This paper presents generic methods for verifying online monitoring systems associated with coal-fired power plants. It is applicable to any on-line system. The methods fundamentally recognize that if coal-fired unite are to be understood, that system stoichiometrics must be understood in real-time, this implies that fuel chemistry must be understood in real-time. No accurate boiler efficiency can be determined without fuel chemistry, heating value and boundary conditions. From such fundamentals, four specific techniques are described, all based on an understanding (or not) of real-time system stoichiometrics. The specific techniques include: 1) comparing a computed ambient relative humidity which satisfies system stoichiometrics, to a directly measured value; 2) comparing a computed water/steam soot blowing flow which satisfies system stoichiometrics, to a directly measured value; 3) comparing computed Energy or Flow Compensators (based on computed boiler efficiency, heating value, etc.), to the unit’s DCS values; and 4) comparing a computed fuel flow rate, based on boiler efficiency, to the plant’s indication of fuel flow. Although developed using the Input/Loss Method, the presented methods can be applied to any online monitoring system such that verification of computed results can be had in real-time. If results agree with measured values, within defined error bands, the system is said to be understood and verified; from this, heat rate improvement will follow. This work has demonstrated that use of ambient relative humidity is a viable verification tool. Given its influence on system stoichiometrics, use of relative humidity immediately suggests that effluent (Stack) flow can be verified against an independently measured parameter which has nothing to do with coal-fired combustion per se. Whether an understanding of coal-fired combustion is believed to be in-hand, or not, use of relative humidity (and, indeed, soot blowing flow) provides the means for verifying the actual and absolute carbon and sulfur emission mass flow rates. Such knowledge should prove useful given emission taxes or an imposed cap and trade system. Of the four methods examined, success was not universal; notably any use of plant indicated fuel flow (as would be expected) must be employed with caution. Although applicable to any system, the Input/Loss Method was used for development of these methods. Input/Loss is a unique process which allows for complete understanding of a coal-fired power plant through explicit determinations of fuel chemistry including fuel water and mineral matter, fuel heating (calorific) value, As-Fired fuel flow, effluent flow, boiler efficiency and system heat rate. Input consists of routine plant data and any parameter which effects stoichiometrics, typically: effluent CO2, O2 and, generally, effluent H2O. The base technology of the Input/Loss Method has been documented in companion ASME papers, Parts I thru IV, which addressed topics of base formulations, benchmarking fuel chemistry calculations, high accuracy boiler efficiency methods and correcting instrumentation errors in those terms affecting system stoichiometric (e.g., CEMS and other data).
Neuro-fuzzy Modeling of the Degree of Efficiency of Transformers at Ship Electric Power Plants
