B31G Remaining Strength of Corroded Pipe

1. What is ASME B31G?

ASME B31G, titled "Manual for Determining the Remaining Strength of Corroded Pipelines," is a supplement to the ASME B31 Code for Pressure Piping. It provides a conservative, analytically-based method for evaluating metal-loss anomalies found by in-line inspection (ILI), hydrostatic testing, or direct examination, and determining whether the pipeline can continue to operate safely at its current or reduced pressure.

History and Development

The B31G methodology originated from burst test research conducted at Battelle Memorial Institute in the late 1960s and early 1970s. Researchers performed over 90 full-scale burst tests on pipe specimens with machined and naturally occurring corrosion defects, establishing the empirical relationship between defect geometry and remaining burst pressure.

• 1971-1984 (Original B31G): The first edition used a parabolic (2/3 area) approximation for the metal loss profile and a flow stress of 1.1 x SMYS. This method was intentionally conservative, designed as a screening tool to determine which anomalies required immediate repair.
• 1989 (Modified B31G / RSTRENG): Kiefner and Vieth developed an improved version using an 0.85dL rectangular approximation for the corroded area and an increased flow stress of SMYS + 10,000 psi. They also introduced the RSTRENG effective area method for profile-based Level 2 assessment, significantly reducing unnecessary repairs.
• 2009-2012 (ASME B31G-2012): The current edition incorporates both the Original and Modified methods, adds guidance on assessment levels, clarifies application limits, and references modern ILI technology capabilities.

Regulatory Context

In the United States, pipeline integrity regulations directly reference B31G methods:

Scope and Applicability

2. Pipeline Corrosion Types

Understanding corrosion morphology is essential for selecting the correct B31G assessment approach. The geometry and distribution of metal loss directly affect which assessment level is appropriate and how conservative the results will be.

External Corrosion

External corrosion is metal loss on the outside surface of the pipe, caused by coating degradation that exposes bare steel to the soil or atmospheric environment. It is the most common form of pipeline corrosion and the primary driver for B31G assessments.

• Causes: Coating holidays (disbondment, mechanical damage, aging), inadequate cathodic protection (CP) current, shielding of CP by disbonded coatings (tape wrap, shrink sleeves), high soil corrosivity (low resistivity, acidic pH), and differential aeration cells at soil interfaces.
• Detection: In-line inspection (MFL, ultrasonic), close-interval survey (CIS), direct current voltage gradient (DCVG), alternating current voltage gradient (ACVG), and bell-hole excavation with direct measurement.
• Typical morphology: Broad, shallow general corrosion or localized pitting depending on coating condition and CP effectiveness.

Internal Corrosion

Internal corrosion occurs on the inside surface of the pipe, caused by corrosive constituents in the transported fluid. In gas pipelines, it is most common at low points where water accumulates and at locations with low flow velocity.

• Causes: Free water in the gas stream (below the water dewpoint), CO2 corrosion (sweet corrosion), H2S corrosion (sour corrosion), microbiologically influenced corrosion (MIC), oxygen ingress, and erosion-corrosion at fittings and bends.
• Detection: Internal MFL or ultrasonic ILI tools, corrosion coupons, corrosion monitoring probes (ER, LPR), and internal visual inspection during maintenance shutdowns.
• Critical locations: Low points (sag bends), dead legs, areas downstream of tees, compressor station piping, and meter runs.

Corrosion Morphology Classifications

Interaction of Multiple Defects

When multiple corrosion anomalies are located close together, they may interact and behave as a single larger defect. B31G provides interaction rules based on axial and circumferential spacing:

3. Assessment Levels

ASME B31G defines three assessment levels with increasing complexity and decreasing conservatism. Each level uses more detailed information about the defect, producing a more accurate (less conservative) estimate of remaining strength.

Assessment Level Selection

Level 0 Screening Criteria

4. Key Parameters

Every B31G assessment requires the same fundamental set of input parameters. Understanding each parameter, its source, and its effect on the result is essential for accurate evaluation.

