How Modified B31G works
Modified B31G came out of PRCI project PR-3-805 (Kiefner & Vieth, 1989). It keeps the same idea as the original criterion but fixes its biggest sources of excess conservatism:
- Area: metal loss is taken as 0.85 × depth × length, between a parabola and a full rectangle.
- Flow stress: SMYS + 10,000 psi (68.95 MPa), closer to real line-pipe behaviour.
- Folias factor: a more accurate three-term expression, plus a long-flaw branch, so there is no length cut-off.
Reading the result
If the safe pressure is at or above MAOP (ERF ≤ 1.0), the anomaly meets the criterion at the entered MAOP. DigSheet also runs a basic 49 CFR 192.933 screen on the failure pressure ratio: ≤ 1.1 × MAOP is an immediate condition; in Class 2 below 1.39, and in Class 3/4 below 1.50, is a one-year condition. Class 1 anomalies between 1.1 and 1.39 follow the ASME B31.8S schedule.
Have a detailed depth profile from a UT grid, laser scan or high-resolution ILI? The RSTRENG effective-area calculator usually recovers more pressure on long, irregular corrosion.
When not to use it
- Cracks, crack-like flaws, gouges, laminations, or metal loss in a dent.
- Selective seam corrosion or grooving at ERW/flash-welded seams.
- Significant axial or bending loads (the methods assume internal-pressure hoop stress).
- Metal loss deeper than 80% of wall: repair or replace regardless of length.
- Low-toughness or vintage pipe where fracture may control before plastic collapse.
FAQ
Modified B31G vs. original B31G: which governs?
Neither "governs" by law; 49 CFR 192.712 accepts both, along with RSTRENG. Modified B31G usually predicts a few percent to 15% higher failure pressure, with more scatter on very short deep pits.
Can I use metric units?
Yes. Use the in/psi vs mm/MPa toggle in the header. In metric, the flow stress is SMYS + 68.95 MPa.
Can I check a whole ILI report?
Yes. The whole-report check runs this method on every row of a CSV anomaly list and ranks the results.