New probabilistic method for solving economic dispatch and unit commitment problems incorporating uncertainty due to renewable energy integration. (June 2016)

• Record Type: Journal Article
• Authors: Lujano-Rojas, J.M.; Osório, G.J.; Catalão, J.P.S.
• Is Part Of: International journal of electrical power & energy systems. Volume 78(2016:Jun.)
• Page Start: 61
• Page End: 71
• Publication Date: 2016-06
• Subjects: Forecasting error -- Probabilistic economic dispatch -- Priority list -- Probabilistic unit commitment -- Wind power; Electrical engineering -- Periodicals; Electric power systems -- Periodicals; Électrotechnique -- Périodiques; Réseaux électriques (Énergie) -- Périodiques; Electric power systems; Electrical engineering; Periodicals; 621.3
• DOI: 10.1016/j.ijepes.2015.11.064
• Languages: English
• ISSNs: 0142-0615

Abstract

Highlights: A methodology to solve unit commitment problem from a probabilistic perspective is developed. The method presented is based on solving the probabilistic economic dispatch problem. Three case studies are analysed, with 6, 10 and 110 units, to evaluate the quality of the solutions. Results are thoroughly compared to those offered by a stochastic programming approach. Abstract: In this paper, a methodology to solve Unit Commitment (UC) problem from a probabilistic perspective is developed and illustrated. The method presented is based on solving the Economic Dispatch (ED) problem describing the Probability Distribution Function (PDF) of the output power of thermal generators, energy not supplied, excess of electricity, Generation Cost (GC), and Spinning Reserve (SR). The obtained ED solution is combined with Priority List (PL) method in order to solve UC problem probabilistically, giving especial attention to the probability of providing a determined amount of SR at each time step. Three case studies are analysed; the first case study explains how PDF of SR can be used as a metric to decide the amount of power that should be committed; while in the second and third case studies, two systems of 10-units and 110-units are analysed in order to evaluate the quality of the obtained solution from the proposed approach. Results are thoroughly compared to those offered by a stochastic programming approach based on mixed-integer linear programming formulation, observing a difference on GCs between 1.41% and 1.43% for the 10-units system, and between 3.75% and 4.5% for the 110-units system, depending on the chosen significance level of the probabilistic analysis.