THE OFFICIAL MAGAZINE OF THE SOCIETY OF FIRE PROTECTION ENGINEERS

FALL 2004 Issue No.24

UNIQUE
INTERIORS
ON THE
ALSO:

12 FIRE TESTING OF INTERIOR FINISH

LAS VEGAS 18 THE ROLE OF INTERIOR FINISH IN
STRIP page 4 FIRE DEVELOPMENT

29 ASSESSING THE BURNING

CHARACTERISTICS OF INTERIOR
FINISH MATERIAL
 FIRE PROTECTION

Fire Protection Engineering (ISSN 1524-900X) is

published quarterly by the Society of Fire Protection
Engineers (SFPE). The mission of Fire Protection
Engineering is to advance the practice of fire protection
engineering and to raise its visibility by providing
contents FA L L 2 0 0 4

information to fire protection engineers and allied

professionals. The opinions and positions stated are
4
the authors’ and do not necessarily reflect those of SFPE. COVER STORY
Editorial Advisory Board Unique Interiors on the Las Vegas Strip
Carl F. Baldassarra, P.E., Schirmer Engineering Corporation How fire protection engineers approach design in some of the most diverse
Don Bathurst, P.E. buildings in the world.
Bob Boyer, Edwards Systems Technology Doug Evans, P.E.
Russell P. Fleming, P.E., National Fire Sprinkler Association
Morgan J. Hurley, P.E., Society of Fire Protection Engineers
William E. Koffel, P.E., Koffel Associates 2 Viewpoint
Jane I. Lataille, P.E., Los Alamos National Laboratory
Margaret Law, M.B.E., Arup Fire 3 Flashpoints
Edward Prendergast, P.E., Chicago Fire Dept. (Ret.)
Warren G. Stocker, Jr., Safeway, Inc. 12 Fire Testing of Interior Finish
Beth Tubbs, P.E., International Code Council Traditional and modern methods of testing for interior wall and ceiling
Regional Editors finish and interior floor finish.
U.S. H EARTLAND Marcelo M. Hirschler, Ph.D.
John W. McCormick, P.E., Code Consultants, Inc.
U.S. M ID -ATLANTIC 18 The Role of Interior Finish in Fire Development
Robert F. Gagnon, P.E., Gagnon Engineering, Inc. A review of past fires where interior finish contributed and new ideas for
U.S. N EW E NGLAND evaluating flammability characteristics of combustible interior finishes.
Thomas L. Caisse, P.E., C.S.P., Robert M. Currey &
Associates, Inc. Robert Brady Williamson, Ph.D., P.E., and Frederick W. Mowrer, Ph.D., P.E.
U.S. S OUTHEAST
Jeffrey Harrington, P.E., The Harrington Group, Inc. 29 Assessing the Burning Characteristics of Interior Finish
U.S. W EST C OAST Material – Standard Test Method for Surface-Burning
Michael J. Madden, P.E., Gage-Babcock & Associates, Inc.
Characteristics of Building Materials (ASTM E-84/UL 723)
A SIA
Peter Bressington, P.Eng., Arup Fire
An overview of the Steiner tunnel test method including history,
A USTRALIA
acceptance, advantages, and limitations.
Brian Ashe, Australian Building Codes Board Randy Laymon
C ANADA
J. Kenneth Richardson, P.Eng., Ken Richardson Fire 34 Fire Alarm Systems and Interior Finish – A Balanced Approach
Technologies, Inc. How does a fire detection and alarm system affect the selection of interior
N EW Z EALAND finishes? When interior finish changes, how must the fire alarm system
Carol Caldwell, P.E., Caldwell Consulting change as well?
U NITED K INGDOM National Electrical Manufacturer’s Association
Dr. Louise Jackman, Loss Prevention Council
Personnel
EXECUTIVE DIRECTOR, SFPE
David Evans, P.E.
T ECHNICAL E DITOR
Morgan J. Hurley, P.E., Technical Director, SFPE
P UBLISHER
Terry Tanker, Penton Media, Inc.
A SSOCIATE P UBLISHER
Joe Pulizzi, Penton Custom Media, Penton Media, Inc.
M ANAGING E DITOR
Timothy Stark, Penton Custom Media, Penton Media, Inc.
A CCOUNT C OORDINATOR Cover photo: The Image Bank/Venetian Hotel & Casino, Las Vegas, Nevada
Jillian Lewis, Penton Custom Media, Penton Media, Inc. Online versions of all articles can be accessed at www.sfpe.org.
A RT D IRECTOR
Pat Lang, Penton Custom Media, Penton Media, Inc. Invitation to Submit Articles: For information on article submission to
Fire Protection Engineering, go to http://www.sfpe.org/sfpe30/fpemagsubmit.htm.
M EDIA S ERVICES M ANAGER
Erik Lodermeier, Penton Custom Media, Penton Media, Inc. Subscription and address change correspondence should be sent to: Fire Protection Engineering,
C OVER D ESIGN Penton Media, Inc., 1300 East 9th Street, Cleveland, OH 44114 USA. Tel: 216.931.9180. Fax: 216.931.9969.
Dave Bosak, Penton Custom Media, e-Mail: asanchez@penton.com.
Penton Media, Inc. Copyright © 2004, Society of Fire Protection Engineers. All rights reserved.

www.sfpe.org 1
 viewpoint

Interior Designers Put Fire Protection into Practice

I nterior designers are visionaries. They

transform interior space from the intangi-
ble to the tangible – they introduce a new
aspect, a new purpose, a new function. They
see what is not yet there, and through their
and smoke while occupants escape to safety.
Benchmarks for success are diverse: build-
ing and fire code compliance for life safety;
ADA compliance to remove barriers for uni-
After the ribbon-cutting opens a compliant
facility, whether a business or residence, it is
the inhabitants and those responsible for
maintenance who can unknowingly diminish
versal accessibility; sustainability in the selec- or even negate code-compliant details. Ulti-
ability to convey their vision to the client, tion of materials; adaptive reuse of existing mately, it is the property owner who must
they are given the opportunity to make the space; green design for the preservation of take responsibility for the ongoing safety of a
vision a reality. They orchestrate the elements natural resources; color and visual, tactile and building. This involves fire safety training of
of space, color, texture, and scale, and apply audible details for improved health and well all inhabitants whether they be residents,
them in new and different ways, creating being; space planning for order and paths of customers, employees, or family members.
desired outcomes for the client. egress; recyclability to reduce overflow in Understanding what is involved in fire
There is not one process, element, or task land fills; indoor air quality to prevent off-gas- protection and incorporating safety tech-
involved in the design of interior space that ing that causes sick building syndrome; ease niques as standard business practice is good
does not take into consideration the preser- of maintenance – the list goes on. customer service. Often the last to leave the
vation of the health, safety, and welfare of It is not possible for code officials to be project, interior designers have the opportu-
human life: furniture is selected through aware of every material specified by designers, nity and challenge to educate clients on the
knowledge of ergonomics; interior structure so thoughtful interpretation is required for new care, handling, and ongoing maintenance of
and furniture placement are determined products not yet addressed in the codes. This newly installed interior finishes, fixtures, and
through knowledge of accessibility needs, is the stuff of ongoing code development. Not furnishings.
egress requirements, adjacency preferences; only the product, but its specific application Interior designers pass along to their
color schemes are determined through must be considered. For interior designers clients a binder of the written specifications,
knowledge of their psychological impact on whose projects are located in numerous juris- photos, and testing documents for all fur-
the human experience; wall and ceiling fin- dictions, whether by state, region, or nation- nishings, fixtures, and finishes included in
ishes are selected through knowledge of ally, the lack of one universal code com- their project. This information serves the
acoustics; floor finishes are selected through pounds the research and paper trail dual purpose of confirming that the de-
knowledge of performance, way-finding, and documentation, not to mention the multiple signer has performed his/her responsibili-
surfacing requirements; and lighting is se- finishes that differing code officials might re- ties within the current codes and invites the
lected through knowledge of ambient, task, quire for clients with multiple projects in just end-user to be trained on the importance of
and focal requirements. as many locations. preserving that compliance.
Whether they specialize in office design, In addition, building and fire codes are en- Less than half of the interior designers cer-
hotel design, restaurant design, healthcare forced by unique individuals who do not al- tified by the National Council of Interior De-
design, residential design, etc., every individ- ways share the same interpretation of the sign Qualification (NCIDQ) and meeting all
ual element combines with the whole to cre- codes. Even two officials working in the requirements for education and experience
ate a cohesive unit that enriches the lives of same office can have differing judgments on are able to provide services directly to clients
the inhabitants and achieves the client’s a code’s intent and application. Designers due to the lack of state registration. Only
strategic goals. may choose to work exclusively with a single twenty-four states in the U.S. currently ac-
Projects begin with the assemblage of in- code official throughout their project, re- knowledge the profession of interior design
formation pertinent to the project. Once pro- questing documentation of the code interpre- through licensure that identifies “registered
ject goals and objectives are clearly outlined, tations in writing. For example, in some com- interior designers” as “design professionals”
the steps that lead the project to a successful munities where CAL 117 has been adopted, along with engineers and architects.
completion are put in place. Parameters for compliance with CAL 133 can be required in Support of interior design legislation is
design are drawn through careful communi- certain situations, completely at the discre- needed to ensure the application of codes by
cation and input from the end-user(s), the tion of the code official. professional interior designers who, along
client, and other project team members. Acquiring and compiling approvals docu- with engineers and architects, champion the
Initially, research of all codes that pertain mentation is the responsibility of the interior health, safety, and welfare of the public.
to the project are considered and applied to designer. Since product is tested by the com-
every element involved in the facilitation of posite piece, even though the disparate parts Lisa Bonneville, ASID, is principal of Bon-
the design: furniture, finishes and coverings, may already have performed to code, this neville Design, an interior design firm serv-
fixtures and equipment, accessories and dec- process can be quite involved. Depending ing residential, retail, business, and health-
orations. This is comprehensive and time- on the scale of a project, the cost of sacrific- care clients. She is also a professional
consuming. It requires meeting with the local ing product to testing may limit the available member of the American Society of Interior
plan reviewers and/or code officials who options and, ultimately, the design itself. Of- Designers (ASID), Certified by NCIDQ, mem-
preside in the jurisdiction where the project ten included in a new or renovated interior ber of the NFPA Technical Committee on Fur-
is located in order to understand together the are materials that the client has expressed a nishings and Contents representing ASID,
full scope of the project, the goal of the de- desire to reuse. If these materials have served and chairperson of the Fundraising Com-
sign concept, and from that, interpret the in- well in their previous capacity and still have mittee for the Massachusetts Interior Design
tent of the codes. Their ultimate goal is to good, code-compliant components, they may Coalition (MIDC) supporting House Bill
preserve life by retarding the spread of fire be refurbished for reapplication. #2592 for licensure for interior designers.

2 Fire Protection Engineering N UMBER 24

 flashpoints
fire protection industry news

Sunderland Joins
Universityof Maryland The SFPE Corporate 100 Program was founded in
1976 to strengthen the relationship between industry
Dr. Peter B. Sunderland has joined the Department of Fire Protec- and the fire protection engineering community.
tion Engineering, University of Maryland, as Assistant Professor. He Membership in the program recognizes those who
support the objectives of SFPE and have a genuine
was previously at the National Center for Microgravity Research at the concern for the safety of life and property from fire.
NASA Glenn Research Center in Cleveland, OH.
BENEFACTORS
FM Global Corporation
Professor Sunderland’s degrees are from Cornell University (B.S.), Koffel Associates, Inc.
the University of Massachusetts (M.S.), and the University of Michigan Rolf Jensen & Associates, Inc.
SimplexGrinnell
(Ph.D.). His research interests are in combustion and fire protection.
His specializations include soot formation, microgravity combustion, PATRONS
laminar diffusion flames, oxygen-enhanced combustion, and experi- Code Consultants, Inc.
Gage-Babcock & Associates, Inc.
mental methods in combustion. Hughes Associates, Inc.
For more information visit National Fire Protection Association
The Reliable Automatic Sprinkler Company
www.eng.umd.edu. Schirmer Engineering Corporation
Specified Technologies, Inc.
Tyco Fire and Building Products, Inc.

MEMBERS
NIST Provides Fire Resistance Data Altronix Corporation
on WTC Floor Systems Ansul, Inc.
Arup Fire
Four fire resistance tests conducted on composite concrete-steel Automatic Fire Alarm Association
trussed floor systems typical of those used in the World Trade Center Cybor Fire Protection Company
Edwards Systems Technology
(WTC) towers showed the test structures were able to withstand stan- Fike Corporation
dard fire conditions for between one and two hours, according to the GE Global Asset Protection Services
Harrington Group, Inc.
National Institute of Standards and Technology (NIST). HSB Professional Loss Control
James W. Nolan Company (Emeritus)
Marsh Risk Consulting
The 1968 New York City building code – the code the towers were National Fire Sprinkler Association
intended but not required to meet when they were built – required a Nuclear Energy Institute
two-hour fire rating for the floor system. The Protectowire Co., Inc.
Reliable Fire Equipment Company
TVA Fire and Lifesafety, Inc.
Shyam Sunder, lead investigator, explains that the tests provide Underwriters Laboratories, Inc.
Wheelock, Inc.
only a means for evaluating the relative fire resistance rating of the Williams Fire & Hazard Control, Inc.
floor systems under standard fire conditions and according to ac-
SMALL BUSINESS MEMBERS
cepted test procedures. Sunder cautions, “These tests alone cannot
Beall & Associates, Inc.
be used to determine the actual performance of the floor systems in Bourgeois & Associates, Inc.
the collapse of the towers. However, they are already providing valu- Davidson and Associates
Demers Associates, Inc.
able insight into the role that the floors may have played in causing Fire Consulting Associates, Inc.
the inward bowing of the perimeter columns minutes Fire Suppression Systems Association
Futrell Fire Consult and Design, Inc.
before both buildings collapsed.” Gagnon Engineering, Inc.
More information visit Grainger Consulting, Inc.
J.M Cholin Consultants, Inc.
http://wtc.nist.gov. Poole Fire Protection Engineering, Inc.
Risk Logic, Inc.
Risk Technologies LLC
Slicer and Associates, LLC
S.S. Dannaway & Associates, Inc.
The Code Consortium, Inc.
Worcester Polytechnic Institute

3 Fire Protection Engineering N UMBER 24

 Unique Interiors on the
Las Vegas Strip
By Doug Evans, P.E. In Las Vegas, one can travel to Paris,
New York, Venice, Egypt, or a tropical
INTRODUCTION island, as well as another solar system,
all within a few blocks. It is possible to

M
take a trip back in time to the Wild West,
any of the largest and most unique buildings in the the Roman Empire, or medieval Europe.
world are located on the Las Vegas Strip. The interi- Fantasy abounds, and to create these
fantasies, the interiors of the facilities are
ors of these facilities gives one the impression of transformed to achieve the desired illu-
being somewhere else and/or in a different time. sion.
The interiors of these facilities contain

