The Spatial Deployment of Renewable Energy Based on China's Coal-heavy Generation Mix and Inter-regional Transmission Grid.
Increasing the utilization of renewable energy is crucial to coping with climate change. As the world's largest C[O.sub.2] emitter, China has realized the importance of developing renewable energy, and in the U.S.-China Joint Announcement on Climate Change set a target of 20% non-fossil fuels in primary energy consumption (renewables and nuclear energy) by 2030 (XNA, 2014). Renewable energy investment is highly capital-intensive, and once built, it creates a "lock-in" effect on the power generation mix for decades. Thus, whether the future low-carbon goals can be achieved depends on the current power system layout. The decision of where to invest in renewable energy, and to what extent, needs to be considered from a forward-looking perspective.
However, there are significant differences in resource endowments and electricity demand from region to region in China. Geographically, the potential for renewable energy is greatest in the north of China, while the power demand in the east and south is higher due to advanced economic development in those regions. The imbalance of resource endowments and electricity loads makes the spatial deployment of renewable energy more difficult.
In the north of China, a coal-heavy power supply structure limits the possibilities for integrating further renewable energy into the power grid. The large-scale application of coal-fired combined heat and power (CHP) has led to serious wind curtailment problems during winter, with a curtailment rate of 20.6% in 2016 (NEA, 2017). Compared with traditional coal-fired units, CHP unit operation is more inflexible because it must generate electricity when providing heating supply. During the winter, high heating demand in the north of China forces the CHP plants in district heating grids into must-run mode. The regional conflicts between the existing power generation mix and renewable energy are significant. Yet, with the increase of ultra-high voltage technology, inter-regional power transmission may be conducive to the consumption of renewable energy. Consequently, both the coal-heavy generation mix and inter-regional transmission planning affect the optimal spatial deployment of renewable energy.
The deployment of renewable energy at the spatial level in a manner that achieves stringent low-carbon target cost efficiency is of particular concern to China. Dong et al. (2016) conducted an empirical analysis of the distribution and cluster patterns of the country's renewable energy industry. Zhang et al. (2017) examined the regional development status of renewable energy, and made projections for the future expansion of renewables in China. Pei et al. (2015) analyzed wind curtailment problems from a temporal-spatial perspective. Their results showed that the present coal-wind-nuclear power structure in the north of China is one of the main reasons for curtailment. Based on the current literature, we find there is a need to implement comprehensive energy planning to analyze the spatial deployment of renewable energy and its related influencing factors.
However, describing fluctuating renewable energy in the energy system model is complicated, because it requires a relatively high-frequency temporal resolution, generally based on hours. Traditional power generation expansion planning (GEP) models are widely applied for energy planning from a bottom-up perspective (Bird et al., 2011; Sullivan et al., 2014; Wright and Kanudia, 2014; Chang and Li, 2015; Li and Chang, 2015). However, these models often simplify the fluctuating and seasonal features of renewable energy, and ignore the dispatch characteristics of generation units, such as ramping limits and start-up and shut-down decisions. These are typically addressed by unit commitment (UC) models (Kaleta and Toczylowski, 2008; Keane et al., 2011; Weigt el al., 2013; Egerer et al., 2015; Moarefdoost el al., 2016). Few studies integrate short-team power dispatch decisions into a long-term GEP framework (Koltsaklis and Georgiadis, 2015; Abrell and Rausch, 2016; Egerer et al., 2016; Bertsch et al., 2017). Pfluger (2014) expanded the electricity system model PowerACE to assess least-cost pathways for decarbonizing Europe's power supply. Scholz et al. (2017) used high temporal and spatial resolution information in an integrated assessment model (IAM) to understand the potential contribution of concentrating solar power to the balancing of renewable energy in Europe. Perez et al. (2016) used a co-optimization planning model to quantify the economic effects of trading renewable energy certificates among the U.S. states.
In the Chinese context, Li et al. (2016) simulated the Chinese power sector in 2030 based on an UC model, which models the dispatch of six regional power grids at hourly intervals. Guo et al. (2017) integrated seasonal and diurnal temporal variations into a GEP framework by dividing one year into twelve time-blocks. Chen et al. (2018) expanded to ninety-six time-blocks based on Guo et al. (2017) to capture the impact of different charging modes of electric vehicles on power system investment. He et al. (2016) presented a high-resolution integrated model that optimizes both the long-term investment and the short-term dispatch, but did not consider the characteristics of CHP.
To better understand the spatial deployment of renewable energy in China, this article makes three significant improvements compared to previous literatures by: (1) integrating hourly features of power dispatch, especially ramping limits, into traditional GEP model, which enables analyzing the impact of short-term power dispatch on long-term capacity investment; (2) modeling China's unique CHP operation mode in the energy system model to characterize the conflict between central heating supply and low-carbon energy transformation; and (3) analyzing the renewable energy deployment from a systematic perspective, which enables us to understand the inherent mechanism between power generation mix, power grid planning, and the large-scale utilization of renewable energy. Based on this framework, this study addresses two research questions: (1) what is the optimal spatial deployment of renewable energy to achieve relevant low-carbon targets? And (2) what are the effects of a coal-heavy generation mix and electricity transmission infrastructure expansion on the spatial deployment of renewable energy in China? Based on the analysis, policy recommendations will be developed for the Chinese government to make long-term strategic decisions about renewable energy deployment.
This paper is organized as follows. Section 2 provides a description of the conceptual framework. The detailed modeling, data, and scenario designs are presented in Section 3. Section 4 discusses the effects of a coal-heavy generation mix and electricity transmission infrastructure expansion on the spatial deployment of renewable energy. Section 5 presents the conclusions and recommendations.
2. CONCEPTUAL FRAMEWORK
2.1 Coal-heavy Generation Mix
Coal is the most important resource for power generation in China. In 2016, the proportion of the installed capacity of coal-fired power units was 57.3%. Coal power plants have more restrictive operational requirements than gas-fired power plants; for example, their ramping is slower. Furthermore, China has a large number of coal-fired CHP units providing district heating that cannot be shut down or easily modulated during the winter heating periods, which limits the flexibility of the power supply. The capacity of CHP reached 283 GW in 2014, accounting for 34% of the total coal-fired capacity (CEPYEB, 2015). In the Northern China and Northeast China power grids, the proportion exceeds 50%.
Figure 1 explains why the above-mentioned inflexible generation mix conflicts with renewable energy through an hourly electricity supply and demand curve. The difference between the daily load curve and the must-run capacity is the potential for other generation units. In an extreme case, even if no other units are dispatched, the amount of available renewable energy alone may exceed the above residual load, causing a renewable energy curtailment problem, especially for wind power. The feed-in profile of wind is typically high in winter mornings, while at this time the load curve is at the lowest point of the day. Due to this hourly mismatch of supply and demand, it is necessary to introduce a power dispatch model integrated with a power expansion model to better characterize renewable energy.
2.2. Inter-regional Electricity Trade
First, a stylized two-region model for electricity trade is presented as a reference for our empirical analysis (see Figure 2). Regions A and B are characterized by marginal cost supply curves ([MC.sub.A] and [MC.sub.B]) and demand curves ([D.sub.A] and [D.sub.A]). The step-wise supply curve is assumed to include two technologies: renewable energy with zero marginal cost and thermal power with different marginal costs [P.sub.A] and [P.sub.B]. Region A approximates the electricity supply characteristics of Northwest China, where thermal power is relatively cheaper and renewable energy has greater potential. In contrast, Region B reflects the characteristics of East China. Through mapping the demand curves to the supply curves, we can find the equilibrium electricity price and trade amount at a given time.
In time [T.sub.1], without regional trade, regions A and B meet demand at the equilibrium prices of zero (region A) and [P.sub.B] (region B). If inter-regional trade is made possible, regardless of the power transmission costs, the equilibrium price becomes [P.sub.A] by mapping [MC.sub.A+B] and [D.sub.A+B]. Region A becomes an electricity exporter and more renewable energy ([Q.sub.2]-[Q.sub.1]) is consumed due to the trade. But, electricity trade can increase the utilization of renewable energy mainly in situations of regional over-supply. For example, there is still a large amount of electricity traded from region A to B in time [T.sub.2], but renewable energy production is unchanged because demand is already higher than renewable energy supply in region A.
Due to central heating demands in the north of China, CHP effectively must operate at must-run capacity during winter. This can be represented in the model by lowering regional electricity demand to represent residual load (i.e. load supplied by non-baseload generators). Such a process is shown in Figure 2 by comparing [T.sub.3] to [T.sub.2] while decreasing demand in region A. The increase of must-run CHP results in renewable energy curtailment if there is no inter-regional transmission infrastructure; electricity trade could potentially lower or avoid this curtailment.
The purpose of this simple analysis is to illustrate the impact of electricity transfer and CHP on renewable energy. However, the influenced areas and the extent of the effects still need to be evaluated from the perspective of the national power system.
3.1. Model Description
3.1.1. Model Description and Assumptions
This study presents a power sector optimization model that integrates unit commitment into a long-term generation expansion planning framework. The illustrative structure of our model is shown in Figure 3. Hourly power dispatch, taking into account characteristics such as ramping limits and start-up and shut-down decisions, is combined with yearly investment decisions for power generation and transmission technologies. In our previous work, we established a multi-region power GEP model to quantify the inter-regional power network (Yi et al., 2016; 2019). However, this model could not describe the fluctuating and seasonal features of renewable energy, and ignored the dispatch of generation units. This study therefore modifies the previous model by adding an hourly-based power dispatch model.
