Steel Silo Weld Map Documentation: The Blueprint Nobody Reads Until Something Breaks

A weld map is a numbered drawing that ties every welded joint in a silo to a specific procedure, welder, and inspection record. Without it, you're building a 5,000-tonne pressure vessel with no way to trace a crack back to the person who laid the bead. I've seen plants lose $180,000 in downtime because a single circumferential seam couldn't be matched to its radiograph. Here's how to get it right.

What a Weld Map Actually Is (and What It Isn't)

More Than a Drawing With Numbers

A weld map isn't a pretty isometric for the client's office wall. It's a forensic tool. Every weld joint on the silo shell, hopper, roof, and skirt gets a unique identifier — typically a combination of ring number, seam position, and azimuth angle. That identifier links to a Welding Procedure Specification (WPS), a Procedure Qualification Record (PQR), the welder's qualification certificate, and the inspection report for that specific joint.

On a 12-ring cement silo with 4 vertical seams per ring and 12 circumferential seams, you're looking at roughly 60 individual weld joints that need tracking. Add the hopper-to-shell connection, the roof rafter attachments, and the anchor chair welds, and you're past 80. Each one needs its own line in the documentation chain.

What Gets Mapped and What Gets Missed

The obvious ones are shell-to-shell circumferential seams and longitudinal (vertical) seams. Those get mapped on every project I've seen. What gets missed — and what causes problems 5 years later — are the secondary attachments. Stiffener-to-shell fillet welds. Ladder bracket penetrations. The 6 mm fillet weld where the dust collector support meets the roof plate. I've watched a 300 mm crack propagate from an unmapped bracket weld on a clinker silo in Vietnam. The bracket held a cable tray. Nobody thought it mattered until the crack reached the main shell seam.

Here's the rule I use: if it's welded and it's structural or pressure-retaining, it goes on the map. If it's welded and it supports something that could fall and kill someone, it goes on the map. That covers about 95% of the joints. The other 5% — temporary lifting lugs, for example — get documented in the as-built record but don't need full traceability.

The Documentation Chain: From Weld Number to Radiograph

Linking the Weld to the Procedure

Each weld number on the map points to a WPS. For a typical carbon steel silo shell using SAW (Submerged Arc Welding) on circumferential seams, the WPS specifies the filler metal classification (AWS A5.17 F7A6-EH14), preheat temperature (minimum 50°C for shell thickness above 25 mm), interpass temperature (maximum 250°C), and heat input range (1.5 to 2.5 kJ/mm). The PQR backs this up with actual test data — tensile strength, bend tests, and Charpy impact values if the design temperature drops below -20°C.

The welder who executes that joint needs to be qualified under EN ISO 9606-1 or AWS D1.1 for the specific position and process. Their qualification number goes next to the weld number on the map. If Welder 047 laid the bead at Joint C-R3-S2 (Circumferential, Ring 3, Seam 2), and that joint later fails ultrasonic testing, you know exactly who to talk to and exactly what procedure they were supposed to follow.

Inspection Records That Actually Mean Something

The inspection side of the chain includes visual inspection (VT), ultrasonic testing (UT), magnetic particle testing (MT), and radiographic testing (RT) — depending on the joint category. EN 14491 and the project specification usually dictate the inspection ratio. For a cement silo operating at near-atmospheric pressure, 10% UT on circumferential seams and 100% VT is common. For a coal silo with explosion risk, I've seen specs demand 100% UT on all shell seams plus MT on all hopper attachment welds.

Each inspection result gets a report number that references the weld map identifier. The report includes the date, the inspector's certification level (PCN or SNT-TC-1A Level II minimum for UT), the equipment serial number, and the acceptance criteria. If Joint C-R3-S2 shows an indication at 14 mm depth with a length of 25 mm, and the acceptance criteria per EN ISO 5817 Level B limits planar indications to 10 mm length, that's a reject. The map tells you which joint. The repair procedure tells you how to fix it. The re-inspection report closes the loop.

Common Mistakes That Turn Weld Maps Into Paperweights

Numbering Systems That Collapse Under Field Conditions

I've seen three different numbering systems on a single project. The fabricator used one system. The inspection company used another. The EPC contractor's as-built drawings used a third. Nobody caught it until the client asked for the UT report on Joint 14-B and three different joints matched that description across three different documents.

The fix is simple: define the numbering convention in the project quality plan before the first plate is cut. I use a system like [Joint Type]-[Ring/Section]-[Seam Number]-[Azimuth]. C-R3-S2-A90 means Circumferential, Ring 3, Seam 2, at 90 degrees azimuth. V-R5-S1 means Vertical, Ring 5, Seam 1. It's not fancy. But at 2am when you're trying to locate a crack indication on a 40-metre-tall silo, you don't want fancy. You want unambiguous.