4 Fire Protection Engineering N UMBER 24

artificial trees, large statues, hand- • Type of substrate and method of tially be conducted to determine if ad-
painted canvas murals adhered to the attachment. verse behavior of the specific material
walls and ceilings, as well as giant • Physical properties of the decorative can be predicted under actual fire condi-
signs/LED screens, and numerous other item (size, thickness, and product type). tions.
types of themed façades. These themed • Properties of topical applications This Data Sheet also indicates that
interiors even include faux buildings in- (pigments, varnishes). small-scale testing may not be represen-
side the main facility. • Combustible concealed voids (com- tative of the respective hazard. Failure to
How can the required fire protection partmentation, sprinkler installation, and achieve ignition in small-scale tests is
aspects be incorporated into these plenums). not substantial proof of noncombustibil-
themed facilities and still allow architec- • Fire-retardant applications. ity. Large-scale testing may be necessary
tural freedom to achieve the design con- • Applicability of recognized fire tests. to determine the actual fire characteris-
cept? What level of fire protection is rea- • Whether hazards are temporary or tics of a material. Data Sheet 1-4 states
sonable? permanent. that “many materials incapable of
This article focuses on the two previ- • Proximity to, and significance of, achieving self-supporting fire in bench
ous questions to provide guidance in ignition sources and adjacent fuel pack- test configurations prove to be very
determining what level of protection is ages. combustible when subjected to larger-
reasonable and offers some examples to • Obstructions to occupant evacua- scale testing.”
demonstrate how that level of protection tion. Most fire testing references indicate
can be achieved. When determining a that testing should be performed in ac-
reasonable level of protection, the first APPLICABLE FIRE TESTS cordance with the expected use of the
aspect to consider is the hazard that material being tested. Potential ignition
must be mitigated. There are many different fire tests that sources must also be considered.
can be used to provide an understand-
CONSIDERATION OF THE ing of the burning characteristics of ma- UL 94 Vertical and Horizontal Burn-
POTENTIAL FIRE HAZARD terials and assemblies. Several of the ing, and NFPA 701 Large- and Small-
tests that may be applicable for the fea- Scale Versions
Looking at the subject from a perfor- tures discussed in this article are summa- These types of tests are classic bench-
mance-based viewpoint, the following rized in the following paragraphs. scale test methods that use a Bunsen
aspects need to be considered. burner type of ignition source. Except
• Proximity to fire sprinklers. Bench-Scale Testing vs. Larger Scale for UL 94 HB, the sample is typically
• Obstructions to sprinkler discharge. Factory Mutual Data Sheet 1-4 (Fire vertical, and the burner is exposed to
• Flammability characteristics (igni- Tests)1 provides general information the lower portion of the sample. Visual
tion temperature, flame spread, heat re- about fire testing. It indicates that small- observation of flame spread, char rate,
lease rate). scale “bench-type” testing should ini- and flaming droplets are evaluated. Typ-

FALL 2004 www.sfpe.org 5

 ■ Unique Interiors on the Vas Vegas strip

ically, the burner exposes the sample for 150 kilowatts. Due to the three-quarter The flame-spread index is a numerical
very short periods of time. pound (0.34 kg) wood crib used as an rating applied to tested materials. It is a
The results from these tests are only ignition source, with an approximate calculated value based on the relation-
applicable to very small, transient expo- peak heat release rate of 18 kW, these ship between the distance the flame
sure ignition conditions. tests may be considered small-scale. front extends within the test chamber
and the respective time it took to reach
UL 1975 The Steiner Tunnel Test that distance. A material with a Class A
In the late 1980s, The Society of Plas- This test is known by several designa- flame-spread rating has been tested with
tics Industry, Inc., and Underwriters Lab- tions, including ASTM E-84, NFPA 255, a flame-spread index of 25 or less. Mate-
oratories (UL) developed testing criteria and UL 723. As described in the NFPA rials receiving a flame-spread index
for foam plastics intended for use in ex- Fire Protection Handbook,3 this test was greater than 25 and up to 75 are as-
hibit booths, on film production stages, originally developed at Underwriters signed a Class B flame-spread rating.
and for decorative objects.2 Decorative Laboratories in the 1920s, and the cur- Class C materials have a flame-spread
objects include such objects as man- rent physical design was completed in index greater than 75 and up to 200. For
nequins, murals, and signs. The amount 1948. all classes, the smoke-developed index
of exposed foam plastic is dependent on This test was developed as a basis to is limited to a maximum of 450.
the proposed use and should be tested compare the surface-burning character- One of the most important concepts
at the same thickness and density as the istics of materials that form the exposed to be aware of when using the Steiner
expected application. The size of the ex- interior finishes of walls and ceilings in a Tunnel test method is that the burning
pected application should be limited to building. Reinforced-cement board is characteristics of thin combustible mate-
the size intended by the test. Larger ap- used to establish the zero value, with rials can be affected by the properties of
plications should be tested in accor- red oak flooring being assigned a rating their substrate. The lid of the furnace
dance with a larger-scale test. of 100. All other tested materials are constitutes the substrate for thin materi-
Foam plastics in exhibit booths and compared with these two values. The als tested in the tunnel. This lid is a non-
film production are allowed to have a peak heat release rate of the gas burners combustible refractory liner. As such, re-
maximum heat-release rate of 100 kilo- used as the ignition source is approxi- sults obtained from this test method may
watts. Decorative objects are limited to mately 88 kW. be quite misleading when no substrate,

6 Fire Protection Engineering N UMBER 24

 ■ Unique Interiors on the Vas Vegas strip

or a combustible substrate, is expected orientation, its actual installation, and a Upon a cursory review of these guide-
for the proposed installation. In addi- moderate fire exposure condition. For lines, they may appear more conservative
tion, the ASTM E-84 test standard speci- certain applications, newer versions of than the allowable percentage limita-
fies that the material be tested in the codes and standards allow this test as a tions. Consider a 100,000 sq. ft. (9,000 m2)
manner in which it is to be used. There- substitute for ASTM E-84. casino with the entire allowable percent-
fore, one often-misunderstood require- Other full-scale fire tests such as large age installed in one location. As such,
ment is that this test standard expects a open corner tests (FM 4880 or UL 1040) the Clark County guidelines take a
substrate to be included when thin com- or nonstandard full-scale fire tests that somewhat different approach than the
bustible materials tested in this manner replicate end-use conditions are also allowable percentage option by limiting
are installed within a building. used to provide a more accurate mea- the size of each item and requiring suffi-
Since its inception, ASTM E-84 has sure of fire performance of wall and cient separation between adjacent items
been used to evaluate all interior finish ceiling materials. to consider each such item a separate
materials. Over the years, it has been fuel package.
recognized that the results of the test ORGANIZING THE APPROACH
method may not be indicative of real-life Decorative Wall Applications
fire performance. For example, the There are many ways to organize fire Wall-type applications can include
NFPA 101 Handbook4 discusses tests protection approaches for unique interi- murals, tapestries, pictures, signs, or
conducted at the Fire Research Labora- ors. One way is to break the features other features that are affixed to or sus-
tory of the University of California at into similar concepts that are already ad- pended from facility walls. Draperies
Berkeley and sponsored by the Ameri- dressed in codes and standards. This ap- and other decorative aspects installed in
can Textile Manufacturers’ Institute in proach is outlined in Clark County’s a vertical plane may also be included.
late 1985. This testing demonstrated that Guidelines for Unique Interiors.5 One way to think about wall-type ap-
flame-spread measurements alone might These guidelines consider of the fol- plications is to consider when a picture
not reliably predict the fire behavior of lowing unique interior design elements: becomes a wall. A picture or sign can
textile wall and ceiling coverings. • Trim generally be hung on a wall without
• Wall Applications concern of a fire hazard. As the picture
Room-Corner Tests • Ceiling Applications or sign gets larger, the potential hazard
Over the last 30 years, various room- • Artificial Plants and Statues increases. At a point, the hazard may
corner fire tests have been used and • Decorative Structures within even overwhelm the building’s fire pro-
standardized. One example is the 30-lb Buildings tection systems.
(14 kg) wood crib room-corner test (UL- Paintings hung on walls are typically
1715) used in many U.S. codes. Other Trim Items considered decorative materials. If the
examples include NFPA 265, which was By its very nature, trim is limited in painting is removed from its frame and
developed specifically to address the size and quantity. Features that can be adhered to gypsum wallboard with a
flammability of textile wall coverings, classified as trim typically do not consti- noncombustible adhesive, its potential
and NFPA 286, which was developed to tute sufficient fire hazard to be a con- to burn is reduced, since thin materials
address interior wall and ceiling finish cern. When trim exceeds reasonable tend to take on the burning characteris-
materials. limitations, a greater level of protection tics of the substrate to which they are
These NFPA tests were developed to becomes necessary. The challenge is de- adhered. Eliminating one surface of a
provide additional engineering data termining what constitutes “reasonable” thin material will typically (and some-
such as heat release rate and smoke pro- limitations? times significantly) reduce that material’s
duction as well as providing a visual ob- Trim can include baseboards, chair ability to exhibit significant flame
servation of the extent of burning. These rails, crown mouldings, door/window spread. Additional considerations are
tests use a gas-fired burner placed in a frames, and handrails. The length of the material’s proximity to ignition
corner of the room with a heat output these trim items is not limited, but Clark sources and automatic fire sprinkler sys-
that replicates the fire growth of the 30- County typically limits the height/width tem effectiveness. For example, the
lb (14 kg) wood crib exposure. The to six inches (150 mm). Beyond this, higher up a wall a mural is located, the
burner produces a 40 kW heat output these trim items are considered farther it should be from most significant
for the first five minutes to simulate a wall/ceiling finish. ignition sources and the closer it will be
small fuel package, such as a wastebas- Some codes and standards allow a to sprinklers.
ket. For the following 10 minutes of the small percentage of the walls and ceil- An additional constraint is the size of
15-minute test, the heat output is in- ings to have decorative combustible fea- a mural. When exposed to fire, large
creased to 150 kW (or 160 kW, depend- tures that are considered trim. As such, murals may delaminate, and burning
ing on the test) to simulate a larger fuel these items are less regulated than if scraps of material may fall down to ig-
package, such as a chair. One of the fail- they were classified as building materi- nite one or more fires that exceed the
ure criteria is if flashover occurs. See the als. To allow decorative combustible fea- intent of the sprinkler design and over-
article in this issue on page 16 for addi- tures up to the percentage of the wall or whelm the sprinkler system.
tional information. ceiling area specified by code has been On the Strip, there are several hand-
These tests provide a more appropri- taken into account in the Clark County painted murals adhered to facility walls.
ate evaluation with respect to material guidelines. Many of these have been tested in ac-

7 Fire Protection Engineering N UMBER 24

cordance with ASTM E-84 to meet the
required limitations. Smaller murals are
deemed “pictures,” and E-84 testing is
not required.
One hand-painted mural adhered to a
facility wall is 130 ft. (40 m) long and 30
ft. (9 m) high. In this instance, flame-
spread and smoke-developed ratings
were determined using ASTM E84 and
found to slightly exceed the allowable
limit. An engineered analysis was pre-
pared to determine if an unreasonable
hazard existed. Some of the mitigating
aspects included:
• The mural was at the upper portion
of a high bay space, which placed it
near ceiling sprinklers.
• The height of the mural above the
floor reduced exposure to ignition
sources.
• The openness of the respective facil-
ity allowed the mural to be seen by occu- Figure 1. Themed rotunda
pants throughout most portions of the
space. sprinklers were installed following the Decorative Ceilings
• Exits are located such that occupants sprinkler installation criteria in NFPA 13, Ceiling-type applications can include
need not evacuate beneath the mural. there would be the potential for 5 sprin- umbrellas, awnings, canopies, nonoccu-
In a different facility, themed wall fea- klers flowing water simultaneously. If piable/decorative balconies, interior
tures were constructed off site, trucked sprinklers on both sides of it activate, eaves/projections, lattice ceilings, and
to the building, and hung on the facility that’s 10 sprinklers. Depending on how roofs of interior structures. This includes
walls. The fabricator was of the opinion the sign is situated relative to ceiling all horizontal installations of any mater-
that these features were “pictures” and sprinklers, one on each end might even ial that can cause obstruction to auto-
that ASTM E-84 testing was not neces- activate. Worst case, there could be 10 to matic sprinklers and/or delay activation
sary. After some discussion, ASTM E-84 12 sprinklers flowing water simultane- of the sprinklers.
testing was conducted, and the results ously. This application approaches the Table umbrellas are frequently about
were found to be within acceptable limi- design area of the sprinkler system and 5 ft. (1.5 m) in diameter. Even though
tations. Combustible concealed voids certainly constitutes an unacceptable these umbrellas are slightly more than
created by this application were elimi- hazard. After discussing the proposal the 4-ft. (1.2 m) limitation to obstruction
nated to reduce the potential of fire and the potential fire concerns with the from sprinkler discharge allowed by
spread within the cavities. manufacturer, this plastic sign was not NFPA 13, they may not be considered an
Plastic windows, large signs, and installed. unreasonable hazard. When decorative
extremely large rear-projection/LED When draperies are tested to deter- ceilings are larger than this, they may
televisions may also be considered wall- mine their resistance to ignition, it is typ- create an unacceptable hazard.
type applications. In the themed rotunda ically in accordance with NFPA 701. A There are several unique ceiling-type
portion of Caesar’s Forum Mall, the de- primary concern is defining a reasonable applications on the Las Vegas Strip. A
signers proposed several plastic rear- size for draperies. Some facilities use few such ceilings are fabrics or thin plas-
projection screens approximately 30 ft. draperies to subdivide large spaces. For tics. Thin combustible materials create
(9 m) long and 15 ft. (4.5 m) high (see example, when an arena is only partially various challenges. The primary chal-
Figure 1). After discussing the burning filled with people, unused seats are fre- lenge is the potential for adversely af-
characteristics of the plastic screens and quently curtained off to create a more fecting sprinkler operation. NFPA 13 re-
building fire protection systems, the intimate feeling. These draperies may quires sprinklers to be installed in the
designers opted to make the screens contain several hundred square yards plane of the membrane and at the deck
out of glass. (square meters) of material. This above (within combustible concealed
In another facility, a manufacturer arrangement may actually constitute a voids).
proposed installing a plastic LED sign 75 larger fire and potentially more haz- A concern with this type of arrange-
ft. (23 m) long and 15 ft. (4.5 m) high ardous condition than is reasonable. ment is which set of sprinklers will acti-
with a 6-inch (150mm) hollow interior. Does it seem reasonable that these ma- vate first. If the fire originates between
This type of application creates a com- terials, tested in accordance with NFPA sprinklers, the heat plume may breach
bustible concealed void in which fire 701, provide an adequate level of pro- the thin membrane, causing sprinklers
can spread unchecked, even if ceiling tection? What test is reasonable for these above to activate first. These sprinklers
sprinklers are functioning properly. If applications? can be expected to prewet the mem-

