关 键 词：

发电厂变电所电气设备及运行课件研制 杨小梅071901010730 发电厂 变电所 电气设备 运行 课件 研制 杨小梅 071901010730

资源描述：

发电厂变电所电气设备及运行课件研制 杨小梅071901010730,发电厂变电所电气设备及运行课件研制,杨小梅071901010730,发电厂,变电所,电气设备,运行,课件,研制,杨小梅,071901010730

内容简介：

华 北 电 力 大 学 科 技 学 院毕 业 设 计（论 文）附 件外 文 文 献 翻 译学 号： 071901010730 姓 名： 杨小梅 所在系别： 电力工程系 专业班级： 电气07K6 指导教师： 杨国旺 原文标题： Composite System Reliability Evaluation Incorporating Protection System Failures 2011 年 5月 8日结合保护系统失效的复合系统可靠性评估 原文出处及作者：Chehreghani Bozchalui , Sanaye-Pasand , Fotuhi-Firuzabad .Electrical and Computer Engineering, 2005, Canadian Conference on.摘要在导致电力系统停电过程中，保护系统失效起到重要作用。本文阐述了系统的可靠性评估合成保护系统失效。可靠性模型用于确定在电力系统中保护系统失效的影响。保护的过程和计划及潜在失效的分析基于故障发生后其对联锁停运的贡献。许多可靠性指标如LOLP、EENS和ECI用于描述电力系统可靠性中保护系统失效的影响。关键词：电力系统，可靠性，保护系统，潜在失效，联锁停运1 前言近年来的研究表明电力保护系统在大干扰的产生和发展中承担了重要角色1。在电力系统历史，保护系统信赖性（当被要求时动作的能力）已经优于系统安全性（当不要求时不动作的能力）。另一方面，解除电力系统管制，通过网络可靠地输送电力的能力十分重要。直到保护系统更趋向于信赖性而不是安全性以前，其误动作的概率增加。根据2安全性和信赖性在保护系统结合。现在多地区保护系统偏向于信赖性且被设计为信赖，即使在全球系统安全性成本中3。因此，大部分的继电器误操作是不希望的动作且表明了主要的干扰传播。这些“误操作”是常见的“潜在失效”，即在正常时保持静止且揭示其他系统干扰4。在文献5，潜在失效有明文规定。由北美电气可靠性理事会制定的一个研究显示继电保护涉及大约75%的主要干扰。在大多数可靠性研究中，保护系统通常假设完全可靠。因此，必须发展可靠性研究。本文描述了电力系统可靠性评估结合的保护系统失效。可靠性模型用于确定保护系统故障对电力系统可靠性的影响。一些其他的研究已经使用蒙特卡洛法9。在本文中，我们使用启发式分析方法的状态抽样法。2 保护失效模型 保护系统有两个基本失效模式：“拒动”和“偶然或非正常跳闸”8。电力系统网络是不间断运行状态，因此任何失效即刻表现出来。然而，一个保护系统仍然在静止状态，直至其被要求动作。在这个系统发生故障时，直到要求其动作时静止状态不发生改变，当然，它不会响应。这些失效被定义为失效点。所以，断路器拒动现象是“拒动”模式。因为故障线路将被后备保护切除，所以这种类型的失效将直接导致至少一个母线与系统隔离。由于系统中伪信号导致非正常跳闸，从而导致断路器无意操作发生时时即刻显示。不过，由于保护系统各种潜在失效，非正常跳闸使问题复杂化4。有两种“非正常跳闸”，一种是发生在无任何异常状态时的不必要跳闸，它可通过自动重合闸恢复，另一种是区外保护误动作9。区外保护误动作是联锁停运的主要原因。3 分析模型和假设3.1 元件/保护模型有许多促进可靠性评估的模型，包括保护系统失效9、10、11、12。载流元件模型与和它联系的保护系统配合，这个参考文献9，区分两种保护失效模式代表两种状态：“误动”和“拒动”。在本文中，我们使用的模型，如图1，在以下状态使用：状态1：载流元件和保护系统均良好；状态2：载流元件良好，但保护系统有误动的风险； 元件良好 元件故障图1 马尔可夫模型的组成及其保护系统状态3：载流元件良好，但保护系统会拒动；状态4：载流元件良好，但保护系统正在检修；状态5：载流元件失效且保护系统会误动；状态6：载流元件失效但保护系统良好；状态7：载流元件失效且保护系统会拒动。图1的符号如下所示：i 护系统的检测率I 保护系统的维修率 元件的维修率 元件的故障率P1 保护系统误动的故障率P2 保护系统拒动的故障率当一个元件处于“误动”和“拒动”时，其保护系统潜在失效或导致故障。3.2 保护系统失效性质“拒动”现象并不十分复杂，因为它只是相关的保护装置而不是系统操作条件。图1中状态3的概率是“拒动”的概率。由于“误动”算入保护潜在失效的存在，系统故障和操作条件与联锁停运有关。距离保护3段和过电流保护对故障参数有相近的保护定值。因此，它们对故障和异常操作条件更灵敏。文献3提出了作为距离保护的全阻抗特性的线路跳闸的潜在失效的概率模型。图2显示了该模型的一些修改。它适合故障时的情况。我们距离保护3段的整定阻抗设为线路阻抗的250%。在起始的故障排除后，系统的潮流将随着系统的拓扑结构的改变而变化。这可能会导致某一线路上的过负荷，过负荷有动作的风险。对现在确定故障点的情况，我们提出了一个线路误动作概率模型作为过电流保护的线路潮流函数。这个模型如图3所示。在图2和图3中，PZ和PI是元件或保护模型（图1）中状态2的概率。PZ和PI是保护系统失效性质，它们被分别用于故障期间和故障后周期中。它表明裸导线跳闸概率错误的依赖故障和系统运行条件。每一条线路有不同的误动概率。图2 裸导线的距离保护失效 PZ：距离保护失效概率 Z：被保护阻抗 Z3：距离保护3段整定阻抗图3 裸导线过电流保护失效概率 PI：过电流保护失效概率 I：被保护电流 IS：过电流保护整定值3.2 联锁停运电力系统中由于一个元件的失效一次联锁停运引起一系列跳闸动作。当故障发生时，电力系统中如过电流或低电压可能会导致一些保护设备的误动作。文献6的假设是一条母线上的任何一条输电线路L动作，那么线路L的潜在失效将表现出。那就是，如果一条线路正确动作，则所有与它的末端母线相连的线路在误动作时被显示出来。这些事故的可能性很小，但不应被忽略而应考虑。电力系统中干扰的传播是一维的13，因此在有时很少发生时，这种情况比一条线路动作更多。在仿真时，我们一次只允许仅且一条线路，特殊情况下，如果多于一条线路可能动作，那么高动作率的一条线路将被选入下一条故障线路。在仿真时，我们假定：1） 保护系统的“拒动”和“误动”不重叠；2） 只有一级故障认为是最初的故障；3） 所有的失效相互独立；4） 当载流元件失效时，保护系统不失效；5） 保护系统的检查不会引起元件失效。4 可靠性和脆弱性指数以下指数在文献8中提及4.1 电力不足时间概率(LOLP) LOLP= 上式中 Ci ：仿真i中的可用容量 Li ：仿真i中的负荷容量 Pi (CiLi) ：仿真i中的电力不足概率 一个电力系统能够承受停电不足和安全危害。LOLP反映了仿真中保护系统失效的电力不足结果，电力不足一重现一系列的中断就停止。4.2 电量不足期望值 (EENS) 上式中 Pi 仿真i中联锁停运概率 Lki 仿真i中母线k的切负荷量 以“MWh”为单位的指标反映了系统可靠性中潜在失效的影响。如果这个指数正常，则它被用于和各种电力系统比较。正常的EENS被定义为电力消减指标（ECI）。 (MWh/MW-yr)上式中 LS是系统总负荷大电力系统电力消减指标已经被定义为“严重指标”。以兆瓦-分钟计的缺供总电量除以以兆瓦计的系统负荷，则严重指标以系统分钟计。5 可靠性指标的计算可靠性指标的计算包括两部分：第一种是发电保护系统潜在失效率，第二种是计算可靠性指标。