The power generation expansion module makes an investment decision at the beginning of each year to decide whether to build new generators and transmission lines. The power dispatch module optimizes hourly operations during the year based on the capacity determined by the power generation expansion module. The model minimizes the total system-wide costs of meeting electricity demand and renewable energy development constraints. The examined planning horizon is 2015-2030. Six typical days at an hourly level are considered within one year. The primary outputs include the least-cost power generation mix, inter-regional transmission networks, hourly electricity output, and carbon dioxide emissions.
The typical day method has been widely used in large-scale energy system models to cut back on temporal resolution and coverage. For example, the model applied in Lienert and Lochner (2012) considered four seasons, with each season including three typical days using a time step of two hours. Twelve typical days were applied in Koltsaklis and Georgiadis (2015), each representing a one-month period. Due to computational limitation, one year consists of six typical days in our model. Each typical day represents two consecutive months to reflect seasonal characteristics. Further, each day is subdivided into 24 one-hour periods. Thus, a one-year period consists of 144 temporal steps. These temporal steps are assumed to be consecutive. The committed capacities of the first hour of the year are exogenously given a series of initial values according to the power outputs in the base year.
The conventional power generation technologies in the model are ultra-supercritical coal-fired power (USC), combined cycle gas turbine (CCGT), coal-fired combined heat and power (CHP), and other coal-fired power. The non-fossil power generation technologies are nuclear, hy-dropower, onshore wind, solar photovoltaic (PV), biomass, offshore wind, and concentrating solar power (CSP).
Ten regions are represented, characterized by geographical position, resource endowment, and power grid structure (see Table 1). (1) Each region is assumed to be a node. Consequentially, the intra-regional grid capacity restrictions or congestion is not considered. Various types of historical lines between regions are included as basic data in this model. However, for newly built lines, only ultra-high voltage direct current (UHVDC) technology is considered due to its economic advantages in long-distance power transmission (Wang, 2007). (2) In particular, ultra-high voltage alternating current (UHVAC) technology is excluded because its advantages in building a larger synchronous power grid are beyond the scope of this study.
CHP reduces flexibility of power outputs under high heating loads. Because of the central heating supply, this study assumes that CHP becomes a must-run unit during winter in five regions: the Northeast, North, Jinmengxi, Northwest, and Xinjiang. Depending on the climate, the heating duration varies from region to region. In the North, Jinmengxi, and Northwest heating covers two typical days, while in the Northeast and Xinjiang regions it covers three typical days.
3.1.2. Mathematical Equations
The mathematical equations related to the unit commitment module include objective function, ramp-up and ramp-down restrictions, reserves restrictions, start-up costs, and power output restriction. The formulations related to the generation expansion planning module were discussed in Yi et al. (2016) and are presented in Appendix A.
This model is a large-scale linear programming (LP) problem solved by the CPLEX algorithm embedded in the software platform of General Algebraic Modeling System (GAMS). (3) CPLEX is chosen due to its significantly short computation time when solving large-scale LP problems.
The traditional method used in unit commitment models simulates each unit individually, which, due to limited computational capability, is not suitable when capacity expansion is also modelled for large-scale power systems. Here, generation units are grouped by technology, which could significantly reduce the amount of the commitment state for generation units. Kirchhoff's first law, the conservation of currents in each node of the power transmission network, is respected in our model, while the voltage law is not included.
The dynamics of unit commitment states for fossil fuel technology are shown in Eq. (1). The capacity of committed technology is determined by start-up and shut-down decisions.
[U.sub.i,n,t,h] = [U.sub.i,n,t,h-1]+[SU.sub.i,n,t,h]-[SD.sub.i,n,t,h][for all]N[member of][N.sup.TH] (1)
where the subscripts i,n,t,h represent region, technology, year, and hour respectively; N represents the set of generation technologies; the superscript TP represents thermal power technologies; [U.sub.i,n,t,h] represents the committed capacity; [SU.sub.i,n,t,h] represents the capacity which starts up at time h in year t; and [SD.sub.i,n,t,h] represents the capacity that shuts down at time h in year t.
The maximum and minimum power output constraints for the committed technology are shown in Eq. (2).
[U.sub.i,n,t,h]*[P.sup.min.sub.n] [less than or equal to] [U.sub.i,n,t,h]*[P.sup.max.sub.n] [for all]n [member of] [N.sup.T]. (2)
where [P.sub.i,n,t,h] represents the power output and [P.sup.max.sub.n], [P.sup.min.sub.n] represent maximum and minimum power output rates respectively.
The ramping up and ramping down of constraints for generation units are shown in Eq. (3) and Eq. (4), respectively. If a generation unit is started up in time h, its power output in time h is assumed to equal its technical minimum output. If a generation unit is shut down in time h, its power output in time h-1 is also assumed to equal its technical minimum output.
[P.sub.i,n,t,h] - [P.sub.i,n,t,h] [less than or equal to] ([U.sub.i,n,t,h] - [SU.sub.i,n,t,h])*[DELTA][P.sup.up.sub.n] + [SU.sub.i,n,t,h]*[P.sup.min.sub.n] - [SD.sub.i,n,t,h]*[P.sup.min.sub.n] [for all]n[member of][N.sup.TP]. (3)
[P.sub.i,n,t,h] - [P.sub.i,n,t,h+1] [less than or equal to] ([U.sub.i,n,t,h+1] - [SU.sub.i,n,t,h+1])*[DELTA][P.sup.down.sub.n] + [SD.sub.i,n,t,h+1]*[P.sup.min.sub.n] - [SU.sub.i,n,t,h+1]*[P.sup.min.sub.n] [for all]n[member of][N.sup.TP]. (4)
where [DELTA][P.sup.up.sub.n], [DELTA][P.sup.down.sub.n] represent the maximum ramping up and down capability.
The available capacity constraint for thermal power and nuclear is shown in Eq. (5). The committed capacity per hour in year t needs to be less than the total capacity determined at the beginning of year t. This model allows CCGT and other coal-fired power to provide reserves, so the committed capacity also needs to eliminate the capacity for spinning reserve ([SP.sub.i,n,t,h]) and non-spinning reserve ([NSP.sub.i,n,t,h]). For renewable energy technologies, the power output depends not only on the yearly capacity, but also on the hourly capacity factor, as shown in Eq. (6). The power outputs of renewable energy can be less than their available capacity to reflect curtailment (CUR) during certain hours. These two equations reflect the link between the power generation expansion module and the power dispatch module.
[mathematical expression not reproducible] (5)
[CUR.sub.i,n,t,h] = [TC.sub.i,n,t]*[cf.sub.i,n,t,h] - [P.sub.i,n,t,h] [greater than or equal to] 0[for all]n[member of][N.sup.RE]. (6)
where the superscript SC, GAS, USC, CHP, NC and RE represent other coal-fired power, CCGT, USC, coal-fired CHP, nuclear power and renewable energy, respectively; [TC.sub.i,n,t] represents the total capacity; [ava.sub.n] represents the available generation capacity rate, which does not include the units that are in a state of routine maintenance; and [cf.sub.i,n,h] represents the capacity factor for renewable energy technology.
CSP plants could be regarded as electricity storage facilities, so their modeling is different from other renewable power generation technologies. (4) The captured thermal energy can either be used directly in the steam turbine or stored in the thermal storage. This requires more explicit decisions about the utilization of solar resources (see Eqs. (7)-(9)).
[P.sup.direct.sub.i,csp,t,h] + [P.sup.store.sub.i,csp,t,h] = [TC.sub.i,csp,t]*[cf.sub.icsp](7)
[P.sub.i,csp,t,h] + [P.sup.direct.sub.i,csp,t,h] = [P.sub.out]. (8)
[STO.sub.i,csp,t,h] = [STO.sub.i,csp,t,h-1]*[[partial derivative].sub.csp] + [P.sup.store.sub.i,csp,t,h] - [P.sup.out.sub.i,csp,t,h]. (9)
where [P.sup.direct.sub.i,csp,t,h] represents electricity generated in the CSP with heat directly from the collectors; [P.sup.store.sub.i,csp,t,h] represents electric equivalent of the thermal inflow into the storage; [P.sup.out.sub.i,csp,t,h] represents electricity generated in the CSP with heat from the storage; [STO.sub.i,csp,t,h] represents aggregated electricity stored in time h; and [[partial derivative].sub.csp] represents energy loss factor by storing.
Operating reserves are required to respond to short-term fluctuations in load, and equipment failure, including spinning reserve and non-spinning reserve. The total credit capacity at each temporal step should be larger than the demand (PD) plus a safety reserve, as shown in Eq. (10). This model holds a base reserve of 10 percent of load and a variable reserve equal to 10 percent of the wind and solar PV outputs in each hour, at least half of which is spinning, referring to He et al. (2016), as shown in Eq. (11). Variable reserves are required to address the uncertainty imposed by intermittent renewable energy generation. A sensitivity analysis on the variable reserves is presented in Appendix C.
[summation over (n[member of]NI)][TC.sub.i,n,t]*[summation over (n[member of]NI)][TC.sub.i,n,t]*[cc.sub.n] + [SPTI.sub.i,t,h] - [SPTO.sub.i,t,h] [greater than or equal to] (1 + [RC.sub.i,t,h])*[PD.sub.i,t,h]. (10)
[summation over (n[member of]SC[union]GAS)][SP.sub.i,n,t] + [STO.sub.i,csp,t,h] [greater than or equal to] [1/2] * [RC.sub.i,t,h])*[PD.sub.i,t,h]. (11)
where NI and I represent the sets of non-intermittent generators and intermittent generators, respectively; [cc.sub.n] represents the capacity credit rate; [SPTI.sub.i,t,h] represents the total electricity transmitted to region i; [SPTI.sub.i,t,h] represents the total electricity transmitted from region i; and [RC.sub.i,t,h] represents the reserve capacity rate, which is calculated as a percentage of regional power load plus intermittent renewable energy output.
