Reliability Evaluation Of Distribution Networks Containing WTG Considering The Correlation Between Wind Speeds And Load Demands

With the appearance of worldwide problem of environmental pollution and energycrisis, the development of new energy, especially the technology of grid-connected windpower generation(WTG) have become hot issues in recent years. As a distributedgeneration(DG), the connection of WTGs to distribution network makes a greatcontribution to alleviate power shortage and improve energy structure in China. Due tothe randomness of wind energy, the power curve of WTG is quite different fromtraditional generators. The integration of DG has changed the operation characteristicsof traditional radial distribution network to make it a dual or multiport power supplynetwork, so a new reliability evaluation method is needed. Focus on these issues, windspeed and load time series is connected with multivariate statistical analysis techniqueand the wind-load bivariate normal distribution considering the correlations betweenthem is established in this paper. The reliability indices of distribution network includeWTGs are calculated using the Monte Carlo method and the sensitivity analyses ofWTG operation and reliability parameters are made in this paper.A joint distribution function for wind speed series and load demand series ispresented considering the correlation of the two series. The wind speed and loaddemand statistics for an aera can be use to evaluate the normal distribution functions ofwind speeds and load demands, respectively, and the correlation coefficient betweenthem. Based on the evaluated normal distribution for wind speeds and load demands, awind-load bivariate normal distribution is modelled using the techniques and principlesof multivariate statistical analysis. The proposed model was tested in modelling abivariate normal distribution for the wind speeds at Do Kooy wind farm in Holland andthe load of the IEEE RTS test system, and the results show that the algorithm is usefuland has efficiency in engineering practicality.Based on wind-load bivariate normal distribution considering the correlationsbetween wind speed and load, a sample method of wind-load sequence is proposed, thenthe reliability of distribution networks with distributed WTGs is calculated using theMonte Carlo method. The wind-load sequence is produced using the Monte Carlomethod based on bivariate normal distribution, then the probability and supply area ofWPGs for a sampling state can be determined using WPG output model. The load pointreliability indices can be evaluated then the feeders and system reliability indices can be obtained. The proposed algorithm is applied in RBTS-BUS6test system with distributedWTGs and the effects of WTG rated output on system reliability indices are analyzed.The results show that the algorithm is useful.Distribution network reliability is related to the power output of WTG, correlationcoefficient, average wind speed, WTG operation and reliability parameters. In order tofind the relationship between system reliability and these parameters, the impactanalyses between these parameters and reliability indices are made in this paper. Thesystem reliability indices with different wind-load correlation coefficient and WTGparameters are calculated and system expected energy not supplied versus theseparameters is plotted. The impect analyses are applied in RBTS-BUS6test system withdistributed WTGs, the results show that wind-load correlation coefficient, everage windspeed and WTG rated wind speed have appreciable impacts on system reliability levelwhile WTG cut-in, cut-out wind speed and its forced outage rate have minor impacts onsystem reliability level.