首先，由图1中的马尔科夫链，已知状态转换率，我们计算出每一个保护系统的潜在失效率。为了这个目的，我们通过马尔科夫链形成转置矩阵，然后使用频率均衡概念，我们解决发电机方程式。状态2“误动”（PZ和PI）的概率由图2和图3决定。它们被用于计算可靠性指标。可靠性指标的计算，依照以下步骤：1）选择故障线路；2）确定显露误操作的所有线路；3）通过故障计算，计算出裸导线继电器的阻抗；4）由图2找出每一个裸导线的跳闸率；5）确定将要分闸的裸导线，如果没有线路分闸，到第9步；6）更新基于新的分闸线路的裸导线；7）若需要从新组合潮流计算的系统拓扑结构，计算裸导线的电流；8）由图3找出每一条裸导线的分闸率，转至第5步；9）记录联锁停运并确定切负荷量，计算可靠性指标。6 算例分析6.1 测试系统Roy Billinton测试系统（RBTS）作为测试系统（如图3）。RBTS的基本数据源自文献14。此外，元件和保护系统的转换率列于表1中。图4 Roy Billinton测试系统表1 元件和相关保护系统RBTS的数据6.2 结果 系统广义可靠性指标总结于表2。这些指标在系统联锁时反映了不安全水平和切负荷量。表2 系统可靠性指标我们也计算了特殊故障线路的独立可靠性指标。这些结果列于表3。这些指标表明系统的哪一部分薄弱。表3 独立可靠性指标7 结论和未来工作 在本文中，基于元件与其保护系统相配合的马尔科夫模型，分析了联锁停运的电力系统行为。计算可靠性指标如LOLE、EENS和ECI，反映复合系统和联锁停运可靠性中保护系统失效的影响。而且计算每一条线路独立的可靠性指标，这项指标可用于评估系统中的薄弱线路。在本文中，不包含联锁停运的潜在失效的电压影响。而且电力系统中的随机事件如每次切负荷的故障定位、故障切除概率和持续电力不足均已考虑。在未来的工作中，综合以上因素将获得更多的电力系统指标。8COMPOSITE SYSTEM RELIABILITY EVALUATION INCORPORATING PROTECTION SYSTEM FAILURESM. Chehreghani BozchaluiDepartment of ECE, Faculty of Engineering, University of Tehran, Iran m.chehreghaniece.ut.ac.irM. Sanaye-PasandDepartment of ECE, Faculty of Engineering, University of Tehran, Iran msanayeut.ac.irM. Fotuhi-FiruzabadDepartment of Electrical Engineering, Sharif University of Technology, Tehran, Iran Abstract Protection system malfunction play a significant role in the sequence of events leading to power system blackouts. This paper describes power system reliability evaluation incorporating protection system failures. A reliability model is used in order to determine the impact of protection system failure on power system reliability. The mechanism and scheme of protection and their hidden failure are analyzed based on their contribution to the cascading outage after occurrence of a fault . A number of reliability indices such as LOLP, EENS and ECI are calculated to describe the impact of protection system failures on reliability of power system.Keywords: Power System, Reliability, Protection Systems, Hidden Failure, Cascading Outage.1 IntroductionRecent studies show that power protection systems have played significant role in the birth and propagation of major disturbances 1. Historically in power systems, protection system dependability (the ability to trip when required) has taken priority over system security (the ability to refrain from tripping when not called for). On the other hand, in the deregulated power systems, the ability to transfer power reliably through a network becomes very important. Until protection systems prefer dependability over security, the probability of their incorrect operation increase. According to 2 security and dependability are intertwined in a protection system. The existing protection system with its multiple zone of protection is biased toward dependability and is designed to be dependable even at the cost of global system security 3. Hence, vast majorities of relay miss-operations are unwanted trips and have been shown to propagate major disturbance. These “miss-operations” are noted as the “hidden failures”, which remain dormant when every thing is normal and are exposed as a result of other system disturbance 4. In 5, the list of hidden failures is well documented. A study by the North American Electric Reliability Council (NERC) shows that protective relays are involved in about 75 percent of major disturbance 6. Most blackouts are somehow