FALL 2004 www.sfpe.org 8

 ■ Unique Interiors on the Vas Vegas strip

Artificial Plants and Statues

This category can include not only ar-
tificial plants and statues, but preserved
plants, mannequins, models, and small,
nonoccupiable decorative structures.
Most of these features will be consid-
ered fuel loading within the building.
Anyone can go to their local retail
store, purchase a 6-ft. (2 m) tall artificial
plant and bring it into a building without
creating a hazardous condition. When a
plant or mannequin becomes 40-ft. (12
m) tall and may include silk or plastic
leaves creating a 30-ft. (9 m) diameter,
combustible obstruction to sprinkler dis-
charge, the hazard may have increased
to an unacceptable level. In addition to
considering the fire-protection aspects,
these large structures might fall on occu-
pants in the event of a fire. So, not only
the fire protection aspects are of a
concern, but the structural aspects may
also require increased consideration.
The Las Vegas Strip contains numer-
Figure 2. Hand-painted mural. ous statues and artificial trees that are
large enough to warrant increased pro-
brane below and the piping/sprinklers that is 100-ft. (30 m) in diameter (see tection. In most cases, they are con-
that penetrate the membrane. Since wa- Figure 2). The burning characteristics of structed of materials acceptable for the
ter from the sprinklers above the mem- this mural are similar to the mural de- base building. This means that the struc-
brane may not be able to get to the seat scribed previously in the decorative wall tural aspects are minimally noncom-
of the fire and sprinklers that penetrate applications portion of this article. Part bustible and, at times, may even be pro-
the membrane can be expected to be of the mitigating aspects included the tected from fire. Combustible voids are
wet, the fire may spread below. Even if height of the mural above the floor and either eliminated or mitigated. The exte-
lower-level sprinklers activate, these thin the respective separation from ignition riors of these features are frequently
membranes may drape down and restrict sources. noncombustible. If it is necessary to fab-
proper water distribution. ICBO ES AC Other approaches can frequently be ricate the exterior out of combustible
171,6 using a modified version of the used to achieve the design concept. If materials, they are typically required to
room-corner test, was intended to deter- an interior awning is desired, sandwich- meet the ASTM E-84 criteria required for
mine if thin combustible ceilings will cre- ing sheet metal between two layers of the respective facility.
ate an unreasonable hazard and ad- fabric may meet the design goals. This The leaves of artificial trees constitute
versely affect sprinklers. Due to the deck will reduce the potential for ignition of thin combustible materials. A 30-ft. (9.1
height (9-ft. [2.7 m]) relative to ceiling the fabric, as well as allow sprinklers un- m) diameter canopy of these leaves cre-
membrane height (8-ft. [2.4 m]), even der the awning to activate properly. ates a challenge for the automatic sprin-
this test may provide misleading results The Las Vegas Strip is also home to klers. As such, the sprinkler design den-
when applied to differing geometries. one fairly common tropical themed res- sity is frequently increased to
The thin combustible ceiling applica- taurant. The design concept includes compensate.
tions that have been allowed on the walls and ceilings covered with artificial One such feature is a model of a well-
Strip use an engineered approach. Since plants. The facility in which this restau- known starship (see Figure 3). This
it is impossible to determine which rant is located contains a high bay model is constructed out of fire-retar-
sprinklers will activate first, or how space, and to create a more intimate dant fiberglass-reinforced polymers
many, the sprinkler system is designed feeling, the designers wanted drop ceil- (FRP). It is approximately 30-ft. (9.1 m)
to flow all sprinklers in the plane of the ings made out of these artificial plants. in diameter and creates a 5-ft. (1.5 m)
ceiling, as well as above these ceilings To help ensure proper sprinkler activa- deep combustible void. The architec-
simultaneously. Occupant evacuation is tion, inverted noncombustible “boxes,” tural arrangement placed this model in a
an integral part of the analysis. Due to using the panelized construction re- rotunda. As such, sidewall sprinklers
these constraints, these applications are quirements from NFPA 13 for guidance, were installed around the perimeter to
limited in size, frequency, and their loca- were suspended from the deck above protect the model, as well as the space
tion with respect to exits. and then covered with the artificial below. Installing automatic sprinklers in-
One other ceiling-type application on plants. Sprinklers were installed in the side mitigated the combustible void. The
the Strip includes a hand-painted mural boxes to protect the space below. FRP also met the required ASTM E-84

9 Fire Protection Engineering N UMBER 24

flame-spread and smoke-developed rat-
ings, as well as other applicable require-
ments to qualify it as a ceiling.

Decorative Structures within

Buildings
One example of buildings within
buildings may be a small gazebo in a
large room that acts as a bar where
there may be one or two employees
serving cocktails. Glasses may be hung
from wooden slats above the bar. A
wooden lattice may create an architec-
tural vision ceiling. Generally, this
would be considered part of the fuel
load within the room.
As the gazebo becomes larger or the
room becomes smaller, the gazebo be-
comes the room. At this point, the
gazebo must be constructed as required Figure 3. Starship model
for the base building.
However, it is difficult to establish the standards. This can include: piable balconies
cutoff between these two cases. Here • Interior wall/ceiling finishes (as dis- • Nonbearing partitions
again, if decorative structures are broken cussed previously in this article) • Columns and bearing walls
into components, concepts can be used • Interior façade eave overhangs, dec- • Mezzanines and occupiable
that are already addressed in codes and orative ceilings/roofs, and nonoccu- floors/balconies

FALL 2004 www.sfpe.org 10

 ■ Unique Interiors on the Vas Vegas strip

Decorative Ceilings/Roofs There are numerous decora-

Code requirements for protection of tive structures inside the major
roofs may also be more restrictive than facilities on the Las Vegas Strip,
is necessary for interior structures. A some of which are even high-
roof is considered the upper-most por- rise buildings in and of them-
tion of a structure. It is intended to act as selves. Many define specific
a weatherproof membrane and also pro- uses, such as retail sales or
vide some structural stability. The roofs restaurants. They invariably pro-
of interior decorative structures are in- vide some of the theme inside
tended as architectural features and do the base building.
not fulfill the same role as an exterior One such decorative structure
roof. Roofs of interior decorative struc- even looks like a horse. This is
tures can be nothing more than ceilings. the three-story Trojan Horse that
As such, the Decorative Ceilings portion greets patrons at the FAO
of this article can be used for guidance. Schwarz Toy Store inside the
If these “roofs” provide some structural Caesar’s Forum Mall (Figure 4).
stability or are expected to support suffi- The legs are actually fire-rated
cient loads, an increased level of protec- columns. Since it is possible to
tion exceeding that allowed for ceilings walk inside the belly and look
is prudent. down on the main entrance, the
floor is a fire-rated slab. The
Nonbearing Partitions walls, which look like the sides
Bearing walls constructed of metal or of the horse, meet the require-
wood studs support more than 100 ments for interior nonbearing
pounds per lineal foot (0.0445 KN per partitions. Materials that clad the
meter) of superimposed load. If con- “walls” to give the appearance Figure 4. FAO Schwarz Trojan Horse
structed of masonry or concrete, bearing of a wooden horse were con- tive structures in accordance with these
walls support more than 200 pounds per structed as required for interior wall fin- relaxed guidelines. ▲
lineal foot (0.089 KN per meter) of su- ish. Since the head bobs up and down, it
perimposed load. Walls supporting less needed to be light. As such, it is con- Doug Evans is with the Clark County,
than these amounts are considered non- structed of fire-retardant fiberglass-rein- Nevada, Building Department.
bearing. Codes and standards typically forced polymers and contains one sprin-
require bearing walls to have a higher kler inside to provide fire protection. The REFERENCES
level of fire resistance than nonbearing tail is braided rope and was determined
walls. to not be any more of a hazard than a 1 FM Global, Property Loss Prevention Data
Certainly, any supporting element drapery or similar fuel package. Sheet 1-4, Revised May 2000, “Fire Tests,”
should be protected to at least the level In another facility, a foam plastic pi- Factory Mutual Insurance Company,
of that which it supports (such as walls rate boat was installed. This boat sits Norwood, MA.
supporting a floor or roof). above a bar and is approximately 30-ft. 2 Laymon, R.K., “Fact-Finding Report on
(9.1 m) long, 8-ft. (2.4 m) high, and 7.5- Combustibility of Products Containing
Columns and Bearing Walls ft (2.3 m) wide. Wooden masts extend Foamed Plastics used for Decorative and
The structural frame includes beyond these dimensions. This is cer- Display Purposes,” Underwriters
columns, girders, beams, trusses, and tainly a greater quantity of foam plastics Laboratories, Inc., Northbrook, IL, File
spandrels having direct connections to than intended by the UL 1975 fire test NC554 – Project 88NR10518, June 9, 1989.
the columns and all other members that described earlier in this article. The inte- 3 Janssens, M., “Basics of Passive Fire
are essential to the stability of the build- rior constituted a combustible void. As Protection,” Fire Protection Handbook,
ing as a whole. part of the mitigating aspects, both the National Fire Protection Association,
Smaller decorative structures within interior and exterior of the feature were Quincy, MA, 2003.
buildings may not be essential to the sta- fully encapsulated with noncombustible 4 Cote, R., Life Safety Code Handbook,
bility of the base building. The structural coatings. National Fire Protection Association, Page
frame may not be subjected to the same 329, Quincy, MA, 2003.
stresses, such as wind loads, as is the pri- Floors/Mezzanines 5 Clark County, Nevada, “Unique Building
mary framework of the main structure. Mezzanines are intermediate floor lev- Interiors Design Guide,” available at
Therefore, the level of protection may els that do not exceed one-third of the www.co.clark.nv.us/development_services.
not need to be as restrictive as required room in which they are a part. Some
6 “Acceptance Criteria for Vinyl Stretch
for the base building. As the decorative codes and standards allow the floors of Ceiling Systems,” ICBO ES AC 171, ICBO
structure increases in size, the level of mezzanines to be less fire-resistive than Evaluation Service, January, 2001.
passive protection will need to be the other floors. As such, it may be reason- Available at www.icc-es.org/criteria/pdf
same as required for the base building. able to consider multiple-level decora- /ac171.pdf.

11 Fire Protection Engineering N UMBER 24

 By Marcelo M. Hirschler, Ph.D. smoke to other parts of the compartment, normally up to 0.15 m thick), either in
or even to other compartments. There- one unbroken length or in separate sec-
INTRODUCTION fore, the fire performance of such materi- tions joined end to end, is mounted face
als needs to be controlled. downwards so as to form the roof of a

“I nterior Finish” is defined in U.S.

codes in a similar manner. The
National Fire Protection
Association defines “interior finish” as
“the exposed surfaces of walls, ceilings,
STEINER TUNNEL TEST

The Steiner tunnel fire test method for

horizontal tunnel 305 mm high. The fire
source, two gas burners, ignites the sam-
ple from below with an 89 kW intensity
(Figure 2), and the combustion products
surface flame spread and smoke develop- are carried away by a controlled linear air
and floors within buildings,” 1 with the ment remains the traditional test used to velocity of 73 m/min (or, exactly, 240
explanation that “interior finish is not assess fire performance of interior finish ft/min). The normal output is a flame-
intended to apply to surfaces within materials. Developed by Al Steiner for spread index (FSI) and a smoke-devel-
spaces, such as those that are con- testing building materials, such as wood oped index (SDI). Flame spread is as-
cealed or inaccessible. Furnishings that, or gypsum board, at Underwriters Labo- sessed visually by the progression of the
in some cases, might be secured in ratories in 1944 (Figure 1), the Steiner flame front, while measurements of opti-
place for functional reasons should not tunnel test has been standardized by the cal smoke density at the tunnel outlet de-
be considered as interior finish.” NFPA major North American standards writing termine the smoke obscuration. This in-
also considers “interior ceiling finish” as organizations (ASTM E-84, NFPA 255, UL formation is used to plot time-based
“the interior finish of ceilings,” and 723, ULC S102) and widely adopted by graphs of flame-spread distance and of
“interior wall finish” as “the interior fin- every North American building and fire optical density. FSI and SDI are then cal-
ish of columns, fixed or movable walls, code. culated based on the ratio between the
and fixed or movable partitions” and In the test, a specimen (7.3 m x 0.56 m, areas under the curves for the material
“interior floor finish” as “the interior being tested and those for a cementitious
finish of floors, ramps, stair treads and board (assigned FSI and SDI values of 0)
risers, and other walking surfaces.” and for red oak flooring (assigned FSI
The International Code Council states and SDI values of 100).
that “interior finish includes interior wall The building, fire, and life safety codes
and ceiling finish and interior floor fin- (IBC, IFC, NFPA 5000, NFPA 101, and
ish,” that “interior wall and ceiling finish” NFPA 1/UFC) all contain requirements
is “the exposed interior surfaces of build- that limit interior wall and ceiling finish
ings including, but not limited to: fixed to Class A (FSI ≤ 25; SDI ≤ 450), Class B
or movable walls and partitions; (25 ≤ FSI ≤ 75; SDI ≤ 450), or Class C (75
columns; ceilings; and interior wainscot- ≤ FSI ≤ 200; SDI ≤ 450). A major flaw in
ting, paneling, or other finish applied the Steiner test appears in the description
structurally or for decoration, acoustical of the test method above and the results
correction, surface insulation, structural obtained from this fire test: the Steiner
fire resistance, or similar purposes, but Figure 1. Photograph of Steiner Tunnel tunnel test does not provide results in en-
not including trim,” and that “interior gineering units. Consequently, the test re-
floor finish” is “the exposed floor sur- sults cannot be used for a fire hazard
faces of buildings including coverings ap- analysis or a fire risk analysis.
plied over a finished floor or stair, includ- This test continued to be popular
ing risers.” 2 Thus, when dealing with when plastics started to be used in con-
testing of interior finish, a distinction struction and in spite of the fact that the
needs to be drawn between walls (and test is not always appropriate for every
ceilings) and floors. material. Samples that cannot be retained
The fire performance of interior wall in place above the tunnel floor or which
and ceiling finish is critical to the devel- melt and continue burning on the tunnel
opment of a fire: interior finish offers fuel floor (typical behavior for most thermo-
contribution and surfaces through which plastics) are still being tested with this
a fire can spread and transport heat and Figure 2. Flame in Steiner Tunnel Test equipment even though the results are

