logoGlobal Energy Interconnection

Contents

Figure(0

    Tables(0

      Global Energy Interconnection

      Volume 4, Issue 4, Aug 2021, Pages 394-404
      Ref.

      Collaborative optimization model of renewable energy development considering peak shaving costs of various flexibility resources

      Ji Liang1 ,Xingang Zhao1,2 ,Shangdong Yang3
      ( 1. School of Economics and Management, North China Electric Power University, Changping District,Beijing 102206, P.R.China , 2. Beijing Key Laboratory of Renewable energy and Low-Carbon Development (North China Electric Power University), Changping District, Beijing 102206, P.R.China , 3. State Grid Talents Exchange and Service Center Company Limited, Beijing 100761, P.R.China )

      Abstract

      China has set carbon emission goals for 2030 and 2060.Renewable energy sources, primarily wind and photovoltaic power, are being considered as the future of power generation.The major limitation to the development of new energies is the limited flexibility of regulations on power system resources, resulting in insufficient consumption capacity.Thus, the flexible resource costs for peak shaving as well as the reasonable coordinated development and operation optimization of regional renewable energy need to be considered.In this study, a renewable energy development layout configuration analysis method was established by considering the composite cost of a power system, comprehensively analyzing the potential of various flexibility regulation resources for the power system and its composite peak shaving cost,and combining renewable energy output characteristics, load forecasting, grid development, and other factors.For the optimization of various flexible resource utilization methods, a peak shaving cost estimation method from the perspective of the entire power system was established by combining the on-grid electricity prices and operating costs of different power sources.A collaborative optimization model of power system operation that aims at the lowest peak shaving cost and satisfies the constraints of operation, safety, and environmental protection was proposed.Finally, a certain area of Gansu Province was used as an example to perform detailed analysis and calculation, which demonstrated that the model has an optimal effect.This model can provide an analysis method for regional renewable energy development layout configurations and system optimization operations.

      0 Introduction

      Recently, China has committed to reducing carbon dioxide emissions by 2030 and achieving carbon neutrality by 2060.At present, the energy industry in China accounts for approximately 80% of carbon emissions.Therefore,controlling carbon emissions in this industry is a basic prerequisite for achieving these goals, and thus, is crucial.In this context, clean energy will see considerable development in China.Before 2030, China’s energy structure will undergo major changes.The development of traditional coal power will sharply decrease.Natural gas power generation consumes fossil fuels, thus contributing to carbon emissions;this can be expected until the carbon capture technology matures.Nuclear power is expected to be developed stably and orderly development by 2030; however, the development is affected by several factors, such as limited coastal sites for construction, long construction periods, and concerns about safety of inland development.In addition,hydropower is expected to have a limited future growth,as it is affected by factors such as limited resource sites, a complex construction environment, and long development periods.Therefore, in the future, large-scale clean energy needs to be supplemented by photovoltaic and wind power.

      At the end of 2020, the clean energy power installed capacity in China reached 955 million kW, accounting for 43.4% of the total installed capacity.Among the energy sources, wind power and photovoltaic installed capacities reached 280 and 250 million kW, accounting for 12.8% and 11.5%, which represents an increase of 34.6% and 24.1%over the previous year, respectively [1].Before 2030, the renewable energy installed capacity in China is estimated to reach 1.2 billion kW.To achieve the target of 25% nonfossil fuel energy consumption, the renewable energy installed capacity should be 1.6 billion kW [2-3].

      Owing to the ever-increasing renewable energy installations, improving the absorbing capacity of power systems and ensuring stable operation of a system with high renewable energy proportion have become important issues for the development and construction of future power systems.Previous studies have advanced the development of new energies. [4-5] summarized the key technologies for the large-scale development of renewable energies and provide guidance for the development of renewable energy grid connections.Reference [6] proposed a low-carbon dispatch strategy for power systems that consider the uncertainty on both sides of the source network.Reference [7] designed a “source-grid-loadstorage” coordinated optimization operation mode and key technologies.Reference [8] proposed an operation optimization model for an electric and thermal energy integrated system considering thermal grid constraints to analyze the evaluation method for wind power consumption.Reference [9] considered the effect of largescale photovoltaic power generation on the power system.Reference [10] proposed a wind, solar, hydro, gas, and coal storage joint optimization dispatch strategy based on a chance-constrained target, which was proposed to reduce the system operation costs through joint operation.However,regarding adaptation to the large-scale development of renewable energy, how to fully explore and coordinate various flexible peak shaving resources in the power system and achieve a unified and orderly shaving technology should be further investigated.

