Steel Silo Welder Qualification Testing: What It Covers, Why It's Non-Negotiable, and How to Get It Right
Welder qualification testing is the documented proof that a welder can produce sound welds under defined conditions — not a suggestion, not a box-ticking exercise, but a hard requirement. For steel silos storing cement, clinker, coal, fly ash, or slag, a single bad weld can become a crack, a leak, and eventually a structural failure. This article covers what the testing actually involves, which standards apply, how mechanical and NDT evaluation work, and the reasons most welders fail — so you can make sure your next silo project doesn't end up as a cautionary tale.
Why Welder Qualification Matters More for Silos Than Almost Any Other Structure
The Load Environment Nobody Talks About
People treat silos like big bins. They're not. A 3,000-tonne cement silo with a 16-metre diameter and 30-metre height has roughly 800 to 1,200 metres of welded seam — shell circumferential joints, longitudinal seams, hopper connections, roof stiffeners, manhole reinforcements. Every single one of those joints sees pressure cycling every time the silo fills and empties. Cement plants cycle their silos two to four times per day. That's 730 to 1,460 pressure cycles per year. Fatigue is the enemy, and fatigue starts at weld defects.
Then there's the material itself. Clinker at 120°C when it goes into the silo. Fly ash that's abrasive enough to wear down a steel wall over ten years. Coal that generates dust with a KSt value of 150 to 200 bar·m/s — that's a St1 dust explosion class, and if you've got a weld porosity pocket full of methane or CO, you're looking at a problem that doesn't stay small.
The Consequence of a Bad Weld
I was on a site in Indonesia where a 2,500-tonne cement silo developed a 2-metre crack along a circumferential seam after just 18 months of service. The crack started at a lack-of-fusion defect in the root pass — something that should have been caught during qualification testing or radiographic examination. The repair cost $47,000 in crane time, welding, and lost production. The root cause? The welder had qualified on flat plate in a shop. Nobody tested him on the actual curved geometry of the silo shell. The weld pool behaved differently. He couldn't see the root. And the defect sat there, invisible, for a year and a half before it found its way to the surface.
Here's the thing. A building beam can tolerate a 2-mm undercut. Nobody cares. But that same undercut on a silo shell seam, under cyclic pressure loading with abrasive material flowing past it, becomes a stress concentrator. Crack initiation within 10,000 cycles. Crack propagation through the wall within 50,000 cycles. That's six months of normal operation.
The Qualification Framework: WPS, WPQR, and Welder Performance
Welding Procedure Specification (WPS)
Before any welder strikes an arc on your silo, you need an approved WPS. This is the recipe — base metal grade, filler metal classification, shielding gas composition, preheat temperature, interpass temperature limit, amperage range, voltage range, travel speed, joint design, welding position. For a typical Q345B steel silo with shell thickness ranging from 6 mm at the top to 25 mm at the hopper, you might need three to four separate WPS documents covering different thickness ranges and positions.
The WPS is qualified through a Welding Procedure Qualification Record (WPQR). You weld a test plate using the proposed parameters, then mechanically test it. Tensile test, bend tests, impact tests if the design temperature requires it. The WPQR proves the procedure produces welds with the required mechanical properties. Without a valid WPQR, you cannot qualify welders. Period.
Welder Performance Qualification
Once the WPS is approved, individual welders qualify by welding test coupons under that procedure. The test coupon is then evaluated — usually by a combination of visual examination, mechanical testing, and/or non-destructive testing. The welder's qualification is recorded on a Welder Qualification Record (WQR or WPQ), which specifies exactly what they're qualified for: material group, thickness range, welding position, joint type, and process.
The critical standards here are ASME Section IX (common for projects with American design codes), AWS D1.1 (structural welding, frequently referenced for silo support structures), EN ISO 9606-1 (European and increasingly common in international projects), and GB/T 19866 (Chinese national standard for welder qualification). If your project is in Southeast Asia or Africa, you'll often see EN ISO 9606-1 specified even though it's a European standard — it's become the de facto international benchmark.
