13.2 Leak Testing: Vacuum Box, Tracer Gas, Diesel & Air Pressure Tests
Key Takeaways
- Vacuum box testing per API 650 Section 8.6 and ASME Section V Article 10 Appendix II serves as the primary code-mandated method for leak-testing tank bottom lap welds and internal corner fillet welds, operating at a partial vacuum of 21 kPa (3 psi / 6 in. Hg) to 35 kPa (5 psi / 10 in. Hg) gauge with a mandatory minimum 5-second dwell time.
- A corner vacuum box featuring an angled 90-degree cross-section and specialized transparent viewing ports is required to inspect the critical interior shell-to-bottom fillet weld where standard flat-bottom boxes cannot seat.
- Tracer gas leak testing (using helium or sulfur hexafluoride SF6) per ASME Section V Article 10 provides ultra-high sensitivity ($10^{-6}$ to $10^{-9}\text{ std cm}^3/\text{s}$) for bottom plate assemblies, detecting through-thickness micro-fissures underneath floors without individual box repositioning.
- The diesel oil test evaluates unbacked shell-to-bottom fillet welds and nozzle attachments by applying diesel fuel to the exterior weld face for a minimum dwell time of 4 hours, inspecting the opposing interior face for capillary wicking or oil wetting.
- Pneumatic air-and-soap testing of nozzle reinforcing pads is conducted at exactly 15 psig (100 kPa) through the telltale hole using an approved foaming solution; following test completion, the telltale hole must remain open and unobstructed to provide continuous pressure relief and leak indication.
Fundamentals of Leak Testing (LT) in Tank Integrity
While volumetric and surface NDE methods (such as RT, UT, MT, and PT) evaluate the structural soundness, geometry, and mechanical integrity of welded joints, Leak Testing (LT) is uniquely concerned with fluid containment—the absolute absence of through-thickness passages, pinholes, wormholes, or micro-cracks. Under API Standard 653 and API Standard 650, leak testing is mandatory across all bottom lap welds, annular plate joints, shell-to-bottom corner fillets, nozzle reinforcement pads, and roof plate seams.
Leak testing methods applied to atmospheric storage tanks fall under four primary disciplines defined by ASME Section V, Article 10 (Leak Testing):
- Bubble Test – Vacuum Box Technique (Article 10, Appendix II / API 650 Section 8.6).
- Tracer Gas Leak Testing – Sniffer Probe Technique (Article 10, Appendix IV).
- Capillary Penetration – Diesel Oil Test.
- Pneumatic Pressure Bubble Test (Nozzle Reinforcing Pads per API 650 Section 7.3.5 / API 653 Section 12.1.5).
Vacuum Box Testing per API 650 Section 8.6 & ASME Section V Article 10
Vacuum box testing is the ubiquitous field method for verifying the leak-tightness of flat bottom lap welds, sketch plates, and internal shell-to-bottom fillet welds.
+-------------------------------------------------------------------------+
| STANDARD FLAT VACUUM BOX SCHEMATIC |
| |
| Transparent Viewing Window |
| +------------------------+ |
| | [ Clear Acrylic ] | Pressure Gauge |
| Air Evacuation | | (-3 to -5 psig) |
| Valve / Venturi | | | |
| +---+ +------------------------+ +---+ |
| | | Handle | | |
| =====+===+========================================+===+===== |
| | Rigid Metal Frame (approx. 6" wide x 30" long) | |
| +==========================================================+ |
| | Thick Sponge Neoprene Rubber Gasket | |
| +----------------------------------------------------------+ |
| ........... Bubble-Forming Solution Across Weld ........... |
| =======================[ Lap Weld ]========================= |
| Tank Bottom Plate |
+-------------------------------------------------------------------------+
Equipment Construction and Assembly
A standard vacuum box consists of:
- Box Housing: A rigid rectangular box, typically 6 inches (150 mm) wide by 30 inches (750 mm) long, fabricated from aluminum, steel, or molded fiberglass.
- Viewing Window: A sealed, transparent acrylic or polycarbonate top window providing an unobstructed view of the entire test segment.
- Bottom Seal: A thick, flexible, closed-cell sponge neoprene or silicone rubber gasket attached to the bottom rim to form an airtight seal against irregular steel floor plates.
