Yes, ASME B31.3 requires volumetric testing (like x-ray) for piping welds, but the specific percentage of welds that must be examined is determined entirely by the “Fluid Service Category” designated by the system owner.
Examination Percentages by Category
ASME B31.3 establishes minimum baseline examination thresholds. While a facility can elect to apply stricter specifications (such as mandating 100% testing for specific process units), the Code minimums are strictly categorized by the service hazard:
| Fluid Service Category | Volumetric Testing (RT / UT) | Visual Examination (VT) |
| Category D (Innocuous/low pressure) | None required | Required |
| Normal Fluid Service | Minimum 5% random | Required (random & 100% subsets) |
| Category M (Lethal/highly toxic) | Minimum 20% random | 100% required |
| High Pressure | 100% required | 100% required |
| Severe Cyclic Conditions | 100% required | 100% required |
Radiographic Testing (RT) is the traditional standard for finding internal weld defects, but ASME B31.3 explicitly permits and often requires other non-destructive examination (NDE) methods:
- Ultrasonic Testing (UT): B31.3 universally accepts UT as a direct substitute for RT. UT is highly utilized because it avoids the radiation safety hazards, restricted access perimeters, and site evacuations associated with x-raying.
- Surface Examinations: Volumetric testing looks “inside” the weld, but for specific configurations, surface methods are also required. For instance, the Code requires Magnetic Particle (MT) or Liquid Penetrant (PT) testing on socket welds in Category M “Lethal service”, or on the root pass of welds in Severe Cyclic service.
In-Process Examination:
In limited circumstances, B31.3 allows comprehensive, real-time visual examination during the actual welding process to substitute for RT or UT on a weld-by-weld basis.
The “Progressive Sampling” Rule
When dealing with partial testing requirements (like the 5% for Normal or 20% for Category M), ASME B31.3 imposes a strict “progressive examination” protocol. If a weld selected for the random sample fails the RT or UT inspection, the Code requires the inspector to test two (2) additional welds made by the exact same welder. If either of those subsequent welds reveals a defect, the sampling escalates again. This progressive failure tree continues until the entire lot of welds produced by that specific welder is 100% examined. Because of this mechanism, poor initial weld quality can quickly force a baseline 5% requirement into a costly 100% mandatory examination.