Contact ussales@ventavid.com
VentaVid

Racking inspection checklist: from hit upright to decision

Racking inspection checklist: reporting a hit upright with guided photos so the PRRS can classify the damage

In this article

This post in 30 seconds.

  • Three tiers, three owners. HSE's HSG76 sets out immediate reporting by anyone who sees damage, regular visual inspections under the person responsible for racking safety (PRRS), and an expert inspection at intervals of not more than 12 months.
  • Weekly is not the law. HSG76 says "weekly or other regular intervals based on risk assessment". Your PRRS sets the frequency and should be able to say why.
  • Amber is widely misquoted. It does not require immediate offloading of the rack, but a damaged component that has been unloaded stays empty until it is repaired, and amber becomes red after four weeks.
  • The weak link is the report. Classification can only be as fast as the moment damage reaches the PRRS, and most of it arrives late or never.
  • Who this is for: warehouse and operations managers, and the PRRS, at warehouses and distribution centres in the UK and Ireland.

A reach truck clips an upright at ten past two in the morning. The driver gets off, looks, sees a scuff and a slight kink below the first beam, and decides it looks fine. The weekly walk is on Thursday.

On Thursday the PRRS finds a bowed upright with three full pallet levels above it. There's no report, no time and no name, and the night shift has since rotated. Nobody lied. There was just no easy way to say "I hit something" at 2am without it turning into a conversation.

That gap, between the hit and the person who can classify it, is what this racking inspection checklist is built around. For the short definition of the term, we keep one on the racking inspection glossary page. This post is the operational version: who does what at each tier, what green, amber and red require, why damage goes unreported, and the checklists for each level.

A note from us. We build Venta Capture, VentaVid's guided capture product, so that an operator or team leader can report a hit upright with a structured photo set from their own phone the moment it happens. It does not inspect racking and it does not classify damage. A trained PRRS and a technically competent inspector still do that. Read the regime first and decide whether the reporting half is worth fixing.

In this post:

Three tiers, three owners

HSG76 describes racking inspection as "a hierarchical approach using several levels of inspection", and each level has a different owner, a different trigger and a different output. The guidance is HSE's Warehousing and storage: A guide to health and safety (second edition, free to download), and the racking section runs from paragraph 641 to 653.

TierWhoWhenOutput
Immediate reportAnyone who sees itAs soon as observedReport to the PRRS
Visual inspectionPRRS or trained staffWeekly or risk basedFormal written record
Expert inspectionTechnically competent personNot more than 12 monthsWritten report to PRRS

Paragraph 643 is the first tier and the shortest: "As soon as a safety problem or damage is observed by any employee, it should immediately be reported to the PRRS. You should have systems in place for reporting damage and defects." Hold on to that second sentence, because the rest of this post hangs off it.

Paragraph 645 is the second. The PRRS "should ensure that inspections are made at weekly or other regular intervals based on risk assessment. A formal written record should be maintained." That's the visual inspection most sites call the weekly walk.

Paragraphs 646 to 648 cover the third. A technically competent person inspects at intervals of not more than 12 months and sends a written report to the PRRS "with observations and proposals for any action necessary". That person "might be a trained specialist within an organisation, a specialist from the rack supplier, or an independent qualified rack inspector", and HSG76 names SEMA's SARI scheme (SEMA approved rack inspector) as the formal qualification route.

Two details from the same section get lost in most checklists:

  • Ground level is the default. Paragraph 649 says normal inspections are carried out from ground level unless there are signs of a problem higher up. Nobody should be climbing racking to inspect it.
  • Keep a log. Paragraph 651: "You should keep a record of inspections, damage and repairs. This could be done in a logbook." Three things in one record, and repairs are the one that goes missing.

Weekly is a habit, not a legal requirement

HSG76 does not set weekly as a fixed interval, and PUWER talks about suitable intervals. The frequency is a decision the PRRS makes and should be able to defend. Paragraph 642 is explicit that the frequency "depends on a variety of factors that are particular to the site concerned and should be determined by a nominated 'person responsible for racking safety' (PRRS) to suit the operating conditions of the warehouse." It lists what that takes into account: how often and how the racking is worked, the dimensions of the warehouse, the equipment used and the people involved.

