A range of parameters must be considered carefully when modelling AC corrosion. Prediction of induced AC voltage profiles along pipelines due to shared right-of-way with high-voltage power lines has been practiced for decades. Modelling of the cathodic protection level on a pipeline resulting from types, position, and current output from CP sources, pipeline dimensions, coating conditions, soil conditions, isolation, etc., has also been implemented during several years. Computer aided prediction of corrosion rates caused by induced AC as a function of AC and DC current densities, coating fault geometry and thickness, soil resistivity, etc. has been attempted only in recent years. The complexity of the AC corrosion process calls for careful and critical evaluation of the computed results, and their practical applicability.
The present paper presents and discusses the various components contained in an electrical equivalent circuit describing the AC corrosion process from a computer modelling perspective. The effect of the coating defect size and geometry on spread resistance and resulting AC current density, the effect of the kinetics of electrochemical reactions relevant for the corrosion process, the effect of diffusion and diffusion coefficients for active chemical species, as well as the impact of the capacitive effect of the electrochemical double layer as a short circuit of the electrochemical processes, soil chemistry, texture and soil resistivity are all aspects that influences the AC corrosion process and therefore the reliability of a computer model. These aspects will be discussed together with the sensibility of a model and the risk of generating inaccurate results due to missing or erroneous inputs. In addition, different model approaches will be discussed and sustained through examples.
The utilization of mathematical modelling in cathodic protection design as well as interference problem solving has proved to be of great value.1,2 The success of the modelling approach relies on correct initial assumptions of the processes occurring in the specific scenario, the use of correct parameters and magnitudes, and a proper mathematical modelling tool. The modelling cannot be a standalone tool as the results must be validated against real observations and measurements on the system under investigation. Mathematical modelling has also been attempted in relation to AC corrosion, which is a complex scenario in terms of the chemistry and physics involved.3 Nevertheless, a successful model would be beneficial in terms of the understanding of AC corrosion, risk assessment and mitigation. Due to the high level of complexity of the problem, a careful analysis of the basis for the modelling is a clear pre-requisite, and a meaningful way of validating the model results should be defined.