At each time step, power demand for all regions must be met by power supply, which consists of power generation from local power plants and net inter-regional power exchange, as shown in Eq. (12).
[PD.sub.i,t,h] = [SPTI.sub.i,t,h] - [SPTO.sub.i,t,h] + [summation over n] [P.sub.i,n,t,h]. (12)
In terms of the heat balance, this model only considers the heat generated by CHP for China's central heating supply, as shown in Eq. (13). Heat generated by boilers is assumed to remain at the base year level. The matching of regional heat supply and demand is at the level of year due to the lack of data on the hourly heat demand curve. Inter-regional heat transfer is ignored because of its huge losses.
[summation over h] [P.sub.i,chp,t,h]*[gamma]*hpc + [HGB.sub.i] [greater than or equal to] [HD.sub.i,t]. (13)
where [HGB.sub.i,t] represents the total central heating demand; [HGB.sub.i] represents the heat supplied by boilers; [gamma] represents the energy conversion coefficient; hpc represents the heat to power ratio, which is set to 0.7 based on the average level in base year (CEPYEB, 2015).
The extra cost of fossil fuel-based technologies for electricity dispatch is reflected in the start-up cost. It can be calculated as:
[SUC.sub.t] = [summation over (i)][summation over (n)][summation over (h)]([SU.sub.i,n,t,h]*[C.sup.startup.sub.n]). (14)
where [SUC.sub.t] represents total start-up costs and [C.sup.startup.sub.n] represents the unit start-up costs.
The objective function is to minimize the accumulated total costs of the Chinese power sector from the perspective of a social planner. Total costs (TPC) include investment cost ([IC.sub.t]), operation and maintenance (O&M) cost ([OMC.sub.t]), fuel cost ([EC.sub.t]) of power generation technologies, the investment cost ([TIC.sub.t]) and O&M cost ([TOMC.sub.t]) of power transmission technologies, and the start-up and C[O.sub.2] emissions cost ([COC.sub.t]), as shown in Eq. (15).
TPC = [summation over (t)]([IC.sub.t] + [OMC.sub.t] + [EC.sub.t] + [TIC.sub.t] + [TOMC.sub.t] + [SUC.sub.t] + [COC.sub.t]). (15)
The data inputs related to the unit commitment module include hourly load curve, capacity factor for renewable energy, ramp-up rates, and start-up costs. Other data entering the generation expansion planning module, including power demand projection, the generation mix of existing technologies, technical and economic parameters for power transmission, distance between regions, energy price, and carbon price, are presented in Appendix B (CEPYEB, 2015; Yi et al., 2016).
3.2.1. Demand Profiles
Given that hourly power demand data in China is not publicly available, we reconstruct demand profiles from available data that capture important features of the load curve, including seasonal and daily variability for each region. The load curves for summer and winter typical days are shown in Figure 4 (Shu et al., 2005; Lai and Hong, 2007; Wei et al., 2010; Tang and Dong, 2012; Lv et al., 2013; Cheng et al., 2015). Based on monthly load characteristics, we fill in gaps by assuming continuity in load profiles and interpolating values for other typical days. One limitation of our model is that it assumes a constant load curve shape in the future. The corresponding load behavior changes are not considered.
3.2.2. Meteorological Data for Renewable Energy
The hourly capacity factors of solar PV and CSP are based on the PVWatts tool and the System Advisor Model from the National Renewable Energy Laboratory (NREL), which estimates the solar resource resources at a given location.
The hourly capacity factors of onshore and offshore wind are calculated as follows. First, the wind speed data within a six-hour interval is collected from Kalnay et al. (1996), which includes the wind resources at a given location. Second, lagrange interpolation is done to adjust the six-hour interval to hourly wind speed data. Third, a deterministic wind power curve for 2.5 MW turbine is used to convert wind speed into capacity factors of wind power (Zhu and Genton, 2012). Finally, the data are calibrated based on regional average capacity factors in 2016 (NEA, 2017).
Hydropower generation depends on natural rainfall inflows, which vary significantly between seasons. Monthly capacity factors for different regions are from Li et al. (2016). The hourly availability of hydropower is assumed to be the same for a given typical day.
Detailed capacity factors of solar PV, onshore wind, and hydropower at the hourly level are presented in Appendix B.
3.2.3. Key Technical and Economic Parameters
Key technical and economic parameters for the power generation technologies, investment cost, O&M cost, start-up cost, maximum power output, minimum power output, ramping up rate, and ramping down rate, are shown in Table 2.
3.3. Scenario Design
Based on the stylized analysis in Section 2, we find the coal-heavy generation mix (notably CHP) and inter-regional electricity trade significantly affect regional renewable energy development. Therefore, we investigate how to deploy renewable energy at the spatial level to achieve low-carbon targets through six counter-factual scenarios structured along these two dimensions. First, we consider exogenously changing the regional CHP capacity to reflect the extent of the coal-heavy generation mix. The following three cases are chosen to represent the future level of CHP capacity:
(1) Base assumes the regional CHP capacities remain at the 2014 level. No new capacity will be built during the planning horizon.
(2) BAU refers to the projection of the increase of CHP capacity from Davidson et al. (2016). The total CHP capacity in 2030 will reach 385 GW, a 36% increase compared to the base year.
(3) Double assumes that the growth rate of CHP is twice that of the BAU scenario, with a total capacity of 487 GW in 2030, to reflect a higher level of coal-heavy generation mix.
A second dimension explores the role of the inter-regional power transmission grid on electricity trade. The following two cases are chosen to represent the future level of the transmission network:
(1) Frozen assumes Chinese inter-regional transmission infrastructure remains at the 2014 level. No new infrastructure will be built during the planning horizon.
(2) Optimal assumes no restrictions on inter-regional transmission infrastructure construction between any regions. The optimal transmission network during the planning horizon is endogenously calculated by our model.
Two policy targets are assumed to promote renewable energy. The first is that the share of non-fossil energy in total primary energy in 2030 is 20% ([[lambda].sub.pet]), as indicated in the U.S.-China Joint Announcement on Climate Change (see Eq. (16)). The second is that the share of non-hydro renewable energy in total power generation in 2030 is 17% ([[lambda].sub.rps]), referring to Yi et al. (2017), as shown in Eq. (17). (5)
[summation over i][n.summation over (n[member of]NFF)][summation over h][P.sub.i,n,2030,h] [greater than or equal to] [PEC.sub.2030]*[[lambda].sub.pet] (16)
[summation over i][n.summation over (n[member of]NFF)][summation over h][P.sub.i,n,2030,h] [greater than or equal to] [summation over i][n.summation over n][summation over h][PEC.sub.i,n,2030,h]*[[lambda].sub.rps] (17)
where the set NFF and NHR represent non-fossil energy and non-hydro renewable energy respectively; [PEC.sub.2030] represents the total primary energy consumption in 2030.
4. MODELING RESULTS
4.1. Optimal Renewable Energy Spatial Deployment
Figure 5 shows the regional power generation from non-hydro renewable energy in the Optimal+BAU scenario in 2030. We take this scenario as the analysis object in this section because it represents the most likely situation. Not surprisingly, the share of onshore wind in the renewable energy generation mix is the largest, reaching 63.9%. It is followed by solar PV, which generates 316 TWh in 2030, accounting for 22.1% of renewable electricity. Power generations from biomass, offshore wind, and CSP remain relatively small, accounting for 6.5%, 4.1%, and 3.4%, respectively.
There is a serious regional imbalance in the spatial deployment of non-hydro renewable energy. 75.7% is from the north of China, which includes the Northeast, North, Jinmengxi, Northwest, and Xinjiang regions. The increase rates in the Northwest and Xinjiang are relatively higher, leading to a nine-fold increase in 2030 compared with 2014. The non-hydro renewable energy in the south of China is mainly from the East, followed by the Center and South regions. Sichuan shows the fastest relative growth, but its absolute amount remains low.
The difference in resource endowments are the main reason for this imbalance. Both onshore wind and solar PV resources are concentrated in the north of China, especially in the North-west, Xinjiang, and Jinmengxi regions. Only biomass and offshore wind are largely located in the central and south of China, but their absolute volumes are limited.
4.2. Impact Factors of Renewable Energy Spatial Deployment
Before our analysis, we characterized the regions by two dimensions. First, from the view of geographic location, we define the Northeast, North, Jinmengxi, Northwest, and Xinjiang regions as "the north of China," while the other five regions are defined as "the south of China." This is because the central heating supply in the north of China affects the utilization mode of CHP significantly in the winter. Second, from the view of resource endowment of non-hydro renewable energy, we define the Northeast, Jinmengxi, Northwest, and Xinjiang regions as "resource-rich" areas, while the other six are defined as "resource-poor" areas. Figs. 8 and 9 show the renewable energy generation changes from the dimensions of both coal-heavy generation mix and inter-regional power transmission.
4.2.1. Coal-heavy Generation Mix
With the increase in CHP capacity, renewable energy deployment shifts from the north of China to the south of China. This trend is alleviated with inter-regional power transmission. As shown in Figure 6, when CHP capacity is at the Double level, renewable energy production in the south of China increases by 59 TWh in the Frozen scenario, but increases by only 27 TWh in the Optimal scenario.
Not all regions in the north of China are affected by CHP capacity changes. The most affected regions include Northeast, North, and Jinmengxi, while renewable energy production in Northwest and Xinjiang regions vary little, as the shares of CHP in these two regions are relatively small. For the north of China, power transmission could mitigate the transfer of renewable energy caused by high CHP capacity both in resource-rich and resource-poor areas. For the resource-rich areas, such as Jinmengxi, there are a large number of transmission lines connected to the East and Center regions. Excessive thermal power can be exported to the load center to provide more space for local renewable energy. For resource-poor areas such as the North region, little electricity is exported, but power transmission leads to a reduction in renewable energy production, which makes it less susceptible to the change of base load.