Defect Parameters

Pipe Parameters

Operating Parameters

Flow Stress

Flow stress represents the effective stress at which the remaining ligament of corroded pipe yields and begins to bulge outward. It accounts for the fact that pipe steel work-hardens beyond SMYS before reaching ultimate tensile strength (UTS). Different B31G versions use different flow stress definitions:

5. Original B31G Method

The Original B31G method, first published in 1984, is the simplest closed-form approach for estimating the failure pressure of corroded pipe. It uses a parabolic approximation for the corroded area and a flow stress of 1.1 x SMYS.

Area Approximation

The Original B31G method assumes that the longitudinal cross-section of the corroded area (viewed as a depth profile along the pipe axis) can be approximated by a parabola. This yields a metal loss area of 2/3 of the bounding rectangle (d x L):

Failure Pressure Equation

Step-by-Step Procedure

6. Modified B31G Method

The Modified B31G method (also called the 0.85 dL method) was developed by Kiefner and Vieth in 1989 to reduce the excessive conservatism of the Original method while maintaining adequate safety margins. It uses two key improvements: a less conservative area approximation and a higher flow stress.

Key Improvements Over Original B31G

Failure Pressure Equation

Comparison: Original vs Modified B31G

When to Use Modified vs Original

• Use Modified B31G when: You want a more accurate (less conservative) result to potentially avoid costly repairs. The Modified method is the standard practice for most pipeline operators and is accepted by PHMSA as equivalent to the Original method.
• Use Original B31G when: Maximum conservatism is desired, when regulatory or company policy requires the Original method, or for quick conservative screening calculations.
• Use both methods when: The anomaly is borderline (near the acceptance threshold). If both methods agree the anomaly is acceptable, confidence is high. If the Modified method accepts but the Original rejects, the engineer should consider the specific circumstances and potentially apply Level 2 (RSTRENG).

7. Folias Bulging Factor

The Folias bulging factor M is the most important geometric parameter in the B31G equation. It accounts for the stress concentration and outward bulging that occurs at a thinned region of pressurized pipe. Understanding M is essential for interpreting B31G results and assessing the effect of defect length on remaining strength.

Physical Meaning

When internal pressure acts on a pipe with a locally thinned wall, the reduced-thickness region experiences higher hoop stress than the surrounding full-thickness pipe. Additionally, the thinned region bulges outward, creating a local deformation that further concentrates stress. The Folias factor quantifies this combined stress magnification effect.

• M = 1.0: Would correspond to a through-wall slot (no bulging effect) or an infinitely small defect. In practice, M is always greater than 1.0 for finite-length defects.
• M increases with defect length: Longer defects allow more outward bulging under pressure, increasing the stress concentration. This is why long, shallow corrosion can be more critical than short, deep pitting.
• M depends on pipe geometry (D and t): Larger diameter pipe with thinner walls (higher D/t ratio) is more susceptible to bulging. The parameter L^2/(Dt) captures this combined effect.

Folias Factor Equations

Effect of L^2/(Dt) on Remaining Strength

The dimensionless parameter L^2/(Dt) is the single most important variable controlling the Folias factor. It combines defect length, pipe diameter, and wall thickness into one parameter that determines how much the defect weakens the pipe.

Practical Implications

• Short defects (L^2/(Dt) < 4): The Folias factor is relatively small, meaning the defect length has limited impact. The result is dominated by defect depth. Short, deep pits are evaluated primarily by their depth ratio d/t.
• Long defects (L^2/(Dt) > 20): The Folias factor causes substantial reduction in predicted failure pressure. Even moderate-depth corrosion (d/t = 0.30-0.40) over a long axial extent can require repair. This is why axial grooving is the most critical corrosion morphology.
• Pipe size effect: For the same defect length L, larger diameter thin-walled pipe has a higher L^2/(Dt) and therefore a higher Folias factor. A 6-inch long defect on 36" x 0.312" pipe (L^2/(Dt) = 3.2, M = 1.68) is more critical than the same defect on 12" x 0.375" pipe (L^2/(Dt) = 8.0, M = 2.39). Wait -- the 12-inch pipe has the higher L^2/(Dt). This is because smaller diameter pipe has a lower Dt product, making any given defect length relatively more significant.

