Research and analysis

Fire safety: Compartment wall/roof junction experiments (executive summary)

Published 22 December 2025

Applies to England

Background

The work reported in this report was carried out by a BRE Global Project team under a Contract placed by the Ministry of Housing, Communities and Local Government (MHCLG) which was novated to the Health and Safety Executive (HSE).  Its contents, including any opinions and/or conclusions expressed, are those of the authors alone and do not necessarily reflect MHCLG or HSE policy.

This Final report is delivered as part of the Ministry of Housing, Communities and Local Government

(MHCLG) (formerly the Department for Levelling Up, Housing and Communities (DLUHC)) project titled “Compartment wall/roof junction experiments”, MHCLG Contract reference CPD/004/120/206.  This contract was novated to the Building Safety Regulator (hereafter referred to as the client) of the Health and Safety Executive (BSR HSE), HSE contract reference 1.11.4.4436, effective date 1 April 2023.  This work was originally part of the call-off element (Part 2 - Experimental fire testing and reporting) of the “Investigation of Real Fires” Project, MHCLG Contract Reference CCZZ18A04.

Executive summary

This programme of experiments provided indicative evidence of the performance of a compartment party wall and three modern roof construction types with respect to Schedule 1 requirement B3(2) of the Building Regulations 2010 (As amended).  The programme comprised eight experiments.  Each roof construction was formed using timber joists and either an independent timber deck or a composite timber deck bonded to an insulation board.  The compartment party wall was 215 mm wide aerated concrete blockwork.  The fire stopping material, where fitted, was mineral fibre or cellular glass. 

The scenarios included:

1. No fire stopping between the head of the compartment party wall and the underside of a timber deck.
2. Fire stopping between the head of the compartment party wall and underside of the timber deck.
3. Additional fire stopping between top of the timber deck and the underside of the roof membrane.
4. A variant of scenarios 1, 2 and 3 with more insulation between the flat roof joists (creating a hybrid deck construction).
5. Fire stopping (under two different compressions) between the head of the compartment party wall and the underside of a foil faced thermoset composite insulation board / deck.
6. No additional fire stopping above a timber deck of an inverted roof with thermoplastic insulation being allowed to pass across the compartment party wall.
7. Additional fire stopping above a timber deck of an inverted roof, preventing thermoplastic insulation from passing across the compartment party wall.

The study considered three roof construction types:

1. Construction type 1 – thermally insulated deck. Comprising tissue faced rigid thermoset polyisocyanurate insulation board on an independent timber deck

2. Construction type 2 – composite deck. Comprising foil faced rigid thermoset polyisocyanurate insulation board with a composite timber deck 

3. Construction type 2 – inverted roof. Comprising thermoplastic insulation on an independent timber deck   

Where fire stopping is only provided between the head of a compartment party wall and the underside of an independent timber deck and thermoset insulation is allowed to carry across a compartment party wall, there is the potential for elevated temperatures to occur above the centreline of the wall and between the centreline and the unexposed face of the wall. 

The presence of additional fire stopping between the independent timber deck and the underside of the waterproof layer prevented elevated temperatures reaching the centreline of the compartment party wall.  This observation applied to 100 mm thick thermoset insulation in Construction type 1 – thermally insulated deck.

Thermally thick constructions with composite deck (or thermally thick constructions generally) may require further investigation to determine if flaming or smouldering combustion can give rise to elevated temperatures close to or beyond the centre line of compartment party walls or if charring of the insulation prevents this mechanism of fire spread.

Fire stopping between the head of a compartment party wall and the underside of an independent timber deck of an inverted roof, together with additional fire stopping between the independent timber deck and the underside of the waterproof layer, prevented elevated temperatures reaching the centreline of the compartment party wall, when the insulation was thermoplastic.  This observation applied to 160 mm thermoplastic insulation in Construction type 3 – Inverted roof.  

It should not be assumed that this conclusion will apply to thermally thicker constructions without further research.  Evidence should therefore be sought to explore the effect of constructing thermally thick and thermally very thick roof constructions up to and across party walls and whether such constructions under some circumstances may allow fire to spread across the line of a party wall.  If re-visiting this, it would be worthwhile considering the effect of wind in driving fire up to and potentially over party walls

It is not known if fire would spread across a party wall if the insulation used was thinner than the insulation used in these experiments, and this may only become apparent following additional research.