Cathodic Disbondment Failure After Immersion Service Startup
A new carbon steel tank bottom is internally coated with a thin-film epoxy for immersion service and is also protected by an impressed current cathodic protection (ICCP) system. The coating was holiday-tested and passed. Three months after startup, the operator reports blistering and disbondment radiating from several small defects near the sump. The CP system is operating at a pipe-to-soil (structure-to-electrolyte) potential that is more negative than the design setpoint. The coating data sheet lists a cathodic disbondment resistance value but the project specification did not require a cathodic disbondment test on the applied coating. The inspector is asked to determine whether the coating failed or the CP system is at fault.
1. What is the likely mechanism, and how should the inspector frame the investigation?
Read the worked answer
The pattern (disbondment radiating from small defects in an immersed, cathodically protected tank) is classic cathodic disbondment. At a coating holiday, the exposed steel becomes the cathode; the cathodic reaction generates hydroxyl ions at the steel surface, raising pH locally and attacking the coating-to-steel bond. The disbondment front advances outward from the defect. Blistering near the sump is consistent with this, especially where electrolyte is present.
The inspector should not simply blame the coating or the CP system. The investigation should establish:
- Is the potential within the design range? The report says the potential is more negative than the setpoint. Over-protection (excessively negative potential) accelerates cathodic disbondment by generating more hydroxyl and hydrogen. The inspector should obtain CP survey data (structure-to-electrolyte potentials, current output, anode condition) and compare to the design criteria.
- Was the coating qualified for CD? The data sheet value is a material property, but the specification did not require a CD test on the applied coating. The inspector should check whether the specified coating system was the one applied and whether the data sheet CD value meets the service condition.
- Defect population. Holiday testing passed, but small defects can be missed or can develop. The inspector should map the disbondment and confirm the defects are true holidays vs. mechanical damage.
- Electrolyte chemistry and temperature, which affect CD rate.
Frame the conclusion as a system issue: the CP potential, the coating's CD resistance, and the defect population interact. The inspector documents the evidence and recommends a joint review with the CP and coating engineers.
2. What should the inspector recommend regarding the CP setpoint and the coating repair?
Read the worked answer
CP setpoint: The inspector should recommend that the CP engineer review and correct the over-protection. The potential should be brought within the design criteria (typically the applicable criterion for the structure and electrolyte, e.g., a protected potential per the CP standard) rather than left more negative than setpoint. The inspector should not adjust the CP system personally unless authorized; the recommendation goes to the responsible engineer. CP survey records should be re-established and trended.
Coating repair: The disbonded coating must be removed back to sound, well-bonded coating (the disbondment front often extends beyond the visible blister), the steel prepared to the specified surface profile, and the coating repaired per the specification. The repair must be holiday-tested after cure. Because the original specification omitted a CD test requirement, the inspector should recommend that the repair coating (and any future lining) be qualified by a cathodic disbondment test appropriate to the service, and that the specification be updated to require it.
Documentation: Record the failure mode, the CP data, the repair procedure, and the specification gap. The root-cause conclusion should state that over-protection plus a coating with marginal CD resistance at defects drove the failure, and that both the CP operation and the specification need correction.