Optimization of energy storage decisions in power networks with renewable generation

We consider a firm, with an already existing portfolio of renewable generation resources and customers dispersed over a certain geographical region, which is planning to invest in energy storage systems. The firm uses its renewable resources to generate electricity and the spot electricity markets to meet the demand from the customers. As a result, the firm faces a multi-period capacity management problem. The complexity of the problem is further exacerbated by the transmission line constraints in between the supply and demand points. These constrains may also motivate the firm to store electricity when the transmission lines are blocked because of the capacities. The possibility of storing electricity links the operating periods with each other. In particular, the energy produced by the renewable generators can be stored in storage units when the prices are low or the production is high while this energy can be released to the system when either the prices are high or the production is low. In this set up, the optimal investment plan is investigated, i.e., the location, the type and the capacity of the storage facilities and how to optimally operate the whole system. This includes the buy or sell decision at each period to the spot electricity markets, store or release energy from the storage facilities and the transmission of power among the nodes in the network. A mixed integer mathematical formulation is constructed. Besides, to solve this big model a heuristic approach is suggested. In this approach a solution for small periods over one year is obtained, then the solution is improved by local search and simulated annealing algorithms. At the last part of approach, the small periods are getting combined by deciding storage locations, types and capacities. IEEE modified RTS-96 benchmark is used and several managerial insides are offered. The solution get from the heuristic approach is close to optimal by 0.155% for original scenario.