12 Fire Protection Engineering N UMBER 24

 ■ Fire Testing of Interior Finish

not representative of the use of the material in realistic situa- Therefore, all carpets and rugs sold in they United States5 must
tions.3 The same can also be said about thin materials, which of- meet the “methenamine pill” test (ASTM D 2859), which ensures
ten give low FSI values mainly due to insufficient material in the that flame spread will be minimal.
test method to permit flame spread to be assessed properly. An Most codes also regulate interior floor finish (in occupancies
understanding of some of these limitations has caused the codes where fire risk needs to be especially minimized) to be tested
to consider alternatives, either as replacements for the Steiner with the flooring radiant panel (ASTM E 648, NFPA 253, Figure 3)
tunnel or as additional options (see section on heat release). and require a “critical radiant flux” for ignition in excess of 4.5
kW/m2 (Class I) or 2.2 kW/m2 (Class II). In the flooring radiant
FLOOR FINISH TEST METHODS panel, the floor finish (such as a carpet) is exposed to an incident
heat flux from an angled gas-fired radiant panel, with a maxi-
Different challenges face interior floor finish than other inte- mum heat flux of approximately 11 kW/m2 at the farthest end
rior finish because heat and smoke rise in a fire. Thus, floor fin- from the igniter. The test method assesses the critical incident
ish is involved either as the initial material ignited in a fire or as flux (which is measured by comparing the distance between the
an additional fuel once a fire has become uncontrolled. Conse- igniter and the point where flame propagation stops to a calibra-
quently, fire safety requirements typically need to ensure that in- tion curve) required for continued
terior floor finish is relatively difficult to ignite and is not capable flame propagation.
of slowly spreading flame from the compartment of fire origin to This approach (even if it is based
a different one. on old-fashioned tests) is quite suit-
The Steiner tunnel cannot assess ignitability, and its fuel able for interior floor finish. Some ap-
source is not appropriate to assess slow flame spread. Experi- plications, typically in the transporta-
ence has shown that many flooring materials (traditional floor tion vehicle arena, also require
finishes such as wood flooring or resilient materials) will not ig- flooring materials to meet one of a va-
nite unless exposed to an ignition source of well over > 1 riety of smoke obscuration require-
kW/m2, but that some carpet-like or loose-fill materials may ig- ments, often based on a static smoke
nite at such low heat fluxes. A study of precision of the flooring chamber box, either with a traditional
radiant panel test method found carpets with critical radiant heat radiant heater (ASTM E 662) or with a
fluxes well under 2 kW/m2.4 conical heater (ISO 5659-2, IMO Fire
Test Procedures Code part 2, also Figure 3: Flooring
known as ASTM E 1995 and NFPA Radiant Panel Test
270). Apparatus (ASTM E 648)

OTHER TEST METHODS

Of course, the key question to ask in any fire is “how big is the
fire?”, and the answer lies in the rate of heat release.6, 7, 8, 9, 10, 11 A
burning product will spread a fire to nearby products only if it
gives off enough heat to ignite them. Moreover, the heat has to
be released fast enough not to be dissipated or lost while travel-
ing through the cold air surrounding any product that is not on
fire. Therefore, heat-release rate dominates fire hazard, and it has
been shown to be much more important than ease of ignition,
smoke toxicity, or flame spread in controlling the time available
for potential victims of a fire to escape.
The above concepts are now applied to fire testing of interior
(wall and ceiling) finish, and all U.S. codes use a room-corner test
for the purpose. The use of the room-corner test can be an alterna-
tive to the Steiner tunnel test (for most interior finish materials) or

Figure 4: Room-
Corner Fire Test

13 Fire Protection Engineering N UMBER 24

the actual requirement (for foam plastic in- dard doorway, are lined with the mater- fluxes are measured in the room.
sulation and textile wall coverings). Thus, ial to be tested. The ignition source is a The severity of the ignition source
the building, fire, and life safety codes al- gas burner placed in one corner (on the was designed to ensure that the gas
low most interior wall and ceiling finish wall furthest from the doorway) flush burner flame alone reaches the ceiling,
materials to be tested using the NFPA 286 against both walls that generates 40 kW without contribution from the test mate-
room-corner test (Figure 4) instead of the for a 5-minute period, followed by 160 rial (Figure 5). Even though the test
Steiner tunnel test, and the test results kW, for a further 10-minute period. measures heat release, the codes simply
must then show that the test specimen Heat release (based on the principle require assessment of whether flashover
does not cause flashover during the as- of oxygen consumption calorimetry) and occurs during the test, so much of the
sessment and does not emit a total amount smoke release are measured in the ex- information collected during the test is
of smoke exceeding 1,000 m2. The total haust duct, and temperatures and heat not used.
amount of smoke released is a measure of
smoke obscuration (or smoke opacity) cal-
culated as the integral, over time, of the
smoke-release rate across the surface of
the duct used for the measurement. Thus,
the smoke release rate is measured in
units of surface area over time and the to-
tal smoke released in units of surface area
(such as m2).
Special rules apply to some materials
or products, as follows:
• Textile wall and ceiling covering ma-
terials are required to meet a Class A
flame-spread index and smoke-devel-
oped index (using the Steiner tunnel fire
test), and be used in a sprinklered envi-
ronment or have passed a specific room-
corner test for textile wall coverings
(NFPA 265, a less severe test and one
where there are no smoke obscuration
requirements), which requires that
flashover not occur.
• Expanded vinyl wall coverings can
be treated like textile wall coverings (see
above) or like most other interior finish
(use the Steiner tunnel or NFPA 286).
• Cellular or foamed plastic materials
must always meet a Class B flame-spread
index (using the Steiner tunnel test).
They can be used as interior trim if the
density of the material is high enough (>
320 kg/m3), and when used in that way,
the amount of cellular foam is limited to
10% of the wall or ceiling. Alternatively,
cellular or foamed plastic materials must
meet the standard smoke obscuration re-
quirement (smoke-developed index of
<450, using the Steiner tunnel test) and
must either be covered by a thermal bar-
rier or meet a large-scale fire test that
fully represents the fire hazard in the sce-
nario in question. One of such tests is the
NFPA 286 room-corner test.

ROOM-CORNER TEST

In the NFPA 286 room-corner test,

three walls and the ceiling (or ceiling
only, for interior ceiling finish) of a 2.4 m
x 3.7 m x 2.4 m high room, with a stan-

FALL 2004 www.sfpe.org 14

 ■ Fire Testing of Interior Finish

Research was conducted very high heat-release rates (but not quite enough for
to look into issues associ- flashover) should probably be classified separately from materi-
ated with room-corner test- als that have low peak rates of heat release rate.
ing.12, 13, 14 As a result, two ad- NFPA 265 is a somewhat less severe variation of NFPA 286,
ditional important criteria which is applied exclusively to textile wall coverings (and ex-
required by the codes are panded vinyl wall coverings). In NFPA 265 and NFPA 286, iden-
that the flame spread does tical test rooms and identical gas burners are used. There are
not reach any of the extrem- three main differences, however, in the ignition sources used
ities of the test sample and for both tests: 1) in NFPA 265, the burner is placed 51 mm away
that the total smoke release from each of the walls, as opposed to flush against the walls as
cannot exceed 1,000 m2 in NFPA 286 (but it is placed in the same corner as in NFPA
over the entire 15-minute 286); 2) in NFPA 265, after the first 5 minutes at 40 kW, the
test period. burner intensity is raised to 150 kW, as opposed to 160 kW in
If all criteria are met, the NFPA 286; and 3) smoke release measurements using NFPA 265
material is suitable for use in are not required in the codes. It is likely that, eventually, textile
all applications where the wall coverings will be required to be treated similarly to other
codes require a material to interior finish.
Figure 5. Flame in Room- be tested by the Steiner tun-
Corner Test nel and where Class A, B, or ANALYSIS OF TEST METHODS
C requirements exist. In
practice, it is rare for a material to spread flame to the extremi- Now that the actual tests used have been presented, it is im-
ties of the test sample and still not cause flashover, since that portant to discuss the validity of the test methods and whether
would mean that the flame would reach the edge of the door improvements should be put in place. One obvious improve-
and stop without exiting the doorway (one of the criteria for ment which would permit a much more logical approach to us-
flashover). This means that any material that does not cause ing fire safety engineering methods would be to apply a test
flashover and releases < 1,000 m2 of smoke is considered equiv- method based on heat release for testing interior floor finish
alent to a Class A material. In fact, it is likely that materials with (such as the cone calorimeter, a bench-scale test, e.g., for exam-
ple, ASTM E 1354, NFPA 271, ISO 5660).
In fact, two ASTM guides and one NFPA guide addressing
fire hazard assessment, ASTM E 2280 (for healthcare occupan-
cies), ASTM E 2067 (for rail cars), and NFPA 555 (on potential
for flashover), all recommend the use of the cone calorimeter
to assess heat and smoke release of interior floor finish, at in-
cident heat fluxes of 25-30 kW/m2. However, it is true that the
combination of the methenamine pill test and the flooring ra-
diant panel test is sufficient to eliminate the vast majority of
“bad actors.” Thus, the methods being used are fairly ade-
quate for a prescriptive fire safety approach that does not dis-
criminate against materials. The additional smoke release test-
ing (used mostly in transportation environments) is not of
very high value, but it may serve to eliminate some poor per-
forming materials.
There continues to be controversy with regard to smoke re-
lease testing of interior wall and ceiling finish. In a field that
continues to be dominated by the Steiner tunnel test (despite its
well-known inadequacies for testing some materials) the ques-

80

70 SwRI (10)
Eurefic (28)
60 SBI (30)
Percentage of Total

Coast Guard (9)

50
BFG (7)
40 Overall (84)

30

20

10

0
Early Flashover Adequate Heat, Adequate Heat,
Low Smoke High Smoke
Figure 6. Room Corner Testing, Heat & Smoke Release

15 Fire Protection Engineering N UMBER 24

 ■ Fire Testing of Interior Finish

1500 2000
Actually 5771
1800
1250 Pk RHR

Pk RHR in Room-Corner Test (kW)

1600
Flashover
Room-Corner Test TSR (m2)

1400
1000
1200

750 1000

800
500
600

400
250
200

0 0
0 200 400 600 800 1000 1200 0 25 50 75 100 125 150 175 200
Steiner Tunnel SDI FSI in Steiner Tunnel

Figure 7. Smoke Release of Interior Finish Figure 8. Rate of Heat Release vs Flame Spread

tion arises: Is it necessary to test for test. However, the codes allow the NFPA corner test as unacceptable.
smoke release in the room-corner test, or 286 room-corner test results based on the • 14 materials had an FSI of 25 or less
is it enough to just develop low heat re- premise that materials that do not cause (i.e., Class A) in the Steiner tunnel and
lease products? flashover (or high smoke release) in the released less than 400 kW in the room-
Figure 615 shows that of five series of room-corner test are known to also have corner test. Both tests classify them as
tests conducted in room-corner tests, sys- flame spread indices of < 200 and smoke Class A.
tematically some 10% of the materials (10 developed indices of less than 450 in the • 2 materials had an FSI of > 25 and <
of the 84) give low heat release but unac- Steiner tunnel test. 75 (i.e., Class B) in the Steiner tunnel and
ceptably high smoke release. So the im- These provisions work well to a point released less than 400 kW in the room-
portance of assessing smoke release of but need refinement. The Steiner tunnel corner test. The Steiner tunnel test classi-
interior wall and ceiling finish in a large- test is likely to give falsely favorable re- fies them as Class B and the room-corner
scale test is clear. Code writers then saw sults (in fact, this happens often with ma- test as Class A.
that both flame spread and smoke re- terials that melt and drip and with materi- • 2 materials had an FSI of 25 or less
lease (or heat release and smoke release) als that are thin films) but it rarely gives (i.e., Class A) in the Steiner tunnel and
must be assessed to adequately regulate falsely unfavorable results (meaning that a released more than 400 kW but did not
the fire performance of interior wall and high flame-spread index, or FSI, is almost cause flashover in the room corner test.
ceiling finish materials, whatever test always indicative of a material with Both tests classify them as Class A.
method is used. mediocre or poor fire performance). The • 2 materials had an FSI of 200 or less
Thus, a smoke-release criterion room-corner test results are potentially (i.e., Class C) in the Steiner tunnel and al-
needed to be added to fire testing using much more suitable to classification of most caused flashover in the room-cor-
the room-corner test (such as NFPA 286), materials, because the heat release rate ner test. The Steiner tunnel test classifies
for which heat release only used to be history is obtained in the test. However, them as Class C and the room-corner test
measured, while both flame spread and the fact that the heat release rate history is as Class A.
smoke release have always been re- not used for code classification purposes In conclusion, fire testing of interior
quired in the Steiner tunnel test. The data results in some inconsistencies occurring finish is probably adequate to eliminate
in Figure 712, 13, 14 shows that this problem when comparing results from both tests. the poorest performers (both in terms of
can be resolved (and has now made its Therefore, it would be important to use heat release, or flame spread, and smoke
way into codes) by using equivalent cri- the heat release rate history in the room- release). However, in terms of maximiz-
teria in both tests, since materials with corner test in conjunction with testing ing the usefulness of current research
very high smoke-developed index (SDI) whether flashover does or does not occur. and to accommodate modern building
are also likely to have a very high total Figure 8 (based on a survey of pub- materials, the Steiner tunnel test falls
smoke release (TSR) in the room-corner lished data developed for this work) short. In applying a variety of new and
test, which is how the 1,000 m2 pass-fail shows the comparative fire performance specific tests, the full capabilities of the
criterion was developed. of 25 materials tested in the Steiner tun- room-corner test, including the actual
In light of steady research, it becomes nel and in the room-corner, and illus- heat release rates measured, could be in-
clear that for all of its traditional merits, trates the problem: corporated into engineering, and im-
the approach of codes to testing interior • 5 materials had an FSI of 200 or less provements in that area would be wel-
wall and ceiling finish is a slightly flawed (i.e., Class A, B, or C) in the Steiner tun- come. ▲
concept. Clearly the room-corner test is a nel but caused flashover in the room
much more accurate way of assessing corner test. The Steiner tunnel test clas- Marcelo Hirschler is with GBH Interna-
fire performance than the Steiner tunnel sifies them as acceptable and the room- tional.