      In terms of operating costs of flexible resources, reference[11] analyzed the situation of power companies in a lowcarbon economy and proposed relevant countermeasures by considering operation costs.Reference [12] fully examined the factors influencing wind power prices and constructed an electricity price model that included wind power (renewable energy power) and thermal power (traditional energy power).Reference [13] compared the demand, function, and action mode of energy storage in the energy internet.Reference[14] proposed a wind farm cost model that considered the randomness and fluctuation of wind power and introduced it into the traditional dynamic economic dispatch model of the power system.Considering the present background of large-scale renewable energy consumption, research on the optimization of power system operation modes from the perspective of construction and operation costs of flexible peak shaving has not been conducted yet.

      In addition, considering the optimization and coordination of various peak shaving resources from multiple dimensions of technology, economy, and society,research on a collaborative optimization model of power systems after large-scale renewable energy development in the future has not yet been performed.Meanwhile, the increase in the scale of flexible resources and deployment costs will have a significant impact on the overall costs of the power system.Therefore, the economic, safety, and environmental constraints should be comprehensively considered to provide suggestions for the optimization of actual grid dispatching operations.

      To this end, in this paper, the energy development scenarios and renewable energy development demand under the carbon peak and carbon neutrality goals are first presented.Then, the consumption problems associated with large-scale renewable energy and the main peak shaving measures are analyzed, and the potential and operating costs of various peak shaving methods are compared.An optimization model is established for the coordinated development of renewable energy, considering peak shaving costs.Finally, considering the power grid in a certain area in Gansu, the proposal of peak shaving resource development while meeting the carbon emission demand is studied, the comprehensive cost change of the power system is analyzed,and the applicability of the collaborative optimization model is verified.

      1 Prospects and main challenges for new power systems

      1.1 Carbon reduction path

      At present, China’s forests and other ecosystems have an annual carbon sink of approximately 600-700 million tons.The forest carbon sink is expected to increase from 17.56 billion m³ in 2018 to 25 billion m³ in 2060, at which time the ecosystem carbon sink will reach 1 billion tons.According to China’s goal of achieving carbon neutrality by 2060, the net carbon dioxide emissions directly emitted by the entire society to the atmosphere must be less than 1 billion tons, primarily corresponding to the natural carbon emission requirements in biological and agricultural fields.Therefore, the energy sector needs to achieve net zero carbon emissions.

      Under the 2060 carbon neutrality goal, China will have a carbon emission path showing a rapid decline after peaking in the future.This will be achieved by controlling the total amount of fossil fuel consumption, improving energy efficiency, and developing non-fossil energy on a large scale.The estimated peak carbon emission of the energy sector is approximately 1.07 million tons.In sequence,under various comprehensive measures, China’s energy sector carbon emissions will rapidly decrease.By 2050 and 2060, carbon emissions will decrease to 5 and 2 billion tons, and carbon capture will be 0.5 and 2 billion tons,respectively.Thus, net zero emissions in the energy sector will be achieved.

      1.2 Energy transition development path

      Under the carbon neutrality goal, China’s total primary energy consumption will show an initial upward trend,followed by a downward tendency.The proportion of nonfossil energy consumption will continue to increase and is expected to reach 85% by 2060.The total primary energy consumption is expected to peak in 2030 as 5.7 billion tons of standard coal.Then, it is expected to start declining slowly and reach 5.2 billion tons of standard coal by 2060.Among the primary energy sources, the proportion of nonfossil energy consumption will increase from 15.8% in 2020 to 20% in 2025, 25% in 2030, and 83% in 2060.Clean energy such as wind and solar will enter the stage of largescale development.From the perspective of different types of primary energies, coal, oil, and gas consumption will peak during the 14th five-year plan, during the 15th fiveyear plan, and around 2035, respectively.

      The proportion of clean energy and renewable energy power generation in China is expected to reach 93%and 60% in 2060, respectively.The installed capacity of renewable energy will reach 5 billion kW.To adapt to the construction of a new power system with a high proportion of renewable energy, various methods need to be adopted to explore and improve the flexible regulation ability of the power system.Currently, the energy increase in China during the 14th five-year plan will primarily be borne by renewable energy.The traditional power system operation will not be able to satisfy the development requirements of renewable energy.Thus, the current flexible regulation methods and performance and the potential of flexibility resources in the power system are particularly important.

      2 Types and potential analysis of flexibility resources

      2.1 Types of system flexibility resources

      Renewable energy types such as solar and wind cause random fluctuations in the power system.Thus, the system needs to have flexible regulation capabilities to adapt to these random fluctuations to ensure that the power remains balanced and the system stays stable.The flexibility resources of power systems can be divided into the following categories.The first category is based on the transformation of the existing power system,primarily by increasing the regulation range and rate of the existing power supply, or by increasing the scale of the interruptible and adjustable loads.These include thermal power flexibility transformation, hydropower operation optimization, and load demand-side response.The other category involves improving the regulation ability of the system through the construction of new facilities, primarily by building new power supplies with strong regulation ability as well as electrical energy storage facilities with both charging and discharging abilities, and by developing controllable and flexible loads.For example, new gasfired power generation and energy storage and hydrogen production by electrolysis of water can be realized.

