【作者】 张富强；

【导师】 文福拴；

【作者基本信息】 浙江大学 ， 电力系统自动化， 2009， 博士

【摘要】 电力工业的市场化改革打破了其垂直一体化的垄断格局,这给电力系统的运行带来了深刻的影响,无功容量不充裕给系统运行带来的损害以及电价波动带来的金融风险是其中的两个重要问题。2003年美加大停电和2000～2001年的加州电力危机就与这两个问题有关。本文就电力市场环境下的这两个问题进行了比较系统的研究。首先对近年来国内外就电力市场环境下无功问题的研究现状进行了综述。包括无功成本分析、无功定价方式、无功获取与付费、无功调度以及无功市场势力等方面。之后,对电价风险管理方面的研究现状做了综述。在无功市场尚未建立之前,需要对无功源进行成本补偿。因此,对无功源进行成本分析非常重要,它是对无功服务进行定价以及补偿的基础。本文比较了各种无功源(包括同步发电机、电容器、同步调相机、变压器分接头、SVC和STATCOM)的物理特性,并且估计了它们提供无功服务时的固定成本和可变成本,接着详细讨论了发电机在提供无功服务时,固定成本在有功功率和无功功率之间的分摊,指出了使用“元／MVA”的表达式以计算无功容量投资成本的一些错误。举例说明发电机在提供无功服务时机会成本的计算。同时也给出了在考虑资金回报率时,电网侧的无功成本折算到每年度的固定成本表达式。对系统所需无功容量的准确估计能够帮助预防因为无功功率缺乏导致的电压崩溃事故,同时也能避免装设过多无功源而造成的浪费。计及了多个无功源可能同时发生预想故障的风险,提出了一种新的无功容量需求的确定方法,此方法同时考虑了购买无功容量的成本、切负荷的成本以及发电再调度的成本,以这三项成本之和最小化为目标,确定需要购买的无功容量的大小及其分布。系统调度机构可以采用类似思想确定每年需要采购的无功源的数量及其位置分布。由于无功功率的本地特性,无功市场势力成为建设竞争性无功市场的最大障碍,因此分析系统中哪些无功源拥有潜在的市场势力对推进无功市场的建设意义重大。现有的无功市场势力评估指标大多只考虑了电力系统的技术特征,没有考虑市场参与者的策略性投标行为。在此背景下,提出了基于智能体代理的市场仿真方法以研究无功市场中参与者的博弈行为,其中采用了模仿者动态算法来模拟智能体(参与者)追求利润最大化的理性行为。发展了计及发电机强制无功出力要求的统一价格竞价模式和PAB竞价模式,并利用得到的结算电价计算市场势力指标,即价格成本裕度指标。获得准确的现货电价分布形式能够帮助市场参与者确定自身在市场竞争中的位置,特别是有效地分配在现货市场参与竞争的电能比例。现有的研究很多假设现货电价服从正态分布的特征,从浙江电力市场异常的现货电价数据出发,本文基于某一区域的负荷呈正态分布的假设,结合市场总供给曲线在某一点飞升的严重非线性特征,采用概率论从理论上推导了电价作为负荷这种随机变量的函数的分布情况,并以加州、PJM和中国浙江电力市场的实际交易电价数据做了验证。结果表明,只有在电力供求关系很宽松的情况下,电价才近似呈正态分布；而在供求紧张的情况下,电价分布明显偏离正态分布,且具有严重的右偏峰特性；实际电力市场在供求紧张时还表现出狭峰特征。紧接以上对现货电价分布形式的研究,提出了一种新的电力市场环境下的双边合同轮流出价谈判合同。基于谈判双方对自身能够承受的风险和利润底线的假设,采用了一系列动态的讨价还价过程以描述谈判过程模型,允许市场参与者根据偏好自己定义风险和收益指标,在本文中以期望效用理论描述电力市场因电价波动带来的风险,以期望利润定义参与者的收益。采用让步因子描述参与者的谈判策略,并为市场参与者设置了期望效用和期望利润的下限限制,这样,谈判最后能够达成一个对双方而言、风险和收益都可以承受的合同。更多还原

