A Corrective Action Program (CAP) is the systematic process by which a nuclear organisation identifies, evaluates, and resolves conditions adverse to quality (CAQs). It is required under NQA-1 Requirement 16 and 10 CFR 50 Appendix B Criterion XVI. But the CAP is more than a compliance mechanism, it is the feedback loop that tells an organisation whether its quality program is working. A well-functioning CAP surfaces systemic weaknesses before they become regulatory findings. A poorly functioning one allows the same problems to recur indefinitely.
What the CAP must do
At minimum, a nuclear CAP must accomplish four things. First, it must ensure that conditions adverse to quality are identified and documented, which sounds obvious but requires a culture where personnel report issues rather than resolve them informally. Second, it must evaluate the significance of each identified condition, distinguishing between minor issues and significant conditions adverse to quality (SCAQs), which require more rigorous treatment. Third, for significant conditions, it must determine the root cause, not the surface symptom, but the underlying reason the condition occurred. Fourth, it must implement and verify corrective actions that prevent recurrence.
Significant vs. minor conditions: NQA-1 draws an important distinction. Minor conditions adverse to quality may be corrected through normal work controls without formal root cause analysis. Significant conditions require documented root cause analysis and corrective actions specifically designed to prevent recurrence. The CAP must have a defined process for making and documenting this categorisation.
Root cause analysis
Root cause analysis (RCA) is the element of the CAP that regulators scrutinise most carefully, and where many programs struggle. An RCA that identifies the root cause as "personnel error" or "failure to follow procedure" is rarely adequate. These are descriptions of the failure mode, not its cause. A genuine root cause analysis asks why the error occurred: Was the procedure unclear? Was training inadequate? Was there schedule pressure that created an incentive to skip steps? Was the process designed in a way that made errors easy to make?
The NRC and CNSC expect root cause analyses to be proportionate to the significance of the condition, a minor documentation error does not require a full HPES (Human Performance Evaluation System) analysis. But for significant conditions, the analysis should be thorough enough that the corrective actions actually address the cause, not just the symptom.
Programs with high rates of repeat findings, where the same issue appears in multiple inspection cycles, often have a root cause analysis problem: the RCAs are being closed with generic actions that don't address the actual reason for the failure.
Trending and programme health
A mature CAP does more than resolve individual issues, it analyses patterns across issues to identify systemic weaknesses. Trend analysis should be examining whether certain types of conditions are appearing repeatedly, whether certain areas of the program or certain suppliers generate disproportionate numbers of CAP entries, and whether previously closed corrective actions have actually been effective.
This trending function is where the CAP intersects most directly with the NCR program. NCRs that close into the CAP, supplier-generated issues, audit findings, and self-identified conditions should all feed into the same trending analysis. An organisation that trends these in isolation, NCRs in one system, CAP items in another, audit findings in a third, cannot see the patterns that a unified dataset would reveal.
Regulatory guidance from the NRC (NUREG-1649) and CNSC (REGDOC-2.1.1) both identify CAP trending as a key indicator of safety culture maturity. Organisations that can demonstrate they are finding their own problems before the regulator does, and acting on trend data, are treated materially differently from those that wait for inspection findings.
Common CAP failures and how regulators find them
NRC enforcement data and CNSC inspection reports consistently cite CAP deficiencies in the same patterns. Aged-open items, corrective actions that remain open well past their committed resolution dates, are the most common. They signal either that corrective actions were not resourced, or that the program lacks the oversight mechanisms to flag approaching deadlines. Inadequate root cause analyses show up as repeat findings: the same condition reappears in the next inspection cycle because the corrective action addressed the symptom, not the cause. Inadequate issue identification rates, a CAP database that is suspiciously thin relative to the scale and complexity of the program, suggest that personnel are not using the CAP as intended, resolving issues informally rather than through the controlled process.
The corrective action program is, in many ways, the most revealing indicator of safety culture. A program where people identify issues proactively, evaluate them rigorously, and close them reliably is a program where the culture supports the quality system. A program where issues are under-reported, evaluations are perfunctory, and items age out is a program where the culture is working against it.
Forged Operations tracks CAP items end-to-end, initiation, significance categorisation, root cause linkage, corrective action assignment, due date monitoring, and effectiveness reviews. AI surfaces ageing items, identifies repeat patterns, and assembles programme health data before the next inspection arrives.
References
- American Society of Mechanical Engineers. ASME NQA-1-2022: Quality Assurance Requirements for Nuclear Facility Applications, Requirement 16 — Corrective Action. New York: ASME, 2022.
- U.S. Nuclear Regulatory Commission. "Criterion XVI — Corrective Action." Code of Federal Regulations, 10 CFR 50 Appendix B. Washington, D.C.: NRC.
- U.S. Nuclear Regulatory Commission. NUREG-1649: Reactor Oversight Process, Revision 6. Washington, D.C.: NRC.
- Canadian Nuclear Safety Commission. REGDOC-2.1.1: Management System. Ottawa: CNSC, 2019.
- CSA Group. CSA N286:12 (R2018): Management System Requirements for Nuclear Facilities. Toronto, Ontario: CSA Group.