Abstract
The production and usage of combustible thermal insulation like polyisocyanurate (PIR), wood fibre and cellulose in buildings are increasing. Traditionally, PIR has only been used in the ground and on roofs, in combination with non-combustible structures. During the last decade, the field of application for PIR has extended to walls, floors, and roofs, as well as to larger buildings. Wood fibre and cellulose insulation have been used in Norway since the early 1980s in combination with constructions of combustible materials, both as insulation of old constructions and as insulation in walls, roofs and floors in new buildings. During the last decade the area of use is also extended to larger buildings. Even if some fire testing is performed, there is a lack of knowledge about the fire behaviour of these combustible insulation materials and how they affect the fire development and spread, and the vulnerability to damages in the fire protective cladding.
The fire performance of these combustible thermal insulations has, therefore, been studied through fire experiments. Fire performances studied here are the temperature development in the material and the behaviour of the material when exposed to fire in terms of charring and general degradation. Three different fire test series have been conducted to study the performance of combustible insulation in timber frame assemblies with fire protective claddings of gypsum boards or plywood boards. The insulation products tested are Kingspan Therma TW50 PIR, Hunton Nativo® Wood Fibre insulation batts, Hunton Nativo® Wood Fibre loose-fill insulation, Ekovilla batt cellulose insulation and CBI Norge Isocell Evolution loose-fill cellulose insulation. For comparison, the non-combustible insulation products Glava Proff 34 batts (glass wool) and Rockwool Flexi A board (stone wool) were also tested. Two standardised test methods with additional measurements were used: EN 1364-1 for fire resistance of non-loadbearing walls and EN 13823 Single Burning Item (SBI) test for reaction to fire testing. In addition, a nonstandardised method was used to study both wall and roof together, as well as the intersection
between them. In the non-standardised test and the fire resistance test, the temperature development in the furnace followed the standard temperature-time curve given in EN 1991-1-2, while in the SBI test, the applied fire source was as described in EN 13823. The temperature development in the constructions was measured with thermocouples, and the degradation of the insulation was observed visually and measured after the tests.
The main goals, results and conclusions are:
(1) Fire and temperature development in combustible insulation products The goal was to study the fire and temperature development in combustible insulation products made of polyisocyanurate (PIR), wood fibre and cellulose when used in building constructions like walls, floors and roofs, and the effect of damages and penetrations from technical installations in the fire protective cladding. The results from fire experiments on specimens with PIR, wood fibre, and cellulose were compared with results from tests on specimens with glass wool and stone wool. The thermal insulation products based on PIR, wood fibre or cellulose used in these experiments are combustible, with reaction to fire class E according to the classification standard EN 13501-1. Glass wool and stone wool used in these experiments are both noncombustible insulation products with reaction to fire class A1.
The evaluated insulation products reacted differently to high temperatures; the insulation products of PIR, wood fibre and cellulose tended to char and burn. While glass wool became discoloured and melted, stone wool became discoloured and porous, but neither of them burned and contributed to the fire.
The results show that a char layer formed during combustion of PIR, wood fibre and cellulose will slow down the temperature increase inside the insulation. Combustible insulation products, as tested in these constructions with a protective cladding, can provide fire protection by slowing temperature increases inside the insulation and on the surface of the timber structure in the construction. Measurements from the non-standardized fire tests show that constructions with these tested insulation products can also achieve similar or better insulation performance compared to the tested glass wool and stone wool insulation when the insulation thickness is the same. However, in some cases the stone wool and glass wool insulation gave slower temperature development and better fire protection of the timber structure. Because the thermal conductivity of PIR is lower than the other products used in the experiments, it will sometimes be installed in smaller thicknessesthan the other insulation materialsinside wall and roof constructions. The thermal insulation for the constructions with PIR in the current tests is therefore better than the other tested constructions.
The EN 1364-1 tests showed that higher temperatures were registered for batts than for loosefill insulation of the wood fibre and cellulose products used in these test series. Therefore, loosefill insulation was assessed to be more fire resistant, which motivated the present study to consider batts as the more challenging product. Batts were therefore used in the wall and roof constructions tested at RISE Fire Research.
The temperatures in the roof construction increased faster than in the wall due to the temperature and pressure gradients inside the test furnace and the sagging and fall-down of the insulation. This demonstrates the importance of correctly installed insulation. According to the results from the SBI tests, the consequences of damages or holes in the cladding at the start of a fire are mainly limited to local fire damages in the insulation and short flame spread behind the cladding. The combustible insulation materials had limited fire spread laterally from the exposed area. There was, however, extensive damage in the direct fireexposed area, ranging from 40 to 100% charring through the insulation thickness. The specimens with glass wool and stone wool experienced no charring but were discoloured through 70 to 80% of their thickness by soot and possibly combustion of the binder.
Classification according to EN 13501-1 does not reflect the heat transport into and through the construction and insulation. An estimation of thermal insulation products' fire performance is, therefore, not possible based only on this classification. This is, however, not the purpose of the reaction to fire classification given by the standard, but some users might assume that products with the same classification perform similarly in a fire, which is not necessarily true.
Combustible insulation materials can contribute to the development and spread of a fire and cause increased smoke production. Smoke production has not been studied in these
experiments but must be considered when using such products. However, whether the insulation is combustible or not, the required fire resistance for both loadbearing and fire
separating constructions must be ensured to prevent structural collapse and fire spread to other parts of the building.
(2) Fire protection performance of claddings used on combustible insulation
The goal was to determine the effect of fire protective claddings when installed on different types of insulation. The non-standardised tests showed that a 13 mm Norgips Standard gypsum board type A cladding with classification K210 A2-s1,d0 provided approximately 7 minutes longer fire protection compared to a 12 mm plywood board cladding with classification K210 D-s2,d0. The gypsum board and the plywood board protected the insulation for approximately 17 and 10 minutes, respectively.
(3) The fire performance of the transition between the wall and roof
The goal was to investigate the fire performance of the transition between claddings on the wall and in the roof.
The non-standardised tests showed that the transition between the walls and roofs performed well, i.e., this was not a weak spot where the fire burned through to the insulation faster than through the cladding. The results obtained in this research project are based on the insulation and cladding products tested here. Other products of similar materials might have different fire performance.