Enhancing Pipeline Integrity Management: Integrating Probability of Exceedance (PoE) with ILI Vendor Certainty Metrics
Introduction
Pipeline integrity management relies on Inline Inspection (ILI) tools to detect and assess pipeline anomalies such as metal loss, cracks, and mechanical damage. However, these ILI-reported anomalies include measurement uncertainty, typically reported at 80% certainty.
For example, an ILI tool might report a 50% wall thickness (WT) anomaly with a ±10% tolerance at 80% certainty, meaning that in 80% of cases, the actual anomaly depth is between 40% and 60% WT, but in 20% of cases, it could be outside this range.
To ensure compliance with PHMSA regulations (49 CFR §192.933 for gas pipelines and §195.452 for hazardous liquid pipelines), operators assess whether an anomaly exceeds regulatory thresholds.
Since ILI tools do not provide absolute measurements, Probability of Exceedance (PoE) is used to quantify the likelihood that an anomaly surpasses a critical threshold. While a low PoE (e.g., 1-10%) may suggest a low probability of exceedance, failures have occurred within this range due to outliers, material variability, or unexpected growth rates. Operators may consider verifying anomalies in this 1-10% PoE range to assess their significance.
Understanding ILI Vendor Confidence and PHMSA Thresholds
Interpreting “80% Certainty” in ILI Reporting
When an ILI vendor reports an anomaly depth with 80% certainty, it means:
✅ 80% Probability: The true depth is within the reported tolerance range (e.g., 40% to 60% WT for a 50% ±10% WT report).
✅ 20% Probability: The true depth falls outside this range, potentially exceeding critical limits.
Since ILI-reported anomalies are estimates, PHMSA regulations require operators to determine if anomalies exceed specific integrity thresholds.
PHMSA Regulatory Requirements for Anomaly Assessment
Gas Pipelines (49 CFR §192.933):
• Immediate Repair Conditions – Anomalies that could pose an immediate integrity threat.
• Scheduled Evaluation Conditions – Anomalies that require assessment within a specified timeframe.
• Monitoring Conditions – Anomalies that do not require immediate action but need continued observation.
📌 Industry guidelines (ASME B31.8S) commonly apply:
• 80% WT loss for immediate repair.
• 60% WT loss for scheduled evaluation.
• 40% WT loss for monitoring.
Hazardous Liquid Pipelines (49 CFR §195.452):
• Immediate Repair Conditions:
• Metal loss greater than 80% of nominal wall thickness, regardless of dimensions.
• 180-Day Conditions (Scheduled Evaluation):
• General corrosion with a predicted metal loss greater than 50% of nominal wall thickness.
• Metal loss greater than 50% of nominal wall thickness in:
• Pipeline crossings.
• Areas with widespread circumferential corrosion.
• Areas affecting girth welds.
While metal loss exceeding 80% requires immediate remediation, anomalies with over 50% metal loss must be evaluated and remediated within 180 days.
Applying PoE with Vendor Certainty Levels in ILI Assessments
Why PoE is Used in ILI Assessments
• ILI vendor confidence levels (e.g., 80%) do not cover all possible outcomes.
• There is still a probability that an anomaly exceeds a given threshold.
• Outliers within the 20% uncertainty can present integrity risks.
By using PoE, operators can assess the probability that an ILI-reported anomaly exceeds PHMSA regulatory thresholds, even if it falls within the vendor’s tolerance range.
Calculating PoE Beyond the 80% Certainty Range
Since ILI vendors report anomalies with 80% confidence intervals, the data can be modeled using a normal (Gaussian) distribution centered around the reported depth with a specific standard deviation (σ).
Example 1: Determining PoE for a 50% WT Hazardous Liquid Threshold
Given Data:
• ILI-reported depth: 40% WT
• ILI tool tolerance: ±10% WT (at 80% certainty)
• Threshold for 180-day evaluation: 50% WT
Step 1: Determine Standard Deviation (σ)
From normal distribution statistics, an 80% confidence interval corresponds to a Z-score of ±1.28. Using this relationship:
σ = 10% / 1.28 = 7.81%
Step 2: Compute the Z-Score for Exceeding 50% WT
Z = (50% - 40%) / 7.81% = 1.28
Step 3: Find PoE from Standard Normal Distribution
Using a normal distribution table:
P(Z > 1.28) = 10%
📌 Interpretation: There is a 10% probability that the true anomaly depth exceeds 50% WT, meeting the threshold for 180-day evaluation.
Example 2: Determining PoE for an 80% WT Immediate Repair Threshold
Given Data:
• ILI-reported depth: 70% WT
• ILI tool tolerance: ±10% WT (at 80% certainty)
• Threshold for immediate repair: 80% WT
Step 1: Compute the Z-Score for Exceeding 80% WT
Z = (80% - 70%) / 7.81% = 1.28
Step 2: Find PoE from Standard Normal Distribution
P(Z > 1.28) = 10%
📌 Interpretation: There is a 10% probability that the true anomaly depth exceeds 80% WT, meeting the immediate repair threshold.
Considering Outliers: The Role of 1-10% PoE in Integrity Management
Even at low PoE values (1-10%), failures have occurred in cases where:
• ILI tools underreport deeper anomalies due to sensor limitations.
• Multiple anomalies interact, accelerating wall loss.
• Unexpected growth rates occur due to environmental conditions.
• Material property variations affect failure pressure calculations.
📌 Operators should consider verifying anomalies in the 1-10% PoE range to assess their significance.
Key Takeaways on PoE in Integrity Management
✅ PoE quantifies the likelihood that an anomaly exceeds regulatory thresholds.
✅ Even at 1-10% PoE, the risk of exceedance exists.
✅ Operators may consider verifying anomalies in this range.
✅ PoE can be integrated into regulatory compliance and risk assessment programs.
Conclusion
📌 Final Thought: Understanding the probability distribution of ILI-reported anomalies allows operators to evaluate whether further verification is necessary.
The information provided in this blog is for informational purposes only and does not constitute engineering, regulatory, or legal advice. The discussion of Probability of Exceedance (PoE), ILI tool accuracy, and PHMSA regulatory thresholds is based on publicly available industry standards and best practices but may not reflect the specific requirements, methodologies, or risk tolerance of individual pipeline operators.
Pipeline integrity management decisions should be made based on operator-specific data, engineering assessments, regulatory requirements, and expert consultation. Readers should refer to 49 CFR §192.933, 49 CFR §195.452, ASME B31.8S, API 1160, and ILI vendor specifications for official guidance and compliance requirements.
The author assumes no responsibility for how this information is applied. Pipeline operators should conduct their own analysis and verification before making integrity management decisions.
