发电系统可靠性研究.doc

发电系统可靠性研究(二) 摘 要 近些年来，电力系统不断向高电压、远距离、大容量方向发展，在提高经济性的 同时，安全可靠性的问题也随之而来。随着电力系统规模的不断扩大，系统结构日趋 复杂，电力系统的可靠性也受到人们越来越多的重视。发电系统作为电力系统的一个 重要部分.在对发电系统进行可靠性分析时，通常假设电力系统的其余部分均完全可靠， 也就是说在任何时刻，衡量系统是正常还是故障，要看系统发出的电力能否满足系统 负荷的需要。 对发电系统尽行可靠性的定量分析，本文是建立发电容量模型。发电机组或系统 在 T 时刻处于某种容量状态的概率叫做容量状态概率模型。它的概率分布是离散的。 在实际应用中容量概率模型有两种表达式：一种是可用或有效容量概率模型；另一种 是停运容量概率模型。本文使用的是有效容量概率模型。对发电系统进行容量模型的 建立和降额、停运的相关计算。 关键词：发电系统；可靠性评估；容量模型 Abstract In recent years, electric power system continued to high-voltage, long-distance and large-capacity direction of the development. At the same time improve the economy, the issue of safe and reliable follow. With the size of the electric power system continues to expand, the growing complexity of system architecture, the reliability of the electric power system has been the increasing importance. Power generation system as the electric power an important part, in the power generation system reliability analysis, the assumption is usually the remainder of the electric power system are totally reliable. That is, at any time measuring system is normal or fault, it is issued by the electricity system can meet the needs of system load. The power generation system reliability of the firms do quantitative analysis, this paper is to establish generation capacity model. Generating units or systems on the T at a time when the capacity of state called the probability of capacity probability model. It is the probability distribution of discrete. In actual application capacity probabilistic model has two expressions : One is available or effective capacity probability model; Another is the capacity outage probability model. This paper is an effective capacity of a probability model.The power system model for capacity building and landing places, the outage is calculated. Key words: reliability estimate，model of capacity，electric power system 目 录 摘要I I Absteact IIII 1.概述1 1 2.发电系统可靠性的基本概念4 4 2.1 发电系统可靠性的基本概念4 2.1.1 元件可靠性 4 2.1.2 元件状态的马尔科夫过程 7 2.2 发电系统可靠性的评估指标9 2.2.1 发电机组的强迫停运率 9 2.2.2 发电系统可靠性的基本指标和标准 .10 3.发电系统可靠性评估的数学模型1 12 2 3.1 发电系统可靠性估计的应用及计算方法介绍.12 3.1.1 发电系统可靠性估计的应用 .12 3.1.2 发电系统可靠性评估所使用的方法 .12 3.2 发电容量模型.15 3.2.1 一台机容量模型 .16 3.2.2 两台机容量模型 .18 3.2.3 三台机容量模型 .22 3.3 容量模型的递推公式.24 3.4 机组降额运行和停运模型的运算.26 3.4.1 减少机组时停运容量模型 .26 3.4.2 机组降额运行时的容量模型 .27 4. 实例计算2828 4.0 数据28 4.1 四台发电机组全额运行时的可靠性分析28 4.2 四台发电机组降额和停运运行时的可靠性分析32 4.2.1 降额运行时的可靠性分析 36 4.2.2 计划检修情况下的可靠性分析 36 5.电力系统可靠性仿真测试3838 5.1 VB 的简介 38 5.1.1 VB 的发展过程 38 5.1.2 VB 的特点 38 5.2 流程图和程序界面.40 5.2.1 流程图 40 5.2.2 程序界面 40 5.2.3 程序代码 41 5.2.4 仿真结果 46 结论4747 参考文献4848 致 谢4949 附录 1 英文原文 5 50 0 附录 2 中文翻译 6 63 3 1.概述 可靠性是指一个元件设备或系统在预定时间内，在规定的条件下完成规定功能的 能力。可靠度（Reliability)则用来作为可靠度的特性指标，表示元件可靠工作的概率。 电力系统包括的范围很大，研究可靠性时要根据发电、输电、配电等不同环节的 要求，突出主要矛盾，构成不同环节的可靠性计算方法。60 年代，人们开始研究可靠 性时，主要侧重于发电系统或以发电和输电组成的组合电力系统的可靠性评估，而配 电系统可靠性研究远未得到应有的重视。主要原因在于发电设备比配电设备集中，设 备一次性投资大，建设周期长，且发电容量不足造成的停电给社会及环境可能带来严 重和广泛后果。然而，配电系统不可靠造成的损失也是非常大的，据电力公司统计： 大约 80的用户故障缘于配电系统故障。目前，在工业发达国家，可靠性已经成为配 电系统规划决策中一种常规性工作。 在国内，对配电系统可靠性的研究始于 80 年代初期，由于缺乏必要的统计数据和 有效的分析方法，发展较为缓慢，近年来，由于电力供需矛盾日益突出，人们对电能 的质量的要求越来越高，电力系统可靠性在生产管理中的地位越来越重要，从上世纪 七十年代初期以来，许多国家的大电网相继发生大的事故，引起大面积停电，不断造 成巨大的经济损失，而且危机社会秩序。因此，定量评定和改善电力系统可靠性越来 越受到人们的重视。以下举例几个国际大停电事故的事例： （1）：美加大停电 2003 年 8 月 14 日 16 时 11 分，美国和加拿大相邻的一个变电站发生了事故，眨 眼之间，加拿大多伦多、渥太华断电，美国纽约、克利夫兰、底特律也同时停电，酿 成北美历史上最为严重的大停电事故。美国东北部的密歇根、俄亥俄、纽约等 6 个州 以及加拿大的安大略省也受到严重影响。停电波及 9300 平方英里，5000 万人饱受断 电之苦。估计整个经济损失在 250 亿300 亿美元之间。 北俄亥俄州的事故导致线路跳闸，从而引起 200 万千瓦潮流变化，巨大的电力环 流冲击使电网联络线相继跳闸，最后造成各地区电压崩溃，引起了这次大停电。 （2）：伦敦大停电 2003 年 8 月 28 日，英国伦敦和英格兰东南地区发生了大面积的停电事故，伦敦 地铁等交通系统受到严重影响。 这次停电的主要原因是安装了一个错误规格的保险丝，致使自动保护设备被误启 动，自动切断了赫斯特、新克劳斯和威姆别利顿电站与电力传输系统的联系，使伦敦 电力供应量瞬间减少了五分之一。由于电力缺额过大造成了这次大停电。 （3）：莫斯科大停电 2005 年 5 月 25 日 10 时许，俄罗斯首都莫斯科南部、西南和东南城区大面积停电， 市内大约一半地区的工业、商业和交通陷入瘫痪。停电损失至少为 10 亿美元。 停电事故由恰吉诺变电站发生系列爆炸和火灾直接引起。该站建于 1963 年，设备 均已老化。且电网处于超负荷运行状态，运行人员也未引起注意，缺乏严格的操作规 程约束及协调手段。 发电系统是电力系统一个重要的组成部分。在对发电系统进行可靠性分析时，假 设电力系统的其余部分均完全可靠，也就是说，如果发电容量充足，输电和配电系统 可以将发电系统的电能传输到任何负荷点，而不致由于过负荷或母线电压偏移超过允 许值等原因出现电力不足。因此，在任何时刻，衡量系统是正常或故障的判据，是发 出的电力能否满足负荷的需要。为了提高发电系统的可靠性，一方面是尽量提高发电 机组的可用度，另一方面是加大发电备用容量，以备在机组发生故障或维修而退出运 行时使用。显然，可靠性的概念和备用是紧密联系在一起的；同样，可靠性和增加机 组的安装和运行费用也是联系在一起的他们之间还没有一个简单的关系式。 