A stochastic hydro-thermal power system model: A sensitivity study case

A new stochastic hydro-thermal power system model for the optimal mid-term planning of the hydro discharge is briefly presented. A probabilistic model is used to determine the cost of generation and scheduling of the thermal units. Aggregation of reservoir and hydro plant sub-systems is used to reduce the dimensionality of the problem. The load is modeled as a Markov chain process. The optimization is performed using Stochastic discrete differential dynamic programming. The method is applied to a real power system to test its response to change in the cost of thermal energy and maintenance schedules of thermal units.