Coordination in markets with nonconvexities as a mathematical program with equilibrium constraints-Part II: case studies

• Published in: IEEE transactions on power systems Vol. 19; no. 1; pp. 317 - 324
• Main Authors: Motto, A.L., Galiana, F.D.
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.02.2004; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Computational efficiency; Computer aided software engineering; Cost function; Costs; Electric power generation; Generators; Heuristic; Lagrangian functions; Markets; Mathematical analysis; Mathematical models; Optimization; Piecewise linear techniques; Power generation; Power generation economics; Power system economics; Power system modeling; Production; Studies; System testing
• ISSN: 0885-8950; 1558-0679
• DOI: 10.1109/TPWRS.2003.820709

This paper is the second of a two-paper series. It is concerned with the numerical study of the solution procedure derived in to solve the coordination problem that arises in a new equilibrium model , which for the purpose of this presentation applies to a static (no-time coupling costs or constraints) electricity pool market with price inelastic demand and no network. The new equilibrium model has the following main properties: i) every scheduled generator satisfies its minimum surplus (or bid profit) condition; ii) the energy price is a system marginal cost (a Lagrange multiplier associated with the power balance constraint in the related economic dispatch problem where all of the discrete variables are fixed to their optimal values); iii) the power balance and all of the generators' technical constraints are satisfied. We present some numerical results based on three test systems: a simple three-generating unit system that can be solved by hand, a 32-generating unit system that consists of piecewise linear offer curves, and a large system of 768 generating units with monotone and nonmonotone, piecewise linear offer curves, some of which are set as must-run units. The results demonstrate that the proposed procedure is more efficient than a heuristic approach, both in terms of solution quality and computational efficiency.