Weekly is a sensible starting point for a busy pallet store, and plenty of sites will land on it. A narrow aisle operation running three shifts might justify more. A low-throughput store with guided trucks and full column protection might justify less. Write the reasoning into your risk assessment, because "we've always done Thursdays" isn't one.

The legal hook underneath is the Provision and Use of Work Equipment Regulations 1998. Regulation 6(2) of PUWER requires work equipment exposed to conditions causing deterioration to be inspected "at suitable intervals; and each time that exceptional circumstances which are liable to jeopardise the safety of the work equipment have occurred." Regulation 6(3) requires the result to be recorded and kept until the next inspection is recorded.

Read those two limbs side by side and the point of tier one becomes obvious:

  • The interval is your weekly (or whatever your PRRS sets). It catches deterioration and damage nobody reported.
  • The exceptional circumstance is the truck strike. It doesn't wait for Thursday. The immediate report is how you find out it happened at all.

How often you walk the aisles and how fast damage reaches you are separate questions. Most sites have answered the first one. We cover the general idea of setting and justifying intervals on the inspection frequency glossary entry.

Green, amber, red: what each one requires

The traffic lights are a classification of damage against a limit, and each colour carries a specific action and, for amber, a clock. HSG76 Table 11 sets them out, and paragraph 653 points to SEMA's guide to the conduct of pallet racking and shelving surveys for the detail. The measuring and classifying rules for damage to uprights and bracing sit in BS EN 15635:2008, Steel static storage systems: application and maintenance of storage equipment.

ColourHSG76 statusWhat happensClock
GreenSurveillance onlyRecord, re-examine next timeNext inspection
AmberAction as soon as possibleRepair; unloaded parts stay empty4 weeks, then red
RedImmediate actionOffload and isolate nowImmediate

The Table 11 wording is worth reading in full, because the summaries you'll find elsewhere tend to compress it.

  • Green "indicates the limit that does not require a reduction in rack carrying capacity or an immediate repair of the system." Components are recorded as suitable for further service "but should be clearly identified for specific re-examination and reassessment at future inspections." And then the sentence people skip: "Exceeding the green level should be considered damage and causes risk to the racking system."
  • Amber identifies damage "sufficiently severe to warrant remedial work but not so severe as to warrant the immediate offloading of the rack. Once load is removed from a damaged component, the component should not be reloaded until repairs have been carried out." The user should isolate those racks, for example with "dated adhesive labels", and amber damage "should be redesignated red risk if remedial work has not been carried out within four weeks of the original designation."
  • Red is "a critical level of damage" that "warrants an area of racking being immediately offloaded and isolated from future use until repair work is carried out." Repair "would usually be by replacement of the damaged component", and HSG76 gives the example of a bay offloaded in the presence of the inspector and roped off.

Reading Table 11 against the checklists that rank for this search, I found the amber rule summarised as "immediately offloaded". That's not what HSG76 says. Amber means the rack doesn't have to be emptied on the spot, but as pallets come out of the damaged positions they are not put back, and the repair has to happen within four weeks of the date the damage was classified.

SEMA's own description of a SARI report is stricter in practice: it says the pallet positions in an amber area "must then be emptied, and only refilled once repairs are complete", and that repairs "have to be carried out within four weeks, or an Amber risk item automatically becomes a Red risk" (SEMA, What is a SEMA Approved Racking Inspector). If your SARI works to that, follow your SARI.

Two practical consequences for the PRRS:

  • The four weeks start at designation, not when the repair order is raised. Put the designation date on the label and in the log, or the clock is a guess.
  • Classification needs a measurement, not an impression. EN 15635 classifies damage to uprights and bracing by measured deviation, which is why the PRRS or inspector has to see the component. A photo tells them where to look and how urgently. It doesn't give them the number.

We keep a general note on defect classification if you're building the same kind of scheme for other equipment.