16 Fire Protection Engineering N UMBER 24

 ■ Fire Testing of Interior Finish

REFERENCES Consumption Measurements,” Fire and Proceedings – Fire and Materials Conf.,
Materials, 1980, 4, pp. 61-65. Feb. 22-23, 1999, Interscience
1 NFPA 5000, “Building Construction and Communications, London, UK, pp. 83-94.
7 Babrauskas, V., and Grayson, S.J., Eds.,
Safety Code,” National Fire Protection Heat Release in Fires, Elsevier, London, 13 Hirschler, M.M., and Janssens, M.L.,
Association, Quincy, MA, 2003. UK, 1992. “Smoke Obscuration Measurements in the
2 International Building Code, NFPA 265 Room-Corner Test,”
8 Hirschler, M.M., “Heat Release from
International Code Council, Falls Church, Proceedings – Fire and Materials Conf.,
Plastic Materials,” Heat Release in Fires,
VA, 2003. Feb. 22-23, 1999, Interscience
Elsevier, London, UK, Eds. Babrauskas, V.
Communications, London, UK, pp. 179-
3 Belles, D.W., Fisher, F.L., and Williamson, and Grayson, S.J., Eds., pp. 375-422, 1992.
198.
R.B., “How well does the ASTM E-84 pre- 9 Babrauskas, V., and Peacock, R.D., “Heat
dict fire performance of textile wallcover- 14 Janssens, M.L., Dillon, S.E., and Hirschler,
Release Rate: The Single Most Important
ings?” Fire Journal, 82(1), pp. 24-30, 74 M.M., “Using the Cone Calorimeter as a
Variable in Fire Hazard,” Fire Safety
(1988). Screening Tool for the NFPA 265 and
Journal, 1992, 18, pp. 255-72.
NFPA 286 Room Test Procedures,”
4 Lawson, J.R., “Fire tests and flooring 10 Hirschler, M.M., “Use of Heat Release Rate Proceedings – Fire and Materials Conf.,
materials,” Proc. 2nd Fire and Materials to Predict Whether Individual Furnishings Jan. 22-24, 2001, Interscience
Conf., Sept. 22-23, 1992, Interscience Would Cause Self-Propagating Fires,” Fire Communications, London, UK, pp. 529-
Communications, London, UK, pp. 253- Safety Journal, 32, 273-296 (1999). 540.
262.
11 Hirschler, M.M., “Flammability and Fire 15 Hirschler, M.M., “Fire Performance of
5 16 CFR 1630, Code of Federal Performance of Polymers,” Organic Polymers, Thermal
Regulations, Chapter II, Consumer Comprehensive Desk Reference of Polymer Decomposition, and Chemical
Product Safety Commission, Part 1630, Characterization and Analysis, Robert Composition,” American Chemical Society
Standard for the Surface Flammability of Brady, Ed. Amer. Chem. Soc., Washington, Preprints, August 2000 National Meeting,
Carpets and Rugs (FF 1-70). DC, 2003, pp. 700-738. Symposium on Fire and Polymers, Symp.
6 Huggett, C., “Estimation of Rate of Heat 12 Finley, G., Janssens, M.L., and Hirschler, Chair: G.L. Nelson and C. Wilkie,
Release by Means of Oxygen M.M., “Room Fire Testing – Recent Washington, DC.
Experiences and Implications,”

17 Fire Protection Engineering N UMBER 24

 The Role of Interior Finish in
Fire Development
By Robert Brady Williamson, Ph.D., ishes have been regulated for more than LaSalle Hotel fire3 in Chicago, and the
P.E., and Frederick W. Mowrer, 50 years. In this paper, the roles of dif- Hotel Winecoff fire4 in Atlanta, the role of
Ph.D., P.E. ferent interior finishes in fire develop- interior finish in fire development be-
ment are addressed. An historical per- came more widely recognized in the fire
INTRODUCTION spective of the significant fires that have protection engineering and building reg-
shaped the regulation of interior finishes ulation communities than it had been

C ombustible interior finishes,

which include the exposed ceil-
ing, wall, and floor linings in
buildings, are large continuous surfaces
over which fires can spread.1 These fin-
in the United States is presented. The
scientific understanding of fire spread
over interior finishes has developed sig-
nificantly over the past 25 years, along
previously. That these fires occurred in
buildings of so-called “fireproof” con-
struction highlighted the contribution of
the interior finishes and decorations to
these fires.
with the quantitative methods needed to
ishes, along with combustible furnish- evaluate the fundamental flammability Following the Cocoanut Grove fire in
ings and contents, provide the fuels properties of materials. Theoretical con- 1942, but before the LaSalle Hotel and
that can permit the development of cepts associated with flame spread the- Hotel Winecoff fires in 1946, A. J. Steiner
enclosure fires, in many cases, to ory are presented in the following sec- of Underwriters’ Laboratories published a
flashover conditions. Depending on the tion. These concepts demonstrate that, test method to classify the hazards of
flammability characteristics of the interi- to a large extent, flame spread on inte- building materials.5 As noted by Steiner6
or finishes and the fire scenarios in rior finishes can be viewed as a race be- apparently in reference to the Cocoanut
which they are involved, interior finish- tween the ignition and burnout of fuel Grove fire, “Public concern is aroused
es may serve as the primary fuel dri- surface elements. Finally, a new way of periodically when a rapidly spreading
ving a fire to flashover or as a second- evaluating, and perhaps eventually regu- fire kills a large number of people or
ary fuel acting as a “fuse” to spread a lating, the flammability characteristics of produces an extraordinary property loss.
fire between primary fuel packages. combustible interior finishes is pre- This concern prompted the development
Once flashover occurs and all exposed sented. This methodology provides a of a test method whereby the fire haz-
combustible surfaces within an enclo- way to move away from the current em- ards of materials could be measured and
sure ignite, interior finishes may repre- pirical basis for the regulation of interior classified with reference to the rate of
sent the most significant fuel package finish flammability to a more quantita- spread of fire, the amount of fuel con-
contributing to the post-flashover fire tive scientific basis. tributed to the fire, and the production of
because of their large surface areas and objectionable smoke while burning.”
total energy content. HISTORICAL PERSPECTIVE This test method is now widely known
Because of their potential to serve as As a result of a number of major and as the “tunnel test” because of the duct-
the primary fuel driving an enclosure widely publicized building fires in the like configuration of the fire test chamber
fire to flashover, the flammability char- United States during the 1940s, including or as the “Steiner tunnel test” in honor of
acteristics of interior wall and ceiling fin- the Cocoanut Grove fire2 in Boston, the its principal developer.

18 Fire Protection Engineering N UMBER 24

 In 1950, ASTME84-50T, Tentative and ceiling material. 3) It was the largest up and repair our furnace. Need for ac-
Method of Fire Hazard Classification of test exposure that could be used without tion by a fire protection group is essential
Building Materials,7 was first approved producing a life hazard in the test enclo- to control the fire hazard being created.”
by the American Society for Testing and sure by the burning of the enclosure it- Steiner went on to say that “the value of
Materials as a tentative standard. This test self. 4) This exposure ... produced a tem- results of a test are dependent on their
method was adopted by all the model perature of 155°F (68°C) at the breathing significance as related to their use, based
building codes in the United States and level, which was sufficiently below the on actual field fire experience.”
by the NFPA Building Exits Code (now chosen life hazard temperature of 300°F Steiner was a proponent of small-scale
the Life Safety Code), resulting in wide- (150°C) to determine to what extent the tests as “effective instruments for develop-
spread regulation of the “flame spread” wall or ceiling material would contribute ment and research, as well as tools for in-
and “smoke development” of interior a life hazard.” spection,” but he also recognized their
wall and ceiling finishes based on tunnel This FM report is significant for a num- limitations: “The small-scale tests can be
test results. The tunnel test remains the ber of reasons. It represents one of the used in the examination of products to de-
primary fire test method used to regulate first systematic efforts to evaluate the termine whether they provide the same
the flammability of interior wall and ceil- flammability of interior wall and ceiling properties as other materials tested in the
ing finishes in the United States more finishes in an end-use configuration. It same manner ..., but they do not provide
than 50 years later, despite recognition of recognizes that fires located in corners fire protection information on the behav-
its technical shortcomings and the devel- represent a realistic worst-case exposure ior of the product, or of assemblies em-
opment of more realistic fire test methods geometry for wall and ceiling linings. It ploying it, under actual use conditions in
for interior wall and ceiling finishes. establishes a selection process for ignition buildings.” He goes on to say that “the
In 1950, Factory Mutual Laboratories sources that challenge the materials being same fire protection engineering consider-
(FM) published a report8 describing a evaluated but do not overwhelm their ations must be given to all tests, whether
room fire test method to evaluate the life performance. Unfortunately, the room fire small or large. The results must be repre-
hazard of interior finishes. In the after- test method developed by FM to evaluate sentative of actual conditions, the classifi-
math of the large life-loss fires of the the life hazard of interior finishes never cations must be realistic and the require-
1940s identified above, this report noted gained the widespread acceptance within ments consistent.” It is interesting to note
that, “There is considerable agitation at the building regulatory community that that Steiner11 viewed the tunnel test as a
the present time to write regulations gov- the tunnel test did. large-scale test, while others12 have viewed
erning the use of interior wall and ceiling Through the 1950s, the tunnel test the tunnel test as a small-scale test.
finish materials, in the interest of reduc- method became more firmly entrenched In 1961, Wilson13 reviewed a number
ing the life hazard in public areas where as the standard for regulating the flamma- of test methods then being used to evalu-
these materials are used in quantity. Be- bility characteristics of interior finish ma- ate the surface flammability of materials.
fore adequate and equitable regulations terials despite the fact that it only had ten- Wilson noted that “None of the agencies
can be established, fire conditions consti- tative status under ASTM. During this developing these test methods has re-
tuting a life hazard will, of necessity, period, the use of plastics in building ported any relation between their test re-
need to be defined and materials tested construction also started to grow tremen- sults and actual fire conditions. ... There
under such conditions of exposure.” FM dously. Both Steiner9 at UL and Wilson10 at has been nothing reported to indicate
developed a test room approximately 4.2 FM voiced concern with the small-scale that four of the test methods (including
m (14-ft.) by 6.1 m (20-ft.) by 3.7 m laboratory procedures, such as ASTM the tunnel test) have ever been directly
(12-ft.) high. FM experimented with a D635 and D1692, and the terminology, compared with any form of actual fire
number of ignition sources consisting of such as “self-extinguishing,” “slow-burn- condition.” Both Steiner and Wilson
wood cribs weighing from 2.3 kg to 13.6 ing,” and “nonburning,” being used to seemed to agree that the results of fire
kg (5 lb to 30 lb) and/or ethyl alcohol evaluate and describe the flammability tests should be representative of actual
weighing from 0.2 kg to 3.4 kg (1 lb to performance of plastic building products. conditions to be valid.
7.5 lb) placed in a corner of the room. Wilson noted that these small-scale lab- Through the 1960s, some of the techni-
The conclusion of the FM report was oratory tests are “intended solely for com- cal shortcomings associated with the tun-
that the ignition source consisting of 7.5 paring the relative flammability of various nel test began to be recognized more
lb of wood and 0.75 lb of alcohol was plastic materials,” and that they “are nei- widely when the tunnel test was used to
considered to be “the most suitable expo- ther designed nor appropriate for the rat- evaluate the flammability characteristics
sure in this enclosure for establishing the ing of plastic products as building materi- of newly developed foam plastic insula-
extent to which interior wall and ceiling als.” Steiner noted that “the tests which tion products that were starting to be
finish materials contributed to produce a classify plastics as self-extinguishing and used in buildings. Some of these prod-
life hazard. Several factors influenced the slow-burning do not correlate with the ucts received low flame-spread ratings in
selection of this exposure: 1) It was of Fire Hazard Classification. To illustrate, the tunnel test, yet rapidly spread fires
sufficient intensity to ignite materials some time ago a plastic which had been when installed in buildings. This anom-
causing them to burn and contribute to classified as slow-burning was subjected alous propensity for rapid flame-spread
the rise of temperature within the enclo- to the tunnel test, and the results were and fire development on exposed foam
sure. 2) Its location in one corner of the disastrous. The material burned so plastics despite low flame-spread ratings
room adjacent to two walls produced a fiercely and created so much smoke and was demonstrated by newly developed
maximum exposure condition to wall molten residue that it took days to clean open-corner fire tests14 that more realisti-

FALL 2004 www.sfpe.org 19

 ■ The Role of Interior Finish in Fire Development

open-corner fire test of a polyurethane Decree signed by 24 companies and the

foam insulation product with a low re- SPI17. As part of the Consent Decree, the
ported flame-spread rating. respondents agreed to perform many
As a consequence of the little-known activities, which ranged from notifying
Childress residence fire15 in which two all prior purchasers of foam insulation
children died as a result of a fire involv- products of the dangers of the products
ing exposed polyurethane foam insula- to sponsoring and conducting research
tion installed in their home, the Federal into the proper ways to protect foam
Trade Commission (FTC) filed a pro- plastic insulation products. These activi-
posed complaint16 against 27 respon- ties are summarized in the 1980 Final
dents, including 25 manufacturers of Report of the Products Research Com-
foam plastic products and 2 trade orga- mittee,18 which was formed to adminis-
nizations, the Society of the Plastics In- ter a $5 million trust fund established as
dustry (SPI) and the American Society part of the Consent Decree.
for Testing and Materials (ASTM), claim- Between the time when the FTC Con-
ing that the respondents were know- sent Decree was signed in 1974 and the
ingly marketing foam plastic insulation PRC Final Report was issued in 1980, the
products with misleading representa- use of thermal barriers to separate foam
tions that such products were “nonburn- plastic insulation products from occu-
Figure 1. Open-corner test of a foam ing” and “self-extinguishing” on the ba- pied spaces in buildings became the
plastic insulation product with a reported sis of inadequate test methods, including standard practice. For example, the 1973
flame-spread classification of 25. the tunnel test. edition of the Uniform Building Code
There was a great deal of activity dur- (UBC) did not make any reference to
cally simulated the dynamics of enclo- ing the year after the FTC proposed foam plastics while the 1976 edition of
sure fires than the tunnel test did. An ex- complaint was issued, which culminated the UBC included a new section (Section
ample of this anomalous behavior is il- in the “Complaint and Decision” of No- 1717) devoted exclusively to foam plas-
lustrated in Figure 1, which shows an vember 4, 1974, that included a Consent tics. This new section generally required

20 Fire Protection Engineering N UMBER 24

foam plastics to be separated from the tion, specimen description, ignition foam plastic assemblies in a standard
interior of a building by a thermal bar- source, instrumentation, and safety con- room configuration” and thus to serve as
rier, such as 1/2 in. (13 mm) thick gyp- siderations which must be decided upon an approved diversified test for foam
sum wallboard, having a finish rating of in the design of a room fire experiment.” plastics under the UBC. This standard
not less than 15 minutes unless specifi- In 1979, Williamson and Fisher20 de- specified a room 2.4 m (8-ft.) wide by
cally approved on the basis of “ap- scribed efforts then underway at the 3.7 m (12-ft.) long by 2.4 m (8-ft.) high
proved diversified tests,” including “fire University of California, Berkeley, to de- with a doorway 0.8 m (2-ft. 6-in.) wide
tests related to actual end-use such as a velop a standard room fire test method. by 2.1 m (7-ft.) high centered in one of
corner test.” The details of a diversified They subsequently reported21 on their the 2.4 m (8-ft.) long walls of the enclo-
test to be used for evaluating foam plas- efforts to evaluate this room fire test sure. The ignition source specified for
tics were not specified until 1982. method. They used an enclosure with this test method was a 13.6 kg (30 lb)
Room fire test methods were used in- dimensions of 2.4 m (8-ft.) wide by 3.7 wood crib located 25 mm (1 in.) from a
creasingly during the mid- to late-1970s m (12-ft.) long by 2.4 m (8-ft.) high, corner opposite the doorway opening.
as an alternative to the open-corner fire which was becoming the most typical During the 1980s, another series of
tests that had been used during the enclosure size for room fire tests. This hotel fires occurred that was reminiscent
1960s and early 1970s. In 1975, Under- work and related work at other fire re- of those in the 1940s, except that these
writers Laboratories reported19 on a se- search laboratories resulted in a pro- hotel fires involved modern high-rise
ries of flammability studies of interior posed ASTM standard room fire test buildings with interior finish materials
finishes that included room fire tests. In method for wall and ceiling materials that should have met modern regulatory
1977, ASTM first published ASTM E603, and assemblies22 in 1982, but this pro- requirements. The first of these hotel
Standard Guide for Room Fire Experi- posed standard was never adopted by fires was the November 1980 fire at the
ments. This document noted that, ASTM. MGM Grand Hotel23 located along the
“There is no standard room fire test at In 1982, Uniform Building Code Stan- Las Vegas Strip in Clark County, Nevada.
the present time, and this report does dard No. 17-5, Room Fire Test Standard The early development of the MGM
not define one. It does set down many for Interior of Foam Plastic Systems, was Grand fire was on the interior wall and
of the considerations for such a test: for first published to “detail a test method to ceiling finishes of a service side station
example, room size and shape, ventila- evaluate the burning characteristics of in the deli restaurant on the casino

