The Fire Resistance of Medium-Rise Timber
Frame Buildings Summary Report
|Commercial in Confidence
|Report reference: TF2000 - Compartment Fire Test
- Mostyn Bullock, Principal Engineer, Chiltern International
Fire/TRADA Technology Ltd, UK
- Tom Lennon, Senior Engineer, Building Research Establishment
- Vahik Enjily, Programme Manager Timber Engineering,
Building Research Establishment (BRE), UK
As part of the Timber Frame 2000 (TF2000) project
a large-scale compartment fire test has been carried out in the
full scale six storey timber frame building at BRE Cardington
in the U.K. The test took place on the 15th September 1999. The
programme of work is an important aspect of a unique collaborative
project between Government, BRE, TRADA Technology Ltd and the
timber industry, focused on enhancing the UK’s potential
to become the world leader in the provision of medium-rise timber
frame buildings. Changes to the Building Regulations in 1991
have made it possible to construct timber frame buildings of
more than four storeys. The interest generated in the TF2000
project has shown that there is a demand for medium-rise residential
timber frame buildings of five or more storeys. However, regulatory
requirements between Scotland and the remainder of the UK are
inconsistent. In particular the requirements regarding fire safety
would benefit from harmonisation based on a rational approach
to the problem.
As with other materials, fire resistance for
elements of structure is based upon standard fire tests on individual
elements. With new methods of construction and new demands on
the industry there is a need to demonstrate the performance of
medium-rise timber frame buildings subject to real fires. The
construction of the six storey timber frame building at Cardington
represented a unique opportunity to obtain valuable data on the
performance of a complete building subject to a real fire.
The fire test compartment consisted of a single
flat on level 3 of the building. The fire load was provided by
timber cribs spread over the floor area of the flat. Although
the test provided useful information on fire growth and development,
the primary objective was to evaluate the performance of the
structure subject to a fully developed post-flashover fire. Ignition
took place in the living area. Flame spread was completely uninhibited
and the ventilation to the compartment arranged to provide a
worst case scenario for fire severity. One of the test objectives
was to assess whether compartmentation of the building was effective
in preventing fire spread from the flat of origin to adjoining
flats through party walls, windows, floors or communal stair
and lift shafts and in maintaining the integrity of the means
of escape and structural stability.
In order to assess the performance of the structure
in maintaining both the safety of the occupants and the integrity
of the compartmentation for the required period of time the building
was comprehensively instrumented. Instrumentation included thermocouples
to measure the temperatures in the compartment and the heat rise
in the cavities surrounding the compartment. Load cells were
used to yield data on the rate of heat release in the compartment
and automatic fire detection, gas analysis and heat flux meters
were used to provide information on the tenability criteria.
The purpose of this paper is to document the fire test programme
for the timber frame building, to describe the performance of
the structure and to present interim results and conclusions.
The primary objective of the TF2000 compartment
fire test was to evaluate the fire resistance of a medium rise
six storey timber frame building subject to a severe natural
fire exposure. The aspects of fire resistance being assessed
were structural integrity and compartmentation. The test allowed
for a realistic assessment of fire spread and provides a quantitative
appraisal of the true performance of forms of construction tested
to the Fire Resistance test. The test provided the opportunity
to demonstrate that this form of construction can meet the functional
requirements of the Building Regulations for England and Wales
and the Building Standards for Scotland for such buildings.
The fire test took place in the level 3 (2nd
floor) flat in the South West corner of the building. The rooms
of the test flat location are identified in plan in the top left
hand corner of figure 1.
Figure 1: Second floor plan © Chiltern
The fire was ignited in the living area of the
flat and progressed to flashover after approximately 24 minutes.
Initial burning was concentrated in the front of the living area
closest to the ventilation opening. To accelerate the time to
flashover the Fire Brigade was asked to intervene by breaking
a single windowpane in the kitchen area. This took place 21 minutes
and 30 seconds from ignition. Following flashover the Fireline
boards over the windows to the floor above were subject to a
heat flux of approximately 30kW/m² (peak plume temperature
in excess of 500°C). The timber frame of the window would,
if exposed, have ignited. Peak temperatures in the living area
of the fire flat reached approximately 1000°C and remained
at this level until the test was stopped at 64 minutes having
reached one of the planned termination criteria.
Based on measurements taken of fuel mass loss,
the peak rate of heat release has been calculated as approximately
The decision to stop the test was taken on the
basis of predetermined termination criteria, one of which was
related to duration of fire attack on exposed joists. The fire
was extinguished by fire brigade intervention through the front
door of the fire flat.
The atmosphere temperatures inside the fire
flat are shown in figure
2 which includes a comparison with the standard ISO 834 fire
exposure for correlation. The measured values of air temperature
in the lobby, the adjacent flat, the flat above and the flat
below are shown in figure
The test was stopped once the fixity of the
ceiling plasterboards in the living room had been lost and the
joists were directly exposed to fire for a period of approximately
8 minutes. Figure
4 shows the temperature in the structural bays of the ceiling
void in the area most directly affected by flame impingement
5 shows the measured temperature profile on the floor joist
in the centre of the living area.
