
Broof(Tx) tells only part of the story. PV fires start under the panel, not above the roof. Discover why Broof(Tx) is not enough and how AllShield BarrierSheet protects roofs against real downward fire.
Broof classifications have long been used to assess how a roof reacts to fire coming from the outside. Although the system is European, not every country uses the same test method. Some regions adopt Broof(t1), others t2, t3 or t4, depending on local building traditions and climate.
With the rise of photovoltaic systems, however, insurers and risk engineers increasingly face a different reality. Most PV fire incidents do not start outside the building, but inside the narrow air cavity between the solar panel and the roof surface. This so-called downward fire scenario is not covered by any Broof method.
This page explains how Broof works, how different European countries apply the t1–t4 variants, and why additional PV-specific fire tests are essential for real rooftop safety.
Broof comes from EN 13501-5 and uses the CEN/TS 1187 test series. All four test variants look only at external fire exposure on top of the roof surface. They measure flame spread, surface damage and heat influence when the fire approaches the roof from outside the building.
What Broof(Tx) does not measure, is the behaviour of a roof when heat and flames originate beneath a solar panel. In the PV cavity, temperatures can rise rapidly, while convection forces heat downward into the waterproofing and insulation. This is a completely different fire dynamic, and one that the Broof system was never designed to evaluate.
Although Broof is a single classification system, countries adopt different test methods. This is the officially used matrix across Europe:
| Broof version | What does it test? | Scenario | Where used? | Relevance for PV fire safety |
| t1 | Burning fire on the roof surface | External fire without wind | Standard in Central Europe (NL, DE, BE etc.) | Low |
| t2 | Burning fire + wind | Flames blown across the roof surface | Standard in Scandinavia (NO, SE, FI) | Low |
| t3 | Burning fire + wind + radiant heat | Fire with additional heat and airflow | Standard in France and parts of Southern Europe | Low |
| t4 | Large external fire + inclination | Simulated façade fire spilling onto the roof | Standard in the United Kingdom and Ireland | Very low |
These variations reflect regional building traditions, but the key point remains: None of the Broof variants simulate fire under PV panels.

Because the Netherlands - like Germany and Belgium - requires Broof(t1), our systems have been tested to CEN/TS 1187 Test 1 by Kiwa. The AllShield Blue build-up and the AllShield BarrierSheet on PIR both meet the full Broof(t1) criteria. This is the correct and legally required classification for flat commercial roofs in Central Europe.
However, even with Broof(t1) compliance, the PV fire scenario remains untested. That is why we carried out additional, far more realistic tests.
PV fires almost always originate in the cable and connector area directly beneath the module. Heat accumulates in the cavity, is trapped by the panel and is driven downward into the roof system. The waterproofing, the insulation and the mechanical fixings are exposed to a completely different heat pattern than in Broof(Tx) tests.
Brof(Tx) does not introduce heat into this cavity and therefore cannot answer the key question insurers now ask: What happens to the roof when a fire starts under the PV array?
To evaluate the real PV-specific fire scenario, we conducted several dedicated tests.
The first is based on CLC/TR 50670, where a 15 kW gas burner is placed directly under a PV panel. This test creates the downward fire geometry that Broof never simulates. The results show that flames remained above the AllShield BarrierSheet and did not penetrate into the insulation or structural layers.
We also performed large-scale fire tests at ZAG, using wood cribs with a higher and longer heat release than the CLC method. These tests were conducted with PIR and EPS in various thicknesses, several PV brands and real mounting systems. In every configuration, the BarrierSheet prevented flame penetration into the roof build-up, and insulation temperatures remained within safe limits.
Additional full-scale setups with PVC, bitumen, PIR, EPS and steel decks confirmed the same behaviour.
A PV roof has two distinct requirements. The construction must meet the national Broof test method (t1) in the Netherlands, and the PV system must be demonstrably safe in a downward-fire scenario. Many traditional systems satisfy only the first requirement.
With AllShield BarrierSheet, the roof is Broof-compliant and simultaneously protected against downward fire. The combination with PIR results in a lighter, more stable and insurer-approved solution compared with mineral wool alternatives. Large-scale tests at ZAG and the CLC/TR 50670 results provide verifiable evidence that the roof build-up remains protected, even under heavy and realistic PV fire loads.
Broof(Tx) is an essential classification for external fire exposure, but it does not test the most critical PV-specific fire scenario. The only way to evaluate downward fire behaviour is through PV-focused fire tests. The outcomes of CLC/TR 50670 and the ZAG large-scale programme show that AllShield BarrierSheet effectively stops downward fire and protects the roof system under extreme conditions.
This combination of Broof(Tx) compliance and proven PV fire resistance delivers a modern, lightweight and insurer-ready roofing concept for commercial buildings.
Flat roofs – especially those with solar panels – face an increasing fire risk. Even the best fire-retardant membranes offer limited protection against flying sparks or thermal ignition beneath PV panels. That’s why AllShield developed two non-combustible fire protection systems, each tailored to a specific application.