8. RSTRENG Effective Area Method

The RSTRENG (Remaining Strength) effective area method is a Level 2 assessment that uses the full depth profile of the corrosion defect rather than just the maximum depth and overall length. It finds the critical sub-length within the defect that produces the lowest predicted failure pressure, eliminating the conservatism inherent in using maximum depth over the entire defect length.

Concept: River-Bottom Profile

The effective area method discretizes the corrosion defect into a series of depth readings along the axial direction. The depth profile is measured at regular intervals (typically every 0.5 to 1.0 inch) from one end of the defect to the other. This profile represents the maximum depth at each axial location, analogous to following the deepest path along a river bottom.

Why RSTRENG Recovers Pressure Capacity

Level 1 methods (Original and Modified B31G) assume that the maximum depth d applies uniformly across the entire defect length L. For defects with irregular profiles -- especially pitting corrosion where deep pits are surrounded by shallower wall loss -- this assumption is very conservative.

When RSTRENG Provides the Most Benefit

Practical Considerations

• Data source: RSTRENG requires detailed depth profile data. This can come from UT grid scans during excavation, high-resolution ILI data (modern MFL and UT tools report depth profiles), or laser scanning of exposed corrosion during bell-hole examinations.
• Measurement spacing: Closer spacing (0.5 inch or less) provides more accurate results but requires more data. For most applications, 1-inch spacing is adequate. Spacing greater than 2 inches may miss critical features.
• Computational requirement: For n measurement points, RSTRENG evaluates n(n+1)/2 sub-length combinations. For a 24-inch defect with 0.5-inch spacing (n = 49 points), this is 1,225 evaluations -- trivial for a computer but impractical by hand.
• Software: The original RSTRENG software was developed by Kiefner and Associates. Several commercial pipeline integrity management software packages now include RSTRENG or equivalent effective area calculations.

9. Acceptance Criteria

The B31G assessment produces a predicted failure pressure P_f. To determine whether the corroded pipe segment can remain in service, this failure pressure must be compared to the MAOP using appropriate safety factors defined by regulation and the applicable code.

49 CFR 192.485 Requirements

For gas pipelines regulated under 49 CFR Part 192, Section 192.485 prescribes the criteria for evaluating pipe with corrosion metal loss discovered during integrity assessments:

Acceptance Decision Matrix

Repair Methods

When a corrosion anomaly does not meet acceptance criteria, several repair methods are available:

• Full-encirclement steel sleeve (Type B): A pressure-containing steel sleeve welded over the corroded area. This is the most common permanent repair for transmission pipelines. The sleeve restores the pressure capacity of the pipe regardless of the corrosion severity beneath it.
• Composite wrap repair: Fiberglass or carbon fiber composite material wrapped around the corroded section and cured in place. Suitable for corrosion up to approximately 80% wall loss. Lower cost than steel sleeves but has a finite service life (typically 20-50 years).
• Pipe replacement (cut-out): Removing the corroded section and welding in a new pipe segment. The most definitive repair but requires a pipeline shutdown unless hot tapping and plugging are used.
• Pressure reduction: Reducing MAOP to a level where the corroded segment meets acceptance criteria. Used as a temporary measure while planning permanent repair, or as a permanent solution for low-criticality locations.
• Grinding (for shallow defects): Mechanical removal of corrosion products and blending the pit profile to reduce stress concentration. Only applicable for very shallow defects (d/t < 12.5% for some operators) and requires verification that the ground profile does not create a new stress riser.

10. Worked Examples

The following examples demonstrate the complete B31G calculation procedure for a typical transmission pipeline anomaly, using both the Original and Modified methods.

