Fire Testing of Roofing Membranes
Disturbing news from the United States
relating to the fire testing of roofing membranes should be of concern to
roofing professionals including designers, manufacturers, building owners,
contractors, testing agencies and code authorities. The Midwest Roofing Contractors Association
(MRCA) and the National Roofing Contractors Association (NRCA) issued a report
on the subject this past January.
Background
Following a field research study conducted
back in 1991, MRCA and NRCA identified fire resistance as an important
performance attribute of most roofing systems.
In 1996 they followed up with a limited series of fire tests on coated
modified bitumen membranes. The
rationale behind these tests was to determine whether field-obtained or, in
other words, “aged” roofing systems maintained their initial fire-resistance
ratings. Half the specimens tested
passed and the other half failed.
Samples were tested in accordance with ASTM E108, (itals)“Standard Test
Methods for Fire Tests of Roof Coverings” (enditals)which is the basis of
categorizing membranes for their resistance to fire from external sources
(Class A, B, & C) and is similar to our own CAN/ULC S-107-M, “Standard
Methods of Fire Tests of Roof Coverings.”
In 2001, MRCA conducted an additional
series of fire tests on existing roofing membranes. This time the range of materials tested was
expanded to include granular-surfaced modified bitumen membranes, PVC
membranes, EPDM and TPO roofing systems. Again, only half the specimens passed
the fire tests.
In 2002, both the MRCA and NRCA continued
the research by conducting more fire tests, but this time on only new and aged
EPDM and TPO membranes. These membranes
were selected because the results for these systems exhibited the widest
variability in the 2001 testing. The
purpose of this series of tests was to again determine whether the aged roof
systems maintained the initial fire-resistance ratings and evaluate whether the
new roof systems tested exhibited fire characteristics similar to those of aged
roof systems tested. Only 13 of the 34
specimens tested passed the fire tests.
Of particular note with this round of tests, of the 13 that passed, 12
were from systems that included a cover board over the base polyisocyanurate
roof insulation.
Predictably and following the release of
the 2002 testing, single-ply membrane manufacturers and Underwriters
Laboratories (UL) challenged the results.
Independent Consultations & Testing
In late 2002, largely in response to
criticisms raised and the apparent disparity in the test results, MRCA and NRCA
retained a widely recognized, independent fire consultant, Hughes Associates
Inc. (HAI) based out of Baltimore, MD, to develop and oversee an additional
series of fire tests.
HAI called for a series of testing with
identical, “new” mechanically-attached EPDM and TPO membrane roof systems at
four recognized fire test laboratories.
All systems tested were listed and approved by UL and Factory Mutual
(FM) as meeting the requirements of Class A exterior fire resistance. Testing
was again conducted according to ASTM E108.
The results of this round of testing
indicated that none of the EPDM specimens tested and only one-half of the TPO
specimens tested passed the requirement for Class A.
At two of the laboratories, additional fire
tests were also conducted to investigate the effects of different roof slopes,
insulation components and thicknesses, and test specimen edge termination
methods. In all, 52 HAI controlled tests
were conducted with only 13 specimens exhibiting Class A results.
Although the MRCA/NRCA testing has been
harshly criticized, the testing program has raised some disturbing issues. They relate not only to the poor results with
respect to the fire performance of the membranes, but to the very character of
the ASTM E108 test procedure. HAI has
concluded this spread-of-flame test method is “problematic, has significant
shortcomings and can potentially provide misleading results for low-slope
membrane roof systems. Also, the method
may not be appropriate for evaluating the exterior fire performance of
mechanically-attached single-ply membrane roof systems.”
The Canadian Roofing Contractor’s
Association’s (CRCA) Technical Director, Peter Kalinger has stated that “What
is most troubling is the large and unaccounted variability in the test results
among the labs where the tests were conducted. The same test, supposedly
conducted in the same way, yield very different results calling into question
the efficacy of the test method itself.
Before any conclusions can be drawn on the fire performance of the
membranes tested, what accounts for the variability has to be determined. The test may be of little value if the
results are inconsistent from lab to lab.”
MRCA/NRCA Recommendations
Both MRCA and NRCA have expressed concerns
with the results of the fire testing and HAI’s opinions. They have shared these anxieties with ASTM,
FM, UL, various testing laboratories and the manufacturers of the roof systems
tested.
Until such time that ASTM E108 is revised
to address the concerns their research has raised, MRCA and NRCA has announced
some pretty strong recommendations. Roof
system designers must ensure that their designs comply with exterior fire-resistance
ratings as determined by a code-approved testing agency (e.g. ULC, in Canada).
Also, because of the impact cover boards
had on the test results, they are recommending that designers include them in
their roof system designs. Furthermore,
for mechanically-fastened single-ply roof systems, the cover boards must be of
the non-combustible type such as glass-mat-faced gypsum boards and gypsum roof
boards.
Canadian Implications
While these studies will no doubt have a
direct impact on the U.S. roofing industry, it remains to be seen what effect
they will have on the Canadian roofing market.
As indicated, the ASTM E108 test method is similar to our own CAN/ULC
S-107-M.
Most ICI roofing products used in the
Canadian marketplace are of U.S. origin.
Designers must be cautious and determine that these materials and
systems have been tested to the appropriate Canadian standards as dictated by
the Ontario Building Code. That being
said, one must wonder whether the roofing systems we are specifying and
designing are actually performing as the manufacturers’ literature claims.
In light of the U.S. experience, it would
be prudent on behalf of the roof system designer, to as a minimum, follow
MRCA/NRCA’s recommendations and investigate this issue thoroughly.