4.2.2. Inter-regional Power Transmission
With the increase of inter-regional power transmission, renewable energy deployment shifts from resource-poor to resource-rich areas, and this trend is more significant in high CHP scenarios. As shown in Figure 7, if the construction of inter-regional transmission lines is allowed, the renewable energy production in resource-rich areas increases by 44 TWh in the Base scenario, and reaches 69 TWh in the Double scenario. Because all resource-rich areas are located in the north of China, the above-mentioned rule that power transmission could lower the transfer of renewable energy caused by high CHP capacity is still applicable here.
Among the resource-rich areas, the Northwest region is most affected by inter-regional power transmission. Although it has the most abundant renewable energy resources, it has fewer power transmission lines in the base year, so the excess renewable energy could not be consumed in the Frozen scenario. While the Xinjiang region, which also has abundant natural resources, is almost unaffected due to geographical factors, it gains little from power transmission compared to the Northwest and Jinmengxi, as it is far from load centers like the Center and East. Among resource-poor areas, renewable energy production rises slightly in the Sichuan and South, which are rich in hydropower. Power transfer could further promote surplus hydropower consumption, leading to more space for other renewable energies.
4.2.3. Decomposition Effects
Onshore wind and solar PV are used to explore the effects of the coal-heavy generation mix and inter-regional power transmission on the power structure of non-hydro renewable energy. Tables 3 and 4 present the capacity of onshore wind and solar PV in 2030 from the perspective of both geographic location and resource endowment. Detailed results at the regional level are shown in Appendix C.
High CHP capacity induces an increase of solar PV and a decrease of onshore wind. The inverse relationship between wind power and electricity demand makes it more susceptible to an increase in the base load. The decrease of onshore wind is mainly in the north of China, especially in the North and Jinmengxi regions, which are highly affected by CHP. In turn, the increase of solar PV is primarily found in the south of China. The increase in must-run capacity is, to a certain extent, not conducive to the utilization of solar PV in the north of China. In addition, surplus wind power is transmitted to the south of China, which has a negative impact on the utilization of local wind power.
Through the analysis of the power supply structure, we find that transfers of renewable energy deployment caused by high CHP capacity are mainly reflected in the reduction of onshore wind in the north of China and the rise of solar PV in the south of China.
Inter-regional power transmission could lead to a significant increase in onshore wind and a decrease of solar PV. The increase in wind power occurs in the resource-rich areas, while the reduction of solar PV mainly comes from resource-poor areas, although a small percentage is in the resource-rich areas. Inter-regional transmission could lead to more efficient power dispatch, which is equivalent to improving the utilization space for renewable energy in resource-rich areas. Thus, greater amounts of onshore wind power are deployed there due to its relatively low generation cost.
The transfer of renewable energy deployment caused by inter-regional power transmission is mainly reflected in the reduction of solar PV in resource-poor areas and an increase of onshore wind in resource-rich areas.
4.3. Renewable Energy Curtailment
Table 5 shows the renewable energy curtailment rate during the planning horizon. (6) It is highest in the Northeast region; other relatively high areas are Jinmengxi, Xinjiang, Sichuan, and South. The high CHP capacity induces an increase in the curtailment rate, particularly in the north of China, while inter-regional power transmission could reduce it across almost every region.
The sources of renewable energy curtailment differ across regions. In the Northeast, Jinmengxi, and Xinjiang regions, it is primarily wind power that is curtailed, mainly due to the lack of flexibility in power dispatch caused by increased CHP capacity. Meanwhile, hydropower is generally curtailed in Sichuan and South regions. These two areas have abundant hydropower resources, which are not only difficult for the local demand to consume, but also to some extent compress the space for other renewable energy sources. The existing transmission channels are unable to meet the utilization of excess resources.
4.4. Optimal Inter-regional Transmission Grid
Figure 8 shows the volume of optimal inter-regional power transmission flows in 2030. (7) The BAU scenarios show a significant increase in inter-regional power flow in the Optimal case, reaching 1,449 TWh in 2030, 2.4 times higher than the Frozen case. The total capacity of inter-regional power transmission infrastructure reaches 347 GW by 2030. The infrastructure built during 2015-2030 accounts for 214 GW. Electricity export volumes increase significantly in the Jinmengxi, Northwest, and Sichuan regions, while electricity import volumes increase significantly in East, Center, and Guangdong.
It is necessary to coordinate transmission infrastructure extension planning and renewable energy deployment in order to minimize system costs and reduce curtailment. However, transmission grids are not exclusively used to promote the consumption of renewable energy. For example, in the Jinmengxi region, a large percentage of electricity from cheap coal power plants is exported. Thus, the average annual utilization rates of these lines are relatively high, between 50%-67%. In contrast, electricity exports from Sichuan are primarily from hydropower, and inter-regional power transmission leads to an increase in generation of 40%. Electricity exports from the Northwest come from both wind and thermal power. This combination allows for a high utilization rate of the lines, which is at around 55%, thus improving the economy of electricity transmission. The Northwest-East and Northwest-Center patterns of electricity transmission play an important role in the consumption of wind power.
Increasing CHP capacity induces excessive thermal power in the north of China. As a result, more transmission infrastructure connecting the north and the south of China are constructed, such as Jinmengxi-East, Jinmengxi-Center, and Northwest-Center. In the Double scenario, the electricity transfers between north and south increase by approximately 29% compared with the Base scenario.
4.5. Cost Analysis
Table 6 shows cost decomposition during the planning horizon. Increasing inter-regional power transmission can lead to a significant reduction in the total production cost in the power sector. In the BAU case, total costs fall by 2.9%, and the absolute amount is 903.5 billion RMB during 2015-2030. (8) Part of the cost reduction is explained by the transfer of coal consumption from high-price to low-price regions. Cost reduction is also explained by the higher utilization of coal-fired power. More flexible dispatching can increase a unit's annual operation hours to avoid unnecessary investment. Increased CHP capacity leads to an increase in the total cost in the given framework, but because it also provides heating supply, cost effectiveness cannot be accurately analyzed from the perspective of the power sector alone.
We also find that inter-regional power transmission induces higher carbon emissions. The main driver of this result is that power transmission increases economic incentives to export cheap low-efficiency coal-fired electricity. It can also be attributed to our assumption that all scenarios need to achieve the same renewable energy production target; therefore, fossil-based electricity is not replaced by clean renewable energy in the Optimal case. High CHP capacity also induces an increase of carbon emissions. Most CHPs in China are 300 MW large-capacity units and due to the lack of nearby heating demand, the efficiency of such units is lower than that of ultra-supercritical coal-fired power units.
The renewable energy deployment changes caused by high CHP capacity induce an increase in the total generation cost of non-hydro renewable energy, especially under current levels of the transmission network. However, optimal inter-regional transmission network extensions reduce this cost by 1.6% in the BAU case, and the absolute amount is 79 billion RMB. Additionally, the start-up cost could partially reflect the cost of thermal power for adapting fluctuating characteristics of renewable energy. Inter-regional transmission is more conducive to the temporal dispatch of electricity, reducing about 30% of start-up costs.
This paper integrates short-term unit commitment into a long-term generation expansion planning framework to develop a multi-regional power sector optimization model for China. Based on this model, we quantitatively analyze the optimal spatial deployment of renewable energy in China to achieve relevant low-carbon targets, and compare the impact of a coal-heavy generation mix and inter-regional electricity transmission on the spatial deployment of renewable energy.
Our analysis highlights the central role played by the north of China in achieving low-carbon policy targets. Approximately 75% of non-hydro renewable energy should be deployed in these regions under the least-cost cases. The share of onshore wind in the non-hydro renewable energy generation mix is the largest, reaching 63.9% in 2030. It is followed by solar PV, which accounts for 22.1% of renewable electricity.
Another significant finding is that the deployment of renewable energy depends on the regional power structure and inter-regional power transmission network. With an increase of combined heat and power (CHP), more renewable energy facilities, especially solar PV, should be located in the south of China because of the conflict between CHP and onshore wind in the north of China. In contrast, an optimal inter-regional transmission network is beneficial to onshore wind production in resource-rich areas, and induces a transfer of renewable energy deployment from resource-poor areas to resource-rich areas.
The scale of CHP primarily depends on the surrounding heat demand. Excessive CHP capacity increases the total generation costs for achieving the national non-hydro renewable energy target, because it leads to an increase in the curtailment rate of renewable energy, particularly in the north of China. An enhanced inter-regional power transmission network is crucial because, on the one hand, it can cope with the significant regional imbalance of energy resources and electricity load; on the other hand, it can mitigate the conflict between the coal-heavy generation mix and renewable energy. It also has the potential to bring sizeable gains from electricity trade, estimated to be 4.4-11.7 billion RMB per year, which would help reach the energy targets.
Our findings also point to the need to consider the coordination between inter-regional transmission infrastructure planning and the spatial deployment of renewable energy. The Northwest-East and Northwest-Center patterns of electricity transmission play a key role in the consumption of onshore wind power. Our analysis demonstrates that intra-regional power network planning must keep pace with the development of renewable energy. Reducing intra-regional electricity congestion is the premise that inter-regional power transmission can bring the benefits in reducing renewable energy curtailment and generation costs.