【Abstract】 The power industry restructuring breaks the vertically integrated monopoly of the tranditional power industry with the aim of introducing competition, improving efficiency optimizing resource allocation and reducing power price. However, many new problems and challenges have emerged during the restructuring process. Two vital problems are damage to system caused by insufficient reactive power support, as well as financial risks due to fluctuations of electricity market prices. If these new problems are not properly solved, the emerging electricity market may not be able to grow healthily and even the security and stability of power system associated operation could be endangered. This dissertation focuses on the study of reactive power related issues and on the research of price risk management in the electricity market environment and some significant results are obtained.This dissertation first reviews the reactive power related issues, which include costs of reactive power, Var pricing, procurement and payment of Var, Var dispatch and reactive market power. As followed, literatures on price risk management problem are reviewed.Physical characteristics of reactive power equipments, including synchronous generator, shunt capacitor, synchronous compensator, transformer tap changer, SVC, and STATCOM, are compared firstly. Fixed costs and variable costs of different reactive power sources in electricity market environment are estimated. And then the fixed cost allocation between active power and reactive power provided by synchronous generator is put forward. Meanwhile, this paper indicates one mistake in calculating investment costs of reactive capacity with yuan/MVA expression. The calculation of opportunity costs of synchronous generator in providing reactive power services is illustrated. Lastly, the annual component of fixed cost taking account of return on capital is presented.Appropriate determination of reactive power requirements is an important issue for maintaining voltage stability, decreasing transmission losses, and mitigating the potential exercise of reactive market power. A model is developed for annual procurement of reactive power capacity considering the possible outage of multiple Var sources in electricity market environment. The objective is to minimize the total cost, including the cost of purchasing reactive power capacity, load shedding and generation redispatch. A method based on the Monte-Carlo simulation and Differential Evolution algorithm is served for solving the established optimization problem. Simulation results from case studies demonstrate the essential features of the developed model and method.Reactive market power has been regarded as a vital problem for its negative effect on the efficient operation of the market. Up to now, research work on the evaluation index for reactive market power has been mainly based on technical characteristics of power system. However, the strategic bidding between market participants is not taken into account. Given this background, an agent-based market simulation framework is presented for analyzing the game behaviors and resulting reactive market power of market participants. Replicator dynamics algorithm is utilized to simulate agents’ rational behavior of pursuing maximum profits. Two models, in which minimum reactive power requirement is taken into account, are presented for describing uniform-price auction and pas-as-bid (PAB) auction in reactive power market. The reactive market power is calculated using well-know price cost margin index. The CIGRE 32-bus system is used for case studies. Simulation results demonstrate the essential features of the developed model and method. Furthermore, the analysis results indicate that reactive market power in uniform-price auction is more severe than that in PAB auction.The study on probability density function and distribution function of electricity prices contributes to the power generators and purchasers to estimate their own management accurately, and helps the regulator monitor the periods seriously deviating from normal distribution. Based on the assumption that load presents normal distribution, combining serious non-linear characteristic of aggregate supply curve, this paper has theoretically derived the distribution of electricity prices as the function of random variable of load utilizing knowledge of probability theory. And the conclusion has been validated with the electricity price data of California, PJM and Zhejiang markets. The result shows that electricity prices obey normal distribution approximately only when supply-demand relationship is very loose, whereas the prices deviate from normal distribution highly and present seriously right-skewness characteristic. Moreover, the real electricity markets also display the narrow-peak characteristic when undersupply occurs.Participants can sign forward bilateral contracts several months in advance of its delivery for mitigating price risk in electricity markets. Up to now, the research work for bilateral contracts has concentrated on the effect of bilateral contracts on reducing price risk. Given this background, we propose a systematic negotiation scheme through which a generator and distribution company can reach a mutually beneficial and risk tolerable forward bilateral contract. Under this approach, the generator and distribution company respond rationally to a stream of bilateral offers/counter-offers considering their respective benefits while accounting for the risks incurred by the uncertainty in the pool spot price. Each negotiating party can choose its own definition of benefits and risk, which are expected utility in this paper. Concession factor is utilized to illustrate the negotiation strategy of participants. The reservation expected utility and reservation expected profit are added to this approach to model the tolerable limits of risk and benefit. Numerical tests show that this flexible negotiating framework can be readily put into practice.更多还原