Industrial Engineering & Management
322 Engineering North
Oklahoma State University
Stillwater, OK 74078
IEM Seminar Series
Industrial Engineering & Management in conjunction with the OSU INFORMS Student Chapter will be sponsoring a Seminar Series in the fall and spring semester. Various topics will be covered by faculty from the department. The series will be held on Thursdays 3:30pm-5:00pm. Seminar topics, location and a short abstract and bio about the presenters will be available here as it becomes available.
Sponsored by IEM & OSU INFORMS Student Chapter
Date: Thursday, October 8th
Time: 3:30 – 5:00 pm
Location: Engineering North 316
Stochastic Unit Commitment at Scale: Cost Saving Analysis for ISO-NE
Abstract: Studies of the cost savings potential for stochastic unit commitment relative to existing deterministic methods are limited, and it is unclear to what degree results from specific geographic regions (e.g., Ireland and the UK) generalize. Further, the details of production cost simulations are critical in the quantification of cost savings, as many existing simulators are partially prescient, fail to simulate out-of-sample, or both. Building on a DOE ARPA-e funded effort, we have recently developed scalable solution technology for stochastic unit commitment, coupled with accurate scenario generation methods and production cost simulators. In this talk, we apply our technologies to a study of ISO-NE with EWITS wind penetration levels and higher. We analyze the cost savings results for various levels of wind penetration, and find that reliability impacts are more significant than costs – particularly at high penetration levels. Similarly, we show that standard methods of quantifying the value of a stochastic unit commitment solution are not always indicative of simulated, out-of-sample cost savings. Insights into limitations of deterministic reserve carrying rules in the face of high wind penetration levels will also be discussed.
Speaker Bio: Speaker Bio: Dr. Jean-Paul Watson is a Distinguished Member of Technical Staff in the Discrete Math and Optimization Department at Sandia National Laboratories, in Albuquerque, New Mexico. He leads a variety of research efforts in stochastic optimization, ranging from basic algorithm development and analysis to applications, including the electric power system and other critical infrastructures. He is a co-developer of the Pyomo open-source software library (www.pyomo.org) for modeling and solving algebraic optimization models. He has co-authored over 60 journal articles and conference papers, in fields ranging from artificial intelligence and operations research to power systems and physical chemistry.