Why damage goes unreported

Paragraph 643 of HSG76 assumes every employee who sees damage reports it immediately. On a real night shift, most of the reasons not to report are stronger than the reasons to. None of these are exotic, and every warehouse manager will recognise at least three:

  • Blame comes first. If the first question after a report is "who hit it", the rational move for the driver is not to be the one who said something. The damage gets found on the weekly walk with no name attached, which is the outcome everyone was trying to avoid.
  • There's no easy way. The route is a paper form in the transport office, a radio call to a supervisor who's on a break, or an email nobody on the shop floor sends. Anything that means leaving the truck and finding a person loses to the pick rate.
  • It looks fine. A scuffed upright with a small kink doesn't look like a structural problem to someone who isn't trained to classify it. HSG76 paragraph 641 is blunt about why that matters: "Any damage to racking will reduce its load carrying capacity."
  • Nobody saw it. Damage found at the start of a shift was caused on a previous one. Without a time, the PRRS can't tell whether that bay has been carrying load on a bent upright for two hours or five days.
  • The report can't be acted on. "Upright damaged, aisle 14" means the PRRS has to walk to aisle 14 to find out which of forty uprights, how bad, and whether the bay above is loaded.

HSG76's own example racking notice (Figure 48) carries the line "REPORT ALL DAMAGE TO THE 'PERSON RESPONSIBLE FOR SAFETY'". Most sites have a sign like that at the end of every aisle. What they don't have is anything that makes the report take less effort than keeping quiet.

The fix is partly culture and mostly design. Make the report ask where, what and how bad before it asks who. Make it possible from the aisle without finding anyone. Make it carry enough that the PRRS can decide whether to isolate the bay now or visit first.

If your operators currently have to find a supervisor to report a strike, see how a guided damage report works from the aisle.

The racking inspection checklist, tier by tier

A single checklist for "racking inspection" mixes three jobs done by three different people, so split it by tier. Each list below is written for the person who does that step. For general principles on building one, see the inspection checklist entry.

Tier 1: the operator damage report

This is the one most sites don't have on paper. It's short on purpose, because it has to be done by someone who hit something and wants to get back on the truck.

The photo set, in this order:

  1. The location code. The label on the beam or upright that identifies the aisle, bay and level. This one photo removes the "which upright?" walk.
  2. A close-up of the damage. Square on to the dent, bend, split or missing component, close enough to see the shape of it.
  3. The whole bay from the aisle. Floor to the top beam level you can see, so the PRRS can see what's loaded above the damage and whether anything looks out of line.
  4. The load notice. The sign showing the maximum load per level and per bay, so the capacity context travels with the report.

Then three questions and one field:

  • Is there load directly above the damage? Yes or no.
  • Is anything visibly leaning, dislodged or falling? Yes means stop, keep people clear and call the supervisor now as well as reporting.
  • Did you see it happen or find it? Found it is a normal answer, and it should never be treated as a worse one.
  • Anything else we should know? Free text or a spoken note.

No name box at the top. The report can record who submitted it, but it shouldn't open with it.

Tier 2: the PRRS visual inspection

Carried out from ground level (HSG76 paragraph 649), at the interval your risk assessment sets, with a formal written record. Per run:

  • Uprights: impact damage, bends and bows, splits, particularly at the lower levels and on corner uprights, which HSG76 paragraph 639 flags as especially at risk.
  • Baseplates and floor fixings: missing, loose or damaged anchors, cracked floor around them.
  • Bracing: bent, missing or disconnected horizontal and diagonal members.
  • Beams: visible deflection, impact damage, beam connectors seated properly, locking pins or clips present.
  • Column guards and end protection: present, undamaged, not themselves carrying damage that has been passed to the upright.
  • Load notices: present, legible, matching the current beam configuration.
  • Loads: within the notice, pallets in good condition and seated on both beams, nothing overhanging.
  • Alterations: beams moved or levels added without checking against the manufacturer's data.
  • Open reports: every tier 1 report since the last walk checked on the ground, classified, and closed or actioned.
  • Amber clock: every amber label checked against its designation date.

