Uncertainty management of renewable distributed generation

Abstract Power systems are changing from centralized ones, consisting of big producers, to distributed generation ones (small producers). Traditionally, Distributed Generation has been used to produce energy near demand and in isolated areas (islands, rural customers, etc.) but the growth of the new renewable technologies and the decentralization in production have caused an increase in the importance of Distributed Generation. Other causes for this increase are the advantages related to technical, economic and environmental benefits granted by this type of energy production. Distribution companies have to satisfy the demand of the final customers. Hence, they must act to ensure the energy supply. The alternatives consist of investments in the expansion of the distribution network comprising the replacement and addition of feeders, reinforcement of existing substations and construction of new substations, installation of new transformers and new generators, or any combination of them. The most common technologies used for Distributed Generation are wind turbines and photovoltaic modules. These types of technologies have the disadvantage of the uncertainty in production due to the dependence on renewable energy sources Renewable Energy Sources. This paper presents a new approach to manage uncertainty in order to solve the Distributed Generation planning problem Distributed Generation Planning Problem. The work presented compares the proposed (probability-based) approach with a traditional level-based method. The probability-based approach produces a better fit compared to the level-based method. This is reflected in lower operation and maintenance costs, although the investment costs are similar for both methods. A case study is used to compare the results of the two methods and to illustrate how scenarios have a significant impact on Distributed Generation investments. It is observed that the reduction in costs is due to the effect of having more scenarios and their balancing effect on costs (compensating high demand in case of low production and vice versa). This makes the method applicable in real-world cases that need a thorough scenario analysis.