FALL 2004 www.sfpe.org 21

 ■ The Role of Interior Finish in Fire Development

level.24 Once the fire flashed over the exterior windows located in each eleva- oped to flashover conditions on the new
side station, it quickly enveloped the tor lobby. The fire did not reach the 29th furniture being stored in the ballroom as
deli restaurant, feeding on the com- floor because of an architectural detail well as on the textile wall material and
bustible interior finishes and furnishings that deflected the flame out and away foam-insulated movable partitions lining
in the restaurant. The deli restaurant from the lobby windows. the walls of the ballroom. The com-
then flashed over, and the fire spread The Las Vegas Hilton Hotel fire and bustible ceiling in the foyer also con-
into and along the length of the casino, other less-publicized fires involving tex- tributed to the fire development.
which was roughly the size of a football tile materials motivated the textile indus- With the exception of the Las Vegas
field. The fire was confined to the casino try to sponsor research at the University Hilton Hotel fire leading to the develop-
level, but 85 people died as a result of of California, Berkeley, to evaluate how ment of the room fire test method for
this fire, with approximately 68 of the well the tunnel test predicts the perfor- textile wall coverings, the hotel fires of
victims located on the upper floors of mance of textile wall coverings.26 As a the 1980s did not inspire significant
the high-rise portion of the building result of this research project, a room changes to interior finish requirements
above the casino. fire test method for textile wall coverings in the building regulations. Instead,
Three months after the MGM Grand was developed. This room fire test these fires motivated the widespread use
Hotel fire, the Las Vegas Hilton Hotel25 method was adopted as UBC Standard of automatic sprinkler protection in
suffered a devastating fire that killed 8 42-2 in 1988 and is also currently desig- high-rise hotels and other residential
people. This fire started in the 8th floor nated as NFPA 265, which is referenced and commercial buildings where sprin-
elevator lobby in the east wing of the by the Life Safety Code and the Interna- kler protection had not traditionally
30-story building. The walls and ceiling tional Building Code. been installed.
of this elevator lobby, as well as all the The fire at the DuPont Plaza Hotel27 in The fire at the Station nightclub in
other elevator lobbies on floors served San Juan, Puerto Rico, occurred on De- West Warwick, Rhode Island, in Febru-
by these elevators, were lined with a cember 31, 1986. This fire, which ary 2003 provides the latest extreme ex-
textile carpet material. The fire in the 8th claimed the lives of 99 people located in ample of the role of interior finish in fire
floor elevator lobby developed to the hotel’s casino, started in a ballroom development. This fire, which claimed
flashover, then spread from the 8th floor located across a covered foyer from the the lives of 100 victims and injured hun-
to the 28th floor of the building via the casino. The fire in the ballroom devel- dreds more, spread very quickly, pri-

22 Fire Protection Engineering N UMBER 24

marily on the exposed convoluted flexi- of scenarios based on fundamental ma-
ble polyurethane foam material that had terial flammability properties obtained
been installed on the walls and ceiling from quantitative small-scale tests, such
of the bandstand in the nightclub. This as the Cone Calorimeter,31 the LIFT appa-
foam plastic product reportedly was in- ratus,32 and the FM Fire Propagation Ap-
tended for use as a packing material and paratus.33 While considerable progress
therefore did not incorporate even a has been made, there is still a need for
nominal amount of fire retardants. In large-scale testing, both to verify model Figure 2.
light of the widespread recognition of predictions and to evaluate performance Room-corner fire
the fire hazards associated with exposed characteristics of some materials and as- test geometry.
foam plastic interior finishes and the
regulation of the application of these
products since the 1970s, it is difficult to
comprehend how this application could
have existed in 2003. It should serve as
a reminder to fire safety professionals
everywhere of the need for continual
diligence.
Much of the focus on the Station fire
has been on the lack of automatic sprin-
kler protection in the nightclub rather
than on the exposed foam plastic inte-
rior finish that ignited so easily and
spread the fire so quickly. Recent large-
scale experiments conducted at the Na-
tional Institute of Standards and Tech-
nology28 with a wet-pipe sprinkler
system and quick-response sprinklers
suggest that the presence of similar au-
tomatic sprinkler protection in the Sta-
tion may have significantly improved the
outcome of the fire there. While auto-
matic sprinkler protection is widely rec-
ognized to be beneficial for both life
safety and property protection, it should
not be considered as an acceptable
trade-off for unsafe and improper instal-
lations of foam plastic materials as inte-
rior finishes. Where such installations of
exposed foam plastics exist, they should
be removed, regardless of the presence
of automatic sprinkler protection.

FLAME SPREAD THEORY AND

MODELING
Concurrent with the development of
room fire test methods that more accu-
rately portray the performance of build-
ing materials under actual fire condi-
tions, the scientific understanding of
flame spread on solid surfaces has ad-
vanced, and models of the flame spread
process have been developed. For ex-
ample, Quintiere29 has developed a fairly
comprehensive yet relatively simple sim-
ulation model for flame spread that has
been incorporated in the BRANZFIRE
zone fire model.30
The ultimate objective of research on
flame spread is to be able to predict the
development of fire under a full range

FALL 2004 www.sfpe.org 23

 ■ The Role of Interior Finish in Fire Development

semblies that can- The section of lining material directly behind the ignition
not yet be modeled source will be the first to ignite. The flame on this section then
accurately, such as may spread vertically and beneath the ceiling, as indicated by
melting, dripping, the orange arrows in Figure 2, as well as laterally and down-
delamination, and ward, as indicated by the black arrows in Figure 2. In general,
Xf2
warping. the upward flame spread and spread beneath a ceiling are
Consider the sce- known as wind-aided spread because the flame is spreading in
nario depicted in the same direction as the buoyant flow of gases. This wind-
Figure 2, which is aided spread is generally much faster than the lateral and
Xf1 Xp2 representative of downward spread because of the larger sections of wall and
the scenario used in ceiling being heated by the advancing flame front.
most room fire Flame spread on a fuel surface can be considered as a se-
Xfo Xp1
tests. The walls quence of ignitions, as illustrated in Figure 3. An exposure fire
and/or ceiling of an or the flame from a segment of the material that is already burn-
Xpo
enclosure are lined ing imposes a heat flux on a fuel element that has not yet ig-
with a combustible nited. The temperature of this fuel surface element increases
Figure 3. Conceptual illustration of flame interior finish mate- under the imposed heat flux. When a fuel element reaches its
spread as a sequence of ignitions. rial. A section of the ignition temperature, the flame spreads to that fuel element,
lining material is and it begins to burn. With this fuel element now burning, the
subjected to an imposed heat flux from an ignition source fire, flame grows longer and imposes a heat flux on the next fuel
which is normally selected to represent a typical incidental fire, surface element. Some materials, such as thin combustible sur-
such as a small trash receptacle fire.34 Such ignition sources are face coatings or materials adhered to noncombustible sub-
normally selected to realistically challenge the lining materials strates, burn out relatively quickly once ignited. Other materials,
but not overwhelm the performance of the lining materials. In such as some wood products, char and consequently have a
room fire tests, such ignition sources are typically located near burning rate that decreases with time. Under some exposure
the corner of two walls because this represents a realistic “worst- conditions, such materials may not burn with sufficient intensity
case” ignition scenario, as noted in the 1950 FM room fire tests. long enough to ignite subsequent fuel elements.
Upward flame spread on a fuel surface generally requires two
conditions to occur:

1. The flame from the currently burning area of the fuel sur-
face must extend beyond the burning area to expose the
adjacent area to a heat flux high enough to ignite the adja-
cent area; and

2. The heat flux must be applied long enough to ignite the

adjacent fuel surface.

To satisfy the first condition, the heat release rate per unit
area of the burning fuel must be high enough to cause the
flame to extend beyond the burning area. In general, the length
of a flame along a vertical burning surface will be proportional
to its heat release rate per unit width,35 which in turn is propor-
tional to the heat release rate per unit area. This can be ex-
pressed as:

( ) ( )
n n
x f = k f Q˙ ′ = k f Q˙ ′′x p (1)

where xf is the length of the flame (m), measured from the

base of the pyrolysis zone, xp is the length of the pyrolysis zone
(m), kf is an appropriate flame length coefficient ((m/(kW/m)n),
Q˙ ′ is the heat release rate per unit width (kW/m), and Q˙ ′′is the
heat release rate per unit area (kW/m2) of the burning area of
the fuel surface. According to this simple theory, the flame
length must be greater than the pyrolysis length in order for
flame spread to occur. Mathematically, this means that for flame
spread to occur the following relation must hold true:

(Q˙ ′′)
n
(2)
kf 1− n
>1
x p

24 Fire Protection Engineering N UMBER 24

 ■ The Role of Interior Finish in Fire Development

Expressed differently, this also establishes the minimum heat- burning duration for the second condition would be the period
release rate per unit area for upward or wind-aided flame of time during which the heat release-rate per unit area causes
spread to occur: the flame length to exceed the pyrolysis zone length. In general,
1/ n the burning duration can be evaluated as:
 x1p− n  (3)
Q′′ > 
˙
 Q′′ m ′′∆Hc m ′′L (4)
 kf  tb = = =
′′ , p ( ∆Hc / L) q˙net
Q˙ ′′ q˙net ′′ , p
For example, Cleary and Quintiere36 have suggested that
kf =0.01 m2/kW and n=1 can be used to represent the flame where Q′′ is the energy content of the fuel surface per unit
length relationship, with a linear relationship between the flame area (kJ/m2), Q˙ ′′ is the average heat-release rate per unit area
length and the pyrolysis length. Based on these values, a heat (kW/m2), m ′′ is the combustible mass per unit area (kg/m2), L is
release rate per unit area of Q˙ ′′ ≥ 100 kW/m2 would be needed the effective heat of gasification of the combustible mass
for upward flame spread to occur. Tu and Quintiere37 have ′′ , p is the net heat flux to the fuel surface
(kJ/kg), and q˙net
also suggested that kf =0.067 m5/3/kW2/3 and n=2/3 are appropri- (kW/m2) in the pyrolysis zone. For thermally thick surfaces, the
ate values to represent this flame-length relationship. Based time to ignition is generally represented, for a constant net heat
on these values, the minimum heat-release rate per unit area flux at the fuel surface, as:
needed for upward flame spread would be Q˙ ′′ ≥ 58 x p kW/m2. 2
Note that this value is a function of the pyrolysis zone length, π  ∆T  (5)
tig = kρc  ig 
with larger heat-release rates per unit area needed to sustain up- 4 ′′ , f 
 q˙net
ward flame spread for longer pyrolysis zone lengths. This is one
reason why some fires may burn out after spreading some dis- where the product kρc is the thermal inertia of the solid sur-
tance up a wall. These relations are shown in Figure 4. face ((kW/m2K)2s), ∆Tigis the difference between the ignition
To satisfy the second condition, the burning duration, tb , of the temperature and the initial surface temperature (K), and q˙net ′′ , f
burning region must be greater than the ignition time, ti g , of the is the net heat flux to the fuel surface in the flame region
exposed region. More specifically, the burning duration should be (kW/m2). In general, the net heat flux terms in Equations 4 and
evaluated as the period of time that the burning region burns at a 5 will not be equal to each other, but for this discussion they are
rate sufficient to achieve the first condition. In other words, the assumed to be proportional to each other, i.e., q˙net ′′ , f = χ p q˙net
′′ , p .
In general, the net heat flux in the pyrolysis zone is expected to
be greater than the net heat flux in the flame zone, in which case
the proportionality factor, χ p , will have a value of less than one.
The burning duration expressed by Equation 4 can be equated
with the ignition time expressed by Equation 5 to determine the
minimum flame heat flux needed to cause ignition before burn-
out occurs. After some manipulation, this can be expressed as:

′′ , f >
q˙net
(
π kρc∆Tig
2
) (6)
4 χ p ( m ′′L)
Equation 6 would be difficult to evaluate quantitatively, par-
ticularly since the value of the proportionality factor is not
known. Nonetheless, Equation 6 is useful for a number of
reasons. First, it demonstrates that there is expected to be a
minimum heat flux for flame spread for materials where fuel
burnout is significant. Thus, it is important that such materials
be tested under exposure conditions sufficient to exceed this
minimum heat flux; otherwise, anomalous test results can oc-
cur when compared with actual field performance. This be-
havior has been observed for textile wall coverings, as noted
above. Second, Equation 6 shows how different material prop-
erties are expected to influence the minimum heat flux for
flame spread. Higher thermal inertias and larger ignition tem-
peratures would be expected to increase the minimum heat
flux for flame spread, while more fuel per unit area would be
expected to lower it. Third, Equation 6 demonstrates the criti-
cal nature of flame spread, where a slight change in the heat
flux or in the combustible mass per unit area (e.g., another
coat of paint) can spell the difference between burnout and
flame propagation. Finally, Equation 6 also shows that pre-
heating of a fuel surface will tend to decrease the minimum
heat flux for flame spread by decreasing the temperature rise
needed to ignite the surface.

