Hydro-thermal scheduling deals with the problem of finding optimal decisions related to the dispatch of power plants (e.g. hydro and thermal) under the presence of uncertainty. The classic theory in this field focuses on the stochastic hydro inflows. This problem can be solved for real problem instances in reasonable time with the so-called stochastic dual dynamic programming (SDDP) algorithms developed by Pereira and Pinto in 1991. We present several extensions of the problem, mainly focusing on additional uncertainties.
Stochastic fuel prices play an increasingly important role when the installed capacity of the thermal plants increase compared to the hydro-thermal plants. We present a scenario tree approach for modeling the fuel cost uncertainty and adopt the classical SDDP algorithm to cope with this additional uncertainty. The proposed method can also easily be applied to cope with electricity demand uncertainty.
After the Kyoto protocol in 1997, the power industry sees itself faced with the challenges of reducing its CO2 emissions. Therefore, we present a model for CO2 emission quotas in the context of SDDP. The emissions are modeled via reservoirs, allowing a time decomposition and an efficient solution algorithm. This new model is then applied to the optimal expansion problem.
Yet, another uncertainty has been introduced through the Cap-and-Trade mechanism for CO2 emissions. In order to deal with this additional uncertainty in a deregulated electricity market, we present two stochastic models optimizing a hydro-thermal power system; the first from the perspective of a global system and the second from a sub-system’s (country or utility) perspective within a liberalized market. Particularly CO2 emission quotas and CO2 certificate prices are taken into account. The first model seeks to compute the electricity system marginal price as well as the CO2 emissions marginal price by minimizing the expected system’s cost of operation. In the second model, the expected revenues of the sub-system within a liberalized electricity market are maximized.
This research has been done in collaboration with two industrial partners.