Record the result per finding with its colour, location code and action. That's your inspection evidence for the next SARI visit and for anyone who asks later.

Tier 3: before and after the expert inspection

The SARI does the inspection. The PRRS makes it useful.

  • Before: pull the log of every tier 1 report and every tier 2 finding since the last expert inspection, with repairs and dates. Trends by location are what paragraph 646's "proposals for any action necessary" are made from.
  • During: walk it with the inspector where you can, and have red areas offloaded in their presence if they find any.
  • After: file the written report, turn every amber into a dated label and a repair order, and put the four-week dates in the diary.

Truck-side checks are their own discipline. If your pre-use checks are part of the same problem, we cover them under the forklift daily check.

Getting the report from the aisle to the PRRS

The operator report only works if it takes less effort than keeping quiet, and it only helps if the PRRS can act on it without walking the site first. That's the problem Venta Capture is built for, and it's a narrow one.

Here's how it runs. A QR code sits on the load notice at the end of each aisle, or on the side of the truck, or in the shift briefing on screen. The operator scans it with their own phone. It opens in the mobile browser: no app to install, no account to create.

The guided capture flow walks them through the tier 1 report in order:

  • Scan the location code. A scan step reads the barcode or QR on the beam label, with manual typing as a fallback if the label is damaged too.
  • Four labelled photo slots. Each has its own one-line instruction and an example photo of what a good close-up or whole-bay shot looks like, so someone who has never reported damage gets it right the first time.
  • The yes/no questions. Tapped answers stored as fields, not buried in a description box.
  • An optional spoken note. Transcribed into searchable text on the case.
  • Submit. A thank-you page with a reference number the operator can quote at the handover.

Standard flows come ready-made and your own team adjusts the steps, photo slots and wording in the builder. Routing rules, escalations and integrations with your WMS or maintenance system are set up around your process by VentaVid, scoped on a call. The guided video capture page covers the mechanics.

On the PRRS's side, the report arrives as a case in a shared inbox, not as a photo in a group chat:

  • A reference number in your own format, plus a field for the location code or asset number your log already uses.
  • A receipt time recorded on the server. It proves when the report was received, which is the time your clock can start from. It doesn't prove when the truck hit the upright, and nobody should claim it does. The timestamped photo entry explains the difference.
  • Routing and notifications. Reports land with the PRRS or the shift supervisor, with a notification on each new submission and an SLA in hours on the route if you want one.
  • A status and an owner. New, in progress, completed or closed, assigned to whoever is going to look at it.
  • A retake in one click. If the close-up is blurred, a retake request goes back asking for that one photo, and the operator is asked why they're recording again. The review and retake page shows the loop.
  • An audit trail. When the report was opened, received, claimed, assigned and closed, and every internal note, so paragraph 651's record of damage has a timeline behind it.

It's asynchronous by design. The operator reports at 2am, the PRRS reviews at 6am or whenever the notification lands, and nobody has to be on a video call with a phone pointed at an upright.

What the PRRS does with it is the point. With the location, the close-up, the loaded bay and the answers in front of them, they can make the first call from wherever they are: isolate that bay now pending a proper look, or book it into today's walk. Then they go and look, measure, and classify.

And this is the line we'd put on the wall. A photo report never replaces the expert inspection, and it never replaces the PRRS's classification. It can't measure a bow against EN 15635 limits, it can't see the back of an upright, and it can't tell green from amber on a borderline bend. What it does is get the damage in front of a trained person faster, with the information they need to decide where to go first.

Offloading, isolation and the repair record

A classification without a closed repair record is only half a finding. HSG76 paragraph 652 says that where damage affects the safety of the racking system, it "should be offloaded and controls introduced to prevent it being used until remedial work has been carried out."

What that means on the floor:

  • Red: offload the affected area immediately and isolate it. Rope or barrier it off, not just a sticker, and tell every shift.
  • Amber: apply a dated label, stop reloading positions as they empty, raise the repair, and track the four weeks from designation.
  • Green: record it with its location code and a photo, and re-examine it at the next inspection.