      2.2 Potential analysis of system flexibility resources

      In this section, the development potential of the types of flexibility resources indicated in the previous section is analyzed.Considering the coal power flexibility transformation,China currently has coal-fired power generation units with a capacity of 1.07 billion kW, as well as completed units and units undergoing flexibility transformation with a capacity of approximately 100 million kW.The peak shaving ability is in general approximately 50% of the rated capacity for pure condensing units in China in actual operation, and only 20%of the rated capacity for typical condensing units during the heating period.Through flexibility transformation,the thermal power units are expected to increase the peak shaving capacity by 20% of the rated capacity, and their minimum technical output will reach 40%-50% of the rated capacity.The pure condensing units will increase the peak shaving capacity by 15%-20% of the rated capacity, and their minimum technical output will reach 25%-35% of the rated capacity.The deep flexibility transformation and startstop flexibility transformation of the units can be realized by strengthening technical communication and cooperation worldwide, adopting methods such as adding electric boilers, heat storage facilities, and bypass facilities, as well as improving boiler combustion methods.The minimum technical output of the units can be reduced to 10%-25%and 0%-10%, respectively.The regulation ability of the entire system is expected to increase by nearly 150 million kW following the flexibility transformations.

      Regarding the flexibility response on the demand side,in 2020, China’s annual maximum load was approximately 1.33 billion kW, which was still dominated by the industrial load, accounting for 65%.The tertiary industry and residents’ living loads accounted for approximately 30%, and the primary industry load accounted for approximately 5%.Owing to the continuous advancement of smart grids and digitization as well as the growing demand for flexible regulation—observed with the rapid growth of electric vehicles—demand-side management methods can be used to achieve controllable load potential,which is expected to reach 15%, i.e., approximately 150-300 million kW.

      Considering hydropower in actual peak shaving operations, the typical configuration scheme includes hydropower stations with a regulation capacity.In addition,a few hydropower stations reserve foundation pits during construction for system peak shaving or the subsequent installation of additional hydroelectric generators.Therefore,the plan entailing reserved pits for additional generators or construction of pumped storage power stations is also a practical peak shaving retrofit plan.In 2020, China’s installed hydropower capacity was 370 million kW, including 30 million kW of pumped storage.These figures are expected to increase to 500 million kW of installed capacity by 2030,including 100 million kW of pumped storage.Considering the hydropower regulation capacity and the construction of pumped storage power stations, the regulation capacity is estimated to reach approximately 300 million kW.

      In 2020, the installed gas power generation capacity in China was 100 million kW.In addition to a small number of combined heat and power units, gas-fired units have a regulation capacity exceeding 85%.The scale of gas power generation is affected by emission reduction factors such as carbon peak and neutrality.According to the mediumand long-term plans for power development, the installed gas power generation capacity is expected to increase to 200 million kW, and the potential of regulation capacity is expected to be approximately 180 million kW.

      The development potential of conversion methods such as renewable energy storage and hydrogen production by electrolysis of water is mostly based on the demand for peak shaving operation of the power system, and these methods are expected to become the main peak shaving methods in the future.

      3 Estimation methods for system peak shaving costs

      3.1 Costs of flexibility resources

      3.1.1 Cost of flexible transformation of coal power

      The flexible transformation of coal power has multiple technical routes, and the various routes have different construction costs and transformation outcomes.For thermal power units, the cost of flexible transformation can be calculated as follows.

      where Cfar is the cost of flexible transformation of coal-fired power units, Cp is the cost of zero-output transformation of low-pressure cylinders, Ct1 is the construction cost of thermoelectrolytic coupling facilities, and Ct2 is the transformation cost of steam turbines, boilers, and power generation control.

      According to the analysis of the transformation project volume and equipment scale, the zero-output transformation of low-pressure cylinders has the smallest volume.For example, the transformation cost of a 300,000 kW unit—20% of which, i.e., 60,000 kW, is the peak shaving capacity of the unit—is approximately 2 million yuan.The investment per unit of new regulation capacity is approximately 33 yuan/kW.The construction cost of thermoelectrolytic coupling facilities is moderate.Heat storage and electric boilers are used to improve the regulation performance, and the investment per unit of new regulation capacity is approximately 300 to 500 yuan/kW.The transformation cost of steam turbines, boilers, and power generation control is the highest, requiring an investment of 2,000 to 5,000 yuan/kW in coal-fired power units for new regulation capacity.

      3.1.2 Cost of regulation capacity improvement of hydropower

      According to the current typical hydropower station configuration plan, the static investment for a new 1.2 million kW hydropower station is approximately 12 billion yuan.The corresponding construction cost per kW is 10220 yuan/kW.For the transformation plan regarding the reserved foundation pits of existing hydropower stations,the cost per kW is 4730 yuan/kW, which is obtained by separately calculating cost for the generators and related auxiliary equipment.The cost data given by power companies is more optimistic, at only 2000 yuan/kW.For pumped storage power stations, the required investment is approximately 20,000 yuan/kW.In the future, as the number of high-quality sites for hydropower decreases,pumped storage power stations will undergo better development.

