Application of Data Validation and Reconciliation to Improve Measurement Results in the Determination Process of Emission Characteristics in Co-Combustion of Sewage Sludge with Coal

Abstract

One of the actions popularized worldwide to reduce the consumption of fossil fuels is the combustion of renewable fuels and the co-combustion of both of these fuels. To properly implement combustion and co-combustion processes in power-generation installations, operational characteristics, including emission characteristics are required. To determine these characteristics, tests must be conducted, within the scope of which, for individual operating stages of the installation’s work, the readings collected from a relatively large number of control and measurement instruments should be taken into account. All these instruments have different levels of accuracy, which, among other factors, bring about lower adequacy of the characteristics determined on the basis of these measurements. The objective of this study is to present possible adaptations of data validation and reconciliation methods to increase the adequacy of emission characteristics for the process of co-combustion of fuels. The methodology is discussed based on the example of studies on the co-combustion process of sewage sludge with coal in a grate furnace. The aforementioned characteristics were determined based on measurement tests of gaseous emissions of flue gas components. The tests were carried out for various preset operational conditions of the process, such as the thickness of fuel layer on the grate, the share of sludge in the fuel, the humidity of the sludge, the theoretical ratio of excess air to combustion, and the distribution of air stream during the process. The research object is described and detailed research results concerning two exemplary measurement tests are given, as well as the most important results referring to the whole research. The performed calculations indicate the necessity to take into account often significant corrections, which can amount to about 10% of the measured value.