520 IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 26, NO. 2, MAY 2011 Impact of Natural Gas System on Risk-Constrained Midterm Hydrothermal Scheduling Cem Sahin, Zuyi Li, Senior Member, IEEE, Mohammad Shahidehpour, Fellow, IEEE, and Ismet Erkmen Abstract—This paper studies the impact of natural gas (NG) contracts and constraints on a GENCO’s midterm risk-con- strained hydrothermal scheduling problem. The NG contracts and constraints are modeled as a set of linear equations. The proposed model utilizes the stochastic price-based unit commit- ment (PBUC). The PBUC hourly solution considers uncertainties of market prices for energy and ancillary services, uncertainties of natural water inflows, and random NG infrastructure inter- ruptions in Monte Carlo scenarios. Illustrative examples analyze the GENCO’s risk levels when considering midterm schedules for generating units, target payoffs, and usages of water inflow, NG and other thermal resources. Simulation results show that a GENCO’s midterm schedules and financial risks could be impacted significantly with the consideration of NG contracts and constraints. Index Terms—Financial and physical risks, generation compa- nies, interdependency of natural gas and electricity, natural gas contracts, stochastic price-based unit commitment. NOMENCLATURE Indices: Index of hydro units. Index of coal units. Index of NG units. Index of pumped-storage units. Index of hydro catchments. Index of pipelines. Index of NG contracts. Index of power plants. Index of subareas. Index of scenarios. Index of time periods (hour). Index of NG storage facilities. Dimensions: Number of hydro catchments. Number of coal units. Manuscript received May 26, 2009; revised December 08, 2009 and February 19, 2010; accepted May 29, 2010. Date of publication August 03, 2010; date of current version April 22, 2011. This work was supported in part by the NSF under grant ECCS-0801853. Paper no. TPWRS-00342-2009. C. Sahin and I. Erkmen are with the Department of Electrical and Electronic Engineering, Middle East Technical University, Ankara, Turkey (e-mail: cem. sahin@uzay.tubitak.gov.tr; erkmen@metu.edu.tr). Z. Li and M. Shahidehpour are with the Department of Electrical and Com- puter Engineering, Illinois Institute of Technology, Chicago, IL 60616 USA (e-mail: lizu@iit.edu; ms@iit.edu). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TPWRS.2010.2052838 Number of NG gas contracts. Number of NG gas storage facilities. Number of hydro units of a hydro catchment . Number of NG units. Number of NG plants. Number of NG pipelines. Number of pumped-storage hydro units. Number of scenarios. Number of NG subareas. Number of time periods under study. Sets: Set of NG units utilizing a gas contract. Set of firm NG contracts. Set of interruptible NG contracts. Set of NG units in a plant. Set of gas contracts belonging to a pipeline. Set of NG units in a subarea Set of NG units sharing a NG storage facility. Variables: Cost of NG usage from contract in scenario . Cost of NG usage from storage facility in scenario . Fuel usage by unit. Operating reserve. Power generation. GENCO’s payoff in scenario . NG withdrawn from storage facility . NG injected to storage facility . Volume of NG in storage facility at time in scenario . GENCO’s downside risks in scenario . Shutdown cost. Spinning reserve. Startup cost. Lagrange multiplier. Lagrange multiplier for risk constraint in scenario . Constants: Interruption rate of NG contract . 0885-8950/$26.00 © 2010 IEEE