This article is the first step in developing a power sector optimization model that integrates unit commitment with generation expansion planning framework based on the Chinese power sector. Our work suggests several directions for future research. First, the consideration for the uncertainty of intermittent renewable energy is not enough in this study. Stochastic models for dealing with uncertainties were widely used in unit commitment problems (Wang and Hobbs, 2016; Warrington et al., 2016; Morales-Espana et al., 2017). But when short-term unit commitment is combined with long-term large-scale capacity expansion, computational feasibility will be the main difficulty in applying stochastic models to cope with the uncertainty of renewable energy. Second, heating systems other than CHP can be integrated into the power sector optimization model to analyze the interaction between electricity systems and heating systems. Data collections of hourly heat demand curve and existing heat boiler, heat pump, and heat storage capacities in China will be the main challenges for such an expansion.
This work is supported by the National Key Research and Development Program of China under Grant No. 2017 YFE0101800, the National Natural Science Foundation of China under Grant No. 71690245, No. 71673266, No. 71403263 and Beijing Natural Science Foundation under Grant No. 9152019. The authors appreciate the weekly joint seminars at CEEP in CAS and Beihang SEM, from where the earlier draft of the paper got improved.
Abrell, J. and S. Rausch (2016). "Cross-country electricity trade, renewable energy and European transmission infrastructure policy." Journal of Environmental Economics and Management 79: 87-113. https://doi.org/10.1016/jjeem.2016.04.001.
Bertsch, J., T. Brown, S. Hagspiel, and L. Just (2017). "The relevance of grid expansion under zonal markets." The Energy Journal 38(5): 129-152. https://doi.org/10.5547/01956574.38.5.jber.
Bird, L., C. Chapman, J. Logan, J. Sumner, and W. Short (2011). "Evaluating renewable portfolio standards and carbon cap scenarios in the US electric sector." Energy Policy 39(5): 2573-2585. https://doi.org/10.10l6/j.enpol.2011.02.025.
Chang, Y. and Y. Li (2015). "Renewable energy and policy options in an integrated ASEAN electricity market: Quantitative assessments and policy implications." Energy Policy 85: 39-49. https://doi.org/10.1016/j.enpol.2015.05.011.
Chen, Z., L. Ma, R Liu, and Z. Li (2018). "Electric vehicle development in China: A charging behavior and power sector supply balance analysis." Chemical Engineering Research and Design 131: 671-685. https://doi.org/10.1016/j.cherd.2017.11.016.
Cheng, L., Y. Feng, and N. Wang (2015). "Load characteristics analysis and peak margin calculation of Beijing-Tianjin-Northern Hebei power grid." North China Electric Power 11: 1-5 (in Chinese).
Cheng, R., Z. Xu, P. Liu, Z. Wang, Z. Li, and L. Jones (2015). "A multi-region optimization planning model for China's power sector." Applied Energy 137: 413-426. https://doi.org/10.1016/j.apenergy.2014.10.023.
China Electric Power Yearbook Editorial Board (CEPYEB) (2015). "China electric power yearbook 2015." Beijing: China Electric Power Press (in Chinese).
Davidson, M.R., D. Zhang, W. Xiong, X. Zhang, and V. Karplus (2016). "Modelling the potential for wind energy integration on China's coal-heavy electricity grid." Nature Energy 1, 16086. https://doi.org/10.1038/nenergy.2016.86.
Dong, L., H. Liang, Z. Gao, X. Luo, and J. Ren (2016). "Spatial distribution of China's renewable energy industry: Regional features and implications for a harmonious development future." Renewable and Sustainable Energy Reviews 58: 1521-1531. https://doi.org/10.1016/j.rser.2015.12.307.
Egerer, J., C. Gerbaulet, C. and Lorenz (2016). "European electricity grid infrastructure expansion in a 2050 context." The Energy Journal 37 (SI3): 101-124. https://doi.org/10.5547/01956574.37.SI3.jege.
Egerer, J., J. Rosellon, and W.P. Schill (2015). "Power system transformation toward renewables: an evaluation of regulatory approaches for network expansion." The Energy Journal 36(4): 105-128. https://doi.org/10.5547/01956574.36.4.jege.
Guo, Z., R. Cheng, Z. Xu, P. Liu, Z. Wang, Z. Li, L. Jones, and Y. Sun (2017). "A multi-regional load dispatch model for the long-term optimum planning of China's electricity sector." Applied Energy 185: 556-572. https://doi.org/10.1016/j. apenergy.2016. 10. 132.
He, G, A.P. Avrin, J.H. Nelson, J. Johnston, A. Mileva, J. Tian, and D.M. Kammen (2016). "SWITCH-China: a systems approach to decarbonizing China's power system." Environmental Science & Technology 50( II): 5467-5473. https://doi.org/10.1021/acs.est.6b01345.
Kaleta, M. and E. Toczylowski (2008). "Restriction techniques for the unit-commitment problem with total procurement costs." Energy Policy 36(7): 2439-2448. https://doi.org/10.1016/j.enpol.2008.01.039.
Kalnay et al. (1996). "The NCEP/NCAR 40-year reanalysis project." Bull. Amer. Meteor. Soc. 77: 437-470. https://doi.org/10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.
Keane, A., A. Tuohy, P. Meibom, E. Denny, D. Flynn, A. Mullane, and M. O'Malley (2011). "Demand side resource operation on the Irish power system with high wind power penetration." Energy Policy 39(5): 2925-2934. https://doi.org/10.1016/j.enpol.2011.02.071.
Koltsaklis, N.E. and M.C. Georgiadis (2015). "A multi-period, multi-regional generation expansion planning model incorporating unit commitment constraints." Applied Energy 158: 310-331. https://doi.org/10.1016/j.apenergy.2015.08.054.
Lai, M. and B. Hong (2007). "The influences of the air conditioning load on the daily load characteristics of Central China power grid." Electric Power Technologic Economics 19(3): 30-33 (in Chinese).
Li, Y. and Y. Chang (2015). "Infrastructure investments for power trade and transmission in ASEAN+ 2: Costs, benefits, long-term contracts and prioritized developments." Energy Economics 51: 484-492. https://doi.org/10.1016/j.eneco.2015.08.008.
Li, Y, Z. Lukszo, and M. Weijnen (2016). "The impact of inter-regional transmission grid expansion on China's power sector decaxbonization." Applied Energy 183: 853-873. https://doi.org/10.1016/j.apenergy.2016.09.006.
Lienert, M. and S. Lochner (2012). "The importance of market interdependencies in modeling energy system-The case of the European electricity generation market." Electrical Power and Energy Systems 34: 99-113. https://doi.org/10.1016/j.ijepes.2011.09.010.
Lv, Q., W. Wang, S. Han, S. Yuan, J. Zhang, and W. Li (2013). "A new evaluation method for wind power curtailment based on analysis of system regulation capability." Power System Technology 37(7): 1887-1894 (in Chinese).
Ministry of Housing and Urban-Rural Development of the People's Republic of China (MOHURD) (2015). China urban construction statistical yearbook 2014. Available from http://www.mohurd.gov.cn/xytj/tjzljsxytjgb/jstjnj/index.html.
Moarefdoost, M.M., A.J. Lamadrid, and L.F. Zuluaga (2016). "A robust model for the ramp-constrained economic dispatch problem with uncertain renewable energy." Energy Economics 56: 310-325. https://doi.org/10.1016/j.eneco.2015.12.019.
Morales-Espa-a, G., L. Ramirez-Elizondo, and B.F. Hobbs (2017). "Hidden power system inflexibilities imposed by traditional unit commitment formulations." Applied Energy 191: 223-238. https://doi.org/10.1016/j.apenergy.2017.01.089.
National Energy Administration (NEA) (2017). "Wind power generation situation in 2016." Available from http://www.nea.gov.cn/2017-01/26/c_136014615.htm (in Chinese).
Pei, W., Y. Chen, K. Sheng, W. Deng, Y. Du, Z. Qi, and L. Kong (2015). "Temporal-spatial analysis and improvement measures of Chinese power system for wind power curtailment problem." Renewable and Sustainable Energy Reviews 49: 148-168. https://doi.org/10.1016/j.rser.2015.04.106.
Perez, A.P., E.E. Sauma, F.D. Munoz, and B.F. Hobbs (2016) "The economic effects of interregional trading of renewable energy certificates in the U.S. WECC." The Energy Journal 31(A): 267-296. https://doi.org/10.5547/01956574.37.4.aper.
Pfluger, B. (2014). "Assessment of least-cost pathways for decarbonising Europe's power supply: A model-based long-term scenario analysis accounting for the characteristics of renewable energies." KIT Scientific Publishing, Karlsruhe, Germany.
Scholz, Y., H.C. Gils, and R.C. Pietzcher (2017). "Application of a high-detail energy system model to derive power sector characteristics at high wind and solar shares." Energy Economics 64:568-582. https://doi.org/10.1016/j.eneco.2016.06.021.
Shu, J., J. Zhang, and Y. Zheng (2005). "Discussion on rational scale of pumped storage power station in East China power grid in 2010." Water Resources and Power 23(5): 54-57 (in Chinese).
Sullivan, P., C. Uriarte, and W. Short (2014). "Greenhouse gas mitigation options in the U.S. electric sector: a ReEDS analysis." The Energy Journal 35(1): 101-114. https://doi.org/10.5547/01956574.35.SI1.6.
Tang, B. and Z. Dong (2012). "Analysis on load characteristics of China southern power grid in 11th Five Year Plan." Electric Power 45(10): 30-35 (in Chinese).
Wang, B., and B.F. Hobbs (2016). "Real-time markets for flexiramp: A stochastic unit commitment-based analysis." IEEE Transactions on Power Systems 31 (2): 846-860. https://doi.org/10.1109/TPWRS.2015.2411268.
Wang, Z. (2007). "Economy and security analysis of AC UHV in China." Electric Power Automation Equipment 27( 10): 1-4 (in Chinese).