Qualification Ranges and Expiry
A welder qualified on a 12-mm plate in the PA (flat) position is NOT automatically qualified for a 20-mm plate in the PF (vertical-up) position. Each variable matters. Thickness qualification ranges are defined in the standard — under EN ISO 9606-1, a test piece thickness of 10 mm qualifies the welder for thicknesses from 3 mm up to 20 mm. But go outside that range, and you need a new test.
Position matters even more for silo work. Most shell welding is done in the PF (vertical-up) position for circumferential seams and PC (horizontal) for longitudinal seams. A welder qualified only on flat plate will struggle — I've seen it dozens of times. Qualification is typically valid indefinitely as long as the welder is actively working and their employer certifies continued competence at six-month intervals. But if a welder hasn't used the process for more than six months, requalification is required. No exceptions.
Mechanical Testing: What the Lab Actually Checks
Tensile Testing
The transverse tensile test pulls a weld specimen until it breaks. The acceptance criterion is that the tensile strength must meet or exceed the minimum specified for the base metal. For Q345B steel, that's 470 to 630 MPa. If the weld breaks in the fusion zone below that value, the welder fails. If it breaks in the base metal above that value, it passes — the weld is stronger than the parent material, which is what you want.
For silos using higher-strength steels like S355J2 or Q420B, the bar is higher. I've seen a project in Kazakhstan where the client specified S420ML for a 10,000-tonne coal silo — minimum tensile strength of 520 MPa. Two welders failed their first qualification because their weld metal was undermatching. Turned out they were using the wrong filler classification. Switched from ER70S-6 to ER80S-G and both passed on the second attempt.
Guided Bend Testing
This is where most welders actually fail. The transverse side-bend test takes a slice of the weld, bends it 180 degrees around a mandrel (typically 4 times the specimen thickness in diameter), and examines the convex surface for defects. Any crack or lack-of-fusion indication exceeding 3 mm in any direction is a rejection.
The side-bend test is brutal because it stresses the root and the sidewall fusion zones simultaneously. A welder who runs too cold, too fast, or with an inconsistent weave pattern will show lack of fusion at the sidewalls. I've watched a welder in Vietnam produce a perfect-looking external weld — great profile, no undercut, nice ripples — and then fail the side-bend test because he'd been skipping the sidewall pause. The weld looked beautiful. It was hollow on the inside.
For silos with shell thickness above 12 mm, you typically test two side-bend specimens — one from each side of the weld. For thickness above 20 mm, four face-bend and four root-bend specimens may be required under some standards. The testing lab charges roughly $200 to $500 per qualification test plate depending on the scope. Not expensive compared to a field repair.
Impact Testing (When Required)
If your silo is designed for low-temperature service — say, a coal silo in northern China where ambient temperatures hit -30°C — the WPS qualification will require Charpy V-notch impact testing. The test temperature is usually 20°C below the design temperature, and the minimum absorbed energy is typically 27 J (per EN 1993-1-10) or as specified by the design code.
Welder qualification itself doesn't usually include impact testing — that's a WPS-level requirement. But if the WPS requires impact testing and the welder's procedure qualification coupon fails impact, the welder can't be qualified under that WPS. The procedure has to be re-qualified first.
Non-Destructive Testing of Qualification Coupons
Radiographic Examination (RT)
Some qualification schemes allow radiographic examination as a substitute for mechanical bend testing — this is common under EN ISO 9606-1. The test coupon is radiographed to EN ISO 17636 (film or digital), and the film is evaluated to EN ISO 5817 quality levels. For most silo applications, acceptance level B (stringent) or C (intermediate) applies, depending on the design code and the fatigue classification of the joint.
