Observer-based fault detection and moisture estimating in coal mills

In this paper an observer-based method for detecting faults and estimating moisture content in the coal in coal mills is presented. Handling of faults and operation under special conditions, such as high moisture content in the coal, are of growing importance due to the increasing requirements to the general performance of power plants. Detection of faults and moisture content estimation are consequently of high interest in the handling of the problems caused by faults and moisture content. The coal flow out of the mill is the obvious variable to monitor, when detecting non-intended drops in the coal flow out of the coal mill. However, this variable is not measurable. Another estimated variable is the moisture content, which is only “measurable” during steady-state operations of the coal mill. Instead, this paper suggests a method where these unknown variables are estimated based on a simple energy balance model of the coal mill. In the proposed scheme an optimal unknown input observer is designed based on the energy balance model. The designed observer is applied on two data sets covering variating moisture content as well as a data set including a fault in the coal mill. From these experiments it can be concluded that the moisture content is successfully estimated and that the fault is detected as soon as it emerges.