发电系统备用容量的估计，无论是对系统的规划设计或生产设计都是十分重要的 问题，各国现在仍在沿用的和近代估计备用容量的方面有以下 6 种： 百分数备用法；偶然故障备用法；电力不足概率法（LOLP 法） ；频率及持续时间 法（F Small Hydro-Power Plant (SHPP); Economic analysis; Monte Carlo method 1. Introduction Small Hydro-Power Plants (SHPPs) have found special importance due to their relatively low administrative and executive costs, and a short construction time compared to large power plants. These SHPPs are in the run-off river category because their generated capacity is based on the deviated water fl ow of river run-off and consists of a diversion dam, conveyance of water system, headpond, forebay, penstock, power house, and tailrace structure of the body of the SHPP as well as other electrical and mechanical equipment (see Fig. 1).The deviated fl ow of a river reaches the forebay after running in a path to the diverted point, and then enters into the SHPP structure via penstock pipes.Daily regulation of the water volume in the headpond is used to get maximum power from the SHPP during peak hours. The amount of energy generated during different daily hours and/or different seasons of the year are the most important issues worthy of study in the run-off river SHPP studies. In other words,calculating the optimal installation capacity (optimal designed fl ow) is one of the most important factors in planning SHPPs. 2. Determination of the optimal installation capacity To determine the optimal installation capacity of SHPPs all technical, economic and reliability indices are considered in a trade-off relation. Using this approach, the amount of annual energy is determined by using categorized statistics of the fl ow duration curve in different months. Then, after specifying the income and costs of the plant, the economic indices of different alternatives are extracted. The reliability indices are then calculated and ultimately, through comparison of the technical, economic and reliability indices, a superior alternative can be selected, determining the optimal installation capacity. This method of calculating the technical, economic and reliability indices and the subsequent processes used in the planning of an SHPP will be further discussed and described. Fig. 1. Schematic diagram of a typical small hydro-power plant. 2.1. Method of energy calculation After determining the downstream water fl ow and environmental needs and rights, the energy calculation is done with respect to the water fl ow categorized data of the river. To estimate the generated energy, a range of fl ow rate is specifi ed based on a level of adjustment.Then, based on the daily and/or monthly statistics of the fl ow duration curve and the water fl ow with different probability percentages (e.g., 20%, 30%, 40%, 50%,60% and 70%), the monthly optimally generated energy is calculated. The amount of optimal annual energy generated is obtained by determining the sum of the monthly energies. It should be noted that in these calculations, different sizes of headponds are also involved. The locations of projects and river water fl ow are constrained by the availability of established head-ponds whose size is an important determining factor.In addition, with respect to the by-laws of the Ministry of Energy regarding energy purchases, energy generation must be divided into three different types:peak load (4 h a day), normal load (12 h a day) and low load (8 h a day). The high value of energy is categorized based on the peak, normal and low loads, respectively,so the planner can choose different alternatives with the highest energy generation relative to the load. While coordinating between energy in peak and base states, the technical indices such as the plant factor of an SHPP should be within a reasonable and acceptable limit. With respect to the role of an SHPP in the load power system network, the recommended index size with a headpond should be a fi gure between 30% and 45%, and without a headpond, a fi gure between 45% and 60% (Energy Ministry of China, 1990). 2.2. Economic calculation method In this section, the method of evaluation of income and costs and ultimately, the economic analysis of SHPPs are described (Hosseini and Forouzbakhsh,2003). The costs of the project are divided into two categories: investment and annual costs. Investment costs include civil costs, electro-mechanical equipment,power transmission line, and other indirect costs.Annual costs include the depreciation of equipment,operating and maintenance, and replacement costs. The income of the project is based solely on the sale of electrical energy. 