APPA-Funded Study Provides Strategy For Optimal Renewable Power Bids

A new American Public Power Association-funded study has brought small- and medium-size utilities one step closer to confidently offering and purchasing renewable energy products on wholesale power markets.

In many wholesale power markets, utilities with a real time production shortage from cleared day-ahead bids must purchase products in the real-time market to make up the shortage. It’s no surprise, then, that wind and solar resources are not currently eligible to provide ancillary services in many electricity markets: their intermittency and the uncertainty of forecasts surrounding these products leads to expensive shortages for utilities.

This is true in the Southwest Power Pool (SPP), which PhD candidate Anne Stratman has researched as part of her studies in electrical engineering at the University of Nebraska-Lincoln’s Power and Energy Systems Lab. Stratman’s current research focus is refining a model that utility operators with renewable energy resources in their portfolios can use to participate more fully in wholesale power markets. Her recent project, Providing System Reserves with Renewable Resources in the Southwest Power Pool Market, offers a glimpse into a future where renewable-based energy systems can provide system reserves using wind and solar resources. This work was partially funded by the American Public Power Association’s Demonstration of Energy & Efficiency Developments (DEED) program.

As a basis for her research, Stratman pulled from Southwest Power Pool and National Renewable Energy Laboratory databases to gather historical data for day-ahead energy and reserve prices, real-time prices, and wind and solar power production. Data collection centered on a location near Beatrice, NE, an area within Nebraska Public Power District (NPPD)’s service territory. Roman Estrada of NPPD offered a utility perspective to Stratman’s research as her DEED project sponsor.

“Through discussions with my NPPD sponsor, I gained valuable insights into current industry practices and challenges faced by small to medium utilities. These discussions were helpful in bridging the gap between academic theory and industry practice and ensuring the model would be useful to utilities,” said Stratman.

Stratman used Beatrice, NE-area price and production datasets to develop a stochastic optimization model for utilities with wind and solar resources that could calculate curves for bidding into markets for different products: namely, energy, spinning reserves, up regulation, and down regulation. Stochastic optimization was used as a low-cost method to consider many different forecast scenarios, adjust for uncertainty, and formulate an optimal bidding strategy in all possible scenarios.

The resulting model suggested that the best approach for wind and solar resources produced in the region would be to offer, on average, about 5 percent of forecasted power output on the day-ahead market and 95 percent on the real-time market. The average product distribution for wind power, in the day-ahead and real-time markets combined, was 84 percent energy, 0.5 percent spinning reserves, 3.5 percent up regulation, and 12 percent down regulation. For solar power, the average product distribution was 91 percent energy, less than 1 percent spinning reserves, 3 percent up regulation, and 6 percent down regulation, Stratman’s report said.

Additionally, Stratman reviewed case studies and found that by offering several types of products, some with much lower real-time prices than others, a utility would likely be able to avoid real-time penalties by offering less expensive products when forecast uncertainty is high.

“Based on the project results and my dissertation research up to this point, I’ve observed that offering reserve products can allow utilities with highly uncertain generation resources to hedge against the risk of large real-time deviation penalties, compared to only participating in the energy market. Usually, it’s only necessary to offer small quantities of reserve products to reduce risk significantly. However, the tradeoff between profit and the need for reliable reserve commitments should also be considered. Hopefully, this project provides a steppingstone towards greater use of wind and solar resources in reserve markets as forecasting methods improve,” said Stratman.

In the future, Stratman plans to write a conference paper about the model developed in the project that would use a forecasting model to generate prices and wind and solar power scenarios, instead of using historical data as scenarios. There is much less variation in scenarios generated using forecasting models than in scenarios that use historical data, meaning that day-ahead bids would not exceed forecasts, the report said.

DEED members interested in experimenting with Stratman’s model are in luck: as part of her research, she developed a simple MATLAB code and a user guide for use by small and medium utilities with wind and solar resources. This software is applicable for utilities participating in any of the organized independent or regional wholesale markets. The software is available to members of APPA’s R&D community via the DEED Project Library.

Non-DEED members interested in learning more might like to stop by the DEED booth at the American Public Power Association’s Engineering & Operations Conference, where a research poster (and perhaps the researcher herself!) will be available to answer follow-up questions. Additional details about the DEED program are available here.