Comprehensive algorithm for hydrothermal co-ordination

Abstract

The authors present a comprehensive hydrothermal co-ordination algorithm where a new Lagrangian relaxation based hydrothermal co-ordination algorithm is integrated into an expert system. In this algorithm, the problem is decomposed into the scheduling of individual units by relaxing the demand and reserve requirements using Lagrangian multipliers. Dynamic programming is used for solving the thermal subproblems without discretising generation levels. Instead of solving the hydro subproblems independently as in the standard Lagrangian relaxation approach, hydrothermal scheduling is used to solve the output levels of hydro units. Hydrothermal scheduling uses the commitment status of thermal units obtained from the solutions of the thermal subproblems. The expert system takes care of constraints that are difficult or impractical for implementation in the Lagrangian relaxation based hydrothermal co-ordination algorithm, such as cycling of gas and steam turbine units, etc. It is also applied to check the feasibility of the solution. Extensive constraints such as power balance, spinning reserve, minimum up/down time, must run, capacity limits, ramp rate and hydro constraints are considered. Accurate transmission losses are incorporated. Nonlinear cost function is used, and the hydrothermal scheduling is implemented using a fast and efficient algorithm. Numerical results based on a practical utility data show that this new approach provides feasible schedules within a reasonable time.