Steel Silo Weld Repair Procedures: What the Inspection Report Won't Tell You

You found a crack in your silo shell. Now what? First, stop treating it like a scratch on a truck bumper. Second, stop letting the maintenance team grind it out without a Welding Procedure Specification (WPS). Steel silo weld repair isn't about filling a gap with molten metal. It's about restoring structural integrity under dynamic loading, thermal cycling, and abrasive wear. Get it wrong, and you're looking at a catastrophic shell rupture.

Assessing the Damage Before You Strike an Arc

You can't fix what you don't understand. Before anyone touches a grinder, you need to know exactly what you're dealing with. Is it a surface scratch? A toe crack propagating from a circumferential weld? Or is it a through-wall failure caused by internal pressure cycling?

Visual vs. Volumetric Flaw Detection

Visual inspection catches maybe 10% of the problem. I'm talking about obvious tears, heavy corrosion, or blown-out welds from a dust explosion. But the real killers hide underneath. You need volumetric testing. Magnetic Particle Testing (MT) will find surface and slightly subsurface discontinuities—think ISO 17638 standards here. For anything deeper, you're running Ultrasonic Testing (UT). UT maps the depth and orientation of the crack. If you skip UT, you're guessing. And guessing gets people killed.

Mapping the Stress Field Around the Defect

Here's the thing. A crack in a silo shell isn't just a material failure. It's a stress relief mechanism. The steel cracked because the localized stress exceeded the yield strength. If you just fill the crack without addressing the stress concentration, it'll crack again. Right next to your shiny new weld. We use Finite Element Analysis (FEA) to map the stress field, especially around hopper transitions and stiffener zones. Eurocode 3 (EN 1993-1-6) gives you the framework for shell structures, but you have to apply it locally. Calculate the membrane stress and the bending stress at the defect. If the stress concentration factor exceeds 2.5, you don't just repair the weld. You reinforce the panel.

Surface Preparation and Joint Design

I was on a site in Indonesia, and the contractor had already ground out a 2-meter longitudinal crack before calling us. They'd removed 4mm of base metal, turning a repairable defect into a structural nightmare. We had to install a 600mm by 1200mm reinforcement plate over the gouged-out section, adding 120kg of steel to a spot that should have just needed a weld. Don't be that guy.

Gouging vs. Grinding: Choosing the Right Metal Removal Rate

Carbon arc gouging is fast. It'll chew through a defective weld in minutes. But it leaves a high-carbon recast layer on the steel. That layer has a completely different metallurgy than the base metal. If you weld over it, you get micro-cracking in the Heat Affected Zone (HAZ). Every time. You must grind after gouging. Take off another 1.5mm to 3mm of material with a rotary cup wheel until you see clean, bright steel. Grinding alone is slower, but for shallow cracks under 5mm deep, it's the only way to go. It doesn't alter the base metal chemistry.

Bevel Geometry for Thick-Section Repairs

When you're dealing with shell thicknesses over 12mm, your joint design dictates the residual stress in the repair. A single-V bevel with a 60-degree included angle and a 1.6mm root face works for most thicknesses. But for thick shells—say, 25mm or 30mm at the hopper transition—you need a double-V or a U-groove. Why? Because a single-V requires massive weld volume. More weld volume means more heat input. More heat input means more distortion. And a warped silo shell doesn't hold pressure. A double-V splits the weld volume roughly 50/50, balancing the shrinkage forces and keeping the shell flat.

The Repair Weld Execution

This is where the rubber meets the road. You've got your WPS, your qualified welder, and your prepped joint. Now you have to lay the bead without destroying the base metal's mechanical properties. Silos aren't static structures. They fill, they empty, they heat up in the sun, they cool down at night. The repair weld has to survive that fatigue cycle for twenty years.

Pre-heat and Interpass Temperature Control

Pre-heat isn't optional. It's physics. You're bringing the base metal up to a specific temperature—usually between 150°C and 250°C for carbon steel—to slow down the cooling rate of the weld pool. Slow cooling prevents the formation of hard, brittle microstructures like martensite. If the weld cools too fast, it cracks under its own residual stress. Use propane torches or induction heating blankets. Measure the temp with a contact pyrometer or infrared thermometer. Don't guess. I've seen welders check the pre-heat by spitting on the steel. Unacceptable.

Interpass temperature matters just as much. That's the temperature of the base metal between weld passes. If it drops below your minimum pre-heat, you stop. Reheat. If it exceeds 250°C to 300°C (depending on the WPS), you're risking grain growth and losing toughness. You wait. Welding a silo is a stop-and-go process. Rushing the interpass temp is the number one cause of weld failure in the field.

Welding Procedure Specifications and Filler Metal Selection

Your WPS dictates everything. Voltage, amperage, travel speed, shielding gas flow. For vertical-up repairs on a silo wall, you're usually running Shielded Metal Arc Welding (SMAW) with E7018 electrodes, or Flux-Cored Arc Welding (FCAW) with E71T-1 wire. Match the filler metal to the base metal yield strength. If you're repairing a Q345B shell, don't use E6010 filler. You need the tensile strength to carry the load.

