Optimal Generation Expansion Planning Model of a Combined 1 Thermal-wind-PV Power System Considering Multiple Boundary 2 Conditions: A Case Study in Xinjiang, China

8 Background: The scarcity of fossil fuels and their high emissions impact on the 9 environment have forced the rapid development of renewable energy. Wind and 10 photovoltaic power play a more and more important role in today's power system 11 because of their clean and renewable characteristics. However, with the large-scale grid 12 connection of wind and photovoltaic power, the contradiction between renewable 13 energy and thermal power is becoming more and more serious. The unreasonable 14 planning of power generation resources has also caused a lot of waste of electric power. 15 Methods: To solve this problem, this paper comprehensively considers the construction 16 and operation costs of power plants, and constructs an optimal generation expansion 17 planning model of a combined thermal-wind-PV power system with the objective of 18 minimizing total cost. The planning is calculated under the boundary conditions of 19 ensuring the safe operation of the power grid and taking into account the share 20 requirement, utilization requirements and construction requirements for renewable 21 energy. The model is applied to the self-use and external power plants in Xinjiang. 22 Results: The results show that this generation expansion plan can reduce the total cost 23 of the power plants while ensuring the load demand. The renewable energy has also 24 received more consumption share and its abandonment rate has dropped significantly. 25 Numerical examples show that the optimization model has good applicability. 26 Conclusion: The proposed optimization model can realize the coordinated 27 development of three types of power sources under multiple boundary conditions, 28 which can not only guarantee the economy of power construction, but also take into 29 account the optimization of environmental benefits. Reasonable generation expansion 30 planning can improve energy efficiency, achieve clean and low carbon in the process 31 of power generation, and promote the sustainable development of society

abandonment. According to the National Energy Administration, the national 48 abandoned amount of hydropower, wind and PV power reached 69, 28 and 5TWh 49 respectively in 2018 [11]. In provinces with more renewable energy installed capacity, 50 such as Xinjiang and Gansu, the abandonment rate even exceeds 20% [12]. Therefore, 51 how to properly plan the development of renewable energy has become the main 52 problem at present. installed capacity of renewable energy has also brought a lot of problems to the Xinjiang 64 power grid. On the one hand, due to the self-use demand has not been deeply excavated 65 and the external transmission space has not been effectively developed, there is a 66 serious phenomenon of renewable energy abandonment in Xinjiang. The high spillage 67 rate has caused Xinjiang to be included in the red warning zone of developing 68 renewable energy for three consecutive years, suspending the construction of wind and 69 factors. In order to improve the applicability of the model, this paper proposes the 139 following assumptions combined with the energy development status of Xinjiang. 140 (1) The fuel of thermal power only consider coal and the unit does not carry out 141 major technical upgrading during the planning period, so the coal consumption 142 coefficient will not drop significantly; 143 (2) The cost of the thermal power unit includes fuel cost and operation cost 144 excluding coal transportation costs, and the remaining renewable power units only 145 include operation cost; 146 (3) The break-even utilization hours will be affected by the income and 147 expenditure costs for the individual power generation units. Therefore, this paper 148 selects the average break-even utilization hours of each types of power sources in 149 Xinjiang for calculation; 150 (4) Because the self-use and external power supply cannot be used interchangeably, 151 it needs to be separately constrained; 152 (5) The construction of UHV transmission channels needs to be considered from 153 multiple aspects. In the case that there is no new UHV line put into operation in the 154 future, the capacity of the supporting power supply will not be increased during the 155 planning period. 156 Based on the above assumption, the generation planning model is constructed with 157 the minimum total investment and operating cost as the optimization goal. 158

Objective function 160
In the power planning, the main consideration is the investment cost of the new 161 power plants, as well as the operation cost of the generators. Therefore, the optimization 162 goal is to minimize the total cost. 163

Constraints 166
(1) Power balance constraint 167 The power consumption and generation in the power system must be balanced in (3) 170 (2) Reserve capacity constraint 171 In order to achieve stable operation of the power system, it is necessary to select a 172 power supply unit with adjustment capability for peak shaving to cope with random 173 changes in load. In this paper, thermal power is selected as the peaking unit, and it 174 should be able to provide a certain reserve capacity [36]. In order to ensure the economic benefits of power plants, the utilization hours 178 should not be lower than the average break-even utilization hours, but not higher than 179 the theoretical utilization hours. 180 , ** ,