      3.1.3 Construction cost of gas power generation

      Natural gas power generation can improve the flexible regulation capability of the system.According to statistics,the current construction cost of natural gas generator units is approximately 3,500 to 4,000 yuan/kW, which will be stable and even slightly reduced in the future.

      3.1.4 Transformation cost of demand-side response

      The demand-side response transformation comprises installing intelligent terminals and informationized communication equipment.After construction, the system will have a response power load that can be regulated and interrupted.The cost of this part of transformation is relatively low, generally in the range 100 to 300 yuan/kW.

      3.1.5 Construction cost of renewable energy storage facilities

      Lithium iron phosphate battery energy storage,lead-carbon battery energy storage, flow batteries, and compressed air energy storage are the current mainstream renewable energy storage technologies.Among them, the energy storage time is generally less than 4 h for lithium iron phosphate batteries and 4 to 12 h for lead-carbon batteries, flow batteries, and compressed air energy storage batteries.In terms of construction costs, the investment per kW is approximately 1,500 to 3,000 yuan/kW for lithium iron phosphate batteries, and 3,000 yuan/kW, 5,000 yuan/kW, and 10,000 yuan/kW for lead-carbon, flow, and compressed air energy storage batteries, respectively.

      3.2 Cost of operation of flexible resources

      3.2.1 Cost of flexible operation of coal power

      After the flexible transformation of coal power, the operation costs primarily consist of two aspects.One is the increase in coal consumption, i.e., approximately 10% after the transformation of the coal-fired power unit system.The other is the opportunity cost of reduction of on-grid power,as the coal-fired power units can operate under pressure,which will reduce the operation hours and power generation of the units.The calculation equation is as follows.

      where Crun is the flexible operation cost of coal power units,Ccost is the increased cost of coal consumption, Csell is the cost of lost power generation, and Ct2 is the operation and maintenance cost of unit regulation.

      3.2.2 Operation cost of hydropower regulation improvement

      For traditional hydropower, the cost of the unit to provide peak shaving services includes fixed costs—which primarily include the mechanical loss caused by the participation in peak shaving and the cost of various actions required during the peak shaving process—and opportunity costs—which refers to the loss of profits caused by reduction in power generation during participation in peak shaving.

      The actual investigation of hydropower companies helps conclude that the fixed costs are mostly incurred owing to the mechanical loss caused by frequent adjustment of output during the peak shaving process, including increased vibration of the units and shaft wear.However, this portion of the cost is already considered in the depreciation costs.Therefore, the cost of hydropower peak shaving should primarily consider the opportunity cost in the regulation process.As the current medium- and long-term contract power in the electricity market allows rolling adjustments within the contract period, and the completion of the power curve has not been assessed point by point, the opportunity cost of hydropower units mostly refers to the profit loss associated with the reduction in power generation, which is caused by the change in water consumption rate caused by a change in output.Considering that the head reservation in actual hydropower dispatch is reasonable, the opportunity cost can be neglected.

      For the foundation pit reservation transformation plan,the transformation costs need to be recovered through the peak shaving compensation mechanism.This is because the additional hydropower installation does not substantially increase the hydropower storage capacity and incoming hydropower and will not generate additional power revenue.Therefore, the peak shaving cost should primarily be the apportionment of the transformation cost for each year.

      For pumped storage power plants, in addition to the fixed cost brought about by mechanical loss, the 25% loss between the pumped hydropower and the generated power is the inherent power cost of this technology.

      3.2.3 Operation cost of gas power

      In addition to the fixed cost of operation and management, natural gas power generation also has variable fuel costs.The proportion of the fuel cost to the total cost of gas power generation is higher than the fixed cost.

      3.2.4 Operation cost of demand-side response

      The demand-side response operation cost is mostly reflected in the labor cost during system operation and maintenance.The overall operation cost is relatively low.

      3.2.5 Operation cost of renewable energy storage facilities

      In addition to fixed costs, various energy storage facilities are also subject to power loss in charging and discharging during operation.For example, the charge and discharge efficiency is 90% for lithium iron phosphate batteries, 80% for lead-carbon batteries, 70% for flow batteries, and 60% for compressed air energy storage.

      3.3 Composite costs of flexible resources

      3.3.1 Analysis method for composite costs of flexible resources

      Flexibility resources significantly differ because of their technical characteristics, operating modes, and regulation performance; thus, they should be compared under the same conditions.Therefore, in this study, the unit electricity regulation cost was adopted as the unified composite cost for various regulation methods.The main principle is to amortize the depreciation of the initial investment and construction cost to the operation cost and to use the regulation capacity or scope as the regulation composite cost of unit electricity.