Warrington, J., C. Hohl, P. Goulart, and M. Morari (2016). "Rolling unit commitment and dispatch with multi-stage recourse policies for heterogeneous devices." IEEE Transactions on Power Systems 31(1): 187-197. https://doi.org/10.1109/TP-WRS.2015.2391233.
Wei, L., L. Zhang, and N. Jiang (2010). "Analysis of load characteristics for Northwest China power grid." Power System and Clean Energy 26(7): 57-62 (in Chinese).
Weigt, H., D. Ellerman, and E. Delarue (2013). "C[O.sub.2] abatement from renewables in the German electricity sector: Does a C[O.sub.2] price help?" Energy Economics 40: 149-158. https://doi.org/10.1016/j.eneco.2013.09.013.
Wright, E. and A. Kanudia (2014). "Low carbon standard and transmission investment analysis in the new multi-region U.S. power sector model FACETS." Energy Economics 46: 136-150. https://doi.org/10.1016/j.eneco.2014.09.013.
Xinhua News Agency (XNA) (2014). "U.S.-China joint announcement on climate change." Available from: http://www.xinhuanet.com/video/2014-11/13/c_127205443.htm (in Chinese).
Yi, B., J. Xu, and Y. Fan (2016). "Inter-regional power grid planning up to 2030 in China considering renewable energy development and regional pollutant control: A multi-region bottom-up optimization model." Applied Energy 184: 641-658. https://d0i.0rg/10.1016/j.apenergy.2016.11.021.
Yi, B., J. Xu, and Y. Fan (2017). "Optimal pathway and impact of achieving the mid-long term renewable energy standard in China." Journal of Systems Engineering 32(3): 313-324 (in Chinese).
Yi, B., J. Xu, and Y. Fan (2019). "Coordination of policy goals between renewable portfolio standards and carbon caps: A quantitative assessment in China." Applied Energy 237: 25-35. https://doi.org/10.1016/j.apenergy.20l8.12.015.
Zhang, D., J. Wang, Y. Lin, Y. Si, C. Huang, J. Yang, B. Huang, and W. Li (2017). "Present situation and future prospect of renewable energy in China." Renewable and Sustainable Energy Reviews 76: 865-871. https://doi.org/10.1016/j.rser.2017.03.023.
Zhu, X., and M. Genton (2012). "Short-term wind speed forecasting for power system operations." International Statistical Review 80(1): 2-23. https://doi.org/10.l 111/j.1751-5823.2011.00168.x.
APPENDIX A: MATHEMATICAL EQUATIONS
A.1 Power generation module
The calculation of the total capacity of each technology is shown in Eq. (1). Every power generation technology will be decommissioned at the end of its expected lifetime.
[TC.sub.i,n,t] = [t.summation over t'=t - [T.sub.n] + 1] [NC.sub.i,n,t'](1)
where the subscripts i, n, t represent region, technology, and year respectively; Tn represents the lifetime of technology (y); [TC.sub.i,n,t], [NC.sub.i,n,t] represent the total capacity, and the new capacity of technology n in year t respectively (MW).
The investment cost is discounted equally to each year of its lifetime. It means if the lifetime of the technology exceeds the planning horizon, investment cost is scaled down to consider the residual value. The calculation is shown in Eq. (2).
[mathematical expression not reproducible] (2)
where [IC.sub.t] represents the total investment cost in year t (RMB); [CAP.sub.n,t'] represents the unit investment cost of technology in year t (RMB/MW); I represents the discount rate to reflect the expected return rate of investment (%), which is 10% in this model (Yi et al., 2016).
The fixed operation and maintenance (O&M) cost can be calculated as follow:
[OMC.sub.t] = [summation over i][summation over n][TC.sub.i,n,t]*[COM.sub.n,t] (3)
where [OMC.sub.t] represents the total O&M cost in year t (RMB); [COM.sub.n,t] represents the unit O&M cost (RMB/MW).
Fuel cost is proportional to the price and consumption rate of each kind of fuel. Fuels used in power sector include the coal, natural gas, biomass, and uranium. It is a major variable cost in the power generation process, calculated as follow:
[EC.sub.t] = [summation over i][summation over n][summation over h]([P.sub.i,n,t,h]*[es.sub.n] + [SP.sub.i,n,t,h]*[spes.sub.n])*[ep.sub.i,n,t] (4)
where the subscript h represents hour; [es.sub.t] represents the total energy consumption cost in year t (RMB); [P.sub.i,n,t,h] represents the power output of technology n at time h in year t (MWh); [es.sub.n] represents the energy consumption per unit of electricity (t/MWh); [SP.sub.i,n,t,h] represents the capacity for spinning reserve at time h in year t (MW); [spes.sub.n] represents the unit energy consumption for spinning reserve (t/MW); [ep.sub.i,n,t,] represents the energy price (RMB/t).
The emission trading scheme will be introduced in this model. A carbon cap (the upper limit for annual C[O.sub.2] emissions of the power sector) and a C[O.sub.2] price are assumed for each year. If the C[O.sub.2] emissions exceed this cap, it needs to buy allowances from the carbon market. On the other hand, it can also sell its superfluous allowances to get revenue. The C[O.sub.2] price is exogenous. The detailed calculations are shown in Eqs. (5)-(6).
[TCE.sub.t] = [summation over i][summation over n][summation over h][P.sub.i,n,t,h]*[es.sub.n]*[[theta].sub.n] (5)
[COC.sub.t]=[pco.sub.t]*([TCE.sub.t] - [ET.sub.t]) (6)
where [TCE.sub.t] represents the total C[O.sub.2] emissions in year t (t-C[O.sub.2]); [[theta].sub.n] represents the unit C[O.sub.2] emission of the fuel used by technology n (t-C[O.sub.2]/t); [COC.sub.t] represents the total C[O.sub.2] emission cost (or revenue) in year t (RMB); pco, represents the C[O.sub.2] price (RMB/t-C[O.sub.2]); [ET.sub.t] represents the C[O.sub.2] cap in year t (t-C[O.sub.2]).
The annual generation capacity expansion can be constrained by factors such as labors, materials, and capability of facilities manufacturing, a certain value is set for the annual construction speed of each type of power generation technology, as presented by Eq. (7).
[NC.sub.i,n,t] [less than or equal to] Max[NB.sub.i,n] (7)
where Max[NB.sub.i,n] represents the maximum expansion allowed within one year (MW).
A.2 Power transmission module
The calculation of the total capacity of each transmission technology is shown in Eq. (8). Every power transmission technology will be decommissioned at the end of its expected lifetime.
[TTC.sub.i,j,t] = [t.summation over t'=t-T+1][TNC.sub.i,j,t'] (8)
where the subscript j also represent region; [TTC.sub.i,j,t], [TNC.sub.i,j,t] represent the total capacity, and the new capacity of transmission technology in year t respectively (MW).
The investment cost of power transmission technology consists of two parts: the fixed costs and variable costs. Among them, the variable costs are proportional to the transmission distance. It is also discounted equally to each year of its lifetime. The calculation is shown in Eq. (9).
[TIC.sub.t] = [summation over i][summation over j][[t.summation over t' = t-T+1](([TFCAP.sub.t'] + [TVCAP.sub.t']*[dis.sub.i,j])*[TNC.sub.i,j,t']*[I*[(1 + I).sup.-1]/1 - [(1 + I).sup.-T]])] (9)
where [dis.sub.i,j]. represents the distance between region i and j (km); [TIC.sub.t] represents the total investment cost of power transmission technology (RMB); [TFCAP.sub.t] represents the unit fix investment cost of power transmission technology in year t (RMB/MW); [TFCAP.sub.t] represents the unit variable investment cost of power transmission technology in year t (RMB/km/MW).
The total O&M cost of power transmission technology can be calculated as follow:
[TOMC.sub.t] = [summation over i][summation over i][TTC.sub.i,j,t]*[TCOM.sub.i,j]
where [TOMC.sub.t] represents the total O&M cost of power transmission technology in year t (RMB); [TCOM.sub.i,j]. represents the unit O&M cost of power transmission technology (RMB/MW).
At each time step, power demand for all regions should be met by power supply, which consists of power generation from local power plants, and net inter-regional power exchange, as shown in Eqs. (11)-(13).
[PD.sub.i,t,h] = [SPTI.sub.i,t,h] - [SPTO.sub.i,t,h] + [summation over n][P.sub.i,n,t,h] (11)
[SPTI.sub.i,t,h] = [j [not equal to] i.summation over j][PT.sub.i,j,t,h]*(1 - [ptl.sub.j,i]) (12)
[SPTO.sub.i,t,h] = [j [not equal to] i.summation over i][PT.sub.j,i,t,h] (13)
where [PD.sub.i,t,h] represents power demand in time h (MWh); [SPTI.sub.i,t,h] represents the total electricity transmitted to region i in time h (MWh); [SPTO.sub.i,t,h] represents the total electricity transmitted from region i in time h (MWh); [ptl.sub.i,j] represents the transmission loss between region i and j (%); [PT.sub.i,j,t,h] represents the electricity transmitted from region i to j (MWh).
Electricity traded within any hour should be subject to the constraint of available transmission capacity, as shown in Eq. (14).
[PT.sub.i,j,t,h] [less than or equal to] [TTC.sub.i,j,t]*avatr (14)
where avatr represents the available transmission capacity rate (%).
APPENDIX B: DATASETS
B.1 Demand projection
Table B1: Power demand projection during the planning horizon Northeast North Jinmengxi Northwest Xinjiang Growth rate of power 4.9 4.5 4.5 6.8 6.8 demand during 2015-2020 (%) Growth rate of power 2.4 2.2 2.2 3.0 3.0 demand during 2020-2030 (%) East Center Sichuan South Guangdong Growth rate of power 4.3 5.3 5.3 4.9 4.9 demand during 2015-2020 (%) Growth rate of power 2.0 2.4 2.4 2.2 2.2 demand during 2020-2030 (%) Source: Yiet al.(2016).
