The primary impediment to sootblowing optimization hasbeen the difficulty associated with modeling the relationshipbetween sootblowing and boiler efficiency. New advances in neuralnetworking technology now provide an attractive approach toaddress this issue. This project involves a plan to develop anddemonstrate an optimized sootblowing scheduler and closed-loopcontroller for coal-fired boilers. The intelligent controller foroptimized sootblowing (ICOS) will incorporate a neural networkprocess model and optimization algorithm to provide closed-loop,optimized control of steam or compressed air sootblowers forfossil utility boilers. ICOS will address four key sub-problemsto provide optimized sootblowing: (1) calculation of sootblowingperformance measures (i.e., boiler cleanliness factors) usingreadily available process instrumentation, (2) modeling theboiler and the relationship between operator controllableparameters and efficiency using neural networks, (3) calculationof the steady state optimal setpoints for the boiler, and (4)determination of the optimal sootblowing schedule to achieveoptimal boiler performance. In Phase I the optimization approachwill be tested and validated. Phase I will quantify the expectedsavings for the controller and will verify the effectiveness ofthe planned technical approach. In Phase II, the controlalgorithm will be incorporated into a personal computer (PC)based software application, interfaced with a sootblowingcontroller, and demonstrated and tested for closed-loop, optimalsootblower control. The savings achieved through use of the ICOScontroller during testing will also be quantified.Anticipated Results /Potential Commercial Applications as described by the awardee: Efficiency losses of over 200 BTU/KWHhave been attributed to sub-optimal control of sootblowers incoal-fired boilers, frequently accounting for over 80% of thecontrollable losses in a coal-fired boiler. For a 500 MW powerplant, this translates into yearly savings of over $1M. Becauseof the increased use of digital control systems (DCS) in thepower industry and the relative ease of interfacing DCS toPC-based controllers, it is expected that the installed cost ofthe Phase II ICOS product will be under $50,000, providing a veryfavorable cost justification for the product.