      3.3.2 Estimation model of composite costs

      The composite cost of power system regulation capability generally refers to the use of flexible resources such as pumped storage, energy storage, and demandside management to perform auxiliary services for power translation.The use of flexibility regulation resources will also increase the operation cost for system and users.The composite cost assessment method of flexible resources is shown in equation (3).

      where Cpl is the fixed cost of equipment operation,Cfar is the initial investment depreciation cost, Cs is the opportunity cost lost as a result of equipment operation,and Closs is the cost associated with efficiency reduction in equipment operation.

      4 Collaborative optimization model of renewable energy development

      4.1 Collaborative optimization goals considering peak shaving costs

      The collaborative optimization of a power system involves several aspects, such as economy, safety, and the environment.The demand for peak shaving is significant in the construction of a new power system with renewable energy as the main component.Research on the coordination and optimization methods of renewable energy development with peak shaving costs as the goal has important guiding value and practical significance.To this end, the goals of collaborative optimization of a power system are to satisfy the requirements of environmental and safety restrictive standards, minimize the composite costs of peak shaving operation during the collaborative optimization period, and determine optimization configuration schemes with various energy storage scales [24-30].

      where t is the t-th instant during the collaborative optimization period, and n is the total time of the collaborative optimization period.

      4.2 Operational constraint conditions

      The constraints in the system optimization model include system operational, safety, and environmental constraints.

      4.2.1 Operational constraints

      Equation (5) shows the system power equilibrium constraint:

      where p is the output of the i-th conventional unit; wj is the predicted output of the j-th renewable energy unit in time period t; vk is the output of the k-th flexible resource,including the energy storage station and demand-side management; d is the DC transmission power in time period t; and lDF(t) is the predicted load of the system in time period t.

      Equation (6) shows the maximum and minimum output constraints of conventional units.

      Equation (7) shows the climbing constraint per unit power:

      where DRi and URi are the upper and lower limits of the allowable regulation output of unit i in each time period,respectively.

      Equation (8) shows the minimum operation and shutdown duration constraints of units.

      where and are the minimum turn on/off time of unit i, respectively, and(t-1) and t-1) are the on/off duration of unit i before time period t, respectively.

      Equation (9) shows the demand-side management operational constraint:

      where vkmax is the normal load of the k-th demand-side management resource and vkmin is the reducible load of the k-th demand-side management resource.

      Equations (10) to (12) show the operational constraints of energy storage stations:

      where em-pmax(t) is the rated power of the m-th energy storage station, em-p(t) is the output power of the m-th energy storage station in time period t, em-smax is the rated stored power of the m-th energy storage station, and em-s(t) is the stored power of the m-th energy storage station in time period t.

      Among them, equation (12) shows the relationship between charging and discharging power and the amount of power for a certain duration.

      4.2.2 Safety constraints

      Equation (13) shows the power flow constraint of the transmission line:

      where GSF is the generation shift factor [31-32] and Fkmin and Fkmax are the minimum and maximum flow constraints of the transmission equipment, respectively.

      Equation (14) shows the system spinning reservation constraint:

      where S% is the system spinning reservation rate.

      4.2.3 Environmental constraints

      Equation (15) shows the constraint on the proportion of renewable energy:

      where Hclean(t) and Hall(t) are renewable energy power generation and total society power generation in time period t, respectively, and D% is the required proportion of renewable energy.

      Equations (16) and (17) are the constraints of wind and solar curtailment rates:

      where Ha-wind(t) and Hwind(t) are wind power curtailment and generation, respectively; Ha-PV(t) and HPV(t) are solar power curtailment and generation, respectively; and Qwind% and QPV% are the control limits of the wind and solar curtailment rates, respectively.

      Equation (4) is the objective function of the system collaborative optimization model, and equations (5) to (17)are the constraints of the collaborative optimization model.Given the initial value of each variable, the optimization algorithm can be used to calculate and obtain the operation strategy for various power supply combinations and the project of direct current (DC) injection of remote clean energy.

      5 Optimization algorithm for collaborative operation strategy

      To achieve the coordination of multiple system output variables, an adaptive optimization algorithm is necessary to rapidly and accurately obtain the optimal solution [15].In this study, a system operation strategy risk method based on the improved particle swarm optimization (PSO) was adopted.PSO is an evolutionary computing technology used to determine the optimal solution through collaboration and information sharing between individuals in a group.This technology is widely used in the field of genetic algorithms,such as function optimization, neural network training,and fuzzy system control.Based on this algorithm, in the previous section, the objective function with the lowest overall system cost and the constraints on operation, safety,and environmental protection is proposed.

      PSO algorithm uses a loop approximation method to obtain optimization results.Its basic process includes six steps.

