Power Plant production optimization
The power industry needs to both run operations as close as possible to optimum but also minimize the environmental footprint as much as possible. The production of power varies with variation in fuel, in this case municipal waste. The challenge is to maintain a stable high output. The estimated benefit was a 5% improvement.
OPPORTUNITIES & CHALLENGES
- Goal: adapt settings to variations in fuel properties so as to maintain a high, even output
- Continuous replanning of optimum production needed but constrained by delayed effect
- Plant is controlled by user-level settings and a fuzzy-logic layer that needed to be taken into account
- Fuel varies significantly in moisture and energy content
Product: SentianController
Industrial processes can be hard to optimize due to the many variables and complex relationships between the variables. This is normally handled by very experienced operators. Sometimes an advanced process control (APC) is added as in this case to support the operators. AI is a new generation of technology to further improve the optimization. It has the benefit to be able to handle very complicated processes and to learn form the data to continuously improve.
Features:
- System trained to optimize output while honoring key constraints:
- Low output variance
- “Sticky” settings that do not change too often for process operators
- Long-term emission controls
The estimated benefit was a 5% production improvement and more stable production.