An Evaluation of Production Casing Corrosion Due to Ingress of Sour Gas in the Tubing-Casing Annulus

Current technologies for assessing corrosion damage in downhole tubing and casing strings have several limitations. Under certain conditions, mechanical, electromagnetic and ultrasonic tools can be run inside a downhole tubing string to quantify corrosion and wall loss in that string. But these tools, at best, may only be able to qualitatively assess the condition of tubing strings which are in contact or close proximity to the tool. In wells that develop communication between the production/injection tubing and casing and allow ingress of potentially corrosive fluids into the annulus, the ability to effectively assess the condition of the production casing is important. This knowledge can drive critical decisions around well operating limits, surveillance programmes, workovers, or abandonment operations. This paper describes the results of corrosion modelling and testing conducted on carbon steel to understand the extent of internal corrosion damage expected on a production casing string when sour gas enters the tubing-casing annulus through a leak source. A wide range of conditions including various hydrogen sulfide (H2S) and carbon dioxide (CO2) concentrations were modelled using proprietary corrosion modelling software. Laboratory tests on corrosion coupons were also performed and compared to the model results. Key findings around the expected corrosion potential of production casing exposed to sour gas include: • Corrosion rates are generally low over a wide range of H2S concentrations. The presence of H2S reduces the general corrosion rate by forming a protective iron sulfide (FeS) scale. • Corrosion rates are sensitive to the chemical composition of the water in the annulus. Higher bicarbonates levels significantly reduce corrosion rates. • General corrosion rates in a sweet gas environment with CO2 can be very high because of the discontinuous nature of iron carbonate scale formed at test conditions. This case study demonstrates how corrosion modelling can be used with laboratory testing to provide reliable insight about the condition of tubulars which cannot be directly measured.