Demonstration of Heat Rate Increase for a Coal-Fired Boiler Utilizing Novel In-Situ Combustion Sensor Technology in Combination with Neural Net Optimization

75542-Currently, coal-fired utility boilers are poorly-controlled devices that could benefit greatly from closed-loop feedback control for combustion optimization. Newly developed neural net software addresses part of the systems¿ need for closed-loop control; however, the sensors that currently supply the neural network with data are located well downstream of the boiler and are often extractive. This project will develop new, in situ sensor technology that utilizes recent advances in diode laser and fiber-optic technology to provide more effective operation of the neural net. The new sensors will be based upon wavelength-multiplexed, tunable diode laser spectroscopy and will be able to measure O2, CO, and H2O species concentrations and temperature directly in the fireball in multiple locations. In Phase I, quantitative spectroscopy will be performed in order to enable the quantification of species concentration and temperature. Optical and mechanical engineering tasks will be conducted to optimize the wavelength multiplexer design. Mating hardware will be developed to mount launch and receiver optics to the boiler. Finally, processing techniques will be developed to optimize the data handshake between the sensor and the neural net. During Phase II, the new sensor technology will be combined with neural net optimization to enable efficiency closed-loop control. Commercial Applications and Other Benefits as described by the awardee: Adoption of the new combined sensor/neural net technology would save over $4 billion/year in reduced coal costs and reduced NOx emissions for the U.S. coal-fired power generation industry. The technology also should be applicable to the optimization of gas turbine power generation and to gas turbine engines for aero-propulsion.