Example Problem Setup

Example 1: Original B31G Method

Example 2: Modified B31G Method

Comparison of Results

Example 3: Deeper Defect Requiring Repair

11. Limitations

ASME B31G is a powerful and widely-used tool, but it has specific limitations that must be understood to avoid misapplication. Using B31G outside its valid scope can produce unconservative (dangerous) or excessively conservative results.

Defect Types NOT Covered by B31G

Geometric Limitations

• Interacting defects: B31G evaluates individual, isolated defects. When multiple defects are close together (axial spacing less than 3t or circumferential overlap), they may interact and behave as a single larger defect. Interaction rules must be applied before the B31G calculation. Failure to account for interaction can produce unconservative results.
• Circumferential extent: B31G assumes that the defect is narrow relative to the pipe circumference. For very wide defects spanning a significant fraction of the circumference, the axial stress (which is half the hoop stress under internal pressure) may become the controlling failure mode. B31G does not evaluate circumferential failure.
• Complex shapes: B31G assumes a simple axial defect profile. Defects with complex three-dimensional geometry (multiple pits, spiral patterns, branching corrosion) may require finite element analysis (FEA) for accurate assessment.
• Wall thickness variations: B31G uses a single nominal wall thickness. If the pipe has significant wall thickness variation due to manufacturing tolerance or prior general corrosion, the assessment should use the measured wall thickness at the defect location rather than the nominal value.

Loading Limitations

• Internal pressure only: B31G considers failure due to internal pressure (hoop stress) only. It does not account for external loads such as soil loading, thermal expansion, settlement, landslide forces, or seismic loading. Pipe segments subject to significant external loads may require a combined loading assessment per API 579 or finite element analysis.
• Static loading: B31G predicts burst pressure under static (monotonically increasing) internal pressure. It does not predict behavior under cyclic pressure loading (fatigue) or dynamic loading (pressure surge, water hammer). Pipelines with significant pressure cycling may require fatigue life assessment in addition to B31G.
• Temperature: B31G uses room-temperature material properties. For pipelines operating at elevated temperatures (above 250 degF) or very low temperatures (below -20 degF), temperature-adjusted material properties should be used.

Material Limitations

• Carbon and low-alloy steel only: B31G was developed and validated for carbon steel and low-alloy steel pipe manufactured to API 5L specifications. It should not be applied to stainless steel, duplex steel, clad pipe, or non-metallic pipe without engineering justification.
• Grade range: The burst test database used to validate B31G covered grades from Grade B (SMYS = 35,000 psi) through X65 (SMYS = 65,000 psi). Application to higher grades (X70, X80, X100) requires caution because the flow stress assumptions may not be valid for high-strength steels with lower strain-hardening ratios.
• Toughness assumed adequate: B31G assumes that the pipe steel has sufficient toughness to fail by plastic collapse (ductile failure) rather than brittle fracture. For vintage pipe with low Charpy toughness (pre-1970 pipe, especially ERW pipe), a fracture mechanics assessment may be more appropriate.

12. Related Standards & References

B31G exists within a broader ecosystem of pipeline integrity standards. Understanding the relationships between these standards helps engineers select the most appropriate assessment method for each situation.

Primary Standards

Assessment and Inspection Standards

When to Use Alternatives to B31G

Key Technical References

• Kiefner, J.F. and Vieth, P.H. (1989): "A Modified Criterion for Evaluating the Remaining Strength of Corroded Pipe" - AGA Catalog No. L51609. The foundational document for Modified B31G and RSTRENG.
• Kiefner, J.F. et al. (1973): "Failure Stress Levels of Flaws in Pressurized Cylinders" - ASTM STP 536. The original Battelle burst test research that provided the empirical basis for B31G.
• PRCI (Pipeline Research Council International): Ongoing research on pipeline integrity assessment methods, including updates to corrosion assessment criteria, interaction rules, and validation against new burst test data.
• GPSA Engineering Data Book: Contains practical guidance on pipeline integrity including corrosion assessment, ILI tool selection, and repair methods.