CO2 corrosion is observed in the oil and gas industry. It is becoming an increasing problem in the
power industry due to the demand for CO2 capture. The research community has had an increased focus
on the specific problem in order to develop predictive models for forecasting process equipment
corrosion rates.
Several models have been developed since 1970. Mechanistic models were developed through the
1990's. During 2001 a mechanistic model was published by Nesic and co-workers. It incorporates a core
electrochemical model coupled to a diffusion and a bulk equilibrium model.
Diffusion theory developed by Nernst-Planck is being used in the corrosion models. Unfortunately
this theory is only applicable to diluted ideal solutions of low ionic strength and may not be applied to
higher concentrations which are found in wet gas pipelines.
We seek to improve corrosion models by extending the above theory to non-ideal realistic systems.
The drawback and limitation of current models are presented, the theory is extended, and examples will
be shown how to improve the theory to high ionic strength solutions. Calculations will show how large
deviations are generated by assuming non-ideal solutions to be ideal. The theory is general and
applicable to other electrolyte diffusion problems.
Keywords: CO2 corrosion models, wet gas pipelines, Mixed solvent electrolyte diffusion, extension
of Nernst-Planck equation, activity coefficient, extended UNIQUAC, thermodynamic factor,
thermodynamic correction factor, glycol