Online monitoring of flue gas emissions in power plants having multiple fuels

Abstract This paper introduces an online monitoring application for flue gas emission measurements. The monitoring is conducted by analytical redundancy by estimating the monitored measurement variables. The estimated variables are CO 2 , H 2 O, flue gas flow and combustion air flow. Additionally, SO 2 content in the flue gas can be estimated with certain limitations. The monitoring method is based on physical combustion modeling and is therefore boiler structure and fuel independent. The model is valid for multiple fuels as long as the fuel flow measurements, fuel properties and flue gas O 2 measurement are available. The estimates can further be improved by flue gas CO measurement. The monitoring method was successfully tested in an industrial wood, peat, bark and slurry fired power plant. The results verify that the method is able to automatically separate sensor faults and process disturbances.