13.4 B31.3 Pressure Leak Testing: Hydrostatic, Pneumatic & Sensitive Leak Tests

13.4 B31.3 Pressure Leak Testing: Hydrostatic, Pneumatic & Sensitive Leak Tests

Pressure leak testing is the final quality verification milestone required before an ASME B31.3 piping system is certified for industrial operation. Governed by ASME B31.3 Chapter VI (Para 345), pressure testing serves two vital engineering functions:

1. Demonstrating Structural Integrity: Proving that pipe walls, fittings, flanges, and weldments can withstand operational and transient pressure stresses without yielding or rupture.
2. Confirming Leak Tightness: Verifying that welded, flanged, and threaded connections do not leak under pressurized conditions.

Testing must be conducted after all welding, postweld heat treatment, and required non-destructive examinations (RT, UT, MT, PT) have been fully completed and accepted.

---

Hydrostatic Leak Testing (ASME B31.3 Para 345.4)

The Hydrostatic Leak Test is the standard, preferred pressure testing method under B31.3 because liquids are virtually incompressible, storing negligible mechanical energy compared to compressed gases.

The Hydrostatic Test Pressure Formula (Para 345.4.2)

Under Para 345.4.2, the minimum hydrostatic test gauge pressure at any point in the piping system is calculated using the following equation:

+---------------------------------------------------------------------------------------------------+
|                             B31.3 HYDROSTATIC TEST PRESSURE FORMULA                               |
+---------------------------------------------------------------------------------------------------+
|                                                                                                   |
|                                    P_T = 1.5 * P * (S_T / S)                                      |
|                                                                                                   |
|   WHERE:                                                                                          |
|   - P_T = Minimum hydrostatic test gauge pressure                                                 |
|   - P   = Internal design gauge pressure of the piping system                                     |
|   - S_T = Allowable stress value of the material at test temperature (from Table A-1)             |
|   - S   = Allowable stress value of the material at design temperature (from Table A-1)           |
|                                                                                                   |
|   CRITICAL STATUTORY CODE LIMITATION:                                                             |
|   - The ratio (S_T / S) shall NOT exceed 6.5. If calculated S_T / S > 6.5, use S_T / S = 6.5!      |
|   - Para 345.2.1(a): if the test pressure would produce circumferential or longitudinal stress   |
|     (on minimum wall) above the yield strength at test temperature, or exceed 1.5 times the       |
|     component rating at test temperature, P_T may be reduced to the maximum pressure that will    |
|     not exceed the LESSER of those two limits.                                                    |
+---------------------------------------------------------------------------------------------------+

Why the Stress Ratio ($S_T / S$) Exists and Why It Is Capped at 6.5

In high-temperature process piping (such as refinery cracking coils or steam lines operating at 800°F to 1,050°F), the material's allowable stress at operating temperature ($S$) is significantly reduced by elevated-temperature creep. When that same piping is tested at ambient room temperature (70°F), the material possesses substantially greater tensile strength ($S_T$).

Multiplying by $S_T / S$ ensures that the ambient test induces a membrane stress proportional to what the system experiences at elevated operating temperature. However, at extreme temperatures (above 1,150°F), the creep allowable stress $S$ drops to very low numbers, causing the mathematical ratio $S_T / S$ to reach 10, 15, or higher. If unconstrained, this would produce a test pressure high enough to burst the cold pipe. Therefore, B31.3 strictly caps the allowable stress ratio at 6.5.

Step-by-Step Hydrostatic Calculation Example

A 10-inch chrome-moly piping line (ASTM A335 Grade P11) operates at a design pressure $P = 500\text{ psig}$ and design temperature $850^\circ\text{F}$.

• Allowable stress at test temperature ($70^\circ\text{F}$): $S_T = 20.0\text{ ksi}$ (from Table A-1).
• Allowable stress at design temperature ($850^\circ\text{F}$): $S = 14.4\text{ ksi}$ (from Table A-1).

1. Calculate Stress Ratio: $\frac{S_T}{S} = \frac{20.0}{14.4} = 1.389$ (Since $1.389 \le 6.5$, use 1.389).
2. Calculate Minimum Test Pressure: $P_T = 1.5 \times 500 \times 1.389 = 750 \times 1.389 = 1,041.75\text{ psig}$ The minimum hydrostatic test pressure is 1,042 psig.