On the repair itself, SEMA's guidance is clear on the two mistakes that turn a repair into a new hazard. It "does not recommend straightening damaged uprights with hydraulic rams, as the steel may already have been weakened", and it warns against substituting components without the manufacturer's checks (SEMA, The dos and don'ts of racking repairs, March 2026). HSG76's Table 11 already expects red repairs to be "usually by replacement of the damaged component".

Then close the loop in the log:

  • Before and after photos from the same position, tied to the same location code. The repair case links to the original report, so the whole story sits in one file.
  • Who repaired it and with what: installer, component, date.
  • When the bay went back into use, and who authorised it.

That's what makes paragraph 651's "record of inspections, damage and repairs" worth having when the SARI arrives or an insurer asks. It's the same idea as proof of work in field service, applied to a repair contractor working in your aisles. If you want to understand why a guided photo set beats "send us a picture", we wrote it up in guided photo capture. A status you define, such as "amber, repair needed", can also forward the case automatically to the maintenance contractor through workflow automation.

The trucks doing the damage have their own daily check, covered in forklift daily check sheets. For how to specify the photo set an operator should take, see photo evidence for inspections.

Frequently asked questions

Is a weekly racking inspection a legal requirement?

No. HSG76 says the PRRS should ensure inspections "at weekly or other regular intervals based on risk assessment", and PUWER regulation 6 requires inspection at suitable intervals and after exceptional circumstances. Weekly is a common result of that risk assessment, not a fixed legal interval.

Do you have to offload racking with amber damage?

Not immediately, according to HSG76 Table 11: amber is "not so severe as to warrant the immediate offloading of the rack". Once load is removed from the damaged component it must not be reloaded until repaired, and amber becomes red if the repair isn't done within four weeks of designation. Many SARIs require amber positions to be emptied, so follow your inspector's report.

Does the annual racking inspection have to be done by a SEMA approved inspector?

HSG76 requires a technically competent person at intervals of not more than 12 months, which "might be a trained specialist within an organisation, a specialist from the rack supplier, or an independent qualified rack inspector." It names SEMA's SARI scheme as the formal qualification route for expert inspectors.

Who is the person responsible for racking safety?

The PRRS is the person nominated to own racking safety day to day: setting the inspection frequency, running the visual inspections, receiving damage reports and the expert report, and keeping the record. SEMA says the role needs the skills to analyse damage data and identify trends, and the authority to act on them.

Can a bent racking upright be straightened?

SEMA does not recommend straightening damaged uprights with hydraulic rams, because the steel may already be weakened. HSG76 expects red damage to be repaired usually by replacing the damaged component, using parts checked against the manufacturer's data.

Can photos replace a racking inspection?

No. A photo report gets damage to the PRRS faster and shows where to look, but classification against EN 15635 limits needs a trained person measuring the component, and the expert inspection stays with a technically competent inspector. Use photos for the report and the record, not the verdict.

Does this apply to warehouses in Ireland?

HSG76 is HSE guidance for Great Britain, but EN 15635 is a European standard rather than a UK one, so the damage classification it underpins is not a British invention. The HSA's Code of Practice on the Safe Use of Industrial Trucks lists structural damage caused by trucks to warehouse racking as a hazard and expects racking to be protected against accidental damage.

Start with one aisle

Put the reporting flow on the end of your busiest aisle first. Build the four photos and three questions, print or display the QR code on the load notice, and brief one shift that a report asks where and how bad, not who. Then look at what reached your PRRS in the first month compared with what the weekly walk used to find on its own.

What it takes to start.

  • Free plan, no credit card. Build the damage report flow and send it to yourself first.
  • Live in 10 minutes for a first standard flow. Stuck? Book a free setup call and we build your first flow together.
  • Everything goes out in your own branding, in the operator's language, with no app to install.

The VentaVid team builds Venta Capture and runs the free setup call on your own damage report.

See Venta Capture on a racking damage report

Or start for free and build the tier 1 report this afternoon. If you'd rather walk through it against your own site first, book a demo.

Turn any smartphone into your eyes on site

Guided video and photo capture. No app, no account, sealed on receipt.