Modelling of Microbiological Influenced Corrosion – Limitations and Perspectives

Microbiologically influenced corrosion (MIC) research in the oil and gas industry has seen a revolution over the past decade with the increased application of molecular microbiological methods (MMM) and new industry standards; however, MIC modelling is an area that has not been fully developed. Models can provide numerous benefits, e.g., guidance on MIC mitigation selection and prioritization, identification of data gaps, a scientific basis for risk-based inspections, and technical justification for asset design and life-extension. This paper describes trends in MIC modelling; different types of models, future needs, and the utility of MIC models from an end-user perspective. Microorganisms can initiate and promote corrosion different ways, e.g., affecting both charge and mass transfer in corrosion reactions. No mechanistic models currently exist that consider the influence of multiple functional groups of microorganisms on reaction kinetics or the significance of microbial growth kinetics on corrosion. The ability to accurately predict MIC initiation and growth is hampered by knowledge gaps regarding environmental conditions affect corrosion under biofilms. In order to manage the threat of corrosion relative to asset integrity, operators commonly use models to support decision-making. The models use qualitative, semi-quantitative or quantitative measures to help predict the rate of degradation caused by MIC and other threats. A new model that links MIC in topsides oil processing systems with risk based inspection (RBI) through the application of data obtained by MMMs, and its implementation, are presented and discussed. Integrated computational materials engineering (ICME) is a promising future approach for prediction and management of MIC, using translational research to deliver new modeling tools to industry in the shortest development time. ICME development would couple our current understanding of MIC, as represented in models, with experimental data, to build a digital “twin” for optimizing performance of engineering systems, whether in the design phase or operations. Since there are still many uncertainties in MIC mechanisms and rates, and because operations occur across a range of conditions, data ranges can be employed in ICME using a probabilistic approach through a Bayesian network.