Steel Silo Longitudinal Seam Welding: What Actually Happens When Those Plates Meet

Longitudinal seam welding on a steel silo is the vertical weld that joins shell plates along the height of the cylinder. It's the seam that carries the hoop tension from the stored bulk material — cement, clinker, fly ash, coal, slag — and if it fails, you're not looking at a repair job. You're looking at a structural event. Done right, it's invisible and lasts 30+ years. Done wrong, it cracks within the first filling cycle. Here's how to tell the difference.

Why the Longitudinal Seam Matters More Than You Think

The Load Path You Can't Ignore

When you fill a 5,000-tonne cement silo, the material pushes outward against the shell. That's hoop tension — circumferential stress that tries to pull the silo apart like a sausage splitting on a grill. The longitudinal seams run parallel to that stress direction, which means they're carrying the full tensile load across every vertical joint.

For a 30-meter-diameter silo storing clinker at 1.2 t/m³, the hoop tension at the bottom shell course can exceed 2,500 kN/m. That longitudinal weld needs to transfer that load continuously across the joint. Any lack of fusion, any slag inclusion, any undercut deeper than 0.5 mm becomes a stress riser. And stress risers in cyclic loading environments? They crack. Every time.

I've seen a 2,800-tonne fly ash silo in Vietnam where a longitudinal seam cracked along 40% of its length after 18 months. The root cause was incomplete penetration — the welders had run the root pass too fast on a 14 mm plate, leaving a 2 mm unfused gap. Under cyclic filling and emptying, that gap became a fatigue crack that propagated through the entire weld throat. The repair cost $180,000 and took the silo offline for six weeks.

It's Not Just Strength — It's Watertightness

Longitudinal seams on external silos are also your weather barrier. Cement and fly ash are hygroscopic — they absorb moisture and set. If rainwater wicks through a pinhole leak in a longitudinal weld joint, you get a wet pocket inside the silo that hardens into a concrete plug. I've opened up silos in coastal Guangdong where the interior had 200 mm thick rings of hardened cement along the seam lines. The material flow was completely disrupted.

A full-penetration longitudinal weld eliminates this. A fillet weld or partial-penetration joint doesn't. That's not opinion — it's basic physics. Water finds the path of least resistance, and an unfused root face is a highway.

Welding Processes: What Works and What Doesn't

Submerged Arc Welding (SAW) — The Default for Shell Plates

For longitudinal seams on silo shell plates 8 mm and thicker, submerged arc welding is the standard. It gives you deep penetration (typically 6-10 mm in a single pass on 12-16 mm plate), consistent bead profile, and deposition rates of 6-10 kg/hour. That's 3-5 times faster than GMAW or FCAW for the same joint.

The typical setup for a longitudinal seam on a 12 mm shell plate: single-V groove with a 60° included angle, 2 mm root face, 4 mm root opening. Root pass with GMAW (to control the backside), then SAW fill and cap passes. Wire diameter 4.0 mm, flux basicity index above 1.8 for good toughness at low temperatures. Travel speed around 400-500 mm/min on the fill passes.

Here's the thing — SAW is forgiving on clean, flat plate. But silo shell plates aren't always clean. They've been rolled, stored outdoors, maybe rained on. Mill scale, rust, moisture — SAW doesn't tolerate any of it within 50 mm of the joint. If the fitters don't grind the groove faces to bright metal before welding, you get porosity. Guaranteed.

GMAW and FCAW — When SAW Won't Fit

For thinner shell plates (6-8 mm), or for repair welds in tight spaces, GMAW with 1.2 mm ER70S-6 wire at 22-28V and 180-250A is the practical choice. FCAW-G (gas-shielded flux core) is common on site because it handles slightly dirtier plate and works in light wind — up to about 8 km/h with a proper gas shroud.

Some contractors use self-shielded FCAW (FCAW-S) for field welds. I'm not a fan for longitudinal seams on silos storing abrasive materials. The slag system in most FCAW-S wires leaves a slightly rougher bead profile, and under material flow, that roughness becomes a wear point. For coal and slag silos especially, you want a smooth cap pass. SAW or GMAW delivers that. FCAW-S doesn't, consistently.

Electrogas Welding — Niche but Worth Knowing

For very thick shell plates (20+ mm) on large-diameter silos, electrogas welding can do the entire longitudinal seam in a single pass. It's fast — one pass, full penetration, done. But the equipment is specialized, the setup is finicky, and the heat input is massive. You're putting 40-60 kJ/cm into the plate, which means a wide heat-affected zone and potential toughness issues if you're not careful with preheat and interpass temperature.

