ISO 19650‑6 is the first time the BIM rules have properly met health and safety at source. In simple terms, it says risk information should live in the model and the Common Data Environment, not just in scattered PDFs and toolbox talks. It asks the design and delivery team to decide what H&S data is needed, structure it, assign clear responsibilities and keep it current across design, construction and handover. Get it right and you cut duplication, spot clashes between design intent and site reality earlier, and make RAMS and permits less reactive.
TL;DR
/>
– Treat ISO 19650‑6 as a workflow: define risk data needs, structure fields, set responsibilities, and use the CDE states to control change.
– Build risk entries against model elements, locations and tasks so RAMS and permits can pull live context.
– Push responsibilities into appointments: who authors, who verifies, who accepts, and when across RIBA stages.
– Keep it lightweight: a consistent set of risk properties beats bloated templates nobody updates.
– Measure value by fewer late design changes for safety reasons, fewer repeat incidents, and faster, clearer RAMS approvals.
What ISO 19650‑6 really means in plain English
/>
– It deals with foreseeable H&S risks linked to the asset and the works, not every site hazard. Think residual design risks, access constraints, temporary stability, live services and maintenance access.
– Instead of risk notes trapped in drawings, risks become data: each with a unique ID, description, classification, location, affected element(s), work phase, control/mitigation, responsible party and status.
– That data sits in information containers in your CDE, with the usual statuses for Work in Progress, Shared and Published. The same model element can carry geometry plus risk properties.
– Dutyholders don’t change because of BIM, but their information duties become explicit. The principal designer curates design-stage risk information; the principal contractor builds and updates it through construction; the client’s team ensures it lands in the asset model and O&M.
– It ties into the project’s Exchange Information Requirements: you state up front what risk information is needed at each stage and exchange, and how it will be structured and checked.
Turning ISO 19650‑6 into site-level workflows
/> Start with your BIM Execution Plan. Add a section for H&S risk information that sets the fields, coding, model locations, approval path and the specific exchanges tied to stage gates. Keep the property set lean and unambiguous. Define how risks against tasks (temporary works, lifting operations) relate to risks against model elements (penetrations, edges, confined spaces). Confirm what remains in narrative RAMS and what must be data-linked.
Next, map responsibility. Designers log residual risks tied to elements as they close out design packages. The principal contractor aggregates and updates risks with temporary conditions, logistics and access. Subcontractors contribute method-specific risks for their packages. The information manager configures containers and approvals in the CDE, while the digital engineer checks consistency and clash between risk data and geometry. The PD and PC agree acceptance criteria and lock the path for publishing to Shared.
Then put it to work. When the site team raises a temporary works proposal, it references the risk IDs and locations already tied to the structure model. The RAMS form pulls the relevant risks, control measures and exclusions directly from the CDE. Permits-to-work link to the same IDs, so when a lift plan or sequence alters, the status of affected risks flips to “Change pending” and triggers a review. At handover, the client receives an asset model with risk data for maintenance and future refurbishments, not a one-off risk register frozen in time.
Here’s a real UK-leaning scenario. A CAT B fit‑out in a tight city‑centre block is racing to hit a landlord completion date. The principal contractor’s planner is juggling delivery windows, a single goods lift and night shifts to pull MEP first fix forward. The design team has flagged residual risks around live risers and fire-stopping to slab soffits, but those notes are buried in PDFs. A ceiling contractor arrives with a scissor lift and finds a congested ceiling void with unprotected edges around service openings, and a last‑minute RFI changes the grid set‑out. With ISO 19650‑6‑style risk data in the model, the MEP coordinator sees the live services and fall risks tagged to zones, the RAMS form pulls those risks into the permit, and the lift plan includes exclusion zones aligned to the model. When the RFI changes the grid, linked risks near the affected bay are automatically marked for re‑acceptance before night shift starts.
Pitfalls and practical fixes with H&S risk data in BIM
/> The biggest trap is overcomplicating the schema. If people need a PhD to fill in a risk entry, they won’t. Agree a standard set of properties and codes that match your site language. Use drop‑downs where possible and keep free text to the “description” box.
