Energy Storage in the Distribution Grid

The implementation of distributed energy storage will play a vital role in the Smart Grid of the future, which is the merging of IT and the electrical power grid. With the increasing penetration of distributed energy generation (DEG), like wind and solar, the power grid is changing from a vertically integrated structure to a more dynamic bidirectional system. This system is often referred to as the Energy Cloud with reference to the Internet as a system based on distributed resources feeding into one common platform. Due to the intermittent nature of DEG, the system relies on large centralized energy producers to maintain steady and secure energy supply. However, by the implementation of distributed energy storage (DES), a variety of services are introduced that permit further increase of DEG penetration. As well as changes in energy generation, the introduction of high intensity loads are pushing the grid to its limits. The number of electric vehicles (EV), induction based cooking tops and direct water heaters are increasing and often occurs at times of peak load and forces the grid equipment to work above tolerance limits which may create high temperatures, premature ageing or in worst case failure. Traditional methods for handling equipment capacity issues have been to reinforce the grid so that peak load is kept within equipment limits. With high intensity loads, this method would result in a grid that is underutilized in off peak periods, which is most of the day. This is not cost effective, and a socio-economical sub optimal solution. In this report, the research targets peak shaving as a service provided by an Li-Ion battery storage located in the distribution grid, with the purpose of achieving increased flexibility that may prevent grid under-utilization and delay the need for grid reinforcements. In order for this solution to be economically viable, and able to compete with traditional grid reinforcement methods, dimensioning strategies are proposed to find the ideal balance between initial investment cost and service time. With seasonal changes and the subsequent variable energy consumption, the average required battery capacity on an annual basis is used as reference when sizing the battery storage. The dimensioning strategies proposed involve the design of an energy storage that accommodates the average required battery capacity within given depth-of-discharge limits, and the use of buffer capacity to accommodate the energy requirements in times of higher load. The energy consumption data available shows that the number of days that require more and less than the energy consumption average are similar, thus counteracting each other and giving a storage size that is ideally sized with the desired level of depth-of-discharge on an annual basis. To validate the sizing strategies, a simulation model is built using MATLAB Simulink, that incorporates energy consumption data from SFE Smart Valley together with a dynamic Li-Ion battery block from the Simscape library