Silo Internal Inspection Procedures Using Confined Space Protocols

Silo internal inspection requires a written confined space entry permit, continuous atmospheric monitoring, and a dedicated standby person outside the vessel. You're looking for wall thickness loss, weld cracking, bolt loosening, and material buildup that changes discharge patterns. Skip any one of those three requirements and you're gambling with someone's life inside a steel tube full of cement dust.

Why Internal Inspection Isn't Optional

I've stood at the bottom of a 4,000-tonne clinker silo in central Vietnam, headlamp on, respirator sealed, and watched a 300 mm section of wall flake off six metres above me. That flake was 4.2 mm thick. The original plate was 12 mm. Nobody had been inside that silo in four years.

What You're Actually Looking For

Internal inspection isn't a walk-around. It's a systematic examination of every surface a camera can't reach. You're checking wall thickness with ultrasonic testing at grid points — typically on a 500 mm × 500 mm pattern for cement and clinker silos, tighter at the cone-to-shell junction where abrasion concentrates. You're tapping welds with a 500-gram hammer looking for delamination. You're counting bolt heads on flange connections and comparing against the original assembly drawing.

Material buildup is the silent killer. I've seen 80 tonnes of compacted fly ash clinging to one wall of a 2,500-tonne silo, creating asymmetric loading the structural engineer never designed for. That silo developed a 15 mm shell deformation within 18 months. The owner thought the foundation had settled. It hadn't. The weight was hanging on the wall like a barnacle.

The Cost of Waiting

A full internal inspection on a 3,000-tonne cement silo runs between $8,000 and $15,000 depending on height, access, and whether scaffolding or rope access is used. A wall replacement on that same silo — if you catch it early — might cost $25,000 to $40,000. Wait until the shell fails under load and you're looking at $200,000 plus production downtime that can stretch past three weeks.

The math isn't complicated. Inspect every 24 months for cement and clinker service. Every 18 months if you're storing abrasive materials like slag or coal with high silica content. Every 12 months if the silo handles hot material above 80°C — thermal cycling accelerates fatigue cracking at weld toes.

Confined Space Classification: Your Silo Qualifies

Here's the thing. Almost every industrial bulk storage silo meets the legal definition of a confined space under OSHA 29 CFR 1910.146 and equivalent standards globally. It's large enough to enter. It has limited entry and exit points. It's not designed for continuous occupancy. That's the definition. Your silo checks all three boxes.

Permit-Required vs. Non-Permit Entry

If the silo has any recognized serious safety hazard — and material engulfment alone qualifies — it's a permit-required confined space. Full stop. You need a written entry permit signed by the entry supervisor. You need atmospheric testing before entry and continuous monitoring during entry. You need a standby person stationed at the entry point with no other duties.

Some plants try to classify their silos as non-permit spaces after purging. Don't accept that shortcut. Residual material, rust scale, and pyrophoric deposits can generate hazardous atmospheres hours after the silo is emptied. I've measured CO levels at 1,200 ppm inside a silo that had been empty and ventilated for 48 hours. The rust was reacting with residual moisture. Nobody expected it.

Atmospheric Testing Protocols

Test in this order: oxygen, combustible gases, toxic substances. Oxygen must be between 19.5% and 23.5%. Below 19.5% you're in an oxygen-deficient atmosphere. Above 23.5% and every spark becomes a potential ignition source. Combustible gas must be below 10% of the lower explosive limit. For cement dust, the LEL varies but typically falls between 30 and 60 g/m³ depending on fineness and volatile content.

Test at multiple levels. Heavier-than-air gases like CO₂ pool at the bottom. Lighter gases like methane accumulate at the top. A single reading at the entry point tells you almost nothing. I test at the bottom, middle, and top of the silo before anyone goes in. Takes an extra 10 minutes. Has saved at least two lives that I know of.

Pre-Entry Planning: Where Most Programs Fail

The inspection itself is straightforward. The planning is where programs fall apart. I've reviewed confined space programs at 40+ plants across Southeast Asia and the Middle East. The ones that work have one thing in common: they treat pre-entry planning as a checklist that nobody signs off on until every box is ticked. The ones that fail treat it as paperwork.

Isolation and Lockout

Before anyone enters, the silo must be physically isolated from all material feed and discharge systems. That means double block and bleed on pneumatic conveying lines. It means lockout-tagout on all mechanical conveyors, rotary valves, and aeration systems. It means disconnecting and blanking any electrical connections to internal equipment like level sensors or temperature probes that could be energized during the inspection.

I was on a project in Indonesia where the crew had LOTO'd the main feed line but forgot the recirculation line from the dust collector. A operator 60 km away in the control room opened that valve remotely. Nobody was inside at that moment — pure luck. After that, we added a physical blind flange requirement on every connected line, not just valve isolation.

Ventilation Requirements

Mechanical ventilation isn't optional. A minimum of six air changes per hour is the baseline for occupied silos. For a 3,000-tonne cement silo with roughly 2,500 m³ of internal volume, that means pushing 15,000 m³/hr of fresh air through the vessel. Use positive pressure ventilation — blow clean air in at the bottom, let it exhaust through the top. Negative pressure extraction pulls dust through the workspace and reduces visibility to zero within 20 minutes.

Position the intake for your blower upwind and at least 15 metres from any dust source. I've seen crews set up the intake directly below a baghouse discharge point. They were pumping concentrated cement dust straight into the silo. The irony wasn't lost on anyone, but the respiratory hazard was real.

