Energy trading in a competitive electricity market can be modeled as a two-level optimization. At the top le, el a Centralized Economic Dispatch (CED) uses a priority list method to solve the fundamental problem of reliable market clearing with price discovery. The lower level consists of a set of Decentralized Bidding (DB) subproblems. The DB model uses a self-unit scheduling simulator based on parametric dynamic programming to produce hourly bid curves for the central dispatch coordinator. Unit operating constraints and costs such as the unit minimum-up and minimum-down times, ramp rates, and the unit start-up, no-load and sunk capital costs are internalized in the bid curves through the simulator. A special algorithm is presented to solve the revenue adequacy problem for marginal units. Both CED and DB models are based on the revenue maximization in contrast with the cost minimization criteria used in the conventional Unit Commitment (UC), The proposed method has been tested in a study case and some interesting results have been demonstrated.
