Methodology for the analysis of corrosion failures in the maintenance of industrial plants

Development of an advanced methodology for the analysis of
corrosion problems, integrated into a pc software tool,
using knowledge-based system (kbs) technology.

Corrosion-related cracking phenomena are the most insidious forms of failure that can be experienced by process equipment, because they tend to strike without warning in a variety of conditions. At this moment, the accurate prediction of corrosion-related cracking - which would allow improved maintenance and design practices - is essentially non-existent because satisfactory engineering models have not been developed for any forms of corrosion cracking. As a result, the approach developed result of the identification and analysis of parameters is the most important in the corrosion phenomena under study. If left uncontrolled, this corrosion can cause costly equipment replacement, loss of revenue from system shutdown, hazardous leaks causing major safety issues and environmental contamination. The use of predictive tools would permit the accurate pinpointing of cracking-sensitive areas and would enable better inspection management, maintenance, and corrective actions to be taken. METHOD will be developed as an advanced failure retrieval and analysis component that will serve as an aid to determine the type of corrosion, provide appropriate advice on remedial measures (maintenance), and contain a library of case histories (including used remedies) for comparison with the present situation (problem). Therefore, the main objective of this project is the development of an 'intelligent' computerised support tool that will give support in: (1) safety analysis in operating plants: When something has been built and the question is: what is the potential damage and expected risk under the actual process conditions and observed (rates of) corrosion, e.g. as revealed during intermediate inspection or maintenance. This would help for example in setting the appropriate process parameters, as well as in 'predictive maintenance' or scheduling of inspections; (2) failure and damage analysis: When the equipment has deteriorated and requires maintenance, and the questions are: what is the failure mode and how can the problem be overcome. This specific tool to be developed is supported by the work methodology of a corrosion engineer which consists of the following steps: 1. failure/damage description 2. Details about the procedure of taking specimens (flow chart part 1 and 2) 3. Input of data of a present case into the database. Comparison of present case with available similar cases (Atlas of corrosion case studies). 4. Querying relevant data records in the case histories database, using a data filter, predefined by the parameters of the present case. The end-user should be able to modify the filter. 5. Examination of the relevant data sets manually. 6. Comparison of the present case with available background information (could also be executed before 5). 7. Analysis by neural network. 8. Intelligent flow-chart for problem analysis (flow chart part 3). 9. Drawing up of a report: * Short introduction: how and when the corrosion engineer was involved. * Facts which were discovered in an on-site investigation (flow chart 1) * Additional investigations (flow chart 2) * Interpretation, essential assessment of failure/ damage cause. References to other cases or to literature to prove the validity of the corrosion engineer's assessment. * Summary (concerning the failure analysis). 10. Recommendations for remedies.
Project ID: 
1 746
Start date: 
Project Duration: 
Project costs: 
2 100 000.00€
Technological Area: 
Market Area: 

Raising the productivity and competitiveness of European businesses through technology. Boosting national economies on the international market, and strengthening the basis for sustainable prosperity and employment.