The central heating demand projections during the planning horizon are from the Building Energy Research Center of Tsinghua University. This total demand is decomposed into each region based on the heat consumption structure in the base year (MOHURD, 2015).
Table B2: Central heating demand projection
Table B2: Central heating demand projection Region (Million GJ) 2015 2020 2025 2030 Northeast 1066 1153 1266 1376 North 995 1076 1182 1284 Jinmengxi 292 316 348 378 Northwest 201 217 238 259 Xinjiang 152 165 181 197 East 123 133 146 158 Center 89 97 106 115 Sichuan 0 0 0 0 South 0 0 0 0 Guangdong 0 0 0 0
B.2 Generation mix of existing technologies
Table B3: Regional installed capacity of power generation technologies until 2014 (GW) Northeast North Jinmengxi Northwest Xinjiang SC 12.5 21.5 8.7 5.0 3.0 CCGT 0.6 9.1 2.1 0.7 0.2 CHP 54.5 77.6 42.5 15.7 8.8 Other coal-fired 16.9 29.8 51.7 40.4 23.3 Nuclear 2.0 0.0 0.0 0.0 0.0 Hydro power 8.1 3.5 3.3 22.5 5.9 Onshore wind 23.0 16.2 17.1 15.5 7.7 Solar PV 0.6 1.5 2.8 11.4 3.3 Biomass 1.6 1.8 0.1 0.2 0.4 Offshore wind 0.0 0.0 0.0 0.0 0.0 CSP 0.0 0.0 0.0 0.0 0.0 East Center Sichuan South Guangdong SC 94.2 43.5 0.0 4.4 20.0 CCGT 26.3 2.9 0.2 0.2 11.6 CHP 63.7 18.5 1.4 0.0 0.0 Other coal-fired 38.2 62.8 14.3 50.8 34.1 Nuclear 10.9 0.0 0.0 0.0 7.3 Hydro power 26.8 66.9 62.6 89.0 13.1 Onshore wind 5.8 2.3 0.3 5.3 2.2 Solar PV 3.6 0.6 0.1 0.3 0.5 Biomass 2.3 2.0 0.1 0.4 0.4 Offshore wind 0.6 0.0 0.0 0.0 0.0 CSP 0.0 0.0 0.0 0.0 0.0 Sources: Yi et al. (2016); CEPYEB (2015).
B.3 Key technical and economic parameters for power transmission
Table B4: Technical and economic parameters for power transmission technology Technology Fix investment cost Variable investment Lifetime (RMB/kW) cost (RMB/km/kW) (year) UHVDC 1491 0.786 40 Technology Transmission loss Annual decrease rate of investment (%/1000km) cost (%) UHVDC 2.75 0.01 Source: Yi et al. (2016).
B.4 Distance between regions
Table B5: Distance of inter-regional transmission lines (km) Northeast North Jinmengxi Northwest Xinjiang East Northeast 0 1078 1283 2286 3118 1940 North 1078 0 697 1539 2727 861 Jinmengxi 1283 697 0 1008 2049 1391 Northwest 2286 1539 1008 0 1370 1824 Xinjiang 3118 2727 2049 1370 0 3176 East 1940 861 1391 1824 3176 0 Center 2085 1006 1261 1393 2763 537 Sichuan 2640 1561 1319 693 1958 1484 South 2943 1864 1836 1347 2549 1451 Guangdong 2773 1694 1999 1952 3287 935 Center Sichuan South Guangdong Northeast 2085 2640 2943 2773 North 1006 1561 1864 1694 Jinmengxi 1261 1319 1836 1999 Northwest 1393 693 1347 1952 Xinjiang 2763 1958 2549 3287 East 537 1484 1451 935 Center 0 959 945 740 Sichuan 959 0 655 1334 South 945 655 0 891 Guangdong 740 1334 891 0 Source: Authors' estimation.
B.5 Energy prices
Table B6: Regional energy prices in the base year Northeast North Jinmengxi Northwest Xinjiang Coal (RMB/tce) 518 522 370 377 346 Natural gas 3000 3650 2800 2300 2110 (RMB/1000[m.sup.3]) Biomass (RMB/t) 304 328 303 265 320 East Center Sichuan South Guangdong Coal (RMB/tce) 608 571 604 587 703 Natural gas 4000 3600 3250 4000 4850 (RMB/1000[m.sup.3]) Biomass (RMB/t) 351 295 280 273 350 Source: Yi et al. (2016).
B.6 Capacity factor
Table B7: Capacity factor of hydropower Typical day Northeast North Jinmengxi Northwest Xinjiang East Day 1 0.171 0.057 0.114 0.228 0.205 0.143 Day 2 0.278 0.093 0.186 0.371 0.334 0.232 Day 3 0.471 0.157 0.314 0.628 0.565 0.392 Day 4 0.514 0.171 0.342 0.685 0.617 0.428 Day 5 0.407 0.136 0.271 0.543 0.488 0.339 Day 6 0.214 0.071 0.143 0.285 0.257 0.179 Typical day Center Sichuan South Guangdong Day 1 0.177 0.251 0.200 0.103 Day 2 0.288 0.409 0.325 0.167 Day 3 0.486 0.691 0.549 0.282 Day 4 0.531 0.753 0.599 0.308 Day 5 0.420 0.597 0.475 0.244 Day 6 0.221 0.314 0.250 0.128
APPENDIX C: DETAIL RESULTS
C.1 Regional power generation mix
Figure C1 shows the regional power generation mix for 2030 in the Optimal+BAU scenario. The share of thermal power in total generation is the largest, reaching 53.2%. Power generation from hydropower, non-hydro renewable energy, and nuclear power remains relatively low, accounting for 18.5%, 17.0%, and 11.3%, respectively.
Our results reveal a significant regional difference in the power generation mix in 2030. Hydropower is mainly located in the Sichuan and South regions, while nuclear power is distributed in the East, Guangdong, and Center. Thermal power occupies the largest share of power generation in all regions except Sichuan, South, and Guangdong. The share of non-hydro renewable energy in regional power generation also varies greatly. This share is higher than 20% in Xinjiang, Northwest, Jinmengxi, and Northeast. However, it is lower than 10% in the South, Guangdong, Sichuan, and Center regions.
Figure C1: Power generation mix in the "Optimal+BAU" scenario in 2030 Non-hydro RE Hydro power Nuclear power Thermal power Northeast 22.8% North 17.2% Jinmengxi 25.3% Northwest 32.4% Xinjiang 42.5% East 11.7% Center 5.4% Sichuan 6.0% South 8.9% Guangdong 6.0% Notes: The numbers above the pillars represent the share of non-hydro renewable energy in regional total power generation. "RE" in the legend is the abbreviation of renewable energy. Note: Table made from bar graph.
C.2 Regional non-hydro renewable energy generation
Table C1: Regional non-hydro renewable energy generation in 2030 Region (TWh) Frozen Optimal Base BAU Double Base BAU Double Northeast 131 131 127 145 141 136 North 213 202 195 196 196 185 Jinmengxi 259 247 233 247 248 255 Northwest 272 272 274 311 311 311 Xinjiang 185 191 185 187 187 185 East 191 196 204 168 172 175 Center 58 67 74 52 52 58 Sichuan 32 32 35 35 35 35 South 54 57 64 60 60 60 Guangdong 35 35 39 28 28 28
C.3 Regional capacity of onshore wind and solar PV
Table C2: Regional capacity of onshore wind and solar PV in 2030 Onshore Wind Solar PV Frozen Optimal Frozen Region (GW) Base BAU Double Base BAU Double Base BAU Northeast 47.7 47.2 50.1 53.4 51.2 51.2 14.6 18.1 North 56.2 51.2 49.7 56.2 56.2 50.8 43.5 42.8 Jinmengxi 80.3 75.9 70.4 80.3 80.3 80.3 42.8 42.8 Northwest 65.9 65.9 66.7 83.9 83.9 83.9 66.3 66.3 Xinjiang 60.4 62.0 58.7 62.5 62.5 62.5 27.4 30.3 East 55.3 55.3 55.3 48.8 48.8 48.8 31.5 31.5 Center 10.9 13.1 13.7 10.9 10.9 10.9 21.6 21.6 Sichuan 8.3 8.3 8.3 8.3 8.3 8.3 8.1 8.1 South 19.3 19.3 22.1 24.9 24.9 24.9 13.3 13.3 Guangdong 15.5 15.5 17.5 13.5 13.5 13.5 6.5 6.5 National total 419.8 413.7 412.5 442.7 440.5 435.1 275.6 281.3 Solar PV Frozen Optimal Region (GW) Double Base BAU Double Northeast 18.1 14.6 14.6 14.6 North 45.0 29.5 29.5 29.5 Jinmengxi 42.8 34.8 34.8 40.5 Northwest 66.3 66.3 66.3 66.3 Xinjiang 32.4 27.3 27.3 27.3 East 38.5 24.5 28.7 31.5 Center 21.6 14.6 14.6 21.6 Sichuan 10.1 10.1 10.1 10.1 South 15.9 10.7 10.7 10.7 Guangdong 6.5 3.5 3.5 3.5 National total 297.2 235.9 240.1 255.6
C.4 Key sensitivity analysis
The spatial deployment of renewable energy may be influenced by many factors, including future fuel prices, cost reduction rate of technologies, and errors in renewable forecasting. Regulation uncertainties, such as environmental policy, electricity market reform, and emission trading scheme, will also impact the evolution pathway of technologies.