- Vacuum Gauge: A calibrated differential vacuum gauge reading in psi, kPa, or inches of mercury (Hg), with a range of 0 to -15 psig (0 to -100 kPa).
- Evacuation Mechanism: An air ejector (pneumatic venturi aspirator driven by compressed plant air) or a portable electric vacuum pump, connected through an evacuation valve and equipped with an adjustable vacuum relief valve.
Operating Parameters & Code Constraints (API 650 Section 8.6)
- Differential Pressure: The vacuum box must be evacuated to a partial vacuum of 21 kPa (3 psi / 6 in. Hg) to 35 kPa (5 psi / 10 in. Hg) gauge pressure.
- Bubble-Forming Solution: An approved commercial foaming liquid (conforming to ASME Section V Article 10 and ASTM E515) is applied immediately prior to setting the box. The solution must produce durable, stable bubbles that do not rapidly burst or dry out under vacuum. Ordinary dish soap is technically prohibited unless qualified for bubble longevity and temperature stability.
- Dwell Time: Once full test vacuum (21 to 35 kPa) is achieved, the weld must be visually observed through the transparent window for a minimum dwell time of 5 seconds. Any stream or cluster of growing bubbles indicates through-thickness air leakage.
- Linear Overlap Requirement: To prevent untested gaps between adjacent box placements, each successive vacuum box placement must overlap the previously tested segment by at least 2 inches (50 mm).
- Surface Temperature Envelope: Standard bubble solutions operate between 40°F and 125°F (5°C to 52°C). Testing outside this window requires certified low-temperature (antifreeze-doped) or high-temperature foaming solutions qualified per ASME Section V Article 10.
+-------------------------------------------------------------------------+
| CORNER VACUUM BOX FOR SHELL-TO-BOTTOM |
| |
| Vertical Tank Shell Plate |
| | |
| | Corner Vacuum Box (L-Shaped Profile) |
| | +-----------------------------+ |
| | | Angled Transparent Window | |
| | | | |
| | ===+-+ [Viewing Interior Fillet] | |
| | |G | | |
| | |A | | |
| | |S | | |
| | |K | | |
| | |E | | |
| | |T | | |
| | ===+-------------------------------+=== |
| | | Foam Gasket |G| |
| =========+=====+===============================+=+============== |
| Bottom | ^ Inside Shell-to-Bottom Corner |A| |
| Plate | +-- Fillet Weld |S| |
| | |K| |
+--------------+-------------------------------------+-------------------+
Specialized Corner Vacuum Box
The internal shell-to-bottom fillet weld presents a 90-degree geometry where a flat box cannot seal. A specialized corner vacuum box is utilized. It features an L-shaped or angled profile with perpendicular sealing gaskets that conform simultaneously to the vertical shell and horizontal bottom plates. Due to restricted optical angles, the corner box requires integrated internal lighting or high-intensity external inspection lamps to clearly illuminate the fillet weld root and toes.
Tracer Gas Leak Testing per ASME Section V, Article 10
When tanks feature secondary containment liners, double bottoms, or thick bottom sketch plates where vacuum box testing is physically restricted or prone to false results, Tracer Gas Leak Testing provides an exceptionally sensitive alternative.
+-------------------------------------------------------------------------+
| TRACER GAS TESTING CONFIGURATION |
| |
| Sniffer Probe (< 1 in/sec scan) |
| \ / |
| \ / |
| Mass Spectrometer | | |
| Leak Detector =======+ +===== |
| (MSLD Console) v |
| =======================[ Lap Weld ]========================= |
| Tank Bottom Plate | | |
| v v Through-Wall Micro-fissures |
| ............................................................ |
| Helium / SF6 Tracer Gas Plenum Under Floor (2 to 5 psig) |
| ============================================================ |
| Foundation / Impervious Secondary Containment Liner |
+-------------------------------------------------------------------------+
Tracer Gas Methodology & Physics
Governed by ASME Section V, Article 10, Appendix IV (Helium Mass Spectrometer – Detector Probe Technique):
- Gas Injection: A tracer gas—predominantly Helium (He) or Sulfur Hexafluoride ($SF_6$) diluted with dry air or nitrogen to a concentration of 5% to 10% by volume—is injected into the interstitial space beneath the tank floor via temporary sub-floor pipes or foundation telltale ports.
