Short-term hydropower planning and bidding under uncertainty is a complicated task. The problem became more challenging with the liberalized market environment within the last two decades. Apart from this new reform taking place in the electricity market, the electricity market participants including hydropower producers experienced the second change in the form of intermittent wind power integration into power systems. Thus, previous decision support tools are not capable of fulfilling market participants’ expectations in the new competitive and highly uncertain environment. Intermittent power sources, namely wind power, increase the imbalances in the power system, which in turn increases the need of the regulating power sources. Being a flexible energy source, hydropower can provide regulating power. For this purpose, new hydropower planning and bidding models must be developed, capable of addressing uncertainties and the dynamics existing within market places.

In this dissertation, a set of new short-term hydropower planning and bidding models are developed for sequential electricity markets under price uncertainties. Developed stochastic coordinated hydropower planning and bidding tools can be classified into two classes, as models with exogenousand endogenous prices.

In the first class, developed coordinated bidding tools address the price uncertainties using scenario trees, which are built based on the distribution function of the unknown variables. Thus, the proposed coordinated bidding and planning tools consider all possible future prices and market outcomes together with the likelihood of these market outcomes. To reflect the continuously clearing nature of intra-day and real-time markets rolling planning is applied. In addition, models apply risk measures as another way to hedge against uncertain prices.

In the second class, hydropower stochastic strategic bidding models are developed using stochastic bi-level optimization methodology. Here market prices are calculated internally as dual variables of the load balance constraints in the lower level ED problems. To be able to solve the stochastic bilevel optimization problem, KKT optimality conditions are applied. By this transformation the problem is converted to a single-level stochastic program, which is simplified further using a corresponding discretization technique.