Welder Qualification Gaps

A welder qualified for 1G (flat position) plate welding isn't qualified for 5G (horizontal fixed position) pipe welding — and the same logic applies to silo shell welds. Circumferential seams on a vertical silo shell are essentially 1G if the shell rotates on a positioner. But vertical seams on a bolted-constructed silo where the rings are welded in place? That's 3G or 4G depending on orientation. I've caught welders on site in Indonesia who had valid 1G qualifications but were running vertical-up beads on longitudinal seams with no 3G or 4G endorsement. The welds looked fine. They weren't qualified. Every joint they'd laid had to be re-inspected — 34 joints on a single silo, at roughly $120 per joint for UT. That's $4,080 in re-inspection costs because nobody checked the qualification scope against the joint position.

Digital Weld Mapping: Where the Industry Is Heading

From Paper Tags to QR Codes

The old method: a welder stamps a metal tag next to each joint with their welder number, and a QC inspector manually transcribes it onto a paper weld map. The error rate on manual transcription is around 3-5% in my experience. On a silo with 80 joints, that's 2-4 errors. Each error is a joint you can't trace.

The newer approach uses QR codes or RFID tags attached to each joint location. The welder scans the code with a tablet, which logs the joint ID, the welder's badge number (read from an RFID chip in their helmet or wristband), the timestamp, and the WPS they're working under. The data feeds directly into a database that generates the weld map in real time. I used this system on a project in Saudi Arabia — 14 silos, roughly 1,100 weld joints total. Zero transcription errors. The database flagged two instances where a welder's qualification was about to expire mid-project, giving us 30 days to arrange re-testing before the welder was locked out of the system.

Integration With BIM and As-Built Models

If the silo was designed in a 3D model — and most are now, even if it's just a structural frame in Tekla or Advance Steel — the weld map data can be embedded directly into the model. Click on Joint C-R3-S2 in the 3D viewer and you see the WPS, the welder's name, the UT report, the RT film, and the repair history if any. This is what the client actually needs 10 years later when they're planning a repair or an upgrade. I've handed over as-built models to clients in Oman and Colombia that included full weld traceability. The feedback was consistent: "This is the first time we've gotten documentation we can actually use."

Case Study: The Clinker Silo in Central Java

A 4,500-tonne clinker silo, 32 metres tall, 18-metre diameter, built in 2019. The fabricator was experienced — they'd built silos across Southeast Asia. But their weld map system was a spreadsheet. No unique joint identifiers. Just "Ring 3, Seam 2" with no azimuth reference. On a 18-metre diameter shell, "Seam 2" could be anywhere along a 56-metre circumference.

Six months after commissioning, a hairline crack appeared at the toe of a hopper-to-shell weld. The client wanted to know which welder laid that bead, what procedure they used, and whether adjacent joints were at risk. The spreadsheet couldn't answer. The fabricator's solution was to UT-test every hopper attachment weld — 96 joints — at a cost of roughly $15,000 and two weeks of downtime. With a proper weld map, we could have identified the exact joint, checked the specific welder's other work on that silo (4 joints), and targeted the inspection. The targeted approach would have cost under $2,000 and taken two days.

We rebuilt the weld map retrospectively using the fabricator's daily logs, welder time cards, and the original RT films. It took three weeks. The crack was traced to a joint where the preheat had been insufficient — the ambient temperature was 8°C, the WPS required 50°C minimum preheat for 30 mm plate, and the welder had used a single propane torch instead of the required four-torch array. The hardness reading at the crack toe was 382 HV10. The WPS limit was 280 HV10. That's a hydrogen-induced cracking scenario, and it was entirely preventable with proper documentation that linked the joint to the procedure and the procedure to the preheat requirement.

FAQ

Is a weld map legally required for steel silos?

Not as a standalone document in most jurisdictions. But the underlying requirements — welder qualification per EN ISO 9606-1 or AWS D1.1, WPS qualification per EN ISO 15614-1, and inspection per EN ISO 5817 or ASME Section VIII — are mandatory in most countries. The weld map is the tool that proves compliance with those requirements. Without it, you can't demonstrate traceability during an audit or after an incident. I've seen insurance claims denied because the plant couldn't produce weld documentation linking a failure to a specific procedure. The weld map is your evidence.

Who is responsible for creating and maintaining the weld map?

The fabricator creates it. The EPC contractor reviews and approves it. The third-party inspection company verifies it against the actual welds. The client maintains it for the life of the asset. In practice, the fabricator's QC department does the heavy lifting — assigning joint numbers, linking welders to joints, collecting inspection reports. But the EPC contractor needs to audit it. I've caught errors on 30% of the weld maps I've audited. Most were minor — a transposed digit, a missing azimuth. One was a systematic error where the fabricator had numbered the rings from top to bottom instead of bottom to top, which meant every joint identifier was wrong. That silo had 74 joints. All 74 identifiers had to be corrected, and all linked documents had to be reissued.

How detailed should the weld map be for a small silo — say, 500 tonnes?

The same level of detail as a 5,000-tonne silo. The number of joints is smaller — maybe 25-30 instead of 80-100 — but the consequences of a failure aren't proportionally smaller. A 500-tonne fly ash silo that ruptures still releases 500 tonnes of material, still creates a dust cloud, still kills anyone in the blast zone. The documentation effort for a small silo is maybe 20% of the effort for a large one. The cost of doing it wrong is the same. I've never understood why some fabricators treat small silos as "no-documentation-needed" projects. That's not a risk assessment. That's a gamble.