Rural Co-ops Turn to AI to Optimize Boiler Efficiency
- Aug 30, 2019 1:07 pm GMT
- 736 views
Purdue University is working with the National Rural Electric Cooperative Association, Great River Energy and Pacific Northwest National Laboratory on a project to develop resources and tools that will allow utilities to determine the costs of operating their large coal boilers at reduced capacity.
The project team is using artificial intelligence to examine coal-driven power plants that have been called on in recent years to operate at partial load, resulting in a loss of efficiency.
The per unit operating cost has increased at coal-burning plants because they aren’t operating at peak efficiency.
“If you tune your car for 65 mph and then run at 30 mph, you’ll only get some fuel efficiency,” one researcher says. “With this project we’re saying let’s tune it for 30 mph and then your efficiency will be higher.”
The goal of the AI research is to create a data-driven model showing the best way to run the partial load boilers at peak efficiency. The resulting software will be provided to the 800 rural cooperatives that are members of National Rural Electric Cooperative Association.
NRECA is a partner in the Department of Energy’s Grid Resilience and Intelligence Project (GRIP), which is advancing the use of AI.
For example, one rural electric co-op, Powder River Energy Corp. in Wyoming, is experimenting with using neural networks to improve load forecasting. That work is still in its infancy and the co-op is not using neural networks at present, in part because some federal and state agencies are unfamiliar with the approach.
The Purdue team is receiving $1.2 million for its work. The project, which is sponsored by the U.S. Department of Energy, received $2.5 million in funds overall and is expected to take a little less than two and a half years to complete.
The team will start receiving data from a North Dakota coal power plant in October.
As part of the project, the research team can’t use complex simulations that involve supercomputers and complicated equations. Instead, research work must be done with the same data the coal-burning power plant operators already have.
The research results also may be used to adjust boiler efficiency to track with the load, allowing the efficiency to adjust to an increase in load.
The computer model for the boilers must be ready in the first year as part of the first phase of the project. Purdue’s role in the project’s second phase will be deployment of the model, and software and troubleshooting.