TSO-STO: A two-step approach to the optimal operation of heat storage systems with variable temperature tanks

Mixed-Integer Linear Programming (MILP) has been generally used in the recent past to evaluate the optimal operation of heat storage systems for district heating. In fact, model equations and constraints can be linearized to strongly reduce the computational time without a significant loss in accuracy, and other simplifying hypotheses can be introduced, such as the constant value of the heat storage temperature.This paper presents instead a non-linear model of a Combined Heat and Power (CHP) system with a variable temperature heat storage serving a district heating network. Optimal operation for a fixed time-dependent demand is searched by varying CHP system loads. The objective is the maximization of management profit in a deregulated electricity market, taking into account investment (CHP and heat storage systems) and operating costs. The nature of the problem is investigated and a new approach for the decomposition of the objective function is proposed to simplify the solution procedure. The impact of different fuel costs and average electricity prices on the results is also analyzed.

► An NLP model of a CHP system with variable temperature heat storage is presented.
► Profit is to optimized for a fixed thermal demand.
► The original dynamic problem is decomposed sequentially.
► System operation and storage size are optimized in two non-dynamic problems.
► The impact of other scenarios is analyzed using the unit cost of thermal energy.