25 Fire Protection Engineering N UMBER 24

 ■ The Role of Interior Finish in Fire Development

The relatively simple theoretical fluence flame spread on wall and ceiling ber of complex interrelated processes,
analysis presented here has identified a finishes include: even for relatively simple geometries,
number of material properties and envi- • The heat flux imposed on the fuel homogeneous fuels, and well-character-
ronmental conditions that are expected surface by an exposure fire. This will in- ized exposure conditions. It is for this
to influence flame spread on interior fluence the burning rate and the size of reason that individual fire tests of inte-
wall and ceiling finishes. The material the fuel area first ignited, and conse- rior finish materials may not be able to
properties include: quently the flame length extending from characterize their performance under a
• The thermal inertia of the material. this area and exposing adjacent fuel ele- full range of field-use conditions.
As shown in Equation 5, the thermal in- ments. By influencing the burning rate
ertia of a material is directly proportional of the fuel, this parameter also influ- PROPOSED EVALUATION
to the ignition time. Low-density materi- ences the burning duration of this area. METHODOLOGY
als tend to also have low thermal con- Ironically, a higher imposed heat flux
ductivities and consequently have very may cause earlier burnout that a lower In 1978, Williamson and coworkers34
low thermal inertias. This is the primary heat flux and consequently not cause suggested that “a standard room fire test
reason why flame spread can be very flame spread under some conditions that could be used both as a development
rapid on exposed foam plastic products. a lower heat does. tool and a performance evaluation
• The ignition temperature of the ma- • The heat flux imposed by burning method until such time as a series of
terial. Although Equation 5 shows that surface flames on adjacent fuel ele- smaller, less expensive tests has been
the time to ignition varies with the ments. This will influence the time to ig- proven. Even then, new materials and
square of the ignition temperature rise, nition of these adjacent fuel surface ele- systems which are different in principle
ignition temperatures for most building ments and consequently the speed of from those already validated for small-
materials fall within a relatively small flame spread. scale fire tests would still require the
range, so differences in ignition temper- • The gas temperatures within the en- full-scale test to show the applicability
atures among materials do not affect closure. The accumulation of hot gases of small-scale tests.” This is similar in
flame spread nearly as much as the or- beneath the ceiling as a result of a fire concept to the evaluation methodology
der of magnitude differences in thermal causes preheating of the fuel surfaces in proposed here.
inertia do. contact with the hot gases. As these sur- The evaluation methodology pro-
• The combustible mass per unit area faces heat up, the temperature rise posed here includes a preliminary
of the material. This parameter is most needed to cause ignition decreases, re- screening/qualification step, followed by
significant for relatively thin coatings sulting in shorter ignition times and a more detailed analysis step. In the
and materials on noncombustible sub- lower minimum heat fluxes for fire screening/qualification step, the flamma-
strates, such as painted or unpainted pa- spread. This effect will be most pro- bility characteristics of a material are
per facers on gypsum wallboard or tex- nounced for materials that are good in- evaluated using a quantitative small-
tile wall coverings adhered to gypsum sulators, such as foam plastics and other scale fire test method, such as the Cone
wallboard, but is also important for ma- low-density materials, because their Calorimeter or the FM Fire Propagation
terials that tend to char. Such materials good insulating qualities will result in Apparatus. One of three outcomes will
are more likely to burn out locally and higher gas temperatures as well as occur, depending on the performance of
not spread a fire than materials with higher surface temperatures than more the material in the bench-scale test.
more combustible mass per unit area. conductive materials will. These outcomes include:
• The ratio between the heat of com- Based on this analysis, it should be • the material will be screened from
bustion and the heat of gasification apparent that flame spread on interior any further consideration if it exhibits
(∆Hc / L) of the material. As demon- wall and ceiling finishes involves a num- flammability characteristics recognized
strated in Equation 4, this “combustibil-
ity ratio” is directly proportional to the
heat-release rate per unit area of a mate-
rial and consequently has an influence Good performance – Intermediate performance Poor performance –
on the flame length as well as on the to- product qualified – further testing required product screened
tal heat-release rate of the fire, which
will have an influence on the preheating Low heat release rate per Intermediate heat release High heat release rate
of fuel surfaces as well as the potential unit area (e.g., rate per unit area per unit area (e.g.,
for flame extension beyond the room of Q˙ ′′ < 65 kw / m 2 ) or Q˙ ′′ > 200 kw / m 2 )
Time to ignition similar
origin. or to burning duration or
• The heat of gasification of a mater- or
ial. While this property individually is Time to ignition much Burning characteristics Time to ignition much
not as significant as the “combustibility greater than burning that cannot be evaluated less than burning
ratio,” it does have an influence on the duration in small-scale tests duration
burning duration and consequently on (e.g., tig > 2 × tb ) (e.g., tig < 0.5tb )
the minimum heat flux for flame spread,
as demonstrated by Equation 6.
The environmental parameters that in- Figure 4. General concept for flammability testing and evaluation.

26 Fire Protection Engineering N UMBER 24

to be unacceptable for the anticipated REFERENCES 7 ASTM E-84-50T, Tentative Method of Fire
conditions of use; Hazard Classification of Building
• the material will be qualified as ac- 1 Belles, D.W., “Interior Finish,” Section Materials, American Society for Testing
ceptable if it exhibits flammability charac- 12/Chapter 3, Fire Protection Handbook, and Materials, 1950.
teristics considered to be fully acceptable 19th Edition, National Fire Protection 8 “Life Hazard of Interior Finishes
for the anticipated conditions of use; Association, Inc., 2003. (Development of Method),” Laboratory
• the material will need to be subjected 2 Moulton, R.S., “The Cocoanut Grove Report No. 11760, Factory Mutual
to additional large-scale testing, such as Nightclub Fire, Boston, Massachusetts, Laboratories, June 1, 1950.
room fire testing, if its expected field per- November 28, 1942,” National Fire 9 Steiner, A.J., “Testing for Fire Safety,”
formance cannot be adequately judged Protection Association, 1962. Quarterly of the National Fire Protection
based on its bench-scale performance. 3 McElroy, J.K., “The LaSalle Hotel Fire,” Association, Vol. 50, No. 3, January 1957,
This concept is illustrated in Figure 4. Quarterly of the National Fire Protection pp. 195-204.
The specific values of the different para- Association, Vol. 40, No. 1, July 1946, pp. 10 Wilson, J.A., “A Different View on –
meters to be used for screening or quali- 4-18. Plastic Fire Hazard Classifications,”
fication will need to be evaluated along 4 McElroy, J.K., “The Hotel Winecoff Quarterly of the National Fire Protection
with the exposure conditions under Disaster,” Quarterly of the National Fire Association, Vol. 56, No. 2, October
which these parameters are evaluated. Protection Association, Vol. 40, No. 3, 1962, pp. 162-164.
The values identified in Figure 4 are in- January 1947, pp. 140-159. 11 Steiner, A.J., “Fire Tests and Fire
tended only as examples, although they 5 Steiner, A.J., “Fire Hazard Tests of Protection Engineering,” Quarterly of the
are consistent with expected perfor- Building Materials,” Quarterly of the National Fire Protection Association, Vol.
mance. ▲ National Fire Protection Association, Vol. 52, No. 3, January 1959, pp. 209-220.
37, No. 1, July 1943, pp. 69-78. 12 “Fire Research on Cellular Plastics: The
Robert Brady Williamson is with the 6 Steiner, A.J., “Fire Hazard Classification of Final Report of the Products Research
University of California at Berkeley, and Building Materials,” Bulletin of Research Committee,” 1980.
Frederick W. Mowrer is with the Univer- No. 32, Underwriters’ Laboratories, Inc., 13 Wilson, J.A., “Surface Flammability of
sity of Maryland. September 1944. Materials: A Survey of Test Methods and

FALL 2004 www.sfpe.org 27

 ■ The Role of Interior Finish in Fire Development

Comparison of Results,” ASTM Special 26 Belles, D.W., Fisher, F.L., and

Technical Publication No. 301, Williamson, R.B., “How Well Does ASTM
Symposium on Fire Test Methods, E-84 Predict Fire Performance of Textile
American Society for Testing and Wallcoverings?,” Fire Journal,
Materials, 1961, pp. 61-82. January/February 1988, p. 24.
14 Williamson, R.B., and Baron, F.M., “A 27 Klem, T.J., “Investigation Report on the
Corner Test to Simulate Residential Dupont Plaza Hotel Fire, December 31,
Fires,” Journal of Fire and Flammability, 1986, San Juan, Puerto Rico,” National
Vol. 4, April 1973, pp. 99-105. Fire Protection Association, undated.
15 Childress vs. Cook Paint & Varnish 28 Madrzykowski, D., Bryner, N.,
Company, In the Circuit Court of Carlyle Grosshandler, W., and Stroup, D., “Fire
County, Missouri, Case No. 11077, Spread Through a Room with
October 6, 1971. Polyurethane Foam-Covered Walls”, NIST
Special Publication 1000-101, National
16 Federal Trade Commission Complaint on
Institute of Standards and Technology,
the Flammability of Plastic Products, File
Gaithersburg, MD, June 2004.
No. 732-3040, May 1973.
29 Quintiere, J.G., “A Simulation Model for
17 Docket C-2596, Complaint and Decision,
Fire Growth on Materials Subject to a
Nov. 4, 1974, printed on pages 1253 to
Room-Corner Test,” Fire Safety Journal,
1279 in the “Federal Trade Commission
Vol. 20, 1993, pp. 313-339.
Decisions.”
30 Wade, C.A., “BRANZFIRE Technical
18 Fire Research on Cellular Plastics: The
Reference Guide,” BRANZ Study Report
Final Report of the Products Research
92 (revised), Building Research
Committee, 1980.
Association of New Zealand, 2003.
19 Castino, G.T., Beyreis, J.R., and Metes,
31 ASTM E1354, “Standard Test Method for
W.S., “Flammability Studies of Cellular
Heat and Visible Smoke Release Rates
Plastics and Other Building Materials
for Materials and Products Using an
used for Interior Finishes,” Subject 723,
Oxygen Consumption Calorimeter,”
Underwriters Laboratories, Inc., June 13,
ASTM International, West
1975.
Conshohocken, PA, 2002.
20 Williamson, R.B., and Fisher, F.L., “Fire
32 ASTM E1321, “Standard Test Method for
Growth Experiments – Toward a
Determining Material Ignition and Flame
Standard Room Fire Test,” Paper No. 79-
Spread Properties,” ASTM International,
48, 1979 Fall Meeting, Western State
West Conshohocken, PA, 1997.
Section of the Combustion Institute.
33 ASTM E2058, “Standard Test Methods for
21 Fisher, F.L., and Williamson, R.B.,
Measurement of Synthetic Polymer
“Intralaboratory Evaluation of a Room
Material Flammability Using a Fire
Fire Test Method,” UCB FRG 82-1, Final
Propagation Apparatus,” ASTM
Report on NBS Grant NB80NADA1072,
International, West Conshohocken, PA,
University of California, Berkeley, August
2002.
1982.
34 Van Volkinburg, D.R., Williamson, R.B.,
22 “Proposed Standard Method for Room
Fisher, F.L., and Hasegawa, H., “Toward
Fire Test of Wall and Ceiling Materials
a Standard Ignition Source,” Paper No.
and Assemblies,” 1982 Annual Book of
78-64, 1978 Fall Meeting, Western States
Standards, Part 18, American Society for
Section of the Combustion Institute,
Testing and Materials, 1982.
1978.
23 Best, R., and Demers, D.P., “Investigation
35 Tu, K.M., and Quintiere, J.G., “Wall
Report on the MGM Grand Hotel Fire,”
Flame Heights with External Radiation,”
National Fire Protection Association,
Fire Technology, Vol. 27, No. 3, 1991,
January 15, 1982.
pp. 195-203.
24 Mowrer, F.W., Williamson, R.B., and
36 Cleary, T.G., and Quintiere, J.G., “A
Fisher, F.L., “Analysis of the Early Fire
Framework for Utilizing Fire Property
Development at the MGM Grand Hotel,”
Tests,” Proceedings of the 3rd
Proceedings of the Second International
International Symposium on Fire Safety
Conference on Fire Research and
Science, pp. 647-656.
Engineering, Society of Fire Protection
Engineers, 1997. 37 Tu, K.M., and Quintiere, J.G., “Wall
Flame Heights with External Radiation,”
25 “Investigation Report on the Las Vegas
Fire Technology, Vol. 27, No. 3, 1991,
Hilton Hotel,” Fire Journal, Vol. 76,
pp. 195-203.
No. 1, January 1982, p. 52.

28 Fire Protection Engineering N UMBER 24

 Assessing the Burning Characteristics of Interior Finish Material

Standard Test Method for Surface-Burning

Characteristics of
Building Materials ASTM E-84/UL 723

By Randy Laymon The Steiner tunnel is a furnace cham- board. Red oak propagates flames to the
Underwriters Laboratories Inc. ber that measures flame spread and end of the tunnel in 5 minutes 30 sec-
smoke development. Its prominence in onds ± 15 seconds and generates a
INTRODUCTION the fire protection community was based flame-spread index of approximately 90.
on its ability to provide cost-effective, A smoke-developed index of 100 is as-

T hroughout history, structural fires

have caused massive destruction
and countless injuries and fatali-
ties. Although the flammability character-
repetitive testing and use a sample size
that could better characterize interior fin-
ish materials used in actual installations.
This method is currently described in UL
signed for red oak. Inorganic reinforced
cement board generates flame-spread
and smoke-developed indices of zero.

istics of interior finish within these struc- 723,1 Test for Surface Burning Charac- EARLY HISTORY AND DEVELOPMENT
tures has played a major role in many of teristics of Building Materials, as well as
these losses, prior to the middle of the ASTM E-842 and NFPA 255.3 The initial version of the tunnel fur-
20th century, fire protection of buildings nace was developed in 1922 when Mr.
focused primarily on: 1) the prevention SUMMARY OF TEST METHOD Steiner, an engineer in UL’s Fire Protec-
of fire occurrence, 2) early detection and tion Department, assessed the effective-
warning, 3) automatic or manual extin- The Steiner tunnel is used to assess ness of a “fireproof” paint. The proto-
guishment, and 4) confinement with the comparative surface-burning charac- type test method consisted of a long
fire-resistant structural components, such teristics of building material samples wooden bench measuring approxi-
as floors, ceilings, walls and partitions, with the exposed area measuring 18 in. mately 18 in. (460 mm) in width and
columns, roofs, and doors. (460 mm) wide by 24 ft. (7.3 m) long, up depth and 16 ft. (4.9 meters) long with a
The occurrence of major fires in indi- to a thickness of approximately 5-6 in. noncombustible top. The interior of the
vidual buildings, distinguished by the (125-150 mm). The test is conducted tunnel was coated with the paint under
rapid flame spread of interior finish mate- with the sample mounted in the “ceil- investigation and ignited with a given
rials, aroused public concern and demon- ing” position of an enclosed tunnel fur- quantity of wood at one end. The extent
strated the need to address and regulate nace measuring 18 in. (460 mm) wide of the spread of flame was compared
the burning characteristics of these mate- by 12 in. (300 mm) deep by 25 ft. long with an unpainted replica, and the flame
rials. Specific material characteristics of (7.6 m). A nominal 5000 Btu/min. (88 retardancy of the coating was thus
concern included the spread of flame kW), 4-1/2 ft. (1.4 m) flame provides an evaluated.
and the amount of heat generated and ignition source to the underside of the In the late 1920s, the development of
smoke developed. This led to the re- mounted specimen for a 10-minute du- pressure-impregnated fire-retardant lum-
search and development of various test- ration. A controlled inlet draft of 240 feet ber, in conjunction with further research
ing protocols, most of which were small, per minute (1.2 meters/second) facili- at UL, led to modifications to the test
laboratory-scale tests. However, based on tates horizontal flame propagation method in which the test sample formed
work conducted by Albert J. Steiner at throughout the test. A light and photo- the top of a 36 in. (91 mm) wide by 13
Underwriters Laboratories Inc., from the electric cell mounted in the exhaust duct in. (330 mm) deep by 23 ft. (7.0 m) long
early 1920s through the 1940s, the 25 ft. record smoke obscuration during the chamber. The use of untreated red oak
(7.6 m) long Steiner tunnel emerged as test. Flame-spread and smoke-devel- and maple flooring in this investigation
the predominant method to characterize oped indices are reported in comparison was a major factor in the selection of red
and regulate the surface-burning charac- with calibration materials of red oak oak as one of the calibration materials
teristics of interior finish materials. lumber and inorganic reinforced cement for the test method.

