462056 Exploiting Dynamic Flexibility to Enable Participation in Multi-Scale Electricity Markets

Wednesday, November 16, 2016: 1:46 PM
Carmel I (Hotel Nikko San Francisco)
Alexander W. Dowling, Ranjeet Kumar and Victor M. Zavala, Department of Chemical and Biological Engineering, University of Wisconsin, Madison, WI

There is significant interest in re-designing the control architectures for a broad spectrum of energy and manufacturing systems to exploit time-varying electricity prices. For example, the Alcoa Point Comfort Power Plant, which is a utility plant that provides electricity and steam to the adjacent aluminum manufacturing facility, re-optimizes its operations every 15 minutes in response to electricity and natural gas price. These new energy-oriented architectures provide load flexibility to the power grid in exchange for monetary payments. Most academic literature on industrial demand response optimization focuses on scheduling and planning decision layers and considers coarse (e.g., 1 hour) time discretizations with multi-day planning horizons. Emphasis has been placed on optimal planning under time-vary electricity prices, although recent work considers providing reserve capacity and interruptible loads. This emphasis missing flexibility at fast timescales (e.g., minutes to seconds) that may be monetized through participating in real-time energy markets and/or providing regulation capacity and other ancillary services.

We develop a mathematical model for multi-scale electricity markets run by the California Independent System Operator (CAISO). The model incorporates prices from both the day-ahead (1 hour intervals) and real-time markets (5 – 15 minute intervals), and considers both energy and ancillary service (regulation, reserves, etc.) sales. Market rules are expressed as linear algebraic expressions with continuous and optionally discrete variables. The models are general and applicable to both conventional thermal generators and storage devices (e.g., batteries).

Using the market model, we optimize the operation policy of a combined heat and power utility system using real price data for all of 2015. We observe participating in energy markets run at multi-times decreases net operating costs (fuel minus market proceeds) by 14 – 54%, depending on fuel price, without degrading satisfaction of onsite steam and electricity demand. The savings increase if the onsite steam and/or electricity demands are made flexible. Approximately one-quarter to one-half of the anticipated savings are from regulation capacity sales to the market. In CAISO, the automatic generator control (AGC) system sends new electricity production setpoints to generators providing regulation every 5 seconds that are within a specified flexibility band (units: MWe). Capacity payments are based on the size of this band. Thus regulation offers a way to monetize the inherent flexibility in manufacturing/energy systems at faster timescales (seconds compared to electricity price variations that occur at slower timescales (minutes to hours). In practice exploiting this flexibility and realizing revenues from electricity markets requires sophisticated controls strategies that manage multiple time-scales and consider uncertainty in market prices.


Extended Abstract: File Not Uploaded
See more of this Session: Energy Systems Design and Operations II
See more of this Group/Topical: Computing and Systems Technology Division