related to the protection system hidden failures. Large-scale power system blackout is a rare event. However, when it occurs, the impact on the system is catastrophic 7. In spite of its importance, the impact of protection system malfunction on overall system reliability has not been well studied. In most reliability studies, protection systems are generally assumed fully reliable. It is therefore necessary to develop reliability study concerning the protection system. This paper describes power system reliability evaluation incorporating protection system failures. A reliability model is developed in order to determine the impact of protection system failure on power system reliability. Some other researches have used Monte Carlo approach 9. In this paper, we use state selection method with a heuristic analytical approach. 2 Protection Failure Modes Protection systems have two basic failure modes: “failure to operate” and “inadvertent or undesired tripping” 8. A power system network is in a continually operating state and hence any failure manifests itself immediately. A protection system, however, remains in a dormant state until it is called on to operate. Failures which occur in this system during the dormant state do not manifest themselves until the operating request is made when, of course, it will fail to response. These failures have been defined as unrevealed faults. So phenomenon of stuck breaker is included in “failure to operate” mode. This type of failure will directly cause at least one bus isolation in the system, because the faulted line will be isolated by backup protection. Undesired tripping, which is due to a spurious signal being developed in the system, thus causing breakers to operate inadvertently , manifest itself immediately when it occurs. Undesired tripping, however, makes the problem complicated due to various protection system hidden failures 4. There are two types of “undesired tripping”, one is unwanted tripping that occurs in the absence of any abnormal state, that can be remedied immediately by auto-reclosure, and another is tripping for faults outside the protection zone 9. Tripping for faults outside the protection zone is the main cause of cascading outages. 3 Analysis Models and Assumptions 3.1Component / Protection Model There have been a number of models developed to facilitate reliability evaluation including protection system failure 9, 10, 11, 12. Model of current-carrying component paired with its associated protection system, which is proposed in 0-7803-8886-0/05/$20.00 2005 IEEECCECE/CCGEI, Saskatoon, May 2005486Authorized licensed use limited to: NORTH CHINA ELECTRIC POWER UNIVERSITY. Downloaded on May 16,2010 at 08:16:44 UTC from IEEE Xplore. Restrictions apply. Incorrect setting & miscalibration Unwanted tripping in absence of any abnormal state Z33*Z3ZPZUnrevealed failure in correctly set Fig.1. Markov model of component and its protection system reference 9 differentiates the two protection failure modes and represents them as two states: “undesired trip” and “failure to trip”. In this paper, we use this model as shown in Fig. 1, where the following states are used: State 1: The current-carrying component and protection system are both good. State 2: The component is good but the protection