Pressure to Pass: A 35% Girth Weld Call in an HCA
An inline inspection (ILI) report lands on your desk for a liquid pipeline segment crossing a High Consequence Area (HCA).
Among the hundreds of reported anomalies, one stands out:
Girth Weld (GWD) — 35% wall loss, slotting characterization ±20–30% accuracy
Reported Location: within the heat-affected zone (HAZ) of a mainline circumferential girth weld
Previous tool runs either didn’t identify this feature or classified it as a minor surface irregularity. The difference now?
This latest run was completed using a high-resolution axial magnetic flux leakage (MFL) tool, capable of detecting narrower and more complex weld-adjacent features than earlier standard-resolution tools.
The integrity team takes notice, but not everyone is equally concerned.
Recent excavations from the same run — all pitting features — matched the tool’s sizing within about 10%. Confidence in the tool is high, and management begins to push back on another dig.
“The tool’s been accurate,” someone says. “It’s only 35%. Let’s move on.”
This is where confidence quietly turns into bias.
When Confidence Becomes Bias
Consistency in prior results often builds trust — and sometimes, overconfidence.
Those earlier digs validated pitting performance, not girth weld slotting features. That’s an important distinction:
Pitting features are well-characterized and typically produce clear magnetic signatures.
Girth weld features, however, involve signal distortion from weld geometry, lift-off variation, and magnetization effects that can understate or exaggerate true depth.
Such features can also represent grooving, grinding marks, or crack-like behavior, which behave differently than uniform corrosion.
Assuming accuracy from one feature type applies to another is a textbook example of confirmation bias — and in integrity management, it can lead to underestimated risk.
PHMSA’s Expectation: Account for Tolerance and Feature Type
Under 49 CFR §195.452(c)(1)(i)(A) and §195.416(c), PHMSA requires operators to explicitly consider uncertainties in reported results — including ILI tool tolerance — when identifying and evaluating anomalies. These uncertainties must also be reflected in engineering assessments, risk evaluations, and repair decisions.
In other words, “within tolerance” does not mean “within safety.” Reported anomaly depths are estimates within a defined error range — and PHMSA expects operators to treat that uncertainty conservatively, especially in High Consequence Areas (HCAs).
Industry standards reinforce this expectation:
API 1163 – ILI systems must be qualified and validated by feature type to demonstrate sizing accuracy for specific morphologies.
NACE SP0102 – Repair prioritization must integrate tool accuracy, uncertainty, and consequence of failure.
ASME B31G and API 579-1/ASME FFS-1 – Provide defect assessment methodologies that explicitly account for measurement error and material variability when determining remaining strength.
In practice, a 35% wall loss indication ±30% could represent up to 65% wall loss, particularly near a weld. In an HCA, that level of uncertainty demands conservative evaluation — not deferral based on optimistic assumptions.
Typical Classification: A 180-Day Condition
According to ASME B31.8S [3] and PHMSA integrity guidance [6], a 35% GWD feature with ±30% uncertainty in an HCA generally qualifies as a 180-day repair condition, depending on operating stress and wall thickness.
This classification accounts for:
Uncertainty in ILI depth sizing
Stress concentration at the weld
Consequence weighting due to HCA location
Yet, management bias — strengthened by prior successes — can sometimes override engineering caution. The danger lies not in the data, but in how it’s interpreted.
ILI Tool Accuracy and Its Role in POF
In probabilistic integrity models, Probability of Failure (POF) calculations depend directly on ILI data accuracy.
A tool’s stated performance — say, ±10% for 80% of metal-loss features for pitting or general corrosion — is a population statistic, not a guarantee for every feature.
For complex morphologies like on girth welds:
Bias (systematic over- or under-calling) and
Scatter (random error)
can be substantially higher than for pitting or general corrosion.
When incorporated into a POF model, a 30% increase in true depth can raise failure probability by an order of magnitude or more.