29 Fire Protection Engineering N UMBER 24

 By the beginning of World War II, current method is still based on a time- COMMUNITY ACCEPTANCE
there was growing interest in reducing distance area calculation but incorpo-
the combustibility of existing materials rates a rate of flame travel as well. Just as the test method developed
through various treatments and in mea- Prior to 1978, a Fuel Contributed In- gradually over a period of years, so did
suring the flammable properties of new dex was reported. This index was based its acceptance. The test method was first
materials. In addition, by the mid-1940s, on air temperatures developed within published in 1950 by Underwriters Lab-
a number of disastrous fires occurred, the tunnel furnace during testing. In oratories Inc. as Standard UL 723. ASTM
including the Cocoanut Grove nightclub 1978, the Fuel Contributed Index was followed by publishing the test method
fire in Boston in 1942 and the Chicago deleted from the method since it was as a tentative standard in 1950 and as a
LaSalle Street Hotel and Atlanta recognized that the value did not provide formal Standard, ASTM E-84, in 1961.
Winecoff Hotel fires, both in 1946. In all, a valid measure of fuel contribution. NFPA adopted the test method as NFPA
670 people perished in these three fires
alone. The magnitude of the fire fatali-
ties in each of these fires was directly re-
lated to the rapid flame spread and
smoke development of the interior fin-
ish materials. These findings highlighted
the need to test and classify materials on
a scale that would measure the three es-
sential material characteristics previously
identified: flame spread, fuel con-
tributed, and smoke developed. All
these factors led to the evolution of the
current tunnel apparatus. It was at this
time that the Surface-Burning Character-
istics Classification Scale was first de-
fined. It was essential that, in order to
classify materials according to the prop-
erties of flame spread, fuel contributed
and smoke developed, as well as to
have this information be of value, a
comparative scale was required. Accord-
ingly, the test initially developed a clas-
sification for each of these properties for
a sample material on a comparative
scale with a combustible (red oak lum-
ber) defined as 100 and a noncom-
bustible cement board as zero.
The current physical version of the
tunnel was completed in the late 1940s.
Many controls were implemented to en-
hance repeatability and reproducibility.
The standard specimen size became 20
in. (510 mm) wide by 25 ft. (7.6 m) long.
The ignition source was adjusted to ob-
tain a nominal 4-1/2 ft. (1.4m) long,
5000 Btu/min. (88 kW) test flame that
generates gas temperatures of approxi-
mately 1200°F to 1600°F (650°C-870°C)
near the specimen surface at the ignition
end of the test sample. The inlet draft
was established at 240 feet per minute
(1.2 meters/second).
The method used to calculate Flame
Spread Index (FSI) has undergone some
modifications over the years. Originally,
the FSI was based on the ratio of the
time at which flames traveled the full
tunnel length or the partial flame travel
distance relative to that of red oak. In
1976, the FSI was changed to a time-
flame spread distance area basis. The

FALL 2004 www.sfpe.org 30

 ■ Standard Test Method for Surface-Burning

255 in 1955. It was adopted by ANSI in promoting more-consistent results by thereby increasing the flame spread in-
1963 as American National Standard A2.5. various laboratories. Recently, a more dex; or the material may sag or drop to
Although the tunnel test provides for comprehensive approach toward the the furnace floor, which may impede
a Classification protocol and is recog- standardization of mounting practices further flame propagation.
nized by standards-developing organiza- has led to the development of ASTM • Thermoplastic materials may be dif-
tions, it does not establish limitations for E2231, Standard Practice for Specimen ficult to evaluate in this as well as other
building codes. The intent of the test Preparation and Mounting of Pipe and standardized fire test procedures and re-
method is to provide a tool for those Duct Insulation Materials to Assess Sur- quire careful interpretation of the test re-
with the responsibility of regulating ma- face-Burning Characteristics. Similar sults. These materials tend to melt and
terials used as interior finish in build- practices for other material types are drip to the floor of the furnace, and may
ings. Widespread reliance on the tunnel currently being considered under the generate potentially misleadingly low
test method by the regulatory commu- ASTM standard-development process. flame-spread values.
nity as an acceptable criterion to assess • Some research has indicated that
interior finish and other materials has ADVANTAGES AND LIMITATIONS some types of thermosetting cellular
been in place for decades. Factors that plastics yielding low flame-spread val-
have contributed to this reliance include: • Certain relationships have been ob- ues may generate flameover conditions
• Support by standards-developing served between Steiner tunnel test re- during certain large-scale room test sce-
organizations, including UL, ASTM, and sults and performance of some materials narios, when utilizing igniting sources of
NFPA. during building fires.2 sufficient heat flux levels.4
• The test method utilizes a large • The test method provides for a real- No single test method provides the to-
sample size and an ignition source rep- istic fire scenario by presenting a sample tal information necessary to completely
resentative of a moderately developed of sufficient size to allow for progressive evaluate the potential for fire develop-
fire scenario. surface burning over a large exposed ment in a building, yet each makes
• The ability of the test method to area. some contribution to the total body of
characterize both high and low flame • A wide range of materials may be knowledge required. The Steiner tunnel
spread materials. tested, including composite construc- test method is the most extensively used
• Research that demonstrates a rela- tions, coatings, faced products, loose-fill and referenced test method to assess
tionship between tunnel test results and materials, sandwich panels, and many flammability of interior finish materials.
certain large-scale test protocols.4 others. UL currently classifies over thirty The results form a basic element in reg-
Interior finish requirements are cur- different product types in accordance ulation of these materials by providing
rently defined in Chapter 8 of the Inter- with the test method. an identification system for inspection
national Building Code 5 and Chapter 10 • The test method provides a means and enforcement authorities. ▲
of NFPA 5000, Building Construction to discriminate products yielding a wide
and Safety Code.6 Interior finishes are range of flame-spread and smoke-devel- Randy Laymon with Underwriters
grouped in the following classes in ac- oped characteristics, allowing for the de- Laboratories Inc.
cordance with their flame-spread and velopment of codes and standards.
smoke-developed indices. • Some research conducted has REFERENCES
Class A: Flame Spread 0-25; Smoke demonstrated useful relationships be-
Developed 0-450. tween Steiner tunnel flame-spread val- 1 UL723, Test for Surface-Burning
Class B: Flame Spread 26-75; Smoke ues and fire performance of materials in Characteristics of Building Materials,
Developed 0-450. large-scale corner configurations using a Underwriters Laboratories, Inc.,
Class C: Flame Spread 76-200; Smoke 20-pound ignition source wood crib.4 Northbrook, IL, 2001.
Developed 0-450. • The horizontal specimen orientation 2 ASTM E-84, Standard Test Method for
Prior to 1960, the tunnel test method places some limitation on the type of Surface-Burning Characteristics of
was used primarily for the evaluation of material that can be realistically Building Materials, American Society for
the surface-burning characteristics of ho- mounted. Depending on the particular Testing and Materials, West
mogenous compositions of ceiling and material being tested, specimens requir- Conshohocken, PA, 2003.
wall finishes, such as acoustical tiles, ing additional support may yield low 3 NFPA 255, Standard Method of Test for
wall coverings, coatings, and various flame-spread values due to the support- Surface-Burning Characteristics of
types of decorative paneling. Through ing material restricting flame propaga- Building Materials, National Fire Protection
inclusion of the Guide to Mounting tion or high-flame spread values be- Association, Inc., Quincy, MA, 2000.
Methods Appendix in the late 1960s, the cause the additional support retains the 4 “Flammability Studies of Cellular Plastics
method was expanded to include the specimen in the ceiling position rather and Other Building Materials Used For
evaluation of composites and assem- than allowing the specimen to fall away Interior Finishes,” Underwriters
blies. Sample mounting techniques can from the area of flame impingement. Laboratories Inc., Northbrook, IL, 1975
have a significant influence on the fire- • Some materials, such as faced com- 5 International Building Code, International
performance indices developed by the posite samples, may delaminate during Code Council, Falls Church, VA, 2003.
test method. While the Appendix is not testing, which may result in one of two 6 NFPA 5000, Building Construction and
considered a mandatory part of the stan- possible responses: the material may ex- Safety Code, National Fire Protection
dard, the Guide has proven useful in pose two or more surfaces to the flame, Association, Quincy, MA, 2003.

31 Fire Protection Engineering N UMBER 24

 Resources
New Additions to the

Publications Catalog
The Code Official’s Guide to Performance-Based Evaluation of Fire Safety
Design Review In a comprehensive treatment of the subject unavailable else-
The use of performance-based design is becoming more where, this book describes in detail the applications of hazard
prevalent as new performance-based codes and guidelines and risk analysis to fire safety, going on to developing and
are developed and adopted. This can create challenges for applying quantification methods. It also gives an explanation
enforcement officials or other stakeholders if they do not in quantitative terms of improvements in fire safety in associa-
have a strong background in performance-based design or if tion with the costs that are expended in their achievement.
their resources are already stretched thin. This guide identifies Furthermore, a quantitative approach is applied to major fire
the types of items that an enforcement official should con- and explosion disasters to demonstrate crucial faults and
sider when reviewing a performance-based design. The con- events. ISBN 0-471-49382-1, 462 pages.
cepts identified in this guide are applicable to performance- $136 SFPE Members/$160 Nonmembers
based fire protection designs that are prepared to meet
performance-based codes, designs that are prepared as equiv-
alencies to prescriptive-based code requirements, and designs
that are intended to meet objectives that exceed those con-
Performance-Based Building Design Concepts:
tained in a code or standard (e.g., business interruption, pro- A Companion Document to the ICC PC
tection of contents, etc.) ISBN 1-58001-202-7, 113 pages. This companion document to the ICC Performance Code for
Buildings and Facilities (ICC PC) leads the reader along a
$40.00 SFPE Members/$49.00 Nonmembers
path from a discussion of the origins of performance-based
building codes and the background behind the ICC PC to dis-
cussion of the basics of performance design. It was designed
SFPE Engineering Guide to Fire Exposures to to assist the reader in learning more about performance, ap-
Structural Elements plying performance concepts appropriately, and knowing
Designing fire resistance of structures in a performance envi- what to look for in the review of designs that have been de-
ronment is a three-step process: defining the fire boundary veloped using the ICC PC.
conditions that the structure will be exposed to, and deter-
mining its thermal and then structural response. SFPE’s The ten chapters cover history and overview, regulatory sys-
newest engineering guide provides the information necessary tems, understanding administrative issues, risk characteriza-
to the fire protection engineer to aid in the first important step tion and performance concepts, fire, structural design, pedes-
of this process: estimating the fire boundary conditions. De- trian movement and safety, building envelope, maintaining
sign methods, their limitations, and examples of their applica- building performance throughout a building’s life cycle, and
tion are provided for fully developed exposure fires and for prospects for the future. ISBN 1-58001-182-9, 300 pages.
fire plumes, the two fire exposures of most importance in the $70.00 SFPE Members/$87.00 Nonmembers
design of structures for fire, 146 pages.
$40.00 SFPE Members/$80 Nonmembers

32 Fire Protection Engineering N UMBER 24

 UPCOMING EVENTS
October 20, 2004 December 6, 2004
Computational Simulation Models in Fire Symposium on Firestopping
Engineering and Research
Washington, DC
Santander, Spain Info: www.astm.org
Info: grupos.unican.es/gidai
Through December 2, 2004
November 16-17, 2004
Fire, Blast, Progressive Collapse Workshops
International Symposium on Tunnel Safety and Security
In various cities in the U.S., please see Web site for dates, times,
Greenbelt, MD
and places.
Info: www.ni2cie.org
http://www.aisc.org/Template.cfm?Section=Events_Calendar&Tem
plate=/Calendar/CalendarEventList.cfm&List=True&TPLID=2&AreaI
November 29-30, 2004
D=47
Fire Risk Evaluation to European Cultural Heritage
Ghent, Belgium
Info: www.firetech.be

FALL 2004 www.sfpe.org 33

 Fire Alarm
Systems
and Interior Finish –

A Balanced Approach
W hat is the relation-
ship between a fire
detection and
alarm system and the interior
finish of a space? Fire protec-
In this article, interior finish is used as
a variable; see how fire detection and
alarm systems must change to maintain
Class B (flame spread 26-75; smoke de-
veloped 0-450)1. Class A (flame spread
0-25; smoke developed 0-450) materials
balance when some other system (inte- are required in Assembly occupancies
tion is not any one system rior finish) changes. Though the link is while Class C (flame spread 76-200;
not a strong one, it is useful to demon- smoke developed 0-450) materials are
but a balance between many strate how different systems interact with permitted in one- and two-family
systems and concepts. This each other. How does a fire detection dwelling units.
article looks at the role of fire and alarm system affect the selection of
interior finishes? What effect does inte- The restrictions on interior finish are
detection and alarm systems rior finish have on the design of a fire based, in part, on:
as a part of a balanced fire detection and alarm system? • the expected/permitted occupant
protection system. From an load
BALANCED PREVENTION AND • occupant mobility
analysis point of view, when PROTECTION • the maximum permitted travel dis-
one facet of fire protection tance
changes, the performance of Model building and fire codes contain • the degree of compartmentation
specific limitations on interior finish. • the presence or lack of automatic
other systems may be affect- Some occupancies or use groups are suppression systems
ed. From a design perspec- permitted to have combustible interior • the presence or lack of automatic
finishes with higher flame spread and detection and alarm systems
tive, if the expected perfor- smoke production characteristics than The above list can be transformed,
mance of one system other occupancies or use groups. For placing any one of the bullet items at
changes, then others may be example, the 2000 International Build- the top as the dependent variable. For
ing Code (IBC) restricts the flame spread example:
required to change in order to and smoke production ratings of interior The requirements for fire detection
maintain the expected level of finish used in the egress components of and alarm in a building are based, in
prevention or protection. unsprinklered apartment buildings to part, on:

34 Fire Protection Engineering N UMBER 24

 Fire Safety Objective(s)

Prevent Fire Ignition Manage Fire Impact

Control Control
Control
Heat-Energy Source-Fuel
Fuel
Source(s) Interactions
Manage Fire Manage Exposed

Yellow events contain interior finish

subevents necessary for success.

Control Control Fire Limit

Suppress Safeguard
Combustion By Amount
File Exposed
Process Construction Exposed

Orange events contain fire detection and

