the long-term development of China’s power generation sector.

2. Methodology

2.1 Model structures and assumptions

Based on the current physical structure of grids, China is divided into seven regions: Northeast, North,

Central, East, South, Northwest and Tibet (Zhou et al., 2010). Among these grids, Tibet grid, as well as

Hainan, Hong Kong, Macau and Taiwan, are relatively independent ones and their load are insignificant

compared to the others, thus they are not considered in this model. In addition, natural resource

440

endowment are important. According to the Ultra-high Voltage (UHV) transmission projects proposed by

the State Grid Corporation of China (SGCC, 2014), Inner Mongolia and Xinjiang will become main

electricity exporting provinces due to their abundant coal reserves and developable renewable resources.

Owing to these considerations, China is modelled as 8 regions reflecting physical grid structures as well as

natural resources. As shown in Figure 1, 8 regions are as follows: Northeast (Heilongjiang, Jilin and

Liaoning), North (Beijing, Tianjin, Hebei, Shanxi and Shandong), Inner Mongolia, Central (Jiangxi, Hubei,

Hunan, Henan, Sichuan and Chongqing), East (Shanghai, Jiangsu, Zhejiang, Anhui and Fujian), South

(Yunnan, Guizhou, Guangxi and Guangdong), Northwest (Shaanxi, Gansu, Ningxia and Qinghai), and

Xinjiang.

Figure 1: Regional division of China based on grid structures and natural resources

In terms of power generation technologies, 7 types are considered: pulverised coal (PC), ultra-supercritical

coal (USC), natural gas combined cycle (NGCC), nuclear (NU), hydroelectric (HD), wind power (WD) and

photovoltaic (PV). These technologies have the potential to be largely deployed across the country (The

State Council, 2013). Based on their different feed-in tariff policies, they are divided into two categories.

The first one includes PC, USC, NGCC and HD, whose on-grid prices are regulated by the government at

any time. The second category includes NU, WD and PV, whose on-grid prices are either a strike price or

a feed-in tariff decided by the time when the plant is put into operation. Due to the lack of CCS

experiences and current governmental policies nationwide, CCS equipment is not considered in this

model.

In order to introduce power transmission between regions, this model considers currently built extra-high

and ultra-high voltage power transmission lines, proposed ultra-high voltage power transmission lines and

their transmission losses. International transmission is neglect due to its relatively low amount compared to

national power demand (NBSC, 2012).

2.2 Mathematical equations
Mathematical equations of the model are presented in this section. These equations can be classified into

two groups. The first group includes objective function and expressions of its relative variables. The

second group indicates physical constrains. Four sets, t, r, g and f, stand for time, region, power

generation type and fuel type, respectively. Meanwhile, t and t’, r and r’ share the same set in the

equations.

2.2.1 Objective function
The objective function of this model is to maximise accumulated profits gained by power sector from 2013

to 2050. As expressed in Eq(1), the sum of regional profits is leveraged to 2013 and totted up.

, ,2050

( 2013)
2013

( )

(1 )

t r t r

t
t

rreve rc

atc
I










 (1)

Profits equals to revenues less costs. Costs consist of three parts: capital costs for constructing new

capacity, operation and maintenance costs, and fuel costs, and can be calculated by Eq(2).

, , , ,rc tinv tom tfct r t r t r t r   (2)

2.2.1.1 Regional revenues
For the first and second groups of technologies, revenues can be calculated by Eq(3) and Eq(4).

Operating hours are referred to statistical materials (SERC, 2011) and their on-grid prices consider

441

governmental documents, including the latest feed-in tariff policies for wind (NDRC, 2014b) and solar PV

(NDRC, 2013). On-grid prices for the first group are decided by the time when new capacity is added.

Those for the second group are regulated by the government at any time.

