The expert system for fouling assessment development is one of the pos
sible approachs to the mitigation of deposit formation on boiler heat
transfer surfaces. This paper presents the evaluation of the diagnosti
c variables and their domain based on the criteria for the fouling ass
essment. In agreement with the boiler furnace design, respective field
s for the deposit control are selected to be instrumented with the 'cl
ean' radiation heat flux and 'not-cleaned' radiation heat flux for the
on-line reading. The ratio of the 'clean' and 'not-clean' radiation h
eat flux is proved to be a sensitive diagnostic variable for the depos
it thickness. The rate of change of radiation heat flux ratio, togethe
r with the efficiency of the heat transfer surface, are diagnostic var
iables used for the design of the expert system knowledge base. The kn
owledge base structure is based on the object oriented concept. In thi
s respect the FOULING object is defined. Its structure is composed of
the respective attributes and values linked with the number of situati
ons to be selected for the assessment. The examples given are related
to a 300 MWe power-plant boiler furnace. It includes the number of sit
uations reflecting different deposit formation situations: total and u
pper level of side surface of fouling deposit. Copyright (C) 1996 Else
vier Science Ltd.