Accurate POF characterization means quantifying that uncertainty — not ignoring it. PHMSA expects operators to show how ILI tool accuracy and tolerance directly feed into both engineering critical assessments (ECA) and risk models.
When Similar Digs Don’t Exist
If no comparable girth weld excavations exist, API 1163 [4] allows for statistical validation:
Analyze prior weld features to determine bias and scatter between reported and field depths.
Evaluate how the tool performs near weld signal interference.
Use those findings to calibrate performance expectations and refine ECAs.
The key is documentation — demonstrating that you’ve quantified what you don’t know.
Data Integration for Defensible Decisions
Before deferring or reprioritizing, operators should consolidate data sources to better understand the feature’s context:
Weld inspection / NDE records – Tie-in or repair weld? Any signs of undercut, porosity, or grinding?
Construction documentation – Counterbores, transition welds, or thickness offsets?
CP and coating data – Any shielding or disbondment near the joint?
IMU / geometry data – Strain, misalignment, or bending stress?
Historical ILI runs – Did previous tools miss or undercall the same feature?
A thorough, multi-dataset assessment not only improves understanding — it strengthens the defensibility of any decision made.
Working With PHMSA and Industry Experts
When management skepticism meets engineering uncertainty, the best path forward is collaboration, not conflict.
PHMSA encourages early communication when uncertainty, tool limitations, or internal debate could delay or affect decision-making. Early notification shows transparency, control, and proactive management.
Engaging industry experts — including ILI vendors or independent consultants — can also help validate morphology interpretation, tool characterization, or ECA assumptions. Peer review adds credibility, builds regulatory trust, and reduces hindsight risk.
The Real Risk Question
If this 35% GWD call is an outlier, it could already be near failure.
The potential cost of being wrong far exceeds the cost of validation:
Environmental: Cleanup and remediation in an HCA.
Public: Safety and community trust.
Regulatory: Audits or Corrective Action Orders.
Operational: Downtime, pressure restrictions, and reassessments.
Integrity management isn’t just about preventing leaks — it’s about defending the decisions you make before they happen.
The Integrity Takeaway
ILI tools are generally within spec 80% of the time, but not equally across all feature types.
A 35% girth weld call in an HCA, identified by a high-resolution axial MFL tool, sits at the intersection of uncertainty, bias, and consequence.
PHMSA expects operators to:
Account for tool accuracy and tolerance,
Integrate uncertainty into both ECA and POF modeling,
Communicate early with regulators when plans are debated, and
Validate conclusions with independent or industry expert review.
Integrity isn’t about proving the tool is right — it’s about ensuring your process is.
Confidence without validation isn’t integrity — it’s assumption.
References
PHMSA, 49 CFR Part 195 – Transportation of Hazardous Liquids by Pipeline, Subpart F, §195.452 – Integrity Management Program.
PHMSA Advisory Bulletin ADB-10-05 – Clarification of terms in 49 CFR Parts 192 and 195 for “within the tolerance of the ILI tool.” (June 28, 2010).
ASME B31.8S – Managing System Integrity of Gas Pipelines, 2004 (reaffirmed 2018).
API 1163 – In-line Inspection Systems Qualification Standard, 2021.
NACE SP0102 – In-Line Inspection of Pipelines, 2020.
PHMSA Integrity Management Guidance for Hazardous Liquid Operators, available at https://www.phmsa.dot.gov.
Penspen, Pipeline Defect Assessment Manual (PDAM), 2003. Available at https://penspen.com/wp-content/uploads/2014/09/pdam.pdf.
Disclaimer
This article is based on a scenario similar to one I have experienced; however, portions of the information presented — including data values, locations, and organizational context — have been fictionalized or generalized for educational purposes.
The content is intended to illustrate key principles of pipeline integrity management, engineering judgment, and regulatory expectations, but does not constitute engineering analysis, professional engineering advice, or specific recommendations.
Readers are encouraged to apply their own company procedures, standards, and regulatory requirements, and to consult qualified professionals when evaluating inline inspection (ILI) data or making repair decisions.
