United StatesEnergy and utilities2020 – ongoingVistra
Neural-network heat-rate optimizer for power plants
Vistra's Martin Lake plant improved efficiency by more than 2%, according to McKinsey
Problem
Power-plant operators adjusted many interacting settings to reduce fuel consumption. Weather and operating conditions complicated those decisions.
Approach
Vistra and McKinsey built a neural-network heat-rate optimizer using historical plant data. Martin Lake deployed it in October 2020. It generated operating recommendations every 30 minutes.
Results
Takeaway
Operators acted on recommendations rather than handing plant control to the model.
Categories
- Industry
- Energy and utilities
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and MLOptimization
- Adoption
- Partner co-development
- Evidence
- Vendor case study
Sources
Source summaries
McKinsey described joint model development and deployment at Martin Lake. It reported the efficiency gain and annual savings, followed by deployment at additional generating units.
Full details are in the original sources.