is at risk for “undesired trip”. State 3: The component is good but the protection is exposed to “failure to trip”. State 4: The component is good and the protection system is being inspected. State 5: The component is failed while the protection system is still under “undesired trip”. State 6: The component is failed but the protection system is good. State 7: The component is failed while the protection system has experienced “failure to trip”. The notations in Fig. 1 are as below: i inspection rate of protective system. I repair rate of protection system. repair rate of component. failure rate of component. 1P failure rate of protection system to exposure to “undesired trip”. 2P failure rate of protection system to state of “failure to operate”. When a component is in “undesired trip” and “failure to trip”, its protection system is suffering from hidden failure and cause malfunction. 3.2Protection System Failure Properties The “failure to operate” phenomenon is not very complicated because it is only related to the protection device itself rather than system operating condition. The probability of state 3 in Fig. 1 is the probability of “failure to operate”. In the case of “undesired trip” in addition to existence of protection hidden failures, system fault and operating conditions are also related to cascading outage. Distance zone 3 and over current protection have the nearest setting value to a faulty parameter; therefore, they are more sensitive to fault and abnormal operating conditions. Reference 3 proposed a model of hidden failure probability Fig. 2. Distance protection failure probability of exposed line PZ : distance protection failure probability, Z: impedance seen by relay, Z3: zone 3 impedance setting of tripping the exposed line as a function of impedance seen by distance relay. This model with some modifications is shown in Fig. 2. It suits for the situation during fault. We choose zone 3 impedance setting as 250% of line impedance. After the initial fault cleared, power flow in the system would change due to the changing system topology. This might lead to over loading problem on certain lines, which are at risk to trip. To present the post fault situations, we propose a probability model of an exposed line tripping incorrectly as function of line flow for over current relay. This model is shown in Fig. 3. In Figs. 2 and 3, PZ and PI are the probability of state 2 in the component/protection model (Fig. 1). PZ and PI are protection system failure properties which are used “during fault” and “post fault” periods, respectively. It shows that the probability of exposed line tripping incorrectly depends on the fault and system operating condition. Each line has a different probability for incorrect trip. 3.3Cascading Outage A cascading outage refers to a series of tripping initiated by one component failure in the power system. When a fault occurs, the impact to the power system such as over-current or voltage dip may cause some protection devices to miss-operate. The assumption by 6 is that if any line sharing a bus with a transmission line L trips, then hidden failure in line L are exposed. That is, if one line trips correctly, then all line connected to its end bus are exposed to the incorrect tripping. The probability of such