Comparisons of laboratory-derived gaseous corrosion data with observations from industrial plant

Corrosive attack of metallic components exposed in high-temperature process atmospheres is well known but the factors governing such degradation are often less well understood. The nature and extent of attack is, of course, influenced by specific operating conditions such as temperature, composition of the atmosphere, flow rates, thermal cycling, etc. Radical differences exist between the various plant atmospheres and therefore in carrying out fundamental laboratory studies it is often unrealistic simply to copy a specific process atmosphere. A wider and more meaningful approach is preferred in which the environments for a given type of process are first characterised in terms of the activities of the reacting species, ( pO 2, pS 2, a c etc.) and then reproduced in a matrix of controlled experiments. A vital step in validating such studies is achieved by exposing identically prepared specimens in plant situations and comparing the results with the laboratory-derived data. This paper presents test results obtained in laboratory-based H 2COH 2OH 2S-containing gas mixtures having a low oxygen potential and varying sulphur activity which model the types of environment likely in coal conversion plant involving fluidised bed combustion or gasification. An example is given of observations made on laboratory-prepared samples exposed in a real plant situation and a brief reference is also made to a screening test developed for characterising plant atmospheres using an appropriate range of pure metals and compounds.