      First step: initialize the position and velocity of a group of particles with a group size of N.Select a group of random particles that meet the constraint conditions with amount r,and then determine the optimal solution through iterative optimization.In each iteration, each particle updates its velocity and position according to the equations (18-19):

      where the subscript d is the number of iterations, xd is the particle spatial position during the d-th iteration, vd+1 is the particle velocity during the d+1-th iteration, ω is the constant of inertia, φ1 and φ2 are learning factors, and rand()is a random number between (0,1).

      Second step: evaluate the fitness of each particle in the initial state.pBest and gBest in equation (18) are the local and global optimal positions of the particle swarm,respectively.

      Third step: for each particle, compare its fitness value with the best position of pBest.If it is good, use it as the current best position pBest.

      Fourth step: for each particle, compare its fitness value with the best position gBest.If it is good, use it as the current best position gBest.

      Fifth step: adjust the particle velocity and position according to equation (18).

      Sixth step: if the end condition is not satisfied, go to the second step.The iteration end condition is generally selected as the maximum number of iterations Gk or/and the optimal position the particle swarm has searched thus far to meet the predetermined minimum adaptation threshold according to the specific problem [33-34].

      6 Case analysis

      A certain region in Gansu Province is used here as an example to conduct a case analysis.This region has a planned annual load of 4700 MW and planned growth of 5200 MW, with an 8000 MW ultra-high voltage DC power injection line; wind power installed capacity of 10000 MW and planned capacity of 2000 MW; photovoltaic installed capacity of 800 MW and planned capacity of MW; coal power installed capacity of 9000 MW, which will remain constant; and hydropower installed capacity of 1000 MW with no new planned hydropower station, but conditions to expand for one 200 MW unit.The potential peak shaving methods, developable scope, and composite costs of the system are as summarized in Table 1.

      Table 1 Adjusting capacity in power system and comprehensive cost

      Supply type Installed capacity/MW Investment/(yuan/kW)Operation cost/(yuan/kW)Composite cost/(yuan/kW)Thermal power flexibility transformation 1 0~800 33 0.05 0.06 Thermal power flexibility transformation 2 0~1600 400 0.1 0.15 Hydropower expansion plan 200 2000 0 0.20 Electrochemical energy storage 0~∞ 1500 0.03 0.45 Demand-side management and regulation 0~700 200 0 0.04 Pumped storage 0~8000 20000 0.075 0.31 Natural gas power 0~500 3000 0.01 0.32

      The full-time sequence production simulation model and collaborative operation strategy optimization algorithm were used to analyze the case, and the analysis results ofpeak shaving resource allocation and peak shaving cost under different scales of renewable energy were obtained.

      From the perspective of construction costs of peak shaving resource, pumped storage has the highest installed cost.Investment for thermal power flexibility transformation and demand-side management is relatively small.The cost of hydropower expansion and natural gas power generation is relatively low, but natural gas power generation will increase carbon emissions, which will adversely affect carbon emission reduction.The construction cost of pumped storage is high, but the overall cost is low, as its storage capacity is large.However, because of the high current cost of electrochemical energy storage, the overall cost is still relatively high.

      To meet the requirements of power system operation and renewable energy curtailment, renewable energy and the newly added peak shaving capacity need to continuously grow.Considering that different peak shaving technology routes have different comprehensive costs, with the lowest comprehensive cost of the system peak shaving as the optimization goal, the optimization analysis of various peak shaving technology scales was performed.Under various renewable energy development scales, the optimization suggestions for newly added peak shaving capacity are shown in Fig.1.When the development of renewable energy is low, the priority is to adopt the thermal power flexibility transformation and load demand-side response of the low-pressure cylinder method.Subsequently,methods such as hydropower expansion, pumped storage,and natural gas power generation can be implemented.Finally, electrochemical energy storage and deep thermal power flexibility transformation can be implemented.The construction of system peak shaving capacity should prioritize technologies with low composite costs and good implementation effects.After the peak shaving potentials of the technologies are fully developed, other technologies with high composite costs can be selected.

      Fig.1 Relationship between renewable energy plant and the adjusting capacity in the power system based on optimization

      Figs.2 and 3 show the construction and composite costs of system peak shaving under various renewable energy scales.Fig.2 shows that, as the scale of renewable energy grows, the construction cost per unit power of peak shaving facilities gradually increases.Particularly after the start of pumped storage construction, the investment scale increases sharply.Fig.3 shows that the composite cost of unit peak shaving power exhibits a steady upward tendency,and the final system peak shaving power cost rises to 0.27 yuan/kWh.After the integration of renewable energy, the system composite peak shaving cost is at a relatively high level.A renewable energy scale of 28000 MW by 2025 is recommended for this region, to ensure the peak shaving cost remains low.

