289206 Real Time Performance Improvement for a Coal-Based Thermal Power Plant Using Artificial Intelligence Techniques

Monday, October 29, 2012: 9:35 AM
303 (Convention Center )
Bhaskar D. Kulkarni, Chemical Engineering & Process Development Division, National Chemical Laboratory, Pune, India and Sanjeev Tambe, National Chemical Laboratory, Pune, India

Real time performance improvement for a coal-based thermal power plant using Artificial Intelligence techniques

Himanshu Pant§, Jatinder Singh Chandok§, Sanjeev S. Tambe# and Bhaskar D. Kulkarni#*


§NTPC Energy Technology Research Alliance, National Thermal Power Corporation, Greater Noida, U.P, 201307, India

#Chemical Engineering and Process Development Division, National Chemical Laboratory, Pune 411008, Maharashtra, India

Abstract: Thermal power plant processes are complex, nonlinear and dynamic which behave differently depending upon variations in the operating and ambient conditions. The complexity aggravates due to the factors such as variations in the quality and ash content of the coal, as also in the load demand, thus  posing significant challenges to maintain the design efficiency of the plant.

Online plant optimization systems can address the said challenges by simultaneous evaluation of parameters resulting in the real-time heat rate improvement through an effective plant operation. The real time performance improvement can be better achieved through data-driven modeling and optimization using Artificial Intelligence (AI) techniques. Accordingly, an adaptive AI-based modeling and optimization system, based on specific plant data from a 500 MW power station has been developed which in real-time provides an advisory to the plant operator for certain controllable boiler parameters viz. excess air, burner tilt, and air distribution. The objective of the system is to maximize boiler efficiency in real time while keeping steam parameters and sprays under feasible zone thus improving the overall unit heat rate. Various AI-based nonlinear modeling and optimization approaches such as artificial neural networks and genetic algorithms have been employed for tasks such as online and offline process modeling, process parameter optimization  and data reconciliation.

Extended Abstract: File Uploaded
See more of this Session: Memorial Session in Honor of Prof. LK Doraiswamy
See more of this Group/Topical: Catalysis and Reaction Engineering Division