- Positive Pressure: The under-floor cavity is pressurized to a low positive pressure, typically 2 to 5 in. water column (up to 2 psig / 14 kPa), ensuring the tracer gas permeates across the entire foundation contact surface.
- Sniffer Probe Scanning: An operator sweeps a calibrated sniffer probe connected to a Mass Spectrometer Leak Detector (MSLD) or electron-capture detector along 100% of the bottom weld seams on the tank interior.
- Scanning Parameters:
- Scanning Speed: Must not exceed 1 inch per second (25 mm/s).
- Probe Distance: The probe tip must remain within $1/8\text{ to }1/4\text{ in.}$ ($3\text{ to }6\text{ mm}$) of the weld seam.
- Sensitivity: MSLD systems detect leak rates down to $10^{-6}$ to $10^{-9}\text{ std cm}^3/\text{s}$, several orders of magnitude more sensitive than bubble testing.
Operational Advantages of Tracer Gas
- Micro-fissure Detection: Detects microscopic, labyrinthine leak paths through plate laminations and microporosities that cannot generate bubbles under 5 psi vacuum.
- Full-Floor Coverage: Tests 100% of floor seams without repeated box repositioning, dramatically accelerating large-diameter turnaround schedules.
- Zero Surface Residue: Eliminates soapy bubble solutions that must otherwise be meticulously scrubbed off before applying tank bottom epoxy liners.
Diesel Oil Capillary Testing & Pneumatic Pad Testing
In addition to vacuum and tracer methods, API 653 and API 650 mandate two specialized leak testing techniques for specific welded components:
1. The Diesel Oil Test (Capillary Wicking)
The diesel oil test utilizes the low surface tension, high wetting ability, and low viscosity of light petroleum distillates (diesel fuel or kerosene) to reveal through-wall capillary leaks in unbacked fillet and lap welds.
- Target Welds: Primarily used on newly welded or repaired shell-to-bottom corner fillet welds (before the second pass or where backing prevents vacuum testing) and temporary attachment gouge excavations.
- Execution Protocol: Clean diesel fuel is brushed or sprayed generously onto the exterior weld face and adjacent plate.
- Dwell Time: The diesel must remain in continuous contact with the weld for a minimum dwell time of 4 hours (and preferably up to 12 hours in cold weather or heavy-wall assemblies).
- Examination & Evaluation: The interior weld face is meticulously inspected under intense visual illumination ($\ge 100\text{ fc} / 1000\text{ lux}$). Any dark oil wicking, dampness, or fluorescent sheen bleeding through to the inside indicates a rejectable through-wall defect.
- Post-Cleaning: After testing, all residual diesel must be thoroughly flushed with volatile solvents to prevent contamination of subsequent coating systems.
2. Pneumatic Reinforcing Pad Testing (API 650 Section 7.3.5 & API 653 Section 12.1.5)
All newly installed or altered shell nozzle reinforcing pads and manway repads must undergo a pneumatic bubble test before tank commissioning.
+-------------------------------------------------------------------------+
| REINFORCING PAD PNEUMATIC TEST SETUP |
| |
| Tank Shell Plate |
| | |
| | Reinforcing Pad Plate (Repad) |
| | +-------------------------+ |
| | | Telltale Hole (NPT) | |
| | | | | |
| | | +-v-+ | |
| | | | | <--- Regulated Compressed Air (15 psig) |
| | ===+======+===+==============+=== |
| | | Interstitial Annular Space | | |
| | +----------------------------+ | |
| | | Outer Fillet Weld | Nozzle Neck |
| | | [Soaped for Bubbles] | Weld |
| ======+==+===============================+==================== |
| | |
+-------------------------------------------------------------------------+
- Test Setup: Clean compressed air is introduced into the interstitial space between the shell and reinforcing pad via the threaded telltale inspection hole (typically $1/4\text{-in.}$ or $1/2\text{-in.}$ NPT).
- Test Pressure: Air pressure is regulated to exactly 15 psig (100 kPa / 1.0 bar). A calibrated pressure gauge is installed in the test manifold.