      Fig.2 Cost of the adjusting capacity in the power system with various renewable energy plants

      Fig.3 Price of the adjusting capacity in the power system with various renewable energy plants

      Overall, by using the collaborative optimization model that considers the composite peak shaving cost proposed in this study, development and construction plans can be developed for various peak shaving resources to facilitate power system development planning and operation strategies.

      7 Conclusion

      According to China’s carbon peak and carbon neutrality goals and the development needs of new power systems with renewable energy as the primary component,the determination of flexible resources is extremely important for the construction of new power systems.This study listed the development potential, construction costs, and operating costs of coal power, hydropower,natural gas, electrochemical energy storage, demandside management, pumped storage, and natural gas power generation; established a comprehensive cost evaluation model; considered operational, safety, and environmental constraints; and established a power system collaborative optimization model considering the composite cost of peak shaving.Finally, considering a certain region in Gansu Province as the object of analysis, a construction method of peak shaving resources under various renewable energy development scales was proposed by utilizing the collaborative optimization model to rationally facilitate the development of new energies.

      The exploration and construction of flexibility resources for the power system are affected by several factors,such as market environment, production relations, and development conditions; thus, a large amount of data and variable conditions should be obtained.In future research,the accumulation of flexibility resource cost data and the evaluation of the composite cost and benefit of the peak shaving system with respect to the auxiliary services of the power market should be considered, to provide basic technical support for more accurate coordinated optimization of the power system.

      Acknowledgements

      This work was supported by the National Natural Science Foundation of China (No.71273088).

      Declaration of Competing Interest

      We declare that we have no conflict of interest.

      References

      1. [1]

        Chen J, Yang R, et al (2015) Optimization design of combined tower structure for large wind turbines.Journal of Hunan University(Natural Sciences), 42(05):29-35 [百度学术]

      2. [2]

        Wang N, Shao X, et al (2015) Prediction of average annual power generation of wind farms based on regional information fusion.Journal of Hunan University(Natural Sciences), 42(08):81-85 [百度学术]

      3. [3]

        Bi P,Xu X, et al (2019) Research on risk assessment method and delivery capability of chain off-network in wind power base.Power System Technology, 43(03):165-172 [百度学术]

      4. [4]

        Kang C, Yao L (2017) Key scientific issues and theoretical research framework for high proportion renewable energy power systems.Automation of Electric Power Systems, 41(9):2-11 [百度学术]

      5. [5]

        Chi Y, Liang W, et al (2016) Summary of research on key technologies of large-scale offshore wind power transmission and grid connection.Proceeding of the CSEE, 36(14):3758-3770 [百度学术]

      6. [6]

        Liu H, Sun J (2014) Voltage stability and control of offshore wind farms with AC collection and HVDC transmission.IEEE Journal of Emerging & Selected Topics in Power Electronics,2(4):1181-1189 [百度学术]

      7. [7]

        Jiang W, Deng W, et al (2015) RMC-based offshore wind power-HVDC system power coordinated control.Automation of Electric Power Systems, 27(10):1-8 [百度学术]

      8. [8]

        Zhao D, Ma J, et al (2017) Reactive power configuration and coordinated control strategy for offshore wind farms sent out via AC cable.Power System Technology, 41(05):1412-1421 [百度学术]

      9. [9]

        Zhang L, Xiong J, et al (2015) Cross-regional energy allocation optimization model considering clean energy participation.Journal of Hunan University(Natural Sciences), 42(04):71-77 [百度学术]

      10. [10]

        Li L, Zheng Y, et al (2016) Wind power fluctuation cost calculation and application research.Proceeding of the CSEE,36(19):5155-5163 [百度学术]

      11. [11]

        Jiang Y, Yang X, et al (2016) Ultra-short-term wind power prediction based on EEMD-IGSA-LSSVM.Journal of Hunan University(Natural Sciences), 43(10):70-78 [百度学术]

      12. [12]

        Yuan Ti, Chao Q, et al (2010) Modeling of dynamic clean economy optimization scheduling for large-scale wind power grid-connected power systems.Proceeding of the CSEE,30(31):7-13 [百度学术]

      13. [13]

        Zhang Z, Guo J, et al (2017) Optimal analysis of offshore wind power grid-connected scheme based on full life cycle cost.Power System Protection and Control, 45(21):51-57 [百度学术]

      14. [14]

        Morandi A, Gholizad B, et al (2016) Technical and economical evaluation of DC high-temperature superconductor solutions for the grid connection of offshore wind parks.IEEE Transactions on Applied Superconductivity, 26(6):1-10 [百度学术]

      15. [15]

        Chang J, Ummels B C, et al (2009) Economic evaluation of offshore wind power in the liberalized Dutch power market.Wind Energy, 12(5):507-523 [百度学术]

      16. [16]

        Tao Y, Shi J (2013) China’s offshore wind power development situation and electricity price policy research.Energy of China,35(6):25-29 [百度学术]

      17. [17]