Hydrostatic Test Medium and Temperature Safeguards (Para 345.4.1 & 345.4.3)

• Test Liquid: Clean water is the standard medium. If freezing conditions exist, a glycol-water mixture or other non-toxic, non-flammable liquid may be used.
• Flammable Test Liquids: Para 345.4.1 requires water unless there is a possibility of damage from freezing or adverse effects of water on the piping or the process; another suitable nontoxic liquid may then be used, and if the liquid is flammable its flash point shall be at least 120°F (49°C), with consideration given to the test environment.
• Stainless Steel Chloride Control: B31.3 does not publish a numeric chloride limit in Para 345. Chloride control on austenitic stainless systems comes from the project specification and from the guidance in Appendix F; verify the specified limit rather than quoting a number from memory. The mechanism is real — chloride residue trapped in crevices initiates transgranular stress corrosion cracking — but the number is not a Code value.
• Brittle Fracture Prevention: Para 345.2.2(c) requires that the possibility of brittle fracture shall be considered when conducting leak tests at metal temperatures near the ductile-brittle transition temperature. B31.3 sets no general numeric minimum test temperature for Chapter II piping; the explicit numeric rule lives in Chapter IX (K345.2.2), where the minimum metal temperature during testing shall be not less than the impact test temperature.
• Holding Time & Examination: Full test pressure $P_T$ must be held for a minimum of 10 minutes. Following this hold, the pressure may be reduced to design pressure $P$, at which point inspectors visually examine all joints, connections, and packing glands for leaks.

---

Pneumatic Leak Testing (ASME B31.3 Para 345.5)

A Pneumatic Leak Test utilizes a compressible gas (dry oil-free air or nitrogen) as the test medium. Because compressed gas stores immense potential energy ($PV$ energy), an accidental rupture during a pneumatic test produces a violent explosion with blast shockwaves and lethal shrapnel.

Restrictions on Pneumatic Testing

Under Para 345.5.1, pneumatic testing is permitted only when:

1. The piping system is designed or supported such that it cannot safely support the weight of water; OR
2. The process fluid cannot tolerate traces of moisture (such as refrigeration systems, anhydrous hydrogen chloride, or catalyst beds).

Pneumatic Test Pressure & Relief Requirements (Paras 345.5.2 and 345.5.4)

1. Test Pressure Window (Para 345.5.4): The pneumatic test pressure is a range, not a single value. It shall be not less than $1.1 \times P$ and shall not exceed the lesser of:

   • $1.33 \times P$ (1.33 times the design pressure); or
   • the pressure that would exceed 90% of the pressure described in Para 345.2.1(a) — that is, 90% of the pressure limited by yield strength at test temperature or 1.5 times the component rating at test temperature.

   $1.1\,P \le P_T \le \min\!\left(1.33\,P,\; 0.9 \times P_{345.2.1(a)}\right)$

2. Overpressure Relief Device (Para 345.5.2): A pressure relief device shall be provided with a set pressure not higher than the test pressure plus the lesser of 50 psi (345 kPa) or 10% of the test pressure.

3. Test Fluid (Para 345.5.3): The gas used as test fluid, if not air, shall be nonflammable and nontoxic.

Stepwise Pressurization Procedure (Para 345.5.5)

To safeguard personnel against sudden rupture, ASME B31.3 mandates a strict, staged pressurization protocol:

                                PNEUMATIC PRESSURIZATION SEQUENCE (PARA 345.5.5)
                                
  0 psig            lesser of 0.5 PT or 25 psi          Gradual steps, strain holds        PT
    |--------------------------------|--------------------------------|----------------------------|
             Preliminary Check              Hold at each step long              Hold >= 10 min
        (examine joints per 341.4.1(a))     enough to equalize strains         (Para 345.2.2(a))
                                                                                      |
                                                                                      v
                                                                          REDUCE TO DESIGN PRESSURE
                                                                             (Close-Up Inspection)

1. Preliminary Check: Gradually raise the pressure until a gage pressure equal to the lesser of one-half the test pressure or 25 psi (170 kPa) is attained, at which time a preliminary check is made, including examination of joints in accordance with Para 341.4.1(a).
2. Incremental Stepping: Thereafter the pressure shall be gradually increased in steps until the test pressure is reached, holding the pressure at each step long enough to equalize piping strains. B31.3 does not prescribe a fixed percentage for those steps — the one-tenth increments are a B31.1 Para 137.5.5 rule, not a B31.3 rule.
3. Full Pressure Hold: Under Para 345.2.2(a) the leak test pressure shall be maintained for at least 10 minutes.
4. Examination Pressure: The pressure shall then be reduced to the design pressure (and not less than it) before examining all joints and connections for leaks.

