Most permit offices have some version of a work-type compatibility chart taped up near the issuing desk — hot work against confined space entry, excavation against lifting, radiography against almost everything else. It tells the issuer what shouldn't run at the same time in the same area. What it usually doesn't do is know what's actually running at the same time in the same area.
That gap is where a permit interaction matrix earns its keep, and where most explanations of it stop short. It gets mentioned as a feature of good SIMOPS management, shown as a static table, and left there. Building one that actually holds up on a live asset takes more thought than that.
What a permit interaction matrix is
A permit interaction matrix is a structured reference that sets out, for every pair of work types an asset performs, whether they can run concurrently in the same area without restriction, run only under stated conditions, or can't run together at all. It's a decision aid used at permit issue — not a substitute for a site-specific SIMOPS assessment.
The distinction matters more than it sounds. The matrix answers a generic question: are hot work and confined space entry, as categories of activity, normally compatible in the same vicinity? A SIMOPS assessment answers a specific one: given exactly where this hot work is happening, this confined space is located, and what's between them, is this particular combination acceptable today? The matrix narrows the judgement call. It doesn't replace it.
Classifying the work types before you classify anything else
The matrix is only as good as the categories feeding into it, and this is usually where the effort goes missing. "Hot work" on most assets covers everything from welding to grinding to using non-intrinsically-safe equipment, and each carries a different ignition risk profile. Lumping them into a single category produces a matrix that's either too permissive for the worst case or too restrictive for the mildest one.
A workable starting set typically separates out hot work, cold work, confined space entry, excavation, lifting operations, electrical isolation and live working, and radiography — each with sub-categories where the risk genuinely differs (for example, radiography usually needs its own exclusion-zone logic that doesn't map cleanly onto anything else on the list). Asset-specific work types — diving, ROV operations, well intervention — get added where they apply. The categories should reflect how permits are actually written on that asset, not a generic list lifted from a template.
Defining compatibility between work types
Once the categories exist, each pairing gets a compatibility rating. Three levels cover most cases:
- **Compatible** — the two activities can run concurrently in the same area without additional controls. Cold work and most planned preventative maintenance tasks often sit here.
- **Conditional** — concurrent working is permitted, but only with stated additional controls: a minimum separation distance, a fire watch, continuous gas monitoring, or a requirement that one activity stops during specific phases of the other. Hot work and cold work in adjacent areas frequently falls into this category, with the condition being distance and barrier requirements rather than an outright block.
- **Incompatible** — the two activities cannot run in the same area at the same time under any conditions the site is prepared to accept. Radiography and most other permitted work nearby is the clearest example, given the exclusion zone it demands. Hot work and confined space entry involving a flammable atmosphere is another.
The conditional category is where most of the real work sits, and where a lot of matrices get vague. "Additional precautions required" isn't a rule an issuer can apply consistently — it needs to say what the precaution actually is, and who decides it's been met.
Where the matrix stops being enough
A matrix built this way tells an issuer that hot work and confined space entry are generally incompatible. It doesn't tell them whether the confined space entry planned for bay 4 and the hot work planned for bay 6 are close enough to matter, whether the ventilation between them changes the answer, or whether a temporary bulkhead removes the conflict entirely. HSG250 sets out the principle that a PTW system should identify and manage interactions between simultaneous activities — it doesn't specify the matrix as the mechanism, and it isn't a substitute for judgement on the specific case in front of you. We've written before about why isolation is the weak point in permit to work, and the same principle applies here — a general rule about work-type compatibility can miss a shared isolation or a specific site layout that only a proper SIMOPS review would catch.
Treat the matrix as the first filter, not the final answer. It should flag potential conflicts fast enough that someone can decide whether the specific case needs escalation — not tell them the decision has already been made.
Keeping the matrix current
A matrix reflects a moment in an asset's life — its layout, its equipment, its normal operating envelope. Any of those changing can quietly invalidate a rating that was correct when it was written. A new confined space created by a modification, a change to gas detection coverage in an area, a permanent change to how radiography is scheduled — all of these can shift a pairing from conditional to incompatible, or the reverse.
In practice this means the matrix needs an owner and a trigger for review, tied into management of change rather than a fixed annual date. If nobody's accountable for updating it when the plant changes, it drifts out of date quietly, and nobody notices until two permits that shouldn't have been compatible are running side by side.
Putting the matrix to work at the point of issue
A matrix on a laminated sheet is only as useful as the issuer's memory and the time they have to check it. On a shift with several permits live and more coming in, cross-checking a new hot work permit against everything else currently authorised in the same area is exactly the kind of task that gets harder, not easier, under pressure — and a static chart can't tell the issuer what else is live right now, because it has no connection to the permit register.
A digital permit system that holds the interaction matrix as structured data can do something the laminated version can't: check a new permit request against every other live permit in the same location automatically, and flag a conflict — or a conditional pairing needing a control check — before the permit is issued, not after. The same logic applies to electrical isolation work, which has its own conflict patterns worth understanding on its own terms — see our piece on when live working is actually permitted. If you're evaluating systems on this basis, our buyer's guide to permit to work software sets out questions worth asking a vendor about how interaction checks actually run at issue, not just how they're described in a demo.
If you're building or refreshing an interaction matrix and want to work through how it should sit alongside your existing SIMOPS process, we're happy to talk through what that looks like on your asset.