Under EN ISO 5817 Level B, the maximum allowable porosity is tight — individual gas pores cannot exceed 3 mm diameter, and the total projected area of porosity in a 100 mm length of weld cannot exceed 2%. Lack of fusion, incomplete penetration, and cracks are zero-acceptance defects. None. A single crack indication, 1 mm long, and the welder fails.
We had a project in Shandong where the client specified RT evaluation for all welder qualification coupons on a 5,000-tonne fly ash silo. Of the first 12 welders tested, 4 failed — 2 for porosity clusters, 1 for incomplete root penetration, 1 for a crack in the heat-affected zone. That's a 33% failure rate on the first pass. After retraining and re-qualification, the failure rate dropped to under 10%. The cost of catching those 4 bad welders in the qualification lab was about $1,200 in testing fees. The cost of discovering them in the field — after the silo was erected and under pressure — would have been $30,000 to $80,000 in repair, plus schedule delay.
Ultrasonic Testing (UT) and Magnetic Particle Inspection (MT)
UT is increasingly used for qualification coupons on thicker sections — above 25 mm — where radiography becomes less sensitive to planar defects like lack of fusion. The acceptance criteria follow EN ISO 17640 or ASME Section V, with evaluation to the same quality levels as RT.
MT is typically used as a supplementary technique for surface-breaking defects. It's fast, cheap ($40 to $80 per coupon), and catches cracks that RT might miss if they're oriented parallel to the beam. For hopper welds on abrasion-resistant steel (Hardox 400 or NM400), MT is almost mandatory because these steels are prone to hydrogen-induced cracking in the HAZ, especially if preheat isn't maintained above 100°C.
Common Failure Reasons and How to Prevent Them
The Top 5 Reasons Welders Fail Silo Qualification
After reviewing qualification results across dozens of projects, the failure pattern is remarkably consistent:
1. Lack of fusion at the root (35-40% of failures). Usually caused by insufficient heat input, incorrect electrode angle, or excessive travel speed on the root pass. Fix: increase amperage by 10-15%, reduce travel speed by 20%, and make sure the welder can actually see the root opening.
2. Porosity (20-25% of failures). Moisture on the base metal, contaminated shielding gas (flow rate below 15 L/min or above 25 L/min causes turbulence and air entrapment), or filler wire with damaged copper coating. On one project in Malaysia, we traced a cluster of porosity failures to a gas hose that had been left in the sun — the hose permeability increased, and moisture migrated into the gas stream. Replaced the hose, problem disappeared.
3. Undercut (15-20% of failures). Excessive amperage or travel speed on the fill and cap passes. Undercut deeper than 0.5 mm is a rejection under most silo welding standards. The fix is training — welders need to slow down at the toes and hold the arc slightly longer to allow the weld pool to fill the groove.
4. Incomplete penetration (10-15% of failures). Root opening too narrow, land too thick, or welder not reaching the root. For single-V butt joints on silo shell, the root opening should be 3 to 4 mm with a 1 to 2 mm land. Anything less than 2.5 mm root opening and you're asking for trouble.
5. Crack in the HAZ (5-10% of failures). Hydrogen-induced cracking, usually from inadequate preheat. For Q345B above 20 mm thickness, preheat to at least 80°C. For Q420B or S355J2 above 25 mm, preheat to 100-120°C and maintain interpass temperature below 250°C. Check with a contact thermometer, not by touching the plate with your hand.
The Geometry Problem Nobody Accounts For
Here's where the industry gets it wrong. Most welder qualification is done on flat test plates in a shop. But silo shell is curved. A 16-metre diameter silo has a radius of curvature of 8,000 mm — it looks flat, but it's not. The weld pool behaves differently on a curved surface. Gravity pulls the pool slightly toward the centre of curvature. The welder's hand position changes. The interpass temperature distribution changes.
I always recommend — and we specify this on every Manxing project — that at least one qualification test per welder be performed on a curved test piece matching the shell radius. It costs an extra $150 to $300 per welder to fabricate the curved fixture. It saves thousands in field rejection rates.