Got a call at 2am from a plant in Shandong — the night crew had let the interpass temperature drop to 15°C on a 20mm shell repair because they were rushing to beat a monsoon. We scrapped the whole weld. 14 hours of work, gone. They had to re-gouge, re-grind, and re-weld the entire joint. The delay cost them three days of clinker production. At $15,000 a day in lost revenue, that cold weld was a $45,000 mistake.

Post-Weld Verification and Testing

The weld looks good. The slag is chipped. The brush is shining. But you're not done. Not even close. You have to verify that the repair actually holds. And you have to verify it without putting the silo back into service prematurely.

Non-Destructive Testing (NDT) Hold Points

After the weld cools to ambient temperature, you wait. Hydrogen-induced cracking can take 24 hours to manifest. So you wait 24 hours minimum. Then you run NDT again. Visual Testing (VT) first. Then MT or Penetrant Testing (PT) on the surface. If the repair is structural—load-bearing—you run UT on the internal flaws. Acceptance criteria should follow ISO 5817, Level B for cyclic loading applications. Level B is strict. It doesn't allow for slag inclusions over 1.5mm or porosity clusters. If you find defects, you mark them, grind them out, and start the whole repair sequence over again.

Pressure and Leak Testing Protocols

Once the weld passes NDT, you test the silo. For a cement or fly ash silo, you're usually doing a pneumatic pressure test. You seal the silo, pressurize it to 1.3 times the design pressure—but never exceeding the Maximum Allowable Working Pressure (MAWP) plus 10%—and check every weld with a soap solution. Watch for bubbles. If you're dealing with a 2,000-tonne silo with a 1,500 Pa design pressure, you're testing at 1,950 Pa. Hold it for 30 minutes. No pressure drop. No bubbles. Then you can introduce the bulk material.

Case Study: The 4,000-Tonne Clinker Silo in Vietnam

A client called us about a 24-meter diameter clinker silo. They'd found a 450mm circumferential crack right at the hopper transition joint. The shell was 8mm thick, but the transition ring was 16mm. Dissimilar thicknesses. High stress concentration. The original EPC contractor had used a single-V bevel and filled it with a massive weld deposit. The residual stress pulled the 8mm shell out of plane by 12mm. Clinker was leaking through the crack.

We mapped the crack with UT. It ran 3mm deep into the 8mm shell. We set up a localized weather enclosure because the site humidity was hitting 95%—ambient moisture is a hydrogen source, and hydrogen kills welds. We pre-heated the joint to 200°C using induction blankets. We gouged out the entire defective weld, ground the recast layer away, and re-beveled it to a 60-degree double-V. We ran SMAW with E7018 electrodes, strictly controlling the interpass temperature between 200°C and 230°C. The repair took 14 hours of continuous welding. We waited 36 hours before running MT and UT. Zero defects. We pressurized the silo to 2,100 Pa. Held it. No leaks. That silo has been running at full capacity for three years now.

Frequently Asked Questions

Can we weld repair a silo while it's full of material?

No. Absolutely not. First, the bulk material acts as a massive heat sink. Your weld pool won't fuse properly to the base metal because the cement or clinker is sucking the heat away. You'll get cold lap and lack of fusion defects. Second, welding on a pressurized, loaded shell changes the stress distribution dynamically. You risk a localized buckling failure. Empty the silo. Clean the repair zone. Then weld.

How much does a typical steel silo shell repair cost?

Your mileage varies wildly depending on access and location. A simple 300mm surface crack repair on a ground-level silo might cost you $3,000 to $5,000, including NDT and scaffolding. But a structural repair at a 30-meter elevation on a 20mm thick shell? You're looking at $25,000 to $80,000. That includes specialized access scaffolding, induction heating setup, qualified welders, and full volumetric NDT. The steel itself is cheap. The access and verification are what you're paying for.

What is the biggest mistake engineers make when specifying silo weld repairs?

They treat it like a static tank repair. Tanks hold liquid. The load is predictable. Silos hold bulk solids. The load is chaotic. Asymmetric filling and unloading create localized bending moments. Friction between the bulk material and the shell wall creates massive vertical compressive stresses. If your repair weld is harder than the base metal—because you used the wrong filler or skipped the pre-heat—it won't yield when the shell flexes. It'll crack. The repair needs to be ductile. It needs to move with the shell.

Trust Your Silo Repairs to Manxing

At Manxing, we don't just design and build bulk storage silos—we fix them when the unexpected happens. With over 200 silo repair and retrofit projects across 30 countries, our EPC teams bring field-tested weld procedures, rigorous NDT protocols, and induction heating setups to your facility. Whether it's a hairline crack in a fly ash silo or a structural failure in a slag storage system, we deliver repairs that hold. Contact our engineering team today for a site assessment and a repair procedure that actually works.