Energy Reports (Nov 2021)
Distributionally robust chance constrained planning model for energy storage plants based on Kullback–Leibler divergence
Abstract
Nowadays, the high penetration of renewable energy, with variable and unpredictable nature, poses major challenges to operation and planning studies of power systems. Employing energy storage plants has been introduced as an effective solution to alleviate these challenges. Several studies have been presented in the literature to provide a framework for planning studies of energy storage plants. However, they usually have two main drawbacks: (i) ignoring the lifetime varying characteristic of energy storage, (ii) inability to model the charge/discharge schedule of energy storage accurately. This paper amends the abovementioned shortcomings by proposing a distributionally robust planning method based on Kullback–Leibler divergence. According to the power function of lifespan of electrochemical energy storage, the lifespan model of energy storage plants with equivalent full cycles times is established. Considering the lifespan model constraints of the energy storage plant and system operating constraints, the planning model of energy storage plants is constructed with the lifespan cycle cost and units’ operation cost as the objective. Furthermore, the ambiguity set of wind farm output based on Kullback–Leibler divergence is embedded into the planning model of energy storage plants, and the distributionally robust planning model of energy storage plant is transformed into mixed integer linear programming model by sample average approximation method. A modified IEEE-30 bus system with two wind farms is studied to demonstrate the effectiveness. The results show distributionally robust planning model reduces about 10% planning cost compared with the traditional robust planning model, and increases the energy storage capacity by 12MW compared with the stochastic planning method. Other, the planning cost of energy storage plants will decrease by 5% ∼23% when the wind abandoned rate increases by 2% and decreases slowly with the increase of abandoned wind rate.