Based on the focus of this study, we examine three key factors: the renewable costs, the variable reserves, and the capacity factors, using the Optimal+BAU scenario, as shown in Figure C2. In the first case, the investment costs of wind and solar in 2030 will be 20% lower or 10% higher than the Optimal+BAU scenario. In the second case, the variable reserves are set to 0%, 10%, 20%, and 50% of the intermittent wind and solar outputs, respectively. These changes lead to a certain impact on system costs, and natural gas capacity. But, they have little impact on the spatial deployment of renewable energy.
In the third case, the capacity factors of wind and solar will be 10% lower or 10% higher than the Optimal+BAU scenario during the planning horizon. The share of non-hydro renewable energy generation in the north of China varies between 75.1% and 76.6%. The decrease of capacity factor requires more renewable units to be built, resulting in less vulnerable to the impact of the historical power structure distribution.
Bo-Wen Yi, (a) Wolfgang Eichhammer, (b) Benjamin Pfluger, (c) Ying Fan, (d) and Jin-Hua Xu (e)
(a) School of Economics & Management, Beihang University, Beijing 100191, China, E-mail: email@example.com.
(b) Fraunhofer Institute for Systems and Innovation Research, Karlsruhe 76139, Germany, E-mail: Wolfgang.Eichhammer@isi.fraunhofer.de; and Utrecht University, Copernicus Institute of Sustainable Development, 3584 CB Utrecht, The Netherlands.
(c) Fraunhofer Institute for Systems and Innovation Research, Karlsruhe 76139, Germany, E-mail: Benjamin.Pfluger@isi.fraunhofer.de.
(d) Corresponding author. School of Economics & Management, Beihang University, Beijing 100191, China. E-mail: firstname.lastname@example.org.
(e) Corresponding author. Center for Energy and Environmental Policy Research, Institutes of Science and Development, Chinese Academy of Sciences, Beijing 100190, China. E-mail: xujinhual email@example.com.
(1.) Xinjiang is separated from the Northwest power grid due to its remote location. Other power grids are further divided into multiple regions mainly because of the differences in resource endowments. Regional geographic distribution and resource endowment are presented in Appendix B.
(2.) In our previous work, the power transmission technology also includes high voltage direct current (HVDC). The results show that almost all new lines are UHVDC; therefore, this study ignores HVDC in order to simplify our model.
(3.) Our previous model is a nonlinear programming model that incorporates a coal transportation module. In this study, the coal transport module is replaced by exogenous coal prices in order to simplify our model into a linear programming problem.
(4.) Our model assumes that a CSP plant has a six hour storage system.
(5.) Non-hydro renewable energy includes onshore wind, offshore wind, biomass, CSP, and solar PV.
(6.) The results of our simulation are lower than the current curtailment rate in China. This is because our study does not consider the uncertainty of renewable energy or intra-regional power congestion. In addition, the matching between power supply and demand is based on a perfect system perspective, which can also lead to a reduction in curtailment rate. Nonetheless, we can still find the patterns of the changes by comparing various scenarios.
(7.) As the power transmission networks in all Frozen scenarios remain at the level of base year, the differences of power flow between each are very small. Therefore, we only present the results of the Frozen+BAU scenario.
(8.) All costs in this study are 2014 constant price.
Table 1: Regional Division in This Study Regional symbol Provinces, municipalities Regional power grids and autonomous regions included Northeast Liaoning, Jilin, Heilongjiang, Northeast power grid East Inner Mongolia North Beijing, Tianjin, North power grid Hebei, Shandong Jinmengxi Shanxi, West North power grid Inner Mongolia Northwest Gansu, Qinghai, Northwest power grid Ningxia, Shaanxi Xinjiang Xinjiang Northwest power grid East Zhejiang, Shanghai, East power grid Jiangsu, Fujian, Anhui Center Hubei, Hunan, Jiangxi, Central power grid Chongqing, Henan Sichuan Sichuan Central power grid South Guangxi, Yunnan, Guizhou Southern power grid Guangdong Guangdong Southern power grid Table 2: Technical and Economic Parameters for Power Generation Technologies Investment cost O&M cost Start-up cost Technology (RMB/KW) (b) (RMB/KW) (b) (RMB/time/KW) (c) USC 3915 290 0.70 CCGT 2830 80 0.25 CHP 4090 350 0.42 Other coal-fired 4880 300 0.42 Nuclear (a) 15500 1000 -- Hydro 10000 46 -- Onshore wind 7750 253 -- Solar PV 8600 128 -- Biomass 10470 300 -- Offshore wind 13000 600 -- CSP 20000 250 -- Maximum Minimum Ramping Ramping power output power output up rate down rate Technology rate (%) (c) rate (%) (c) (%/h) (c) (%/h) (c) USC 100 40 10 10 CCGT 100 25 40 40 CHP 100 40 (d) 10 10 Other coal-fired 100 40 10 10 Nuclear (a) -- -- -- -- Hydro -- -- -- -- Onshore wind -- -- -- -- Solar PV -- -- -- -- Biomass -- -- -- -- Offshore wind -- -- -- -- CSP -- -- -- -- Notes: The symbol "--" means that the data those are not needed in this model. (a) This study assumes that the output of nuclear power is always at the maximum level. (b) Yi et al. (2016), Davidson et al. (2016), Cheng et al. (2015), Yi et al. (2017). (c) Li et al. (2016), Koltsaklis and Georgiadis (2015). (d) During winter time, the minimum power output rate of CHP in the north of China is assumed to increase to 55%, referring to Davidson et al. (2016). Table 3: Capacity of Onshore Wind and Solar PV in 2030 from a Geographic Location View Onshore Wind Frozen Optimal Region (GW) Base BAU Double Base BAU Double Base The north of 310.5 302.2 295.6 336.3 334.1 328.7 194.6 China The south of 109.3 111.5 116.9 106.4 106.4 106.4 81.0 China National total 419.8 413.7 412.5 442.7 440.5 435.1 275.6 Solar PV Frozen Optimal Region (GW) BAU Double Base BAU Double The north of 200.3 204.6 172.5 172.5 178.2 China The south of 81.0 92.6 63.4 67.6 77.4 China National total 281.3 297.2 235.9 240.1 255.6 Table 4: Capacity of Onshore Wind and Solar PV in 2030 from a Resource Endowment View Onshore Wind Base BAU Double Region (GW) Frozen Optimal Frozen Optimal Frozen Optimal Resource-rich 254.3 280.1 251.0 277.9 245.9 277.9 areas Resource-poor 165.5 162.6 162.7 162.6 166.6 157.2 areas National total 419.8 442.7 413.7 440.5 412.5 435.1 Solar PV Base BAU Double Region (GW) Frozen Optimal Frozen Optimal Frozen Optimal Resource-rich 151.1 143.0 157.5 143.0 159.6 148.7 areas Resource-poor 124.5 92.9 123.8 97.1 137.6 106.9 areas National total 275.6 235.9 281.3 240.1 297.2 255.6 Table 5: Renewable Energy Curtailment During the Planning Horizon Frozen Optimal Region (%) Base BAU Double Base BAU Double Northeast 7.20 8.06 11.66 3.51 3.61 4.64 North 0.03 0.01 0.82 0.01 0.04 0.01 Jinmengxi 1.76 2.01 3.09 0.81 0.86 0.89 Northwest 0.93 1.05 1.39 0.48 0.45 0.42 Xinjiang 1.07 1.40 1.64 0.93 1.12 1.27 East 0.49 0.48 0.43 0.05 0.03 0.03 Center 0.03 0.05 0.04 0.06 0.05 0.05 Sichuan 3.65 4.03 4.61 0.13 0.14 0.14 South 1.85 1.87 1.81 0.16 0.16 0.14 Guangdong 0.07 0.05 0.55 0.00 0.00 0.00 National average 1.81 2.00 2.46 0.43 0.45 0.49 Notes: This curtailment rate is the sum of hydro power, onshore wind, offshore wind, and solar PV during the planning horizon. Table 6: Costs Decomposition During the Planning Horizon Frozen Optimal Costs (Billion RMB) Base BAU Double Base BAU Investment cost for 7,607.9 7,643.2 7,832.2 7,114.8 7,138.4 power generation O&M cost for 8,604.6 8,673.8 8,811.1 8,160.1 8,219.6 power generation Investment cost for 0.0 0.0 0.0 362.8 402.3 power transmission O&M cost for power 137.3 140.5 143.5 242.4 262.9 transmission Start-up cost 225.6 230.4 228.0 152.4 157.1 Energy consumption 14,322.0 14,371.8 14,365.5 13,991.7 13,925.3 cost C[O.sub.2] emission 320.3 337.1 340.6 370.8 387.7 cost Total costs 31,217.7 31,396.8 31,720.9 30,395.0 30,493.3 Total generation 5,015.7 5,042.5 5,181.6 4,946.1 4,963.5 costs for non-hydro renewable energy Optimal Costs (Billion RMB) Double Investment cost for 7,187.7 power generation O&M cost for 8,276.5 power generation Investment cost for 416.9 power transmission O&M cost for power 271.7 transmission Start-up cost 161.2 Energy consumption 13,923.5 cost C[O.sub.2] emission 403.6 cost Total costs 30,641.1 Total generation 4,973.7 costs for non-hydro renewable energy
|Printer friendly Cite/link Email Feedback|
|Author:||Yi, Bo-Wen; Eichhammer, Wolfgang; Pfluger, Benjamin; Fan, Ying; Xu, Jin-Hua|
|Publication:||The Energy Journal|
|Date:||Jul 1, 2019|
|Previous Article:||Investing in Smart Grids: Assessing the Influence of Regulatory and Market Factors on Investment Level.|
|Next Article:||Co-firing Coal with Biomass under Mandatory Obligation for Renewable Electricity: Implication for the Electricity Mix.|