Don’t let it drift. Risk data is living information and needs the same CDE discipline as drawings. Set cadence: weekly package reviews with the PD/PC lead; change notifications linked to model revisions; and acceptance steps before anything is Published. Small beats perfect: start with the top 10 recurring hazards per package and grow from there.
Linking to model geometry is non‑negotiable if you want value. Tag risk data to model element IDs, spaces or grids. If the model is immature, use agreed location codes or zones and update links as the model matures. Visualisation matters: teach supervisors how to filter models by risk category and by status in the viewer, not just in spreadsheets.
Finally, make it useful downstream. Write the properties so they can feed RAMS, permits and O&M without retyping. For example, “Control: edge protection to TWC-approved design” is better than “be careful at edges”. Reference drawing numbers, permits or method statements by unique IDs, not filenames on someone’s desktop.
# Common mistakes
/>
– Treating the risk register as a static spreadsheet detached from the model. That breaks traceability and invites version drift.
– Dumping design-stage placeholders into Published without acceptance. It clogs the system with noise and erodes trust.
– Leaving subcontractors out of authoring. You lose method-level insights and end up firefighting on site.
– Using vague location fields like “Level 3 area” with no link to model zones or coordinates. Nobody can filter or act on that.
# Checklist for making H&S risk data usable in the CDE
/>
– Define a minimal property set: ID, description, hazard class, location/element link, phase, control, responsible party, status, related docs.
– Configure CDE workflows so risks can only move to Shared when accepted by the named PD/PC roles.
– Establish coding for locations (grids/zones/rooms) and insist risks use those codes or model GUID links.
– Train the site team to navigate model filters by risk category and export RAMS‑ready summaries.
– Make subcontractor appointments explicit: who authors which risks, at what stage, and with what check.
– Tie change control to model revisions so risk statuses auto‑flag when affected geometry or sequences change.
– At completion, export risk data with the asset information model in an agreed open format and a human‑readable register.
What to watch next in the UK is tighter client expectations: more employers are asking for structured H&S risk information at tender and asset handover, and want to see it used in permits and maintenance. The sites that win time back will be those that make risk data visible in day-to-day decisions, not just in compliance deliverables.
FAQ
/>
How does ISO 19650‑6 relate to CDM roles on a UK project?
It doesn’t create new dutyholders, but it clarifies how information from existing roles is structured and exchanged. The principal designer leads the curation of design-stage risks; the principal contractor drives construction-stage updates and controls. The client specifies what risk information they want at each stage. Everyone uses the CDE to make those exchanges auditable.
# What minimum data should a risk entry include in BIM?
/> Aim for a unique ID, a clear description of the hazard, the location or model element link, the phase of work it affects, and the proposed control. Add the responsible party and a status such as proposed, accepted, implemented or closed. If relevant, reference related documents like drawings, RAMS or permits by their unique IDs. Keep fields consistent so you can filter and report reliably.
# Do we need a new CDE or specialist tool to implement this?
/> Most mainstream CDEs and model viewers can handle structured properties and simple workflows. You may choose add‑ons for visualising risks in 3D or for linking to permits, but they’re not a prerequisite. Focus first on defining the fields, the approval path and the exchange points. If your current stack can’t handle property filtering and status control, then consider an upgrade as part of your information strategy.
# How can subcontractors add risk information without BIM authors?
/> Give them a simple template with the required fields and clear location codes, and set up a gateway for import into the CDE. A coordinator can then map entries to model elements or zones if the subbie isn’t authoring directly. Short training sessions in the model viewer help supervisors see the impact of their entries. Make contribution and acceptance a condition of their appointment and programme.
# How is change controlled when a design or method changes on site?
/> Tie risk entries to model revisions and to the RFI or change reference, so the status flips when an affected element or sequence changes. The PD/PC acceptance step should be required before a risk moves to Shared again. Update linked RAMS and permits with the same IDs to keep everything in step. Keep a brief audit trail in the CDE so you can see who changed what and when.