, , , ', , , , '

' 1g

t r g g t r r g g r t

t t TLT

rreve nbc OH OGP
  

  

(3)

, , , , , , ,t r g g t r r g g r t
rreve ic OH OGP   (4)

2.2.1.2 Regional costs
Three parts of costs can be calculated by Eqs(5) to (7), respectively. Capital costs referring to OECD’s

report (OECD, 2010) are leveraged to annum. Total lifetimes (TLT) are referred to IEA’s report (OECD,

2010). O&M costs are assumed to be part of capital costs. Fuel costs are decided by Fuel Price (FP), Fuel

Consumption Rate (FCR) and the power generated.

 
, , , ',, '
(1 ) 1 (1 )' 1

t I
inv CAP nbg t r g t rg t TLTgI It t TLTg

 
 

    


       
 

(5)

 , , , , ,tom om ict r g t r g g t r
g g

   

(5)

   , , , , , , , , , , , , , ,, , ,,tfc fc FP rfd FP fd FP pg FCRf t r f t r f t r f t r f t r g t r f g tg f t rt r
f f ff g g

   
              
   
   

(7)

2.2.2 Physical constrains

2.2.2.1 Power demand and supply
Regional Power Demand (PD) is met by electricity generated in its own region and transmitted in or out

(losses are included). Ideal transmission power from r’ to r equals to the negative value of that from r’ to r,

and is limited by transmission capacity(SGCC, 2014). These expressions can be written as Eqs(8) to (11).

, , , , , , , ,PD pg ttr ic OH ttrt r g t r t r g t r r g t r
g g

     
(8)

, , '

, , ' [1 ], ' , , '

itrt r r
ttr

t r r TRLOSS itrr r t r r


 





(9)

, , ' , ',
itr itr

t r r t r r
 

(10)

, , ' , ',
itr TRLIMIT

t r r r r t


(11)

Regional power demand in year t is extrapolated based on historical power demand (NBSC, 2012) as well

as governmental projections (Hu et al., 2009). Transmission loss rates between regions are calculated

referring studies on UHV transmission (Zhao et al., 2009) and actual distances.

2.2.2.2 Installed capacity
This model assumes that all types of technologies will be decommissioned at the end of their lifespan.

Thus the installed capacity of type g in region r in year t can be expressed as Eq(12).

, , , ',
' 1

t
ic nbc

g t r g t r
t t TLT

 
  

(12)

Installed capacity for existing plants refer to statistical books (China Electricity Council, 2009) and annual

reports (EBCEPY, 2013). Considering resource endowment, maximum regional installed capacity for

renewable (CAE, 2010) limits their deployment, as shown in Eq(13). New built capacity is constrained by

construction practicality, referring to historical experiences (EIA, 2011), as noted in Eq(14). In addition,

targets for clean energy (NDRC, 2014a) are included. Furthermore, it is assumed no nuclear plants will be

added in Northwest, Xinjiang and Inner Mongolia due to its abundant coal reserves.

, , ,
ub

ic ICg t r g r (13)

(r is the exporting region)

(r is the importing region)

442

, ,
ub

nbc NBg t r g
r

 (14)

2.2.2.3 Fuel supply

Annual total fuel demand must not exceed Fuel Supply Capacity (FSC), as expressed in Eq(15). The

capacity is both limited by domestic productivity (NBSC, 2013) and international availability (IAEA, 2009).

,
ub

tfd FSCf t f
r


(15)

2.2.2.4 Assumptions for economical parameters

Capital costs for technologies are expected to decline gradually through to 2050 with different rates. The

model assumes that on-gird prices for coal-fired plants will remain stable, and those for others will be

regulated by policies (The State Council, 2014). Fuel prices take the current coal and gas prices (China's

Chemical Products Web, 2015) and are set to be flat. Discount rate is assumed to be at 7 %.

3. Results and discussions

The Linear Programming Solver of the General Algebraic Modelling System (GAMS) was used in this

paper for modelling and optimisation.

3.1 Installed capacity for technologies in regions
Installed capacity in North, East, Central and South regions will share some similarities as shown in Figure

2. Hydroelectric power will be exploited, notably in Central and South. Solar and wind will not be well

deployed due to insufficient subsidy policies. In order to meet increasing power demand, thermal power

plants will remain the majority and USC plants will gradually replace PC plants through to 2050. NGCC will

not be deployed in these areas concerning high gas prices. Nuclear power will have the priority to develop

in the East to fulfil the national clean energy targets as well as meet its demand by 2050.