occurrence is small but not negligible and is considered. Spread of disturbance is one-dimensional in power system 13, hence the case that more than one line trip at sometime rarely happens. In the simulation, we let one and only one line trip at one time, in specific, if more than one line might trip, the one with higher tripping probability is selected to be next tripping line. In simulation, we also assume that: 1)“failure to trip” and “undesired trip” of the protection system failure doesnt overlap. 2)Only first-order contingency is considered to initial fault. 3)All failures are mutually independent. 487Authorized licensed use limited to: NORTH CHINA ELECTRIC POWER UNIVERSITY. Downloaded on May 16,2010 at 08:16:44 UTC from IEEE Xplore. Restrictions apply. Fig. 3. Over current protection failure probability of exposed line. PI: over current protection failure probability I: current seen by relay IS: over current relay setting 4)When the current-carrying component is in failure state, the protection system does not fail. 5)Inspection of protection system dose not leads to component failure. 4 Reliability and Vulnerability Indices The following indices are defined and used 8. 4.1Loss of Load Probability (LOLP) ()?=iiiiLCPLOLPwhere iC: Available capacity in simulationi.iL: Load in simulation i ()iiiLCP: Probability of loss of load on simulationi.A power system can withstand one outage without adequacy and security violation. LOLP presents the loss of load resulting from protection system failure in simulation; the series of outage is stopped as soon as loss of load occurs. 4.2Expected Energy Not Supplied (EENS)?=ikiikLPEENS8760where iP: is the probability of existence of cascading outage in simulation i,kiL: is load curtailment at bus k in simulationiThis index with units of “MWh” shows the impact of hidden failure on system reliability. If this index is normalized, it can be used to compare various power systems. Normalized EENS is defined as Energy Curtailment Index (ECI): ?=ikiikLSLPECI8760 (MWh/MW-yr) whereLSis the system total load. This bulk power energy curtailment index has also been designed as the “Severity Index”. The total energy not supplied in MW-Minute is derived is divided by system load in MW, severity is therefore expressed in system minute. 5 Calculation of Reliability Indices The calculation of reliability indices contains two parts. The first one is generating protection system hidden failure probability and second is calculating reliability indices. For the first one, from Markov chain in Fig. 1, with known state transition rates, we figure out the hidden failure probability of each protection system. For this purpose, we form the transition matrix according to Markov chain, then using frequency balance concept we solve the generated equation. Probability of state 2 “undesired tripping” (PZ and PI) is determined using Figs. 2 and 3. They will be used in calculation of reliability indices. In calculation of reliability indices, following steps are performed: 1)Select faulted line 2)Determine all lines that are exposed to miss-operation 3)From fault calculations, compute the impedance seen by relay for the exposed lines. 4)Find the probability of tripping for each exposed line using Fig. 2. 