        Weng Z, Shi L, et al (2014) Dynamic economic dispatch of power system considering wind power cost.Proceeding of the CSEE, 34(4):514-523 [百度学术]

      18. [18]

        Zhang Z, Sun Y, Gao D, et al (2013) A Versatile Probability Distribution Model for Wind Power Forecast Errors and Its Application in Economic Dispatch.IEEE Transactions on Power Systems, 28(3):3114-3125 [百度学术]

      19. [19]

        Li P, Wang J, et al (2014) Comparative analysis of small disturbance stability of doubly-fed and direct-drive wind turbines.Journal of Hunan University(Natural Sciences), 41(01):92-97 [百度学术]

      20. [20]

        Ai X, Liu X (2011) Dynamic economic dispatch of power system with wind farm based on credibility theory.Proceeding of the CSEE, 31(S1):12-18 [百度学术]

      21. [21]

        CAO Y, LI Q, et al (2018) A comprehensive review of Energy Internet: basic concept, operation and planning methods, and research prospects.Journal of Modern Power Systems and Clean Energy, 6(3): 399-411 [百度学术]

      22. [22]

        Ji F, Cai X, et al (2014) Fuzzy robust dispatch for power systems with wind farms.Zhongguo Dianji Gongcheng Xuebao/Proceedings of the Chinese Society of Electrical Engineering,34(28):4791-4798 [百度学术]

      23. [23]

        Yuan X, Fan Y (2013) China power structure prediction based on logistic and learning curve model.Journal of Hunan University(Natural Sciences), 27(04):51-55 [百度学术]

      24. [24]

        Ai X, Liu X (2011) Dynamic economic dispatch of power system containing wind farm based on credibility theory.Proceeding of the CSEE, 31(S1):12-18 [百度学术]

      25. [25]

        Weng Z, Shi L, et al (2014) Dynamic economic dispatch of power system considering wind power cost.Proceeding of the CSEE, 34(4):514-523 [百度学术]

      26. [26]

        Zhou W, Sun H, et al (2011) Review of economic dispatching of power systems with wind farms.Power System Protection and Control, 39(24):148-154 [百度学术]

      27. [27]

        Chang J, Ummels B, et al (2009) Economic evaluation of offshore wind power in the liberalized Dutch power market.Wind Energy, 12(5):507-523 [百度学术]

      28. [28]

        Chen G, Chen J (2013) Modeling and algorithm for environmental economic dynamic scheduling of wind farm electric power system.Proceeding of the CSEE, 33(10):27-35 [百度学术]

      29. [29]

        E J, Li G, et al (2011) Simulation analysis of planetary gear fatigue in megawatt wind power yaw reducer.Journal of Hunan University(Natural Sciences), 38(09):32-38 [百度学术]

      30. [30]

        Zhang K (2016) Summary of research on economic dispatching of power system with wind power grid-connected.Mechanical and Electrical Information, (27):40-41 [百度学术]

      31. [31]

        Sun Y, Xiao J, et al (2016) Future transmission grid utilization assessment considering safety constraints.Power System Technology, 40(12):3818-3824 [百度学术]

      32. [32]

        Tian Y (2015) Improvement of particle swarm optimization algorithm and its application in power system.Electronic Test,(19):47-48 [百度学术]

      33. [33]

        Hou Y, Lu L, et al (2004) Improved particle swarm optimization algorithm and its application in power system economic load distribution.Proceeding of the CSEE, 24(07):99-104 [百度学术]

      34. [34]

        Wang Y, Cai Z (2011) A hybrid multi-swarm particle swarm optimization to solve constrained optimization problems.Journal of Taiyuan Heavy Machinery Institute, 11(1):295-304 [百度学术]

      Fund Information

      supported by the National Natural Science Foundation of China (No.71273088);

      supported by the National Natural Science Foundation of China (No.71273088);

      Author

      • Ji Liang

        Ji Liang (1972), Male, Ph.D., Senior Engineer.His research interests include power markets,economic dispatching, and power dispatching operation management.

      • Xingang Zhao

        Xingang Zhao (1971), Male, Ph.D., Professor,Doctoral Supervisor.His research interests include renewable energy industry theory and policy.

      • Shangdong Yang

        Shangdong Yang (1980), Male, Ph.D., Senior Engineer.His research interests include power markerts, economic dispatching, strategy and plan of power company management.

      Publish Info

      Received:2021-01-11

      Accepted:2021-07-01

      Pubulished:2021-08-25

      Reference: Ji Liang,Xingang Zhao,Shangdong Yang,(2021) Collaborative optimization model of renewable energy development considering peak shaving costs of various flexibility resources.Global Energy Interconnection,4(4):394-404.

      (Editor Yanbo Wang)
      Share to WeChat friends or circle of friends

      Use the WeChat “Scan” function to share this article with
      your WeChat friends or circle of friends