[!NOTE] A preliminary pneumatic test can precede a hydrostatic test too. Para 345.2.1(c) permits a preliminary test using air at no more than 25 psi (170 kPa) gage prior to hydrostatic testing to locate major leaks.

---

Other B31.3 Leak Test Methods

1. Initial Service Leak Test (Para 345.7)

• Exclusive Application: Permitted exclusively for Category D Fluid Service (nonflammable, nontoxic, $P \le 150\text{ psi}$, $-20^\circ\text{F}$ to $366^\circ\text{F}$).
• Procedure: The piping is pressurized with the actual process fluid during initial commissioning. The pressure is gradually increased to operating pressure and held for at least 10 minutes while all joints are visually examined.
• Prohibition: Using an Initial Service Leak Test on Normal Fluid Service or Category M is a severe code violation.

2. Sensitive Leak Test (Para 345.8)

• Application: Used for Category M toxic piping or where microscopic leaks cannot be tolerated.
• Execution: Conducted in accordance with ASME BPVC Section V, Article 10 (Leak Testing), utilizing a helium mass spectrometer or halogen diode detector.
• Method (Para 345.8.2): The Bubble Test — Direct Pressure Technique per Section V, Article 10, Mandatory Appendix I, or another method with demonstrated sensitivity not less than $10^{-3}$ std mL/s under test conditions.
• Pressure Level: When the Bubble Test — Direct Pressure Technique is used, the test pressure shall be at least the lesser of 15 psi (105 kPa) or 25% of the design pressure, and the pressure is raised to the lesser of one-half the test pressure or 25 psi (170 kPa) for a preliminary check before being increased in steps with strain-equalizing holds.

3. Alternative Leak Test (Para 345.9)

If both hydrostatic testing (due to damage to internal refractory or risk of brittle fracture) and pneumatic testing (due to extreme explosive danger) are judged unfeasible by the owner and designer, an Alternative Leak Test may be utilized. It requires:

1. 100% Volumetric Examination (RT or UT) of all circumferential and longitudinal butt welds.
2. 100% Surface Examination (MT or PT) of all other welds (fillet, socket, branch).
3. A formal flexibility analysis under Para 319.
4. A Sensitive Leak Test under Para 345.8.

---

Practical Inspection Scenarios & High-Frequency Exam Traps

Scenario: The High-Temperature Hydrostatic Calculation Trap

A power boiler cogeneration piping line is designed for $P = 400\text{ psig}$ at $1,100^\circ\text{F}$. From Table A-1, the material's allowable stress at test temperature ($70^\circ\text{F}$) is $S_T = 18.0\text{ ksi}$, while at design temperature ($1,100^\circ\text{F}$) the allowable stress is $S = 2.0\text{ ksi}$. The contractor calculates: $\frac{S_T}{S} = \frac{18.0}{2.0} = 9.0$ $P_T = 1.5 \times 400 \times 9.0 = 5,400\text{ psig}$ The contractor prepares to pump the system to 5,400 psig.

The CWI Audit Finding: The Owner's Inspector intervenes and prevents the test. Under Para 345.4.2, the ratio $S_T / S$ shall not exceed 6.5. Substituting 6.5: $P_T = 1.5 \times 400 \times 6.5 = 3,900\text{ psig}$ Testing at 5,400 psig would have permanently deformed or ruptured the piping.

Common Exam Traps to Avoid

• The Pneumatic Multiplier Trap: Hydrostatic is at least $1.5 \times P$ (temperature-corrected by $S_T/S$); pneumatic is a window of at least $1.1 \times P$ and at most the lesser of $1.33 \times P$ or 90% of the Para 345.2.1(a) pressure. Treating $1.1P$ as a fixed value rather than a floor is the more common version of this error.
• The Cross-Code Staging Trap: B31.3 Para 345.5.5 says only "in steps... long enough to equalize piping strains." The explicit "one-half, then one-tenth increments" staging is B31.1 Para 137.5.5, and the "relief device at 1-1/3 times the test pressure" recommendation is B31.1 Para 137.2.
• The Initial Service Category Trap: An exam question might ask if an Initial Service Leak Test can be used on a 100 psi diesel fuel line. Because diesel is flammable, it is Normal Fluid Service, not Category D. An Initial Service Leak Test is strictly illegal.
• The Close-Up Inspection Pressure Trap: Visual examination of piping joints during a pneumatic test is never conducted at full test pressure ($P_T$). Pressure must be reduced to the design pressure ($P$) before examiners approach the line.