The predictive assessment of the optimum replacement time due to the corrosion damage accumulation allows
the control of associated operating expenditures in oilfield operations. To such purpose a model was developed
for the assessment of CO2 corrosion related damage accumulation.
The model proposed is based on the calculation of the CO2 corrosion damage accumulation and failure
probability as a sequence of random processes. The distribution of failure likelihood is then assessed in terms of a
Markov random process from both an Ergodic, or repair, and an Absorbent, or unhindered damage accumulation,
scenarios.
The value of the optimum replacement time estimated by the model proposed was compared with an actual
documented case from a mature oilfield development. The modelled results indicate that the mean transition time
for failure without mitigation is around 6 years which coincide with the observed behaviour from the field. The
modelled optimum replacement time interval for this case was estimated as 5 years of service time.
Keywords: CO2 corrosion, stochastical models, localized corrosion failure, reliability prediction