Fire Performance: It’s About the System, Not the Material

The Insulation Manufacturers Association (IMA) is the Trade Association that represents both the polyisocyanurate (PIR) and polyurethane (PUR) insulation industry in the UK. The organisation’s constituent members manufacture rigid insulation that provides around 40% of the total thermal insulation market on home shores.

Recently, the IMA commissioned the Fire Protection Association to carry out two comparative fire tests on flat roof assemblies fitted with identical reinforced bituminous membranes and mounted PV arrays. One assembly used PIR insulation classified as combustible under EN 13501-1. The other used stone wool classified as being non-combustible.

Everything else in the build-up – including the roof deck, the membrane and the PV configuration – was kept identical so any difference in fire behaviour could be attributed to the insulation layer.

During the testing procedure, a propane burner was used to simulate a localised fire starting beneath one PV panel. The fire was allowed to develop freely for almost 22 minutes before being extinguished.

Test results: the reality

The results were clear. In both tests, fire performance across the roof was broadly similar. Neither form of insulation allowed fire to penetrate through its full depth, while the air and vapour control layer remained intact underneath both. If anything, the damaged surface area was smaller on the PIR assembly than on the stone wool configuration.

That result runs contrary to the instinct that a non-combustible product must always outperform a combustible one. It’s a difficult message to land as it sounds counterintuitive to argue that a combustible material can be just as fire safe as one that isn’t, but that instinct misunderstands what fire classification actually measures.

A given product’s performance in isolation tells you very little about how it behaves once fitted within a complete system alongside a membrane, a deck and everything above it. We accept this readily enough with timber, which burns, yet remains a trusted structural material due to how it’s designed and detailed within a building. The same logic applies to insulation beneath PV panels.

This isn’t an argument for complacency. Far from it, but empirical evidence must outweigh simple opinion. Fire safety in any roof build-up depends on making sure the whole system is right: from the deck up to the waterproofing and everything mounted above it.

Rather, it’s an argument against decisions that fixate on one material’s classification, while ignoring how that material actually performs once installed.

Policy built on assumption

This matters. Why? Policy is increasingly being shaped by exactly the assumption this form of testing challenges and can lead to poorer performing buildings overall. Proposals such as those within the Draft London Plan, which sets out a presumption against combustible materials in certain buildings, respond to public concern that’s understandable, but not always well founded.

A given building is required to perform in many different ways. Fire safety is only one category. For its part, system-level testing is a more reliable guide than material classification alone.

Rooftop PV installations are only going to become more common on flat roofs as decarbonisation targets begin to bite. That makes evidence from test procedures like that described above more valuable, not less.

Ultimately, it’s very much the case that architects, specifiers, insurers and regulators all need a system-level understanding of fire risk rather than relying on assumptions about individual materials.

Simon Storer is CEO of the Insulation Manufacturers Association (www.insulationmanufacturers.org.uk)

*Copies of the 12-page report ‘Fire Performance of Thermal Insulation Products in End-Use Conditions’ are available online