We used electrogas on a 45-meter-diameter slag silo in Saudi Arabia — 22 mm shell plates at the bottom course. The welds passed UT inspection, but we had to run a full PWHT (post-weld heat treatment) cycle on every seam to bring the HAZ hardness back below 250 HV10. That added $12 per meter of seam. On 200 meters of longitudinal welds, it's not nothing.

Welding Procedure and Qualification — The Paperwork That Saves Your Silo

WPS and PQR: Non-Negotiable

Every longitudinal seam weld on an industrial silo needs a qualified Welding Procedure Specification (WPS) backed by a Procedure Qualification Record (PQR). The applicable standards depend on the project location: ASME Section IX for North American projects, EN ISO 15614-1 for European and most international work, GB/T 19866 for Chinese domestic projects.

The PQR test coupon must match the production conditions — same base metal grade (Q235B, Q345B, S355JR are common for silo shells), same thickness range, same welding process, same heat input range. The test lab runs tensile bends, macro sections, and impact tests on the coupon. If the longitudinal weld on the PQR fails a side bend test, the whole procedure gets rejected. No exceptions.

I've watched contractors try to use a generic structural steel PQR for silo shell welding. It doesn't work. The loading conditions are different — silo shells see cyclic hoop tension, not static compression. The joint geometry is different — vertical-up welding on a curved plate behaves differently than flat-position structural welding. Get a silo-specific PQR. It costs $2,000-4,000 to qualify. A failed seam costs 100 times that.

Welder Qualification — The Human Factor

A qualified WPS is worthless if the welder running the bead isn't qualified for that position and process. Welder qualification per ASME Section IX or EN ISO 9606-1 requires the welder to make a test coupon in the same position as the production weld — typically 3G (vertical up) or 4G (overhead) for longitudinal seams on a silo shell.

Here's a detail that catches people: the qualification thickness range. If a welder qualifies on a 12 mm coupon, they're typically qualified for 3 mm to 24 mm (per ASME, the qualified thickness is 2t for groove welds, where t is the coupon thickness). But if the silo shell has a 6 mm top course and a 20 mm bottom course, you need to verify the welder's qualification covers both. I've seen a project in Indonesia where the contractor's welders were qualified on 10 mm plate but were welding 18 mm bottom-course seams. The inspector caught it. Half the seams had to be cut out and re-welded.

Distortion Control — The Warping Problem Nobody Talks About

Why Shell Plates Bow After Welding

Welding puts heat into metal. Metal expands when hot and contracts when cool. A longitudinal seam on a 12 mm × 2,000 mm wide shell plate might see peak temperatures of 1,500°C at the weld pool, dropping to ambient over a 100 mm wide zone. That uneven heating and cooling creates residual stresses — and those stresses pull the plate out of flat.

On a typical silo shell, longitudinal welding can cause 5-15 mm of angular distortion across a 2-meter-wide plate. That doesn't sound like much until you're trying to fit the next shell course and the bolt holes don't line up. Or until the silo wall has visible waviness that the client's engineer flags as non-conforming.

What Actually Works to Control It

Sequence matters more than anything. Weld the longitudinal seams from the center of each shell course outward, alternating sides. This balances the shrinkage forces. Backstep welding — running each bead segment in the opposite direction to the overall progression — helps too. Segment length of 150-200 mm, backstep distance of 50 mm.

Pre-setting the plates (pre-bending them in the opposite direction of expected distortion) works for shop fabrication but is nearly impossible in field erection. For field work, the best tool is a rigid backing bar — a copper or steel bar clamped behind the root opening that acts as both a heat sink and a mechanical restraint. It reduces root pass distortion by 40-60%.

On a project in Kazakhstan, we were welding longitudinal seams on 16 mm Q345B plates in -15°C ambient. The thermal gradient was brutal — 1,500°C weld pool to -15°C plate. We used ceramic backing bars with integrated copper heat sinks and kept interpass temperature between 100-200°C. Distortion stayed under 8 mm per 2-meter plate. Without the backing bars, we were seeing 20+ mm. The difference was visible from 50 meters away.

Inspection and Acceptance Criteria

Visual Inspection — Your First and Best Tool

Every longitudinal seam gets a 100% visual inspection before any NDT. You're looking for surface-breaking defects: cracks, undercut, porosity, overlap, insufficient throat. Per EN 1090-2 (Execution Class EXC3 for silos), undercut deeper than 0.5 mm is rejectable. Surface porosity with any pore larger than 1.0 mm is rejectable. Cracks of any length are rejectable. Period.

Visual inspection catches maybe 60-70% of the defects that matter. The rest are subsurface — lack of fusion at the root, slag lines in the fill passes, lamellar tearing in the base metal. That's where volumetric NDT comes in.

