Teaching reliability analysis of HV/MV substations and distribution feeders using educational software

D International Journal of Electrical Engineering Education, Volume 49, Number 4 (October 2012), © Manchester University Press http://dx.doi.org/10.7227/IJEEE.49.4.2 Teaching reliability analysis of HV/MV substations and distribution feeders using educational software Pedro J. Martínez-Lacañina, José L. Martínez-Ramos, Alfonso Bachiller-Soler and Darío Monroy-Berjillos Electrical Engineering Department, Escuela Técnica Superior de Ingeniería, University of Seville, Seville, Spain E-mail: [email protected]; [email protected]; [email protected]; [email protected] Abstract In this paper, educational software for the reliability assessment of high/medium-voltage (HV/MV) substations and distribution feeders is presented. It is a useful tool for learning the utility of reliability indices of the HV/MV substations and their distribution feeders, and their influence on the electrical power system operation. By means of an interactive graphical interface, multiple configurations of HV/MV substation layouts can be selected. The Monte Carlo method and the equipment outages have been used to calculate the expected reliability performance of substations and primary distribution feeders. The proposed educational tool has been evaluated to measure students’ satisfaction, and questionnaire and results of this evaluation are presented and discussed. Keywords distribution feeders; engineering education; HV/MV substations; reliability assessment Nowadays, in restructured power systems the deployment of distributed generation and smart grids implies that distribution systems have the overall responsibility to manage power exchange and load curtailment bids to fulfil the reliability commitments and the reliability requirements of customers.1,2 In this new scenario, HV/MV substations play an important role due to meshed networks ends and radial supply system begins (i.e., feeders) in these substations. Future distribution networks require novel concepts and systems for their planning, design, monitoring and control architectures to ensure the security and resilience of the infrastructures supporting this deployment and to evaluate their impact on the reliability of the distribution systems. From the educational point of view, it is necessary to include this new paradigm and concepts in electrical engineering studies. In relation to this approach, educational software is presented in this paper for improving not only the teaching of reliability assessment of HV/MV substations and their distribution feeders, but their effects on the electrical power system and the quality of service. To develop general-purpose software to assess the reliability of substations and their distribution feeders, the educational tool uses various substation components such as breakers, busbars and transformers. By means of a graphical user interface, it is possible to perform failure modes and to provide a concise and orderly description of the various combinations of occurrences.3,4 Once the reliability assessment is performed, the results are checked out to determine the remedial actions that can be carried out to bring the system back to a normal state, and to reach acceptable reliability indices. Besides, it is possible to perform remedial actions to prevent 2_ijeee_4515.indd 365 9/27/2012 8:01:04 PM D 366 P. J. Martínez-Lacañina et al. International Journal of Electrical Engineering Education, Volume 49, Number 4 (October 2012), © Manchester University Press contingencies and to obtain acceptable reliability indices to avoid possible penalties due to the new restructured power systems. A special emphasis has been placed by the authors on attaining the new educational paradigm demanded by the Bologna process.5 In the subject of engineering, not only maintains the technical