2.2.1. Investment costs Civil costs consist of the construction and hydro-structural costs of the project, including a dam,conveyance of water system, the water penstock structure, a headpond, the forebay, the power house,the tailrace structure, the access road and any future unpredicted costs taken from the preliminary designs of a feasibility study. Electro mechanical equipment costs include turbines,generators, governors, gates, control systems, a power substation, electrical and mechanical auxiliary equipment, etc. With respect to the nature of SHPPs (lower than 5 MW), the costs are evaluated to be approximately US$500/kW of power installation. Note that the control system is assumed to be manual, and the costs of using remote control would naturally be higher (Hosseini and Forouzbakhsh, 2003; Aab-niroo Co, Studies Management Offi ce, 2003).Power transmission line costs include a power transmission line for delivering generated energy from power plant to power transmission network. The transmission line cost depends on the location, type of existing system (overhead line or cable system), and capacity of SHPP as well as length of transmission lines, which have a very high affect on project costs. Indirect costs include Engineering and Design (E Department of the Army, 1985). S Department of the Army, 1985).Infl ation costs during construction: To precisely calculate the investment cost of a project, it is necessary to take into consideration the infl ation rate during the course of the project and adjust the investment cost with respect to the infl ation rate. The infl ation rate of future years should be determined by obtaining the average of previous years infl ation rate. There are two benefi ts for the SHPPs: (1) tangible benefi ts and (2) intangible benefi ts. The tangible benefi t is the sale of electrical energy. Based on approval by Iranian regulators, the purchase of electrical energy from SHPPs has been gua2.2.2. Annual costs To obtain the net benefi t of a project, annual costs, in addition to investment costs should be calculated.Annual costs include depreciation of equipment, Operating and Maintenance (O Department of the Army,1985; International Atomic Energy Agency, 1984). Replacement and renovation costs: The main parts of the SHPP, such as generator windings, turbine runners and other parts will eventually need replacement and renovation. With respect to the nature of these SHPPs, the costs of renovation and reconstruction of equipment at year 25 is taken to be approximately equal to the total value of equipment at time of purchase. To estimate the costs for largeand medium sized power plants, the percentage of wear should be determined for different sections separately so that the calculation of these costs can be done in a more precise way (Hosseini and Forouzbakhsh, 2003; Aab-niroo Co, Studies Management Offi ce, 2003). 2.2.3. Income (2) the governmental sector with transmission lines; (3) the private sector without transmission lines; and (4) the private sector with transmission lines. In each, the electrical energy purchasing rates are being provided in different months of the year based on the peak load (4 h a day), normal load (12 h a day) and low load (8 h a day). Meanwhile,for the private sector, different purchase rates are being presented with four options, namely, 100%, 75%, 50%and 25% of private investment. Due to peak hours of energy consumption, the purchasing rate would be more attractive for the producer of energy. The annual infl ation-purchasing rate is being considered to be 5% in the calculation. The intangible benefi ts cover the positive environmental effects, fl ood control, agriculture and irrigation, fi sh farm pools, camps and recreation centers, etc. which eventually turn into quantitative values. The intangible benefi ts are not included in this economic analysis of the project, but naturally a more desirable result will be obtained for the economic indices when taking these factors into account (Aab-niroo Co,Studies Management Offi ce, 2003). 2.2.4. Financial and time specifi cations and methods of capital distribution Capital depreciation period for construction costs: 50 years. Replacement and renovation of electro-mechanical equipment: 25 years. Duration of construction: 3 years. Annual interest rate: 10%. Annual infl ation rate: 5%. Table 1 shows the capital distribution during the investment period. This table presents construction time from 1 to 6 years (Hosseini and Bathaei, 2001; Aabniroo Co, Studies Management Offi ce, 2003). In this table, the construction costs are expensed in the relevant subsequent years. Thus, with the effects of interest and infl ation, the costs of the subsequent years can be predicted. Social and economic factors could