Energy System Economics and Optimization
How much does electricity cost, who decides, and why does the answer change every five minutes? This course answers those questions by building up the optimization problems that sit underneath wholesale electricity markets. Students learn to formulate power system operation and investment decisions as optimization problems, solve them with modern solvers, and read the prices that come out of them as economic signals.
The course moves from microeconomic fundamentals through economic dispatch, unit commitment, and network-constrained markets, and closes on the questions that dominate the field today: how renewables and storage change the shape of the market, and how a participant should bid into it.
What you will learn
By the end of the course, students are able to:
- Define and discuss the major problems in power system economics
- Formulate these problems as optimization problems, and solve simple ones by hand
- Use optimization packages to solve realistically sized problems
- Describe the various types of electricity markets and discuss their purposes
- Discuss bidding strategies under perfect and imperfect competition
- Explain and calculate locational marginal prices
- Explain the economic pros and cons of different energy resources
- Discuss the factors that drive energy system investment
Topics
Economic foundations. Supply and demand, surplus and welfare, market power. How electricity differs from other commodities, and why that difference forces a market design.
Market organization and contracts. How electricity markets are organized, the risks participants carry, and the forward contracts and financial instruments used to hedge them.
Optimization and economic dispatch. Linear programming, duality, and inequality constraints. Solution algorithms for economic dispatch, and where the marginal price comes from.
Unit commitment. Mixed-integer linear programming, start-up and minimum-run constraints, and what non-convexity does to pricing.
Networks and nodal pricing. Introduction to power flow, transmission constraints, congestion, security-constrained economic dispatch, locational marginal prices, and financial transmission rights.
Participation and equilibrium. Bidding into the market as a generator, load, or storage operator. Equilibrium models of strategic behavior.
Operations, ancillary services, renewables, and storage. Reserves and ancillary service markets, and how variable renewables and energy storage reshape both operations and prices.
Course projects
Students work in small teams on a term project, either self-defined or chosen from a set of problems built on real system data. Recent project topics:
- Wind integration in ISO New England. Take the position of a policy maker and study how to best integrate up to 12,000 MW of wind capacity, using an eight-zone ISO-NE model with hourly wind capacity factors and demand.
- Peak shaving on the Columbia campus. Columbia buys electricity under the ConEdison SC9 Rate II tariff, where the demand charge is set by the highest two consecutive 15-minute intervals. Forecast campus demand and design a strategy to shave that peak.
- Energy portfolio management for a sustainable campus. Assemble a day-ahead supply portfolio from at-risk wind, baseload, peak-load, and load-following contracts, and manage the resulting risk exposure.
- Wind farm sizing and siting on the IEEE-14 system. Act as a system planner deciding where and how much wind to build under transmission constraints and multiple demand and wind scenarios.
Self-defined topics are strongly encouraged, and need only be related to energy systems and use the economic and optimization methods developed in the course.
Market experiments
Several class sessions are run as live market experiments: students bid against each other in a simplified electricity market and see the clearing outcome, the prices, and their own profit. The experiments make market power, capacity withholding, and equilibrium behavior something students experience rather than only derive.
Prerequisites
Linear algebra and differential equations (MATH 2030 or APAM 2101), and basic programming background (ENGI 1006E). No prior power systems coursework is assumed.
Assessment
Homework 40%, midterm 30%, final project 20%, and participation in class quizzes and market experiments 10%. Homework is completed in Jupyter notebooks; optimization is required from the midpoint of the course onward. Projects are graded on presentation, problem formulation, code quality, and analysis of results.
Textbooks
Required: Kirschen and Strbac, Fundamentals of Power System Economics, 2nd edition, Wiley, 2018.
Supplemental: Masters, Renewable and Efficient Electric Power Systems, Wiley, 2004; Conejo and Baringo, Power System Operations, Springer, 2018. All are available online through the Columbia Library.