Effect of Time-Variability Weather Conditions on the Reliability of Distribution Systems M. R. ZARE, R. HOOSHMAND, S. ESHTEHARDIHA, M. BAYATI POODEH Islamic Azad University, Majlesi Branch; Electrical Engineering the University of Isfahan IRAN Abstract: -Reliable evaluation of distribution systems is of high importance in the maintenance and expansion of these systems. A time-sequential simulation technique is presented in this paper in which the effects of weather conditions and maintenance methods in the assessment of reliable cost of integrated distribution systems are provided. Time-Varying Weight Factors (TVWF) are defined to investigate the effect of weather conditions and present maintenance methods on Failure rates (FR). In fact, the average Failure Rate (FR) is combined with TVWF to provide time-varying repair times (TVRTs) for each component. Similarly, the average Repair Time (RT) is also combined with TVWF to produce Time- Varying-Repair Time (TVRT). An experimental distribution system showed that TVFR has more effects on the interruption costs of the sensitive costumers. It has also significant effects on the indices of all costumers. So, it is necessary to consider TVRT in evaluating the reliability of the network cost. Key-Words: - Distribution systems, network reinforcement , reliability cost/worth 1 Introduction The subscriber's interruption costs and reinforcement reliability costs of the network are useful indices for the network designers while making optimum designs and administrative decisions [1]–[3]. The reliability of these indices has great effects on the final decisions made by the network designers. The index reliability depends upon the techniques, component parameters, and the models employed in the analysis. The techniques employed for the reliability assessment of the power systems are totally divided into analytic graphs [4]-[5], Monte Carlo [5]–[8], and a combination of simulation and analytical methods [9]. The reliable parameters are mostly considered to be constant in the analytical techniques. So, the Failure Rate and Repair Rate (RR) parameters are constant. Time To Failure (TTF) and Time To Repair (TTR) have exponential distributions. In actual power systems the restoration time can have other distributions such as normal or lognormal. The FR is a function of the weather conditions and environment. In adverse weather conditions, the Failure Rate of a component might be larger than that in the normal weather conditions [10]. The repair times of the components are also affected by variable weather conditions as well as repair methods. Restoration time during a winter season is greater than that considered for a normal amount. The repair time also depends on whether it is daytime, night, weekday or weekend. In a given weather condition the average time required for restoration during a weekend or within some hours at night can be greater than the usual amount during the weekdays or daytime. The equivalent two-state model and the four-state weather model are employed in [10] and [11], respectively. Constant failure rate and repair time are considered in these models. Billinton and Li [12] employed the Monte Carlo sampling technique to incorporate variable weather conditions in a composite form. A uniform distribution for weather condition sampling based on occurrence probability and weather conditions was employed. The weather conditions are divided into normal and adverse in all states. The probabilities of normal and adverse conditions are employed to incorporate the effect of weather. The data analysis of weather conditions and repair methods are not considered in [10] and [12]. The time-varying nature and uncertainty of the system parameters can not be easily considered while employing analytical methods. The result of data analysis and system random behavior by the application of chronological simulation with the assumption of time-variable load and interruption cost models is presented in [8]. Vang and Billinton did not consider the time-varying FR and TR in [8]. Instead of the application of the probabilities of adverse or normal weather conditions, the issues of Time-Varying Failure Rate (TVFR) and Time- Varying Restoration Time (TVRT) are defined in this article to specify the weather condition effects and variable repair methods on the reliability of parameters of a component. The chronological 7th WSEAS International Conference on Electric Power Systems, High Voltages, Electric Machines, Venice, Italy, November 21-23, 2007 160