knowledge base of the field (i.e. mathematical and scientific), but also stresses the implementation of educational tools as a reinforcement of learning done by hand. In contrast to the real world, students using this kind of tools are able to ‘step outside’ of the process to review and understand it better.6–8 Based on the idea presented above, the aim of developing a new educational tool to help the students in the learning of the reliability assessment theory was focussed on a piece of educational software (training simulator). This tool is being applied, with successful results, in the training of undergraduates in the Electrical Engineering Department, and it leads to more interest in the exercises applied in the subject of reliability analysis of HV/MV substations and their distribution feeders. The rest of the paper is structured as follows. In the next section a brief description of reliability analysis is introduced. A section on the educational tool follows, and in a further section the classroom practice is described and the development of the educational experience. Finally conclusions are drawn, and an appendix has been used to include a basic theory of probability, the Monte Carlo method, and the results given by the educational tool. Background theory The application of probabilistic techniques in the analysis of unscheduled and scheduled events provides a quantitative prediction and information of the system performance, and, more important, a way of consistently evaluating both the reliability level of alternative operational arrangements and network reinforcements.9,3 In order to understand the effects of substation component failures, it is necessary to study station component outage processes. For this purpose, the continuous Markov process is the most popular probabilistic method to study a component in continuous discrete states.3,4,9,10 This is a specific stochastic process independent of all the past states except the immediately preceding state. An example of a two-state system, with failure and repair rates λ and μ, respectively, is shown in Fig. 1. Usually, the inverse of λ is known as the Mean Time To Fail (MTTF), and the inverse of μ is known as the Mean Time To Repair (MTTR). In a Markov process, the probability of failure or repair for a fixed interval of time is considered constant. The steady-state probabilities of residing in the Fig. 1 Two-state space diagram of a component. 2_ijeee_4515.indd 366 9/27/2012 8:01:04 PM D Teaching reliability analysis 367 International Journal of Electrical Engineering Education, Volume 49, Number 4 (October 2012), © Manchester University Press operating state, ‘State 0’, and in the failed state, ‘State 1’, are designated as p0 (availability) and p1 (unavailability) respectively. The p1 probability is also called Forced Outage Rate (FOR). In Appendix A a brief explanation to obtain p0 and p1 has been included. It is clear that substation failure assessment is highly dependent on the component outages. Therefore the collection of station component outage data (failure rate and repair rate) and their assembly on a real database is an important and necessary activity for the evaluation of reliability.11,12 In the presented tool, the corresponding data, i.e. failure and repair rates, have been obtained from Refs 13 and 14, including all the necessary information for the calculation of the independent outages of generating units, transmission lines, transformers, busbar and feeders. Typical reliability indices The Monte Carlo technique (see Appendix B) has been implemented in the educational tool, and the calculated probability, frequency and duration indices associated with the states of the substation are the traditional indices LOLP [pu] (Loss of Load Probability), LOLF [occ./y] (Loss of Load Frequency), and EPNS [MW] (Expected Power Not-Supplied). The reliability indices obtained with the proposed tool are related to particular supply and load scenarios of the distribution system in which the analysed substation is located. If a system state S is considered, then the functions representing the LOLP, LOLF and EPNS reliability indices are: F S S S LOLP if is an availability state 1 if is an unavailabil () =0 iity state { F S S S LOLP if is an availability state if is an unavailabil () =0 µ iity state { F S S S LOLP LS if is an availability state P if is an unavailab () =0 iility state { where PLS is the load shedding [MW] in the S unavailability system state. By using the preceding functions, the reliability indices are estimated as follows: LOLP LOLF EPNS LOLP LOLF EPNS = = = ∈ ∈ ∑ ∑ 1 1 1 NF S NF S NF S SS G SS G S ( ); ( ); ( ) SS G∈ ∑ (1) where G is the set of system states and NS is the number of sampled system states. Other reliability indices as LOLE [h/year] (Loss Of Load Expectation), LOLD [h/occ.] (Loss of Load Duration), and EENS [MWh/year] (Expected Energy Not Supplied) can be obtained as: LOLE LOLP T LOLD LOLP LOLF EENS EPNS T= ⋅ = = ⋅; ; (2) 2_ijeee_4515.indd 367 9/27/2012 8:01:04 PM D 368 P. J. Martínez-Lacañina et al. International Journal of Electrical Engineering Education, Volume 49, Number 4 (October 2012), © Manchester University Press where T is the period of time over which the analysis is extended (by default this is one year). Spanish reliability indices Apart from the reliability indices shown above, the tool also gives the TIEPI and NIEPI indices, which are the reference reliability indices in Spanish electrical regulation. The former is the time of the interruption equivalent to the load shedding of all the demand in the studied area, and the latter is the number of interruptions equivalent to the load shedding of all the demand in the studied area. These Spanish reliability indices are calculated as follows: TIEPI LOLP PI NIEPI TIEPI LOLD = =; (3) where PI is the sum of the power of all the MV/LV stations in the area where the TIEPI and NIEPI indices are calculated. The Spanish electrical regulation fixes maximum values for both reliability indices (TIEPI and NIEPI), and, obviously, the corresponding penalties when utilities operate with reliability indices higher than these maximum values. In relation to penalties, the versatility of the proposed educational tool facilitates to teach students in what way it is possible to propose remedial actions to obtain acceptable reliability indices under the point of view of the electrical regulation. Remedial actions are corrective/preventive maintenance programs, and reinforcement operations on weak points of the distribution system. Description of the educational tool The educational tool presented in this paper focuses on the reliability assessment problem faced by electrical utilities when dealing with the meshed sub-transmission networks and the supply points of the distribution facilities.1,2,15,16 The corresponding flowchart used by the tool to systematically obtain the reliability assessment of HV/MV substation is shown in Fig. 2. The part of the software related to Monte Carlo Simulation has been carried out in the preceding section and Appendix B. In this section the graphical user interface will be described. For this purpose, the most significant interactive screens are shown performing the same steps that students do to obtain the reliability indices of the HV/MV substation configuration and their distribution feeders. Before embarking on the determination of reliability indices of substations, it would be helpful to be familiar with some of the common substation layouts and their corresponding names. Certain configurations may be more suited to a specific task, and, therefore, the equipments in each type of substation may vary, but with the exception of switching stations, they generally include transformers, circuit breakers, isolation switches and measurement equipments. Figure 3 shows the interactive screen which permits us to select the basic substation configuration (bus layout) included in the educational tool: Single Bus (with 2_ijeee_4515.indd 368 9/27/2012 8:01:04 PM D Teaching reliability analysis 369 International Journal of Electrical Engineering Education, Volume 49, Number 4 (October 2012), © Manchester University Press switches or breakers), Double Breaker Double Bus (with or without transfer bus), Ring Bus and Breaker-and-a-Half. For example, in the same Fig. 3 a Double Breaker- Double Bus configuration has been selected. Once a basic configuration is selected, in the next phase the student must complete the remainder of substation characteristics by selecting the position of measurement equipments and other electrical components of the substation (number of transformers and their power, circuit breakers, and isolation switches). For this process, the student has to complete various screens. As an example, the selection of the number of transformers and their power assigned in MVA is shown in Fig. 4. The selection of electrical characteristics of the HV lines (supply side of the substation) is presented in Fig. 5, and Fig. 6 shows how the number of substation MV feeders and their length and thermal capacity are selected. In the case of the HV lines and MV feeders, the position of the measurement equipments (current and voltage transformers, etc) also are assigned. All this information is used in order to Fig. 2 Flowchart of the educational tool. 2_ijeee_4515.indd 369 9/27/2012 8:01:04 PM D 370 P. J. Martínez-Lacañina et al. International Journal of Electrical Engineering Education, Volume 49, Number 4 (October 2012), © Manchester University Press Fig. 3 Selection of the basic configuration of a substation. Fig. 4 Selection of transformers and their electrical characteristics. 2_ijeee_4515.indd 370 9/27/2012 8:01:05 PM D Teaching reliability analysis 371 International Journal of Electrical Engineering Education, Volume 49, Number 4 (October 2012), © Manchester University Press obtain the failure rate, repair rate, and load shedding, this last whether it is necessary to be applied in the reliability evaluation. In the next phase, the values of the failure rates (λ) and repair times (r) of the substation equipments must be selected (Figs 7 and 8). These values are included Fig. 5 Selection of HV line parameters in the substation. Fig. 6 Selection of MV line parameters in the substation. 2_ijeee_4515.indd 371 9/27/2012 8:01:05 PM D 372 P. J. Martínez-Lacañina et al. International Journal of Electrical Engineering Education, Volume 49, Number 4 (October 2012), © Manchester University Press by the student in two steps. In the first step, Fig. 7, the student has to complete all the fields relative to the failure rates of whole substation components (HV lines, MV feeders, transformers, measurement equipments, etc), and in the second step, Fig. 8, a similar process has to be repeated for the fields relative to the repair times of whole Fig. 7 Selection of failure rates of substation components. Fig. 8 Selection of repair times of substation components. 2_ijeee_4515.indd 372 9/27/2012 8:01:05 PM D Teaching reliability analysis 373 International Journal of Electrical Engineering Education, Volume 49, Number 4 (October 2012), © Manchester University Press substation components. By default, the tool presents the standard failure and repair rates of CIGRE13 (Figs 7 and 8). It is important to note that, as Fig. 6 shows, the tool includes a module to select the priority order to load shedding for each one of the MV feeders of the substation. The load bus priority order is in accordance with agreed load-shedding protocol. The priority order of the load buses is established based on economic factors that recognise the customer costs associated with failure of supply. The most convenient index for this purpose is the interrupted energy assessment rate (IEAR),16 which measures the monetary loss for the customer in terms of a function of the energy not supplied (ENS). The priority order is related to the TIEPI and NIEPI Spanish reliability indices, and hence, to the penalties that the utilities must pay according to Spanish regulations. There are three levels of priority order to load shedding,16 and each MV feeder connected to the load bus is assigned a priority order. Level 3 implies the first load to shed, level 1 corresponds to the last load to shed, and level 2 is equivalent to indifferent load shedding. By including this information in a corresponding screen, it is possible to study how selective load shedding can modify the reliability indices of the substation MV feeders, assuming all other parameters are unchanged. In the last phase, the reliability indices are carried out, in a first stage, for the HV side of the substation (substation reliability indices), and in a second stage, for the MV-feeder supply points (load reliability indices). The results of the reliability assessment are presented in window form (Fig. 9), which shows the reliability indices of the analysed substation using overall indices and the individual reliability indices of a MV demand point connected to the substation. Fig. 9 Reliability indices of substations and MV feeders. 2_ijeee_4515.indd 373 9/27/2012 8:01:05 PM D 380 P. J. Martínez-Lacañina et al. International Journal of Electrical Engineering Education, Volume 49, Number 4 (October 2012), © Manchester University Press In the Monte Carlo method the accuracy level of the simulation process is expressed by the β variation coefficient9,4 The β coefficient and the required number of samples NS for reliability evaluation in real systems are related to each other as follows ββ = ( ) ≈⋅ V Q Q NQ S ;1 where Q ¯ is the estimate of the system unavailability (failure probability), and V(Q ¯) its variance. It can be observed that for a desired accuracy level, β, the required number of samples NS depends on the system unavailability (Q ¯) but is independent of the size of the system. Appendix C Base case results in format txt *** SUBSTATION BASIC CONFIGURATION *** ⇒ Double Breaker Double Bus *** HV Lines parameters *** ⇒ Length[Km] Line01 HV, 0,4000E+2 ⇒ Length[Km] Line02 HV, 0,3500E+1 *** Transformers HV/MV parameters *** ⇒ Power[MVA] Trafo01, 2,0000E+1 ⇒ Power[MVA] Trafo02, 2,0000E+1 *** MV Lines parameters *** ⇒ Length[Km] Line01 MV, 0,3500E+1 ⇒ Length[Km] Line02 MV, 0,2500E+1 ⇒ Length[Km] Line03 MV, 0,5750E+1 ⇒ Length[Km] Line04 MV, 0,4650E+1 ⇒ Load[MW] Line01 MV, 0,4500E+1 ⇒ Load[MW] Line02 MV, 0,4500E+1 ⇒ Load[MW] Line03 MV, 0,4500E+1 ⇒ Load[MW] Line04 MV, 0,4500E+1 ⇒ Priority order Line01 MV, 1 ⇒ Priority order Line02 MV, 1 ⇒ Priority order Line03 MV, 1 ⇒ Priority order Line04 MV, 1 *** Failure rate (failures / year)] *** ⇒ Measurement Equipment HV, 1,4000E-3 ⇒ Circuit Breaker HV, 4,3000E-3 ⇒ Lighting Breaker HV, 1,4600E-2 ⇒ Disconnect Switch HV, 2,0000E-3 ⇒ Transformer HV/MV, 1,3000E-3 ⇒ Line HV (per 100 km), 2,7000E+0 ⇒ Measurement Equipment HV, 2,0000E-3 ⇒ Circuit Breaker HV, 1,9000E-3 *** Repair time (hours/failure)] *** ⇒ Measurement Equipment HV, 5,000E+0 ⇒ Circuit Breaker HV, 1,200E+1 2 2_ijeee_4515.indd 380 9/27/2012 8:01:06 PM D Teaching reliability analysis 381 International Journal of Electrical Engineering Education, Volume 49, Number 4 (October 2012), © Manchester University Press ⇒ Lighting Breaker HV, 6,000E+0 ⇒ Disconnect Switch HV, 6,000E+0 ⇒ Transformer HV/MV, 1,200E+2 ⇒ Line HV (per 100 km), 5,500E+0 ⇒ Measurement Equipment HV, 4,000E+0 ⇒ Circuit Breaker HV, 3,000E+0 **** Number of stochastic experiments, 2466822 *** SUBSTATION. RELIABILITY INDICES *** ⇒ LOLP[pu] (sample mean), 1,721557E-4 ⇒ LOLP[pu] (confidence interval), 1,506388E-4, 1,936726E-4 ⇒ EPNS[MW] (sample mean), 5,137111E-3 ⇒ EPNS[MW] (confidence interval), 4,490042E-3, 5,784180E-3 ⇒ LOLF[h/f] (sample mean), 4,797454E-4 ⇒ LOLF[h/f] (confidence interval), 4,112965E-4, 5,481943E-4 ⇒ EENS[MWh], 4,500109E+1 ⇒ LOLD[h/f], 3,588480E-1 ⇒ TIEPI[h/year], 1,500036E+0 ⇒ NIEPI[int/year], 4,180144E+0 *** LINE01 MV. RELIABILITY INDICES *** ⇒ LOLP[pu] (sample mean), 3,431994E-4 ⇒ LOLP[pu] (confidence interval), 3,128208E-4, 3,735780E-4 ⇒ EPNS[MW] (sample mean), 6,863989E-4 ⇒ EPNS[MW] (confidence interval), 6,256417E-4, 7,471561E-4 ⇒ LOLF[h/f] (sample mean), 5,096408E-4 ⇒ LOLF[h/f] (confidence interval), 4,410971E-4, 5,781845E-4 ⇒ EENS[MWh], 6,012854E+0 ⇒ LOLD[h/f], 6,734144E-1 ⇒ TIEPI[h/year], 3,006427E+0 ⇒ NIEPI[int/year], 4,464453E+0 *** LINE02 MV. RELIABILITY INDICES *** ⇒ LOLP[pu] (sample mean), 3,482390E-4 ⇒ LOLP[pu] (confidence interval), 3,176418E-4, 3,788363E-4 ⇒ EPNS[MW] (sample mean), 6,964781E-4 ⇒ EPNS[MW] (confidence interval), 6,352835E-4, 7,576726E-4 ⇒ LOLF[h/f] (sample mean), 5,107745E-4 ⇒ LOLF[h/f] (confidence interval), 4,422260E-4, 5,793230E-4 ⇒ EENS[MWh], 6,101148E+0 ⇒ LOLD[h/f], 6,817862E-1 ⇒ TIEPI[h/year], 3,050574E+0 ⇒ NIEPI[int/year], 4,474385E+0 *** LINE03 MV. RELIABILITY INDICES *** ⇒ LOLP[pu] (sample mean), 3,436320E-4 ⇒ LOLP[pu] (confidence interval), 3,132366E-4, 3,740274E-4 ⇒ EPNS[MW] (sample mean), 6,872640E-4 ⇒ EPNS[MW] (confidence interval), 6,264731E-4, 7,480548E-4 ⇒ LOLF[h/f] (sample mean), 5,099194E-4 ⇒ LOLF[h/f] (confidence interval), 4,413758E-4, 5,784629E-4 ⇒ EENS[MWh], 6,020432E+0 ⇒ LOLD[h/f], 6,738948E-1 ⇒ TIEPI[h/year], 3,010216E+0 ⇒ NIEPI[int/year], 4,466894E+0 *** LINE04 MV. RELIABILITY INDICES *** ⇒ LOLP[pu] (sample mean), 3,404136E-4 ⇒ LOLP[pu] (confidence interval), 3,101593E-4, 3,706679E-4 2_ijeee_4515.indd 381 9/27/2012 8:01:06 PM D 382 P. J. Martínez-Lacañina et al. International Journal of Electrical Engineering Education, Volume 49, Number 4 (October 2012), © Manchester University Press ⇒ EPNS[MW] (sample mean), 6,808272E-4 ⇒ EPNS[MW] (confidence interval), 6,203187E-4, 7,413358E-4 ⇒ LOLF[h/f] (sample mean), 5,092436E-4 ⇒ LOLF[h/f] (confidence interval), 4,407016E-4, 5,777855E-4 ⇒ EENS[MWh], 5,964046E+0 ⇒ LOLD[h/f], 6,684691E-1 ⇒ TIEPI[h/year], 2,982023E+0 ⇒ NIEPI[int/year], 4,460974E+ Appendix D Test case 1 results in format txt **** Number of stochastic experiments, 506822 *** SUBSTATION. RELIABILITY INDICES *** ⇒ LOLP[pu] (sample mean), 3,421557E-4 ⇒ LOLP[pu] (confidence interval), 3,006388E-4, 3,836726E-4 ⇒ EPNS[MW] (sample mean), 9,137111E-3 ⇒ EPNS[MW] (confidence interval), 8,490042E-3, 10,784180E-3 ⇒ EENS[MWh], 9,500109E+1 ⇒ LOLD[h/f], 7,088480E-1 ⇒ TIEPI[h/year], 3,250036E+0 ⇒ NIEPI[int/year], 8,380144E+0 Appendix E Test case 2 results in format txt **** Number of stochastic experiments, 496751 *** LINE01 MV. RELIABILITY INDICES *** ⇒ LOLP[pu] (sample mean), 4,431994E-4 ⇒ LOLP[pu] (confidence interval), 4,128208E-4, 4,735780E-4 ⇒ EPNS[MW] (sample mean), 7,863989E-4 ⇒ EPNS[MW] (confidence interval), 7,256417E-4, 8,071561E-4 ⇒ LOLF[h/f] (sample mean), 6,096408E-4 ⇒ LOLF[h/f] (confidence interval), 5,610971E-4, 6,951845E-4 ⇒ EENS[MWh], 7,012854E+0 ⇒ LOLD[h/f], 6,984144E-1 ⇒ TIEPI[h/year], 4,256427E+0 ⇒ NIEPI[int/year], 6,235453E+0 2_ijeee_4515.indd 382 9/27/2012 8:01:06 PM S 1. Author: details of publisher? 2. Author: decimal places are indicated by points not commas in UK. Please clarify use of commas in Appendices C D E and/ or amend to points if necessary AUTHOR QUERY FORM Journal: International Journal of Electrical Engineering Education Article: Dear Author: During the preparation of your manuscript for publication, the questions listed below have arisen. Please attend to these matters and return this form with your proof. Many thanks for your assistance. Query References Query Remarks ijeee_4515 2_ijeee_4515.indd 1 9/27/2012 8:01:06 PM