How to Leak-Test Thin-Sheet Welds in Tower Internals
Many tower internals depend on thin-sheet welds for liquid tightness. Collector decks, distributor troughs, sumps, feed galleries, internal pans, and segmented joints may contain long weld seams fabricated from relatively thin stainless steel or alloy plate.
A weld can appear continuous and still contain pinholes, lack of fusion, crater cracks, or intermittent leakage paths.
Selecting the wrong examination method creates two risks: accepting a leaking internal or damaging the internal during testing.
No single method is ideal for every tower component. The test must match weld access, permitted pressure, material, joint geometry, cleanliness, and the type of defect being sought.
Visual Examination Comes First
Leak testing should not replace good fabrication inspection.
Before any pressure or tracer method is used, the weld should be examined for:
Missed weld length.
Crater cracks.
Arc strikes.
Undercut.
Burn-through.
Excessive reinforcement.
Sharp transitions.
Distortion.
Weld spatter.
Incomplete cleaning.
Visual examination can find obvious defects cheaply and prevent wasting time on testing a visibly unacceptable seam.
However, visual acceptance does not prove liquid tightness. Microscopic pinholes and connected porosity may be invisible.
Liquid-Fill Testing
Filling a distributor or collector with water is intuitive and directly demonstrates whether liquid escapes.
Advantages include:
Simple equipment.
Direct observation.
Realistic liquid loading.
Simultaneous level and deflection check.
Ability to assess drains and overflow paths.
Limitations include:
Very small leaks may be difficult to see.
Surface tension may temporarily seal fine defects.
Evaporation and wet surfaces complicate interpretation.
Water may be incompatible with cleanliness requirements.
Complete drying may be difficult.
Chloride-contaminated water can threaten stainless steel.
The structure may not be designed for unrestricted test depth.
The test liquid quality, maximum fill level, hold time, drainage, and drying procedure should be specified.
Bubble Testing With Low-Pressure Air
A weld can be pressurized gently on one side while a bubble-forming solution is applied to the other. Escaping gas creates visible bubbles.
This method can detect small through-leaks more sensitively than a short water test. However, thin-sheet equipment must not be exposed to uncontrolled air pressure.
Compressed gas stores energy. A small pressure acting over a large panel area can generate enough force to buckle or rupture the internal.
Testing should use:
A calibrated low-pressure regulator.
A relief device.
A sensitive pressure gauge.
Defined maximum test pressure.
Controlled isolation.
Suitable bubble solution.
Clean surfaces.
Safe venting.
Ordinary shop air should not be connected directly without pressure-limiting protection.
Vacuum-Box Testing
A vacuum box is placed over a section of weld coated with bubble solution. Air drawn through a leak forms bubbles beneath the transparent box.
This is useful for flat or gently curved seams when only one side is readily accessible. It avoids pressurizing the entire internal.
Limitations include:
Difficulty sealing over corners and stiffeners.
Limited coverage per box placement.
Missed areas between test positions.
Surface preparation requirements.
Poor suitability for complex pipe intersections.
Operator dependence.
The box overlap and test sequence should be recorded so the full seam is covered.
Dye-Penetrant Examination
Liquid penetrant testing can reveal surface-breaking cracks and porosity. It is valuable for weld-quality assessment, but it is not identical to a through-leak test.
A penetrant indication may not pass completely through the sheet. Conversely, a tortuous through-path may not produce a clear indication on the examined surface.
Penetrant residues can also be difficult to remove from high-purity or chemically sensitive equipment. Cleaning compatibility must be reviewed.
Penetrant examination is best used as a complementary weld-integrity method rather than the only proof of leak tightness.
Helium and Tracer-Gas Methods
Tracer-gas testing can detect extremely small leaks and is valuable for high-integrity or high-purity service.
Helium is introduced on one side while a detector checks the other. The method is sensitive but requires specialized equipment, controlled background conditions, trained operators, and an agreed acceptance threshold.
Using an extremely sensitive method without a service-based acceptance criterion can cause unnecessary repair of leaks that have no practical hydraulic significance.
The specified leak rate should match the consequence of leakage.
Choosing the Correct Method
The selection should consider:
Is the requirement structural integrity, liquid tightness, gas tightness, or all three?
Can both sides of the weld be accessed?
What pressure can the internal safely withstand?
Is water permitted?
Can the component be dried fully?
Are test residues acceptable?
Is the weld flat, curved, or geometrically complex?
What leak size matters to process performance?
Will testing occur in the shop or inside the tower?
Does the method cover field joints?
A distributor trough may be best checked by water fill plus targeted bubble testing. A flat collector seam may suit vacuum-box testing. High-purity gas-tight piping may justify tracer-gas methods.
Avoiding Test-Induced Damage
Test pressure must be based on the internal’s structural capacity, not the available plant-air pressure.
Temporary blanks and seals also carry load. A poorly restrained blank can become a projectile, while an overstressed panel may remain permanently distorted after the test.
Water weight must be included when testing large collector decks. Deflection can change weld stress and produce leakage that would not appear in a small local test.
Repair and Retest
Leaking areas should be marked precisely before draining or depressurizing. Repairs should follow an approved procedure that limits distortion and material damage.
After repair, the affected area must be retested using the original method or an approved equivalent. Visual examination alone is insufficient.
Repeated repairs in the same area may indicate poor joint design, contamination, access limitations, or excessive restraint rather than isolated welder error.
Inspection Documentation
The final record should include:
Internal identification.
Drawing and weld references.
Test method.
Test medium.
Pressure or vacuum level.
Hold time.
Test-liquid quality.
Examined areas.
Leak locations.
Repairs and retest results.
Photographs are useful, but a marked drawing provides clearer evidence that all seams were covered.