Figure 2: Installed capacity for technologies in East, North, Central and South

In terms of regions with abundant wind and solar sources, wind power will develop rapidly by 2020 in

Northeast and Northwest because of the national clean energy targets as well as the cease of feed-in tariff

for wind by 2020, and though to 2050 in Xinjiang owing to declining capital costs (shown in Figure 3). PV

will experience rapid development by 2020 in Xinjiang and Inner Mongolia, but no new addition in

Northeast and Northwest due to insufficient subsidies. These indicate the importance of subsidies for wind

and solar power as well as the strong national targets. PC will remain popular in Northwest, Xinjiang and

Inner Mongolia until 2035 owing to low coal prices. NGCC will expand in the three areas because of low

gas prices and subsidies for gas plants.

Uncertainties for fuel prices and capital costs may largely affect the results. For instance, if gas price could

drop significantly in the future, install capacity for NGCC would increase notably compared to the case

performed above. On the other hand, if capital costs for wind and solar could decline faster than the rate

set in the case, more capacity would be installed. However, the paper aims at performing a case study

based on reasonable assumptions and giving insights into the power sector rather than predicts the future.

443

Figure 3: Installed capacity for technologies in Northeast, Northwest, Xinjiang and Inner Mongolia

3.2 Power transmission

Actual transmitted power is compared to theoretical transmission capacity to examine the utility of

proposed UHV lines. By analysing the results between 2040 and 2050, some transmission lines will reach

its maximum capacity. These are from North to East, Central to East and South, Northwest to East and

North, and Inner Mongolia to Northeast (black arrows in Figure 4). The rest is unused, from Xinjiang to the

other regions, North to Central, Northwest to Central, Northeast to North, and Inner Mongolia to East and

Central. East is load centred, and will receive power from other regions due to its high fuel prices. Central

area tends to export power rather than import due to its abundant hydropower. For the three resource-rich

regions, Northwest is the only one exporting large amount of power. Xinjiang and Inner Mongolia lacks the

momentum to export power due to relatively low on-grid prices and long-distance transmission losses.

Figure 4: Power transmission schematics between 2040 and 2050

4. Conclusions

In this study, a model reflecting natural resource endowment and grid structures is built aiming for

maximising profits for power generation sector in China. Various energy subsidy policies are considered in

order to deliver pathways for the development from 2013 to 2050.

For thermal power, USC will gradually replace PC, but relatively late in Northwest, Xinjiang and Inner

Mongolia where coal is relatively cheap, and NGCC will only develop in these three regions due to low-

cost gas. With respect to clean energy, hydropower will be fully exploited. Energy subsidies and national

targets are important to the deployment of wind and PV both in the short term and long term. Nuclear will

be competitive in the East and Northeast due to high regional prices for fossil fuel.

Utility of transmission lines between regions varies significantly. For load-centred regions with high fuel

prices, such as East, transmission capacity will be fully used. Regions with abundant hydropower like

Central will tend to dispatch its regional power rather than importing power. With respect to resource-rich

regions, subsidies for renewable and incentives for coal plants are required to stimulate power exporting.

444

References

CAE(Renewable energy development strategy research group of Chinese Academic of Engineering),

2010, Books of China's renewable energy development strategy research-comprehensive volume (in

Chinese), China Electric Power Power Press, Beijing, China.

Cai W., Wang C., Wang K., Zhang Y., Chen J., 2007, Scenario analysis on CO2 emissions reduction

potential in China's electricity sector, Energy Policy, 35, 6445-6456.

Cheng R., Xu Z., Liu P., Wang Z., Li Z., Jones I., 2015, A multi-region optimization planning model for

China’s power sector, Applied Energy, 137, 413-426.

China's Chemical Products Web, 2015, Coal Prices for cities

<www.chemcp.com/news/201501/529762.asp>, accessed 20.01.2015.