also be included in this calculation. When execution activities begin, the annual payments should be expensed in the midyear, in order to lessen the effect of infl ation, thus lowering the investment value. For example, according to Table 1, for a 3-year construction project, the percentage of the cost in each year are as following: 37% of capital in the middle of the fi rst year, 56% in the middle of the second year and 7% in the middle of the third year. 2.3. Reliability calculations The reliability index of Loss of Load Expectation (LOLE) is calculated by using the Monte Carlo method (Billinton and Allan, 1987, 1996). The Monte Carloalgorithm is one of the strongest engineering tools that enables us to perform a statistical analysis of the uncertainties involved in engineering problems. This method is very applicable in solving complicated problems where many random variables are involved in non-linear equations. The Monte Carlo analysis can be imagined as a simulation method, which replaces a practical execution with a computer simulation. The basis of the Monte Carlo analysis is to produce a series of random numbers. The produced homogenous random numbers retain the same characteristics of the probability of their occurrences in the selected domain between 0 and 1. In this method, fi rst, n random numbers are produced for each one of the existing random parameters in the given equation and this equation is then solved for each single random selected Fig. 2. The LOLE calculation algorithm, using the Monte Carlo method number. Finally, n values are obtained for the concerned equation by using the related relations to obtain the statistical information of the histogram sample. It should be noted that as the number of iterations increase, the answer would more closely approach the real value. With a decreasing probability of not supplying electricity to subscribers (customers)there is a direct relation to an increase in the number of generation units (generators) and, as a result, an increase in investment design status and/or utilization.More investment will defi nitely lead to an increase in utilization costs, which should be refl ected in the tariff of energy sale. Subsequently, economic limitations would lead to a decrease in the reliability of the system.Therefore, there could be a compromise between reliability and economic restrictions that could lead to diffi cult management decisions in both the design and operational stages (Billinton and Allan, 1996; IEEE,1985). In general, a study of the reliability of the three sections of generation, transmission and distribution of power systems should be done. In this paper, only generation reliability is being considered to meet the load demands. The transmission and distribution reliability have been assumed to be perfect (reliability=100%). The LOLE calculation algorithm, using the Monte Carlo method, is shown in Fig. 2 (Billinton andAllan, 1987). Fig. 2. The LOLE calculation algorithm, using the Monte Carlometho 3. Case study The case study of the SHPP Nari (small hydropower plant) is presented. This SHPP is located in the West Azarbaijan Province of Iran. The SHPP is of run-off river type and the object is to determine the optimal installation capacity. 3.1. Energy calculation of the Nari SHPP In Table 2, the Nari river fl ow duration curve for different months is shown based on the routine daily statistics of the river. After doing feasibility studies in different specialized work groups and specifying the determination of the plant layout in the preliminary phase, a channel with a 3.6-km length and a net head of 300 m is being obtained. Furthermore, there is a suitable position for construction of the regulating daily headpond before the penstock entrance at the end of the channel (Ministry of Energy of Iran, Aab-niroo Company, 2002).There are six alternatives of headpond volumes of 0,5000, 10,000, 15,000, 20,000 and 25,000 m3 with six different fl ow rate probabilities of 20%, 30%, 40%,50%, 60% and 70% on the fl ow duration curve, making 36 alternatives. After surveying the fl ow duration curve and different sizes of headponds, 14 alternatives out of 36 are chosen as the best. Alternatives 1 and 2 have a headpond volume of 5000 m3, a designed fl ow rate of 0.7 m3/s and an installation capacity of 1.75 MW with a fl ow rate probability of 40% and 60%. Alternatives 38 have a headpond volume of 10,000 m3, a designed fl ow rate of 1 m3/s and an installatio