UT vs. RT — Which NDT Method for Longitudinal Seams

Ultrasonic testing (UT) per EN ISO 17640 or ASME Section V Article 4 is the standard volumetric inspection for longitudinal silo welds. It detects lack of fusion, slag inclusions, and cracks in the weld volume. A 45° or 60° shear wave probe scanning from both sides of the weld, with a reference notch or side-drilled hole for calibration.

Radiographic testing (RT) gives you a permanent record — a film or digital image of the weld cross-section. It's better for porosity and slag detection but worse for planar defects (cracks, lack of fusion) that are oriented perpendicular to the beam. For longitudinal seams on silos, I recommend UT as the primary method with RT on 10% of seams as a cross-check.

Acceptance criteria per EN 5817 Level B (stringent) or Level C (moderate) depending on the design specification. Most silo designers specify Level B for longitudinal seams — that means no cracks, no lack of fusion, no incomplete root penetration, and slag inclusions limited to 3 mm length maximum.

Case Study: 6,000-Tonne Clinker Silo — Longitudinal Seam Welding in Practice

Project: 6,000-tonne clinker silo, 28 m diameter × 32 m cylindrical height, located in Central Java, Indonesia. Shell material: Q345B, thickness ranging 10 mm (top course) to 20 mm (bottom course). Total longitudinal seam length: approximately 1,400 meters across 16 shell courses.

Welding specification: SAW for all shell plates 12 mm and thicker (root pass GMAW, fill/cap SAW). GMAW for plates 8-11 mm. WPS qualified per EN ISO 15614-1, PQR tested with tensile, bend, and Charpy V-notch impact tests at -20°C (minimum 27 J). Welder qualification per EN ISO 9606-1, 3G position.

Field conditions: Ambient temperature 28-36°C, relative humidity 75-90%. Monsoon season meant daily rain from 2-5 PM. We scheduled welding for 6 AM to 1 PM, with tarpaulin shelters over the seam areas for the remaining hours. Plate surfaces were wiped with acetone within 30 minutes of welding to remove moisture.

Distortion management: Copper backing bars on all seams 12 mm and thicker. Weld sequence: center-out, alternating sides, backstep technique. Measured distortion after each course — average 6 mm per 2-meter plate, maximum 11 mm on the 20 mm bottom course. All within the 15 mm tolerance per EN 1090-2.

Inspection results: 100% visual inspection, 100% UT per EN ISO 17640 Level 2B, 10% RT per EN ISO 17636-1 Class B. First-pass UT acceptance rate: 94.2%. The 5.8% rejection rate was primarily root pass porosity (trapped moisture) and one instance of lack of fusion on a 16 mm seam where the welder had deviated from the WPS travel speed. All rejected welds were repaired per an approved repair WPS and re-inspected. Final acceptance: 100%.

Timeline: 1,400 meters of longitudinal seam welding completed in 47 working days with a crew of 8 welders and 4 fitters. Average production rate: 30 meters of completed, inspected seam per day. That's about 4.5 minutes per meter of 14 mm seam — realistic for field conditions with fit-up, welding, interpass cleaning, and inspection included.

FAQ

Q: Can longitudinal seams on a steel silo be welded from one side only?

A: Technically yes, with a backing bar. But for full-penetration joints on shell plates 10 mm and thicker, I strongly recommend welding from both sides or using a permanent backing bar with a verified root pass. One-sided welding without backing almost always leaves some root defect — lack of fusion, slag trapped at the root face, or uneven penetration. For a silo carrying 3,000+ tonnes of material, that's a risk you don't need to take.

Q: What's the minimum preheat temperature for longitudinal seam welding on Q345B shell plate?

A: Per EN 1011-2, for Q345B (equivalent to S355), preheat depends on thickness and hydrogen content. For plate thickness 12-25 mm with low-hydrogen processes (SAW with basic flux, GMAW), minimum preheat is typically 25-50°C. For thickness above 25 mm, 75-100°C. In practice, we preheat to at least 50°C on all seams 12 mm and thicker, measured 75 mm from the joint edge. In cold climates (below 5°C ambient), we bump that to 75°C minimum regardless of thickness.

Q: How much does longitudinal seam welding cost per meter on a steel silo?

A: For field-welded longitudinal seams on shell plates 12-16 mm thick, expect $25-45 per meter of completed, inspected seam. That includes fit-up, welding (root + fill + cap), interpass cleaning, visual inspection, and UT. Shop-welded seams on pre-fabricated shell courses run $15-25 per meter because the environment is controlled and the plates are flat. Thicker plates (18-22 mm) push the cost to $40-70 per meter in the field due to additional weld passes and longer inspection times. These numbers are for Southeast Asia and Middle East labor rates — European and North American projects run 2-3x higher.