GENCO's risk-constrained hydrothermal scheduling

Lei Wu, Mohammad Shahidehpour, Zuyi Li

Research output: Contribution to journalArticlepeer-review

88 Scopus citations

Abstract

This paper presents a stochastic midterm risk-constrained hydrothermal scheduling algorithm in a generation company (GENCO). The objective of a GENCO is to maximize payoffs and minimize financial risks when scheduling its midterm generation of thermal, cascaded hydro, and pumped-storage units. The proposed schedule will be used by the GENCO for bidding purposes to the ISO. The optimization model is based on stochastic price-based unit commitment. The proposed GENCO solution may be used to schedule midterm fuel and natural water inflow resources for a few months to a year. The proposed stochastic mixed-integer programming solution considers random market prices for energy and ancillary services, as well as the availability of natural water inflows and generators in Monte Carlo scenarios. Financial risks associated with uncertainties are considered by applying expected downside risks which are incorporated explicitly as constraints. Variable time-steps are adopted to avoid the exponential growth in solution time and memory requirements when considering midterm constraints. A single water-to-power conversion function is used instead of several curves for representing water head and discharge parameters. Piecewise linearized head-dependant water-to-power conversion functions are used for computational efficiency. Illustrative examples examine GENCOs' midterm generation schedules, risk levels, fuel and water usage, and hourly generation dispatches for bidding in energy and ancillary services markets. The paper shows that GENCOs could decrease their financial risks by adjusting expected payoffs.

Original languageEnglish
Pages (from-to)1847-1858
Number of pages12
JournalIEEE Transactions on Power Systems
Volume23
Issue number4
DOIs
StatePublished - 2008

Keywords

  • Financial and physical risks
  • Generating companies
  • Midterm operation
  • Mixed integer programming
  • Monte Carlo simulation
  • Stochastic price-based unit commitment
  • Variable time step

Fingerprint

Dive into the research topics of 'GENCO's risk-constrained hydrothermal scheduling'. Together they form a unique fingerprint.

Cite this