Communication and Rescue

The standby person at the entry point has one job: maintain communication with the entrants and initiate rescue if something goes wrong. They don't assist with the inspection. They don't leave the entry point. They don't answer the radio for anything else.

For silos deeper than 12 metres, a tripod and winch system rated for rescue is mandatory. The winch must be pre-rigged and tested before entry begins. Self-retracting lifelines (SRLs) are acceptable for vertical entry, but they must be inspected before each use — I've seen SRLs with frayed cables that would have snapped under a 100 kg load. Test them. Every time.

Rescue time matters. OSHA's general industry standard doesn't specify a maximum response time, but the practical limit is 4 minutes for an unconscious entrant in an oxygen-deficient atmosphere. If your rescue team can't reach the entry point, set up, and extract a person in 4 minutes, you don't have a rescue plan. You have a body recovery plan.

Inside the Silo: Inspection Execution

You've planned. You've tested. You've ventilated. Now you're going in. The inspection itself follows a structured sequence that balances thoroughness with time — because every hour inside a confined space is an hour of exposure.

Visual Survey and Hammer Testing

Start with a full visual survey before touching anything. Look for standing water, material accumulation, rust staining patterns, and any visible deformation of the shell. Rust staining in vertical streaks usually indicates water ingress through the roof or a leaking seal. Horizontal bands of discoloration often mark old material levels — useful for understanding flow patterns.

Hammer testing comes next. A 500-gram ball-peen hammer, swung with moderate force, will reveal delaminated scale, hollow areas behind the wall, and loose bolt connections. A solid ring means intact steel. A dull thud means something's behind the wall — scale, moisture, or compacted material. Mark every dull spot with chalk. You'll ultrasonic-test those locations specifically.

Ultrasonic Thickness Measurement

Grid-pattern ultrasonic testing is the backbone of any internal inspection. For cement silos, a 500 mm × 500 mm grid is standard. For abrasive service — slag, coal, alumina — tighten to 300 mm × 300 mm in the lower third of the shell and the entire cone section. Record every reading. Compare against the original design thickness.

Here's a number that matters: if wall thickness has reduced to 60% of original design thickness in any area larger than 0.5 m², that section needs replacement. Not monitoring. Not next shutdown. Replacement. I've seen plants push that to 50% and get away with it for a while. I've also seen a 10 mm plate fail at 5.8 mm under normal loading because the corrosion wasn't uniform — it was pitting, and the pit depth was 4 mm in a 6 mm remaining wall.

Weld and Connection Inspection

All shell-to-shell welds, shell-to-cone welds, and shell-to-roof welds get visual inspection for cracking, porosity, and undercut. Pay special attention to the toe of the weld where it meets the parent metal — that's where fatigue cracks initiate. For silos in seismic zones, also inspect anchor bolt connections and base ring welds. A 2 mm crack in a base ring weld under cyclic loading can propagate to failure in under two years.

Flange connections on manways, nozzles, and instrument taps get bolt-count verification and torque spot-checks. A flange that's lost more than 10% of its bolt preload is a leak waiting to happen. For cement service, that leak means material escape, dust generation, and eventually structural loading on the flange itself as material accumulates around the connection.

Case Study: 5,000-Tonne Clinker Silo — Central Vietnam

A cement plant in central Vietnam called us after their annual external inspection showed a 3 mm bulge in the lower shell of a 5,000-tonne clinker silo. The silo was 32 metres tall, 14 metres in diameter, built in 2014. External visual inspection couldn't determine the cause.

We set up confined space entry with full isolation, LOTO on four feed lines, and positive pressure ventilation at 18,000 m³/hr. Atmospheric testing showed O₂ at 20.8%, CO at 12 ppm, LEL at 2%. Entry was cleared.

Inside, we found the bulge was caused by a 2.3 m² area of wall thinning — original 12 mm plate reduced to an average of 6.1 mm, with localized pits down to 4.2 mm. The cause was a failed internal liner plate that had allowed clinker to abrade the shell directly over a four-year period. Ultrasonic testing on a 200 mm grid revealed the affected area was roughly twice what the external bulge suggested.

We replaced the section with 12 mm Q345B plate, welded per EN 1993-4-1 (Eurocode 3 for shell structures), and inspected the welds with magnetic particle testing. Total downtime: 9 days. Cost: $38,000 including materials, labor, and our supervision. The plant manager told me the alternative — a shell failure during operation — would have cost them 6 weeks of production and roughly $180,000 in lost revenue. He wasn't exaggerating.

Post-Entry: Documentation and Repair Prioritization

The inspection isn't over when the crew comes out. The report is what drives action, and most inspection reports I've seen are useless. They describe problems without quantifying them. They recommend "monitoring" without defining intervals. They don't prioritize.

What a Useful Report Contains

Every anomaly gets a location (height, azimuth, and reference point), a measurement (thickness, crack length, bolt count), a photograph, and a severity rating. I use a three-tier system: immediate repair (within 30 days), scheduled repair (next planned shutdown), and monitor (re-inspect at next interval). That's it. No ambiguity.

The report also includes atmospheric test results, entry and exit times, personnel involved, and any deviations from the entry permit conditions. If the ventilation failed for 10 minutes and was restarted, that goes in the report. If an entrant reported dizziness, that goes in the report. These details matter if something goes wrong later — and they matter for regulatory compliance.

Frequently Asked Questions