At the time when the test was stopped both the
compartmentation and structural integrity of the building were
maintained. Maximum temperatures in the structural voids forming
the boundaries of the compartment generally remained below 100°C
with the exception of the localised areas where the timber members
were exposed to fire for a period of time. The fire load in the
bedrooms did not ignite although there was charring of the timber
members of the walls enclosing the corridor in the hot smoke
layer. Higher temperatures were recorded in the bathroom and
had there been any fire load in the bathroom this may have ignited.
6 illustrates the fire dynamics of the test with the times
and effects of key events highlighted. The temperatures shown
were recorded at two different locations below the ceiling
of the living area.
Despite average atmosphere temperatures in excess
of 900°C for 30 minutes there was no evidence of fire spread
outside the compartment of origin during the test. At no time
prior to the intervention of the Fire Brigade did air temperatures
exceed 50°C in the lobby, the adjacent flat, the flat above
or the flat below. With the exception of localised breakdown
of the plasterboard linings between the kitchen area and the
adjacent flat and the loss of plasterboards to the ceiling which
led to the decision to terminate the test, temperatures in the
cavities both internally and externally did not exceed 100°C
and generally remained below 50°C. The cavity barriers remained
effective for the duration of the test.
Tenability conditions in the Protected
Results from the test illustrate a gradual reduction
in visibility to approximately 30m just prior to the Fire Brigade
opening the entrance door to the flat at 59 minutes and 12 seconds
from ignition. During firefighting operations the visibility
dropped to a minimum of approximately 6 metres.
The smoke detector in the lobby was activated
immediately following ignition as the flat entrance door remained
open for some time. The CO detector in the lobby registered an
alarm condition (300ppm) at 28 minutes and 30 seconds from ignition.
The CO level at nominal respiratory height (1.58m above floor
level) close to the stair access door registered only a few parts
per million (ppm) above the ambient level up to the point where
the Fire Brigade gained access to the fire flat.
Pressure measurements indicate that the fire
rapidly established a dynamic positive pressure within the fire
flat reaching a peak at the top of the entrance door of approximately
5.5 Pa post-flashover. This would have had the effect of forcing
smoke and hot gases around door edges but video evidence supported
by visibility, carbon monoxide and temperature measurements indicate
that the smoke and intumescent seals around the door mitigated
The data collected in the protected lobby indicate
that conditions favourable for means of escape would have existed
for the duration of the test. This was borne out by observations
from the Fire Brigade following the test.
The concept of time equivalence relates the
severity of a natural fire exposure to an equivalent period in
a standard ISO 834 fire resistance test. The fire test allowed
an empirical estimate of equivalent duration to be calculated
using data from the compartment test. Temperatures recorded at
the back of the plasterboard in the ceiling void of the living
room area in the test were compared to similar points in a section
of replica ceiling tested in a furnace and exposed to standard
temperature and pressure conditions. Figure
7 shows the average results from the comparison. The temperature
in the TF2000 test exhibits a slower initial rate of heat rise.
This was expected as flashover did not occur until 24 minutes
into the test. At this point there was a severe heat flux which
removed the chemically bonded water from the gypsum board at
a faster rate than in a fire resistance test. The graph indicates
that the TF2000 test was approximately 10% more severe than a
60 minute fire resistance test exposure. Inspection of the remains
of the furnace specimen show very little charring of the timber
joists. The joists in the TF2000 test were charred to a significantly
greater extent in areas of the ceiling close to the window.
A number of softwood and hardwood cubes were
positioned in the fire flat for the test and replicas of these
cubes were subsequently exposed to a standard furnace exposure
for exactly 60 minutes to provide a comparison of the extent
of charring. The results of the charring analysis are summarised
in table 1.
Table 1 The results of the
|* H = hardwood S
The compartment fire test met the stated objectives
of the programme. The following conclusions may be drawn from
an analysis of the data and from observations during and after
- Derived values of time equivalence have demonstrated that
the performance of a complete timber frame building subject
to a real fire is at least equivalent to that obtained from
standard fire tests on individual elements.
- Results indicate that fire conditions in the living room
of the flat represented an exposure approximately 10% more
severe than a standard 60 minute fire resistance test.
- The test has demonstrated that timber frame construction
can meet the functional requirements of the Building Regulations
for England and Wales and the Building Standards for Scotland
in terms of limiting internal fire spread and maintaining
In meeting the requirements of the regulations
and the objectives of the research programme a number of issues
- The standard of workmanship is of crucial importance in
providing the necessary fire resistance performance especially
nailing of plasterboards.
- Correct location of cavity barriers and fire stopping
is important in maintaining the integrity of the structure.
- This type of construction is one that, in the United Kingdom,
has a relatively low market share generally and in medium
rise terms is very recent. For this reason fire brigades
are unlikely to be familiar with the type of construction
details used. Clearly education on timber frame for these
bodies is necessary.
- The issue of vertical flame spread from floor to floor
via the windows needs to be addressed.