Abstract:
Type 11 hot corrosion has been identified as a major life limiting factor of gas turbine
components in the topping cycle of coal fired combined cycle power plant. Impurities in the
coal combustion gases provide the environmental contaminants necessary for type 11 hot
corrosion to occur.
It is the purpose of the present study to develop corrosion lifting models such that
corrosion rates and thus component lives in coal fired combined cycle plant gas turbines may
be accurately predicted thus minimising efficiency losses and plant downtime due to
corrosion related problems.
Type 11 hot corrosion has been shown to follow bi11lodal distributions which cannot
be modelled using the well known mathematical models. It has been shown that a
probabilistic approach to modelling is appropriate and that the Gumbel Type I extreme value
model of maxima can be used to model the maximum extreme corrosion data This is
appropriate as it is the maximum extreme corrosion which in life limiting in the plant gas
turbine.
Basic corrosion data has been generated through a series of laboratory hot corrosion
tests designed to simulate the ambient conditions within the plant gas turbine. The variables
having most influence on the corrosion process have been identified as ; temperature, thermal
cycling, alkali (Na + K) metal sulphate deposition rate, S02 and HCl in the ambient
atmosphere.
The corrosion models have been developed from this data which accurately predict
the type 11 hot corrosion rates observed in the coal fired gas turbine of a combined cycle
power plant .
