Impacts of energy storage on power systems with stochastic generation

Abstract

Planning and operation of electric power systems are increasingly influenced by the presence of stochastic infeeds from converters of renewable energy such as wind and photovoltaic systems. The implications reach from complex voltage band management methods on the distribution system level to more and more demanding power flow management methods on the transmission system level. The volatile character of the stochastic generation is the main reason for such complications, especially since the unrestricted, maximum infeed of as much renewable energy as possible is becoming a declared goal in the energy strategies of many countries worldwide. One possibility to reduce this volatility is the use of energy storage devices, distributed in the system. They may be used to shift energy from periods of high infeed power to periods of low power and hence approach the characteristics of stochastic generation to conventional generation schemes. This work is aimed to be a theoretical contribution to answer the question for the usefulness of distributed energy storage devices in complex systems. It is directly connected to the question on the necessity of new energy storage technologies, both questions being probably one of the best-known chicken-and-egg problem of energy technology. The assessment of systems with a high repartition of stochastic generation, supported by energy storage, has so far been elaborated with the same deterministic methods that are typically used for conventional power systems. Furthermore it should be noted that there have been no concise standard procedures for the integration of storage into stochastically fed, meshed power systems. Especially for unmeshed stand-alone solutions, that is, islanded systems with no connection to national electricity networks, many deterministic dimensioning methods are in use. The concession to the stochastic nature of the infeeds is made only through the utilisation of measured time series of the power generation process and the subsequent calculation of the most inconvenient operation point during the sequence. It can, however, be easily shown that the obtained results may be paradox or at least misleading: the more data material is at hand, the more capital-intensive would be the chosen energy storage asset. The present report shows the theoretical backgrounds of this fact and proposes to invert the energy storage assessment method: Instead of planning the necessary characteristics of the storage assets based on the