Share requirements for non-hydropower renewable energy 197
According to the Notice on Establishing and Improving the Safeguard Mechanism 198 for Renewable Energy Power Consumption, renewable energy power consumption 199 responsibility weights are set by the province's power consumption. The requirements 200 for Xinjiang from 2018 to 2020 are 11.5%, 12% and 13%, respectively. However, since 201 this paper only considers three types of power resources, the weights of non-202 hydropower renewable energy in 2020, 2025, and 2030 are set to 30%, 35%, and 40%, 203 respectively. 204

Utilization requirements for renewable energy 205
According to the Clean Energy Dissipation Action Plan (2018-2020), it is 206 necessary to optimize the power structure, reasonably control the pace of power 207 development, and improve the utilization rate of power supply during the 13th five-208 year Plan period. The target for clean energy consumption in Xinjiang is that the 209 utilization rate of PV power should reach 85% in 2018, 90% in 2019 and 2020, and 210 wind power utilization rates should reach 75%, 80% and 85% in 2018, 2019 and 2020 211 respectively. With the increasingly serious problems of energy and environment, the 212 requirements for energy efficiency will become stricter. Therefore, the utilization rate 213 of wind and PV power is set to 95% and 90% respectively in 2025, and it will reach to 214 95% in 2030 in this paper. 215

Construction requirements for renewable energy 216
According to the Notice on Establishing Monitoring and Early Warning 217

Mechanism to Promote Sustainable and Healthy Development of Wind Power Industry
Industry, Xinjiang has been in the red warning zone for three years, and it is not allowed 221 to increase the installed capacity of renewable energy. Since the red warning requires 222 multiple indicators to judge, in order to simplify the calculation, this paper sets the wind 223 and PV abandonment rates are less than 15% and 10% respectively as the red warning 224 release standard. 225

Problem statement 227
In recent years, due to the energy red warning, the growth rate of renewable energy 228 installed capacity in Xinjiang has slowed down, and the proportion of renewable energy 229 decreased slightly. Although the power supply structure is constantly being optimized,

Break-even hours 244
Under the current average on-grid price level, the break-even hours are calculated 245 according to the cost and benefit data of the power plants (Table 1). 246 Table 1 The current break-even hours of three power recourses 247 Thermal power Wind power PV power Self-use External Self-use External Self-use External renewable energy will fall because of the new electricity price policy, which would 249 reduce the income of three kinds of power plants resulting an increase in their break-250 even hours (Table2). 251

Self-use demand 253
In recent years, the self-use demand in Xinjiang has been increasing, but the

External demand 275
The demand for external power transmission is mainly related to the receiving end. As can be seen from the Table 3, with the increase in the self-use and external 292 demand of Xinjiang, the three types of power operation hours are higher than the break-293 even hours, which can obtain a certain profit. However, the utilization hours of the 294 supporting thermal power in 2020 are low. This is because in the case of insufficient 295 total load demand in Xinjiang, the self-use thermal power can only increase revenue by 296 seizing the share of external thermal power in the UHV transmission channel to achieve 297 profitability. And since the break-even utilization hours of the supporting thermal power 298 plants are relatively low, they can still make a profit after dividing part of the grid share 299 to the self-use thermal power plants. This phenomenon will be alleviated as the total 300 demand increases. Although the power delivery ratio is gradually increasing, this is due Year supply and the reduction in unit cost. 306 Under the optimized generation expansion plan, the operation hour of thermal, 308 wind and PV power plants are high, which can realize the break-even of three kinds of 309 power sources under the condition of coal price fluctuation. The generation planning 310 model has good applicability. 311

Renewable energy efficiency analysis 312
As clean and low-carbon renewable energy, the utilization efficiency of wind and 313 PV power is also a key factor in generation expansion planning. Fig. 5 shows the 314 utilization rate of Xinjiang's renewable energy during the planning period. Influenced 315 by policies and social responsibilities, the utilization rate of renewable energy is 316 increasing year by year, and the development trend is unchangeable. The utilization 317 rate of wind power has gradually increased from 85.85% to 96.42%, and that of PV 318 power is even close to 98%. This may be caused by the following two reasons. On the 319 other hand, the power structure in Xinjiang is also gradually optimized to clean and low 320 carbon. As can be seen from Fig 6,

Sensitivity analysis 329
Total load demand is the decisive factor affecting the installed capacity. Therefore, 330 based on the above plan, the influence of self-use and external demand changes on the 331 optimization model is analyzed. The change rate is set to ±5% and the results are shown 332 in Fig. 7. Thermal power is the most sensitive to total demand changes, followed by 333 wind power, and PV power is the least sensitive. On the one hand, due to the limitation 334 of renewable energy penetration, wind and PV power can't be installed in large 335 quantities. The grid connection of renewable energy requires the power system to 336 provide spare capacity, which will also increase the installed capacity of thermal power. 337