22 November 2019
Yearbook article: Addressing the fire safety challenges of ambitious timber structures
This is an extract from the TRADA 2019 Timber Industry Yearbook.
Timber structures are undergoing a resurgence due to their aesthetic appeal, environmental credentials and economic speed of construction. Engineered wood technologies have made timber structures viable in markets that have typically been dominated by steel and concrete. As such, the complexity of schemes for which timber is adopted and the scale of building in which it is used is rapidly shifting.
The burgeoning tall timber market polarises opinion. It can be difficult to dispute the aesthetic, environmental and construction programme credentials. Nonetheless, timber is combustible and the perception of its performance in fire is rooted in a long and distressing history. A century or more ago various ‘great fires’, such as London (1666), Chicago (1871), Baltimore (1904) and San Francisco (1906), all devastated neighbourhoods that were constructed of timber buildings. Because of the destructive aftermath of such fires, building resilience in fire began to enter the consciousness of the world’s developed societies. The need for ‘fire-proof’ materials became apparent, which instigated an important and familiar concept – fire resistance.
To assess the veracity of a product’s fire-proof credentials, a consistent means of benchmarking was required – a test (Figure 1). The test that came to define the fire resistance of an element of construction is consistent with that which emerged in a period where fire science was in its infancy. The heating regime (Figure 2) was intended to be more severe than any fire that could be reasonably expected using anecdotal evidence from the fire brigades, who were seen as the authority on fire development at the time.
However, shortly after the emergence of the standard fire test, it was quickly appreciated that fires did not behave according to the adopted test heating regime. The work by Simon Ingberg in 1928 led to a concept that allowed real fires to be correlated with the idealised test heating regime according to the non- permanent fire load (furniture, fixtures, stored goods, etc).
This concept allowed fire-resistance durations (or periods) to be defined according to a building’s use, but on the pretence that the structural frame or enclosure does not contribute to the available fuel.
The same principles still define what fire resistance is required for elements of structure in different buildings today, with values proposed in various guidance documents and building codes around the world. Ironically, the standard tests are now adopted to assess the fire-resistant performance of building elements constructed from combustible materials, too. The relevance of fire resistance for timber structures is often debated, what cannot be ignored is the influence that the principle has on design today.
Modern timber buildings and fire design guidance: a divergence
Given the aesthetic qualities of timber, there is often an architectural ambition to have the structural elements expressed. Engineered wood products can provide the visual qualities and durability of traditional solid-sawn timber coupled with a greater strength and reduced production time. Products are composed of boards or lamella bonded together in layers. This approach results in an increased strength over solid timber as imperfections such as knots are not as critical to the product performance. Production times are reduced as it is not necessary to wait for a tree to grow to the same extent as required for a solid timber element of the same dimensions. In addition, engineered wood products can be formed into panels of dimensions not possible using single pieces of sawn timber. Typical engineered wood products include cross- laminated timber (CLT), glued laminated timber (glulam) and laminated veneer lumber (LVL).
The traditional treatment of timber in fire is through charring rates, with elements sized to accommodate the relevant load combination after accounting for the loss of material as the timber combusts. Conventionally, a fire-resistance period is defined according to some recognised guidance, for example Approved Document B or BS 9999, and subsequently elements are sized to achieve, for example, 90 minutes’ structural fire resistance. This sequential approach de-couples the design of the structural element from its impact on fire development, essentially assuming that the contribution of the structure to the fuel is negligible. This is in conformity with the original work of Ingberg and the resulting fire-resistance period recommendations found throughout the world.
Where the structure constitutes a high proportion of an enclosure’s lining, the de-coupling of the combustion of the structural elements from the enclosure fire environment is, at best, crude, and at worst, dangerous. In such a case, the contribution of the structure to the fuel is no longer negligible. The impact of the exposed structure has been shown experimentally to potentially significantly impact fire development and fire severity. In doing so, the core assumptions behind fire-resistance periods cease to be valid, with alternative fire safety engineering approaches likely to be the only practical route to compliance with statutory fire requirements.
