A Probabilistic Approach to Prediction of CO₂-Corrosion, and Its Application to Life Cycle Cost Analyses of Oil and Gas Equipment Available to Purchase

Unduly conservative predictions of CO₂-corrosion may prevent cost-effective solutions for corrosion control. Similarly, field experience has demonstrated that over-optimistic assumptions of the efficiency of planned corrosion mitigation may cause excessive costs not anticipated during the initial design; including monitoring, maintenance, repair and early condemnation.

This paper propounds a probabilistic approach to the design of CO₂-corrosion control where both the prediction of CO₂-corrosion rate and inhibitor efficiency factors are treated as stochastic parameters, i.e. their reliabilities are assessed in quantitative terms. Such an approach allows the probability of achieving efficient corrosion control to be calculated for the specific corrosion modes which are relevant (pitting, preferential weld corrosion, longitudinal grooving and uniform corrosion). When the estimated cost consequences for the associated failure modes (leakage, rupture, collapse, etc.) are combined with the results of the probabilistic calculations, the "Total Risk Cost" associated with the use of C-steel can be included in a Life Cycle Cost Analysis, together with costs for investment and the scheduled operational costs for corrosion control. The penalties and benefits of increasing the wall thickness for C-steel, or the alternative use of CRA’s, can thus be compared in monetary units. Two examples of subsea pipeline cases show how a probabilistic evaluation may result in a different corrosion control strategy than that derived by a traditional deterministic treatment, even though both approaches use the same "nominal" CO₂- corrosion rate predictions and inhibitor treatment efficiencies.