China Electricity Council, 2009, Collection of statistics information on China's power industry in 2008 (in

Chinese), China Electricity Council, Beijing, China.

EBCEPY(Editorial Board of China Electric Power Yearbook), 2013, China electric power yearbook 2011 (in

Chinese), China Electric Power Press, Beijing, China.

EIA(U.S. Energy and Information Administration), 2015, International energy statistics, EIA, Washington

DC, USA<www.eia.gov/cfapps/ipdbproject/IEDIndex3.cfm >, accessed 25.02.2015.

Gnansounou E., Dong J., 2004, Opportunity for inter-regional integration of electricity markets: the case of

Shandong and Shanghai in East China, Energy Policy, 32, 1737-1751.

Hu X., Chen L., Lei H., 2009, China's Low carbon development pathways by 2050-Scenario analysis on

energy demand and carbon emissions (in Chinese), Science Press, Beijing, China.

IAEA (International Atomic Energy Agency), 2009, Uranium: Resources, production and demand, OECD

NEA Publication 6891, 456, <www.oecd-nea.org/ndd/pubs/2014/7209-uranium-2014.pdf>, accessed

01/07/2015.

Koltsaklis N.E., Dagoumas A.S., Kopanos G.M., Pistikopoulos E.N., Georgiadis M.C., 2013, A

Mathematical Programming Approach to the Optimal Long-Term National Energy Planning, Chemical

Engineering Transactions, 35, 625-630.

NBSC (National Bureau of Statistics of China), 2012, China Statistical Yearbook 2012 (in Chinese), China

Statistics Press, Beijing, China.

NBSC, 2013, China Energy Statistical Yearbook 2013 (in Chinese), China Statistics Press, Beijing, China.

NBSC, 2014, China Statistical Yearbook 2012 (in Chinese), China Statistics Press, Beijing, China.

NDRC (National Development and Reform Commission), 2013, Notice on the use of price leverage to

promote the healthy development of solar PV industry (in Chinese), NDRC, Beijing, China.

NDRC, 2014a, China's national plan on climate changes (2014-2020) (in Chinese), NDRC, Beijing, China.

NDRC, 2014b, Notice on properly adjusting wind power feed-in tariff (in Chinese), NDRC, Beijing, China.

OECD, 2010, Projected costs of generating electricity, OECD, Paris, France.

Parikh J., Chattopadhyay D., 1996, A multi-area linear programming approach for analysis of economic

operation of the Indian power system,IEEE Transactions on Power Systems, 11, 52-58.

SERC(State Electricity Regulatory Commission), 2011, Annual Report on Electricity Regulation (2011) (in

Chinese). SERC, Beijing, China.

SGCC (State Grid Corporation of China), 2014, Projects, SGCC, Beijing, China

<www.sgcc.com.cn/ywlm/projects/index.shtml>, accessed 20.02.2015.

The State Council, 2013, The 12th five-year plan for energy development (in Chinese), The State Council,

Beijing, China.

The State Council, 2014, Energy Development Strategy Action Plan(2014-2020) (in Chinese), The State

Council, Beijing, China.

Wang H., Nakata T., 2009, Analysis of the market penetration of clean coal technologies and its impacts in

China's electricity sector, Energy Policy, 37, 338-351.

Zhang D., Liu P., Ma L., Li Z., Ni W., 2012. A multi-period modelling and optimisation approach to the

planning of China's power sector with consideration of carbon dioxide mitigation, Computers &

Chemical Engineering, 37, 227-247.

Zhao B., Shi X.F., Sun K., Zheng Y., Zhang H.Y., 2009, Economic Analysis of the Instance for UHV AC

Transmission, Proceedings of the CSEE, 29(22), 8-11.

Zhou X., Yi J., Song R., Yang X., Li Y., Tang H., 2010, An overview of power transmission systems in

China, Energy, 35, 4302-4312.

Zhu L., Fan Y., 2010. Optimization of China's generating portfolio and policy implications based on

portfolio theory, Energy, 35, 1391-1402.