An engineering methodology for exposed timber enclosures
The starting point for an engineering methodology for structures where enclosures are formed of exposed timber is an acknowledgement that the status quo approach is inappropriate. There must be cognisance of the impact any exposed timber has on the fire development, and subsequently what measures need to be put in place to ensure the structure has an appropriate likelihood of withstanding the burnout of a fire. However, the underlying principle of designing exposed timber buildings that are resilient to fire is not different from more traditional forms of construction. This is to avoid fire spread beyond the compartment of fire origin and provide enough structural resistance until a full burnout of the fire is achieved.
Recently, several projects in the UK have included CLT elements as part of the design of the building structure. CLT is made from planks glued together in layers perpendicular to one another and usually manufactured in large panels several metres in each direction. Various configurations are common such as three, five or seven layers, which can have equal or unequal thicknesses, such that total panel thicknesses can range from 40mm to 300mm. In contrast to solid timber, where the charring rate is taken to be constant through the thickness, CLT panels have the potential for layers to fall off due to delamination/debonding. This is possible if the glueline between layers is subjected to increased temperatures so that there is a softening of the adhesive. If delamination occurs, then an increase in the fire’s heat release rate is likely due to virgin wood behind the delaminated portion being exposed to increased temperatures and subsequently igniting.
There are many variables that can influence the delamination of CLT layers such as:
- the adhesive type
- orientation of the panel
- grain direction
- timber species
- number and variability in thickness of the layers.
The delamination mechanism of CLT is still not fully understood and research is ongoing into each of the aforementioned variables. However, it is known that if the gluelines are subjected to increased temperatures, delamination is likely. Hence, a key design principle when analysing the fire resistance of CLT slabs is to try to prevent delamination by limiting the prospect of the char interface (commonly adopted as the in-depth penetration isotherm at 300°C) from reaching the first glueline.
OFR Consultants, alongside global research institutes, has developed an engineering methodology that provides a robust basis for the design of exposed CLT structures while acknowledging the uncertainty in the causes of delamination. The main foundation of this methodology is to mitigate the prospect of delamination, thus allowing the char that develops to stay in place and insulate the virgin wood below. If the char layer interface does not reach the first glueline, it is assumed that delamination will not occur. As long as this facet holds true, eventually auto-extinction of the timber can be reasonably expected.
This methodology has been applied to UK projects (Figure 3) and uses the latest version of the B-RISK fire zone model developed by BRANZ. This is used to simulate fire development in an enclosure and at each point in time computing the heat flow through exposed internal surfaces. In doing so, the depth of char is computed, with this timber mass added to the available fuel. Depending on the design problem, the designer may choose to place more emphasis on the combustion of the extra fuel, leading to external flaming or a prolonging of the fire’s duration. Once full burnout has been simulated, the depth of char to exposed internal faces is determined, and this can be used to inform what minimum internal lamella thickness might be appropriate to obviate delamination, in the process supporting a greater possibility of the fire undergoing auto-extinction. Further, where depths of char are too deep, and the exposed lamella thickness is impractical to achieve, the method can be adopted to inform the extent to which exposed timber is justifiable.
Adopting the method generally shows that a by-product of the enclosure being constructed of a combustible material is an increase in fire severity. Where it is necessary to have some of the enclosure’s exposed timber encapsulated, any protective linings should be specified to achieve a fire resistance that is capable of withstanding the burnout of the relevant design fire conditions and in consideration of the impact any exposed combustible enclosure material has on fire severity. Armed with charring depths, the extent of protective linings and any constraints on exposed lamella thickness, designers are then in a position to undertake relevant checks on the loadbearing performance of the structural timber elements.
This is an extract from the TRADA 2019 Timber Industry Yearbook. Download the full article, including supporting images, references and further reading, here
Article written by OFR Consultants' Danny Hopkin, Associate Technical Director, Stephen Dickson, Fire Engineer, and Michael Spearpoint, Research Leader. Danny Hopkin presented on the changes to Building Regulations guidance and the subsequent responses by industry sectors and companies at the Better Timber Buildings conference this month.