摘要
本文基于拥有风力发电厂与电力储能设备的电力运营商视角,考虑运营商参与电力交易会对市场电价产生影响的基础上,通过动态规划,以利润最大化为目标,研究其最优联合调度策略。研究表明:在考虑充放电效率以及运营成本时,存在三个最优参考点将储能设备中的可用库存空间分为四个区域,进而对应四个不同的决策活动:将风电和从市场上购买的电力全部存入储能系统;将部分风电存入储能系统并将剩余部分销售到市场;售出全部风电不向储能设备中存入电量;将所有的风电和储能系统中的部分电量销售至电力市场。运营商可根据当前时刻风力发电量和储能设备中的电量与最优参考点之间的大小关系,得到相应的最优电力调度策略及其交易量。此外,本文研究发现随着市场影响强度的增大,运营商的利润随之减小,因此运营商应在其市场影响强度与最优电力交易量之间做好权衡。本研究对运营商的电力调度具有重要的指导意义和实际应用价值。
Wind power generation has increased with the participation of renewable energy power generation in power grid transactions.However,because wind power generation is easily affected by factors such as geography and climate,its high level of uncertainty and intermittent issues have attracted widespread attention.To satisfy the requirements of the power market to maintain a balance between power supply and demand,this study introduces energy storage,a useful tool for mitigating power supply fluctuations through the charging and discharging of energy storage devices to satisfy electricity market demand.By acquiring power to store when prices are low and sell to the market when prices are high,electricity merchants can also use energy storage to engage in energy arbitrage and maximize profit.Co-optimization of renewable power plants with energy storage by storing electricity when the demand for power is lower than the supply of renewable energy can potentially provide future value by mitigating the intermittent nature of renewable energy generation.Most models used in current studies assume that the energy storage capacity is sufficiently small with respect to the wholesale power market and that charging and discharging decisions do not affect the price of electricity.Consequently,a merchant(also known as a price-taker merchant)sells or purchases a certain amount of power at a price that is unrelated to his or her own trading actions.Price arbitrage,in which the merchant′s operational decisions affect electricity prices in the market(i.e.,a price-maker merchant),in turn influences the merchant’s actions,thus illustrating the value of large energy storage,such as pumped storage hydropower.We investigate the best multi-period scheduling decision made by electricity merchants by considering market impact,unreliable wind energy,and energy storage limitations.To facilitate merchant decision-making in considering the impact of the market,we approximate the price of electricity by a linear function of the amount of power that the merchant transacted in the reward function.This research overcomes the challenges in obtaining analytical conclusions when incorporating market impact,because nonlinear reward functions are used instead of the conventional linear reward and value functions.The energy storage capacity,pumping/charging and generating/discharging capacity,and facility pumping/charging and generating/discharging efficiency must all be considered in modeling.To maximize profit,we construct this problem as a Markov decision process and investigate the electricity merchant′s optimal co-optimization operational trading decisions.To address this optimization problem,we first divide the original optimization problem into three sub-optimization problems corresponding to three possible activities of the electricity merchant and conditional value functions in each period.On the basis of the Bellman equation,the best solution for each sub-optimization problem is discussed.Our closed-form analytical results may support merchants’multi-period decision-making for energy storage scheduling.We finally combine and compare the best sub-optimization situation to draw overall conclusions regarding the original problem and determine the best decision rules across the entire optimization horizon.We analytically show that three state of charge(SOC)reference points depend on the energy that is currently available,the predicted price,the market price impact,and the energy of wind generation that is currently available.The storage SOC is divided into four subregions by using three SOC reference points relating to four distinct activities.By simply comparing the current storage SOC and the SOC reference points in the following period,the merchant can obtain the corresponding optimal decisions,which are discharging stored power and selling all wind generated power to the market;selling all wind generated power and storage remaining idle;storing and selling a portion of wind generated power;and storing all wind generated power and purchasing power from the market.The pricing differential between peak and off-peak periods shrinks when the market impact is taken into account,thus raising the cost of acquiring power from the electricity market and lowering the revenue from selling power to the electricity market.We analytically demonstrate that when both the price-maker and the price-taker merchants offer the same generating/discharging and pumping/charging limits to the independent system operator,the market impact may decrease electricity merchants′expected profit and substantially alter the optimal storage policy structure.An electricity merchant who merely has storage is a special case different from that of a merchant who also has a wind farm.The decision-making approach is also supported by a case study of synthesized data and a study of a midcontinent independent system operator′s real data.The relationship between market impact strength and anticipated maximum profit is investigated.With an increase in the strength of the operator′s trading decisions on the market,the operator′s income declines.Therefore,throughout the optimization horizon,merchants should decide to limit the power transaction quantity to lessen the adverse market impact on trade.Our findings demonstrate that,by affecting the values of reference points,market impact substantially alters the best storage scheduling strategy.Furthermore,compared with the use of separate schedules for energy storage and wind farms,the co-optimization scheduling strategy benefits merchants with energy storage and wind power plants.This research may provide guidance and practical application value for merchants participating in wholesale power market scheduling who have energy storage and renewable power sources.
作者
刘健
孙心玥
戴卓妮
欧盟
刘焰焰
张娜
LIU Jian;SUN Xinyue;DAI Zhuoni;OU Meng;LIU Yanyan;ZHANG Na(School of Economics and Management,Nanjing University of Science and Technology,Nanjing 210094,China;Business School,Nanjing University of Information Science and Technology,Nanjing 210044,China)
出处
《管理工程学报》
CSSCI
CSCD
北大核心
2024年第2期206-220,共15页
Journal of Industrial Engineering and Engineering Management
基金
国家自然科学基金项目(71671092、72071112)
南京信息工程大学气候与环境治理研究院开放课题(ZKKT2022A16)。
