Monitoring of heat losses

The main heat losses in a boiler consist of flue gas loss, incomplete combustion loss and imperfect combustion loss, as well as convection and radiation loss. The latter does not depend on the operational conditions, and the imperfect combustion loss in modern boilers is negligible.



Systematic monitoring of the outlet loss of a pulverised fuel boiler
Monitoring of the outlet loss is performed by tracking the working point in a graph with independent coordinates (oxygen in flue gas and flue gas outlet temperature). The graph curves are calculated accordingly to the current chemical composition of flue gas and the moisture content. Monitoring of this biggest heat loss enables the boiler efficiency to be monitored on an ongoing basis. The operators may change the settings of the oxygen set value if it is deemed advisable and practicable.

The second biggest heat loss in the boiler is the incomplete carbon loss that arises from the remains of unburned flammable matter in the form of fly ash and slag. The content of flammable matter in fly ash is not measured by direct methods and therefore its value is laboratory-determined. This is a serious disadvantage as the consequent delay prevents it from being monitored online.



Systematic monitoring of flammable matter in fly ash
In order to eradicate this drawback, a model has been applied in which characteristics are determined with an artificial neural network. As a result, we can obtain a graph of the flammable matter content in fly ash with a time resolution of 10 seconds. An example of the graph is presented in the picture.

Due to changes in the facility properties such as wear and tear of mill pulverising components, periodic fuel changes and other such examples, the other feature of a neural model - which is its continuous "learning" - becomes valuable. The possibility of verifying model properties by means of continuous confrontation with laboratory tests is utilized here. In this case there is no requirement for tests to be performed on a daily basis.



The results of the operations of the trained neural model that determines the carbon content in fly ash are presented in the picture. These laboratory test results are then compared with the results estimated for the times when fly ash was sampled.



Flammable matter in fly ash - a comparison of laboratory test results with the values that have been estimated with the neural model





download brochure "Monitoring of Heat Losses" (0,2 MB)