[!CAUTION] Never exceed 15 psig during repad testing. Because a reinforcing pad presents a large planar surface area ($Force = Pressure \times Area$), excessive pneumatic pressure can easily bow the repad, tear the attachment fillet welds, or permanently deform the shell plate.
- Leak Examination: While maintaining 15 psig, an approved bubble-forming solution is applied over 100% of the nozzle-to-shell weld (interior and exterior) and 100% of the repad-to-shell perimeter fillet weld. Any expanding soap bubbles indicate through-wall leakage.
- Post-Test Telltale Hole Mandate: Upon successful test completion, the air connection is removed, and the telltale hole must remain open and completely unobstructed to the atmosphere. The hole may be lightly packed with soft grease or plugged with an unthreaded plastic dirt cap to prevent rainwater ingress, but it must NEVER be welded shut, plugged with a solid threaded metal plug, or coated over. If the internal shell weld leaks during operation, the open telltale hole safely vents product outward, alerting operators before the repad overpressurizes.
Master Comparison: Leak Testing Methodologies
The following table compares the operational characteristics, code references, differential pressures, dwell times, and sensitivities across the four principal storage tank leak testing methods.
| Leak Testing Method | Code Reference | Test Medium / Vehicle | Pressure / Differential | Minimum Dwell Time | Detection Sensitivity | Primary Application Zones |
|---|---|---|---|---|---|---|
| Standard Vacuum Box | API 650 Sec. 8.6 & ASME V Art. 10 App. II | Air evacuation + foaming bubble solution | 21 to 35 kPa (3 to 5 psi / 6 to 10 in. Hg) partial vacuum | 5 seconds continuous observation | Moderate ($10^{-3}$ to $10^{-4}\text{ std cm}^3/\text{s}$) | Tank bottom lap seams, sketch plates, flat roof seams |
| Corner Vacuum Box | API 650 Sec. 8.6 & ASME V Art. 10 App. II | Air evacuation + foaming bubble solution | 21 to 35 kPa (3 to 5 psi / 6 to 10 in. Hg) partial vacuum | 5 seconds continuous observation | Moderate ($10^{-3}$ to $10^{-4}\text{ std cm}^3/\text{s}$) | Inside shell-to-bottom fillet weld (90° corner geometry) |
| Tracer Gas (Sniffer) | ASME Section V, Article 10, Appendix IV | Helium (He) or $SF_6$ (5-10% in air/N2) | 2 to 5 in. w.c. (up to 2 psig / 14 kPa) under bottom | Scan rate $\le 1\text{ in./sec}$; probe $\le 1/4"$ | Ultra-High ($10^{-6}$ to $10^{-9}\text{ std cm}^3/\text{s}$) | Entire bottom floor, double bottoms, secondary containment |
| Diesel Oil Test | API 650 Sec. 7.3.5 & API 653 Sec. 12.1.6 | Clean light diesel fuel / kerosene | Capillary action (zero applied pressure) | 4 hours minimum (up to 12 hours cold) | High for tight cracks ($10^{-4}\text{ std cm}^3/\text{s}$) | Unbacked corner fillet welds, temporary attachment sites |
| Pneumatic Pad Test | API 650 Sec. 7.3.5 & API 653 Sec. 12.1.5 | Compressed air + bubble-forming solution | Exactly 15 psig (100 kPa / 1.0 bar) | Sufficient to soap & inspect 100% welds | Moderate ($10^{-3}\text{ std cm}^3/\text{s}$) | Shell nozzle reinforcing pads, door sheet repads, manways |
An NDE technician is performing vacuum box testing on newly welded bottom lap seams of an API 650 storage tank. In accordance with API 650 Section 8.6 and ASME Section V Article 10 Appendix II, what is the required partial vacuum pressure range and the minimum observation dwell time per test segment?
Following the installation of a replacement nozzle and reinforcing pad on a storage tank shell, a pneumatic pressure test is conducted through the pad's telltale hole. In accordance with API 650 Section 7.3.5 and API 653 Section 12.1.5, what test pressure is mandated, and what must be done with the telltale hole after testing?
A contractor elects to perform a diesel oil test on a new unbacked shell-to-bottom corner weld prior to internal vacuum box testing. According to API 650 and API 653 repair inspection practices, what is the minimum required contact dwell time for the diesel fuel before final visual inspection of the opposing face?