5)Determine which exposed line will trip. If no line trips, go to 9. 6)Update the exposed lines based on the newly tripped line. 7)Re-array the system topology if necessary and from power flow calculation, calculate the current on the exposed lines. 8)Find the tripping probability for each exposed lines using Fig. 3, go to 5. 9)Record the cascading outage and determine the amount of load curtailment. Calculate the reliability indices. 6 Case Study 6.1Test System The Roy Billinton Test System (RBTS) is used as the test system (Shown in Fig. 3). The basic data for RBTS can be found in 14. In addition, the transition rates of component and protection systems are listed in Table I. Fig. 4. Roy Billinton Test System Unwanted tripping in absence of any abnormal state IS2ISIPIIncorrect setting & miscalibrationUnrevealed failure in correctly set 488Authorized licensed use limited to: NORTH CHINA ELECTRIC POWER UNIVERSITY. Downloaded on May 16,2010 at 08:16:44 UTC from IEEE Xplore. Restrictions apply. Table I Component and Associated Protection System Data for RBTS Line Component Protection System ?(1/yr) (1/yr) 1?P(1/yr) 2?P(1/yr) i(1/yr) I(1/yr) 1,6 1.5 876 .046 .083 4 219 2,7 5 876 .0014 .2 4 219 3 4 876 .0118 .422 4 219 4 1 876 .0165 .166 4 219 5 1 876 .0165 .166 4 219 8 1 876 .0165 .166 4 219 9 1 876 .0165 .166 4 219 6.2ResultsSystem wide reliability indices are summarized in Table II. These indices represent the degree of vulnerability and load curtailment when system suffers from cascading. Table II System reliability indices LOLP EENS (MWh) ECI (MWh/MW-yr) Severity Index (minute) 0.00393 684.323 3.754 0.000428 We also calculate the individual reliability indices with the specified faulted lines. The results are shown in Table III. These indices show that which part of system is the weakest. Table III Individual reliability indices Line LOLP EENS ECI 1 1.67E-7 0.1332 7.2E-4 2 2.05E-10 0.0000144 7.78E-8 3 2.36E-10 0.0001756 9.5E-7 4 1.52E-5 5.3296 2.38E-2 5 1.52E-5 5.3296 2.38E-2 6 1.67E-7 0.1332 7.2E-4 7 2.05E-10 0.0000144 7.78E-8 8 4.29E-7 0.4697 2.53E-3 9 3.9E-3 683.28 3.6934 7 Conclusion and Future Work In this paper, based on a markov model of component paired with its protection system, behavior of power system under cascading outage is analyzed. Reliability indices such as LOLE and EENS and ECI are calculated to describe the impact of protection system failure on reliability of composite systems and cascading outages. Also calculated the reliability indices for each line individually, with this indices we could estimate vulnerable links in the system. In this paper we did not include the influence of voltage based hidden failures in cascading outage. Also the stochastic feature of power system events such as fault location, fault clearing probability and loss of load duration at each curtailment can be considered. In the future work by incorporating above factors more accurate indices for the power system can be obtained. 8 References 1 NERC Disturbance Report; North American Electric Reliability Council; New Jersey; 1984-1988. 2 A. G. Phadke, S. H. Hortowitz, J. S. Throp, “Aspects of power system protection in post-restructuring era”, Proc. 32nd Hawaii International Conference on System Science,Vol. 3, Jan 1999. 3 J. S. Thorp, K. Bae, “An Importance Sampling Application: