31 American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) ISSN (Print) 2313-4410, ISSN (Online) 2313-4402 ยฉ Global Society of Scientific Research and Researchers http://asrjetsjournal.org/ Reliability Improvement of System Reconfiguration by Considering the Time-varying Load with Optimal Scheduling of Distributed Generations Wunna Thua*, Phoo Ngone Sib aWunna Thu,Yangon Technological University, The Union of Myanmar bPhoo Ngone Si,Yangon Technological University, The Union of Myanmar aEmail: wunnathu.ptn@gmail.com bEmail: phoongoneytu@gmail.com Abstract The distributed generation (DG) locations have significant impacts on network configuration and loss. By fixing DGs in suitable optimal locations and by generation power based on the load conditions, the total power loss in the system can be reduced and the system reliability can be improved. Distribution system reliability assessment is able to predict the interruption profile of a distribution system based on system topology and component reliability data. In this paper, Fuzzy algorithm is used to obtain the optimum position and size of DG units in the distribution network in order to reduce network loss at the lowest cost. Also, a time-varying load curve for optimal scheduling of DGs considering active power loss and DG cost is used. The result shows the improvement of bus voltage profile and decrease losses due to install the optimal size of DGs by optimal scheduling of distributed generations. The test system is Yangon 66kv; 45-bus and the results obtained reveal the effectiveness of proposed method. Keywords: Distributed generation; Fuzzy algorithm; Loss reduction; Optimum position; System reliability. ------------------------------------------------------------------------ * Corresponding author. http://asrjetsjournal.org/ American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2016) Volume 26, No 3, pp 31-41 32 1. Introduction The accurate load representation is not easy issue regarding to its changeable nature and variety. It is hard to predict the system behavior in case of outage, voltage or frequency variation or other factors that can affect system stability without proper information, or system behavior data. The powers system are mostly nonlinear and operates in constantly changing conditions, so it is hard to obtain precise information and data about i.e. generators or loads during disturbance. Reliability is one of the most important parameter to analysis the performance of the system. Reliability can be calculated on both customer side and utility side. Reliability analysis is generally done in distribution system because most of fault occurs in distribution system only [7]. The parameters for calculation of reliability improvement are: Momentary interruption of Single operation of an interrupting device is that results in a voltage zero. For example โ€“ two circuit breaker operations equals two momentary interruptions [2]. Sustained interruption of any interruption not classified as a momentary interruption. Interruption is longer than five minutes [2]. Distribution system reliability indices is the most common distribution indices include the System Average Interruption Duration Index (SAIDI), Customer Average Interruption Duration Index (CAIDI), System Average Interruption Frequency Index (SAIFI), Momentary Average Interruption Frequency Index (MAIFI), Customer Average Interruption Frequency Index (CAIDI), Customers Interrupted per Interruption Index (CIII), and the Average Service Availability Index (ASAI) [6]. The performance of Yangon distribution system becomes inefficient due to the reduction in voltage magnitude and increase in distribution losses. With this regard, changing environment of power systems design and operation have necessitated the need to consider active distribution network by incorporating Distributed Generation units (DGs) sources [4]. DGs are grid-connected or stand- alone electric generation units located within the electric distribution system at or near the end user. The case study area is Yangon 66kv distribution system and the total customer of Yangon is 1141097 customers. 2. Research Methodology Start Load Flow Running by using Newton - Raphson Method Finding the Optimal Size and Location of Distributed Generation by using Fuzzy Calculation of Reliability Indices Stop Data Surveying and Situational data analysis Optimal Scheduling of DGs by using Time - varying Load Figure 1: Flow Chart of the proposed method American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2016) Volume 26, No 3, pp 31-41 33 (2.1) Data surveying and situational data analysis (2.2) Load flow running by using Newton-Raphson method (2.3) Finding the optimal size and location of DGs (2.4) Optimal Scheduling of DGs by using time-varying load (2.5) Calculation of Reliability Indices 2.1 Data surveying and situational data analysis The test system used in this paper is Yangon 66kv distribution system. In Yangon, if a transformer or line is failure, the repaired time takes long time that is about 48 hours or more. It is not reasonable for power system improvement. 66/33kV 2 x 30 MVA Seinpanmying Kyaikkasan 66/33kV 2x 30 MVA 66/33kV 2 x 30 MVA Pathein Nyunt 66/11kV 3 0 MVA 66/11kV 20 MVA 66/33kV 30 MVA 230/66 kV 2 x 100 MVA Tharketa E/ Dagon(1) GT + STG 92 MW S/ Dagon Thaketa O/H O/H E/ Dagon(2) 66/11kV 10 MVA Max Power GE 52. 8 MW Mindin 66/11/6. 6 kV Yuzuna 66/33/6. 6 kV 30MVA 66/11kV 10 MVA E/ Dagon 66/11kV 2x 20 MVA N/ Dagon Bailey 66/11kV 20 MVA E/ Dagon 66/11kV E/ Dagon 66/11kV 1. 5 MVA Figure 2: Test system of Yangon distribution system Figure 3: Daily Load of Yangon Distribution System American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2016) Volume 26, No 3, pp 31-41 34 2.2 Load flow by Newton-Raphson method The following table-2 shows the bus voltage (per unit) of each bus of power flow solution by Newton-Raphson Method. Some of the radial bus voltages are out of the permissible range (ยฑ5%). Table-2 indicates that the total load of 989.710MW and the total loss of 6.567MW. For the permissible range of bus voltages, Fuzzy algorithm is used for optimum position of distributed generation. Table 1: Power Flow Solution by Newton-Raphson Method Descriptions MW Total Load 989.710 Total Loss 6.567 2.3 Implementation of Fuzzy Method Two objectives are considered while designing a fuzzy logic for identifying the optimal DG locations. The two objectives are: (i) to minimize the real power loss and (ii) to maintain the voltage within the permissible limits. Voltages and power loss indices of distribution system nodes are modelled by fuzzy membership functions. A fuzzy inference system (FIS) containing a set of rules is then used to determine the DG placement suitability of each node in the distribution system. In the first step, load flow solution for the original system is required to obtain the real and reactive power losses. Again, load flow solutions are required to obtain the power loss reduction by compensating the total active load at every node of the distribution system. Loss Reduction Index (LRI) value for ith node can be obtained using equation 1. ๐‘ƒ๐‘ƒ๐‘ƒ๐‘ƒ๐‘ƒ๐‘ƒ = (๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ(๐ฟ๐ฟ)โˆ’๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ(๐‘š๐‘š๐ฟ๐ฟ๐ฟ๐ฟ)) (๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ(๐‘š๐‘š๐‘š๐‘š๐‘š๐‘š)โˆ’๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ๐ฟ(๐‘š๐‘š๐ฟ๐ฟ๐ฟ๐ฟ)) (1) Seven membership functions are selected for PLI. They are VL, L, ML, M, MH, H and VH. All the seven membership functions are triangular as shown in figure-4. Seven membership functions are selected for Voltage. They are VL, L, ML, M, MH, H and VH. These membership functions are trapezoidal as shown in figure-5. Seven membership functions are selected for DSI. They are VL, L, ML, M, MH, H and VH. These seven membership functions are Gaussian as shown in figure-6. Figure 4: Specifying the first input Variables as PLI American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2016) Volume 26, No 3, pp 31-41 35 Figure 5: Specifying the second input Variables as VOLT Figure 6: Specifying the Output Variable Suitability-Degree IF premise (antecedent), THEN conclusion (consequent). For determining the suitability of DG placement at a particular node, a set of multiple-antecedent fuzzy rules has been established [5]. Table 2: Rule Base for Suitability Index Voltage LRI VL L ML M MH H VH VH VH VH H H MH MH M H VH H H MH MH M ML MH H H MH MH M ML ML M H MH MH M ML ML L ML MH MH M ML ML L L L MH M ML ML L L VL VL M ML ML L L VL VL Total load and line loss are calculated by using MATLAB software. Firstly, load flow is running to obtain network losses. Then power loss reduction is evaluated by compensating the same minus active load at every node, and load flow solutions are required. Power loss index (LRI) can be evaluated by eq.1 that normalizes loss American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2016) Volume 26, No 3, pp 31-41 36 reduction to take values between minimum and maximum such that the highest reduction takes the value of 2.096 and the lowest reduction takes the value of -0.708. Table 3: Calculation Table Bus Ploss_orginal Ploss_new Ploss_reduction(PLI) Volt_ibus LRI Suitability 22 6.567 5.635 0.932 0.94 0.584879 0.667 26 6.567 6.298 0.269 0.945 0.348431 0.61 31 6.567 6.139 0.428 0.946 0.405136 0.601 The suitability indices were calculated by using loss reduction index and voltage. Figure 7: Suitability Index According to the calculation results, the highest suitability indices of buses are bus number 22, 26 and 31. For the improving of bus voltage profile and decrease losses, the optimal sizes of DGs are installed in these buses. 2.4 Optimal Scheduling of DGs by using Time-varying Load Table 4: Maximum Power Rating Line Size Ampere (from table) Power Factor Stability Margin Power (MW) Bus 21-22 605MCM 760A 0.85 0.7 52.6 Bus 25-26 120mm2 460A 0.85 0.7 31.3 Bus 30-31 397.5MCM 590A 0.85 0.7 40.8 0 5 10 15 20 25 30 35 40 45 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Bus Number Su ita bil ity In de x (p u) American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2016) Volume 26, No 3, pp 31-41 37 Table 5: Comparison of the Best Case by Hourly Hours Case Ploss Q loss Ploss_org 1:00 Case 3 5.43 27.113 5.664 2:00 Case 3 5.394 26.743 5.773 3:00 Case2 5.661 28.826 5.717 4:00 Case 3 5.27 26.941 5.912 5:00 Case 3 4.824 23.627 5.71 6:00 Case2 4.497 22.675 5.649 7:00 Case2 3.959 20.58 5.49 8:00 Case 1 3.942 20.395 5.738 9:00 Case 3 3.775 19.383 6.251 10:00 Case2 3.514 17.637 6.311 11:00 Case2 3.509 17.591 6.567 12:00 Case 1 3.983 19.795 6.075 13:00 Case2 3.771 19.496 5.979 14:00 Case 1 3.724 19.06 6.358 15:00 Case 1 3.81 19.831 5.815 16:00 Case 1 3.51 17.715 6.145 17:00 Case 3 3.807 19.472 6.195 18:00 Case2 3.833 19.657 6.014 19:00 Case 1 3.952 20.425 5.71 20:00 Case 1 3.851 19.888 5.816 21:00 Case 3 3.918 19.591 5.794 22:00 Case 3 3.827 19.87 5.614 23:00 Case2 4.173 21.701 5.34 24:00:00 Case 3 4.735 23.142 5.542 Table 6: Comparison of Per Unit Voltage with and without DGs Unit 1:00AM W/O DG Case 1 Case 2 Case 3 0.962 0.976 0.987 0.976 0.956 0.989 1.017 0.989 0.969 0.977 0.983 0.977 0.967 0.975 0.981 0.975 0.964 1.001 0.991 0.991 0.961 1.013 1 1 0.961 0.998 0.977 0.998 0.956 1.005 0.976 1.005 American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2016) Volume 26, No 3, pp 31-41 38 Figure 8: Comparison of Per Unit Voltage Figure 9: Comparison of Losses 2.5 Reliability Indices The sample calculation of radial distribution circuit contains three load points with total load capacity of 15.24MW and total customers connected with system were 43729 customers. Initially all the reliability parameters were calculated for the base case and then the disconnecting switches are connected with load points B, C and D and again all parameters were calculated. Now in case 1, 2 and 3, a DG of capacity 10, 10 and 35MW were connected at the poor bus voltage buses and again all the reliability parameters were calculated. Table 7: Load Point Indices for Base Case (for bus 31) Load Point ฮป (failure/yr) r (Hrs) U (Hrs/yr) B 1.052 2.563 2.696 C 1.052 2.572 2.706 D 1.052 3 3.156 American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2016) Volume 26, No 3, pp 31-41 39 Table 8: Load Point Indices for Case1 and 2 (for bus 31) Load Point ฮป (failure/yr) r (Hrs) U (Hrs/yr) B 1.052 0.131 0.138 C 1.052 0.129 0.136 D 1.052 0.053 0.055 Table 9: Load Point Indices for Case 3 (for bus 31) Load Point ฮป (failure/yr) r (Hrs) U (Hrs/yr) B 1.052 0.230 0.236 C 1.052 0.230 0.235 D 1.052 0.175 0.176 Table 10: Comparison of Reliability Indices for all Three Case (for bus 31) Indices Without DG Case 1 Case 2 Case 3 SAIFI (int/cust/yr) 0.00007217 0.00007217 0.00007217 0.00007217 SAIDI (hrs/cust/yr) 0.0001957 0.00000752 0.00000752 0.00001479 CAIDI) (hr/yr) 2.7116603 0.1042458 0.1042458 0.2050063 ASAI (%) 0.99 0.99 0.99 0.99 ENS(MWh/yr) 130.42 5.014 5.014 9.86 AENS(MWh/yr) 0.00298 0.0001146 0.0001146 0.00022548 Figure 10: Comparison of ENS (for bus 31) American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2016) Volume 26, No 3, pp 31-41 40 Figure 11: Comparison of ENS (for bus 26) Figure 12: Comparison of ENS (for bus 22) From the figure 10, 11 and 12, reliability improves is a good manner for system because of the penetration of DGs to the radial bus with appropriate location and size. As system reliability indices SAIFI and ASAI is not change but the indices of SAIDI, CAIDI and ENS are improved by case 1, 2 and 3. Availability of service is not increased. The indices of CAIDI, SAIDI, ENS and AENS will be decreased after putting the DG sources respectively. 3. Conclusion In this paper, Fuzzy method was implemented by using MATLAB and was tested for a Yangon 66kv, 45-bus test system. This method was compared after connecting one DG, two DGs, and three DGs to the system at different load power values. The suitability indices improvement show the location of radial buses where the DGs should be introduced. By installing DGs at the radial buses, the total power loss of the system has been reduced drastically and the voltage profile of the system was also improved. The calculation results of table-5 and 6 were showed that appropriate size and location of DG units will lead a significant role to minimize the losses in distribution system. The sample calculation of SAIFI, SAIDI, CAIDI, ASAI, ENS and AENS are presented as reliability indices when reliability improvement of Yangon distribution system was evaluated. In this system if the DG of more capacity reliability of system will improves more but cost of DG will much higher than service availability. Moreover, it was found that losses reduction and voltage profile improvement is more in the Yangon network by installing the DG units. American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2016) Volume 26, No 3, pp 31-41 41 Acknowledgements The author wishes to acknowledge especially thankful to Dr. Wunna Swe, Associate Professor and Head of Electrical Power Engineering Department, Yangon Technological University, Dr. Phoo Ngone Si, Associate Professor of Electrical Power Engineering Department, Yangon Technological University, Dr. Than Zaw Htwe, Associate Professor of Electrical Power Engineering Department and all my teachers from Department of Electrical Power Engineering, Yangon Technological University, for their suggestions and correction on this paper. Moreover, the author is also grateful to Ministry of Electric and Energy, Myanmar for giving a chance to study in Yangon. References [1] โ€œIEEE Guide for Electric Power Distribution Reliability Indicesโ€, IEEE Standard 1366, 2003 Edition [2] R. Billinton, โ€œDistribution System Reliability Evaluationโ€, IEEE tutorial course- Power System Reliability Evaluation [3] P. P Barker. and R. W. De Mello, โ€œDetermining the impact of distributed generation on power systems: Part 1- Radial distribution systems,โ€ IEEE Power technology, Inc, Vol. 3, pp. 1645โ€“1656, 2000 [4] T. Gozel and M. H. Hocaoglu, โ€œAn analytical method for the sizing and siting of distributed generators in radial systems,โ€ Electric Power System, Vol. 79, pp. 912โ€“918, 2009. [5] A. Pregelj, M. Begovic, and A. Rohatgi, โ€œRecloser allocation for improved reliability of DG enhanced distribution networks,โ€ IEEE Transaction on Power System, Vol. 20, pp. 1442โ€“1449, 2006. [6] Allan. R.N, R. Billinton and Shale, โ€œBibliography on the Application of Probability Methods in Power System Reliability Evaluation 1977-1982โ€, IEEE Trans. On Power Apparatus and Systems, Vol. PAS- 103, No. 2, Feb. 1984. [7] R.Billinton, J.E.Billinton., โ€˜Distribution System Reliability Indicesโ€™ โ€“ IEEE Transaction on Power Delivery, Vol.4, No.1, January 1989. [8] [8] Andrade WS, Borges CLT, Falca หœo DM. โ€œModeling Reliability Aspects of Distributed Generation Connected to Distribution Systemsโ€. Proceedings of IEEE Power Engineering Society, General Meeting, Montreal, Canada, 2006.