Distributed generation resources (DG) can operate in three modes of peak shaving, backup, and intentional islanding. Optimal operation of DGs has high influence in reliability improvement and cost reduction. Accordingly, in this paper using the index of expected cost of interruption, the optimal operation mode of DGs is identified. The purpose of this optimization is improving the reliability and reducing the cost of distribution companies which own DG resources. Since any method, which considers all operation modes simultaneously, has not been presented, this paper is aimed to present a method to identify operation mode of DG. The presented method uses Monte Carlo simulation in order to model the network component and a seven-state Markov model is presented. Results show that total operation cost of the distribution companies is reduced 256838.489 ($/yr) by determining optimal operation mode for DG. Results also show that operation cost of system reduces drastically when DG operation mode is identified hourly rather than situation when DG operates in only single mode. Accordingly, using the presented method, companies which own DG can choose optimal operation mode of their DGs for each hour so that most benefits are acquired.

Increasing application of distributed generation on distribution networks is the direct impact of development of technology and the energy disasters that the world is encountering. Using these resources in distribution network is one of the most influential solutions to reduce losses, improve voltage profile and improve power quality. To obtain these goals the resources capacity and the installation place are of a crucial importance. In this paper a new method is proposed to find the optimal and simultaneous place and capacity of these resources to reduce losses, improve voltage profile and reduce network harmonics. The proposed method is also capable of identifying the appropriate number of resources. The method is tested on actual power network of Zanjan Province, Iran and the simulation results are presented and discussed. Genetic algorithm is used to obtain the best answers.