Case Study: 6,000-Tonne Clinker Silo — Vietnam
We were the EPC contractor for a 6,000-tonne clinker silo at a cement plant in central Vietnam. Shell diameter: 20 metres. Shell height: 32 metres. Material: Q345B, thickness ranging from 8 mm (top shell) to 22 mm (hopper). Design code: EN 1993-4-1 (Eurocode for silos). Welding standard: EN 1090-2, Execution Class EXC3.
The fabrication subcontractor had 18 welders, all qualified to EN ISO 9606-1 on flat plate. We required re-qualification on curved test pieces matching the 10-metre shell radius. Of the 18 welders, 7 failed their first curved-geometry qualification — 4 for lack of fusion at the root, 2 for porosity, 1 for undercut. That's a 39% failure rate on the first attempt.
After two weeks of retraining — focused on root pass technique, gas flow optimization, and travel speed control — 6 of the 7 passed re-qualification. The 7th welder was reassigned to non-structural work (platform handrails, ladder rungs). Total cost of re-qualification: approximately $4,200 in test plates, lab fees, and welder time. Estimated savings from avoided field repairs: $60,000 to $100,000, based on our experience with similar projects where welders were qualified only on flat plate.
The silo was commissioned in March 2023. First RT survey of production welds (10% sampling per EN 1090-2 EXC3): 2 defects found out of 84 joints examined. 2.4% rejection rate. Both defects were repaired and re-examined with no further issues. The plant manager told us it was the cleanest silo commissioning he'd seen in 25 years. That's what proper welder qualification buys you.
FAQ
How long does a welder qualification remain valid?
Under EN ISO 9606-1, welder qualification is valid indefinitely provided the welder's employer confirms at intervals not exceeding six months that the welder has been actively using the qualified process. If a welder hasn't used the process for more than six months, requalification is required. ASME Section IX follows a similar principle but requires the employer to maintain a continuity log. In practice, most project specifications require re-qualification if the welder has been off the process for more than three months.
Can a welder qualified under one standard be accepted under another?
Not automatically. A qualification to ASME Section IX doesn't directly transfer to EN ISO 9606-1, and vice versa. The testing methods, acceptance criteria, and essential variables differ. However, some third-party inspection bodies offer cross-standard evaluation — they'll review an existing qualification against the requirements of the target standard and issue a supplementary assessment. This costs $300 to $600 per welder and takes one to two weeks. For international projects with mixed design codes, this is often cheaper than full re-qualification.
What percentage of production welds should be RT-tested after welder qualification?
It depends on the execution class and the joint's fatigue classification. Under EN 1090-2 EXC3 (typical for silos), the minimum is 10% for secondary joints and 20-100% for primary joints, depending on the joint category. Under ASME Section VIII (if the silo has pressure-retaining requirements), the joint efficiency selected during design determines the examination ratio — a joint efficiency of 0.85 requires spot RT (minimum 10% per weld), while 1.0 requires 100% RT. For most cement and clinker silos, we specify 20% RT on circumferential shell seams and 10% on longitudinal seams. That's the sweet spot between quality assurance and cost.
Get Your Silo Welding Right From Day One
Welder qualification isn't paperwork. It's the first line of defence against field failures that cost tens of thousands of dollars and weeks of schedule. At Manxing, we've managed silo welding quality across projects in 30+ countries — from 500-tonne specialty powder silos to 40,000-tonne clinker storage. We specify curved-geometry qualification, enforce EN ISO 5817 Level B acceptance, and track every welder's continuity log from day one of fabrication.
If you're planning a steel silo project and want to make sure the welding programme is set up correctly — from WPS development through production NDT — talk to our engineering team. We'll review your specification, identify gaps before fabrication starts, and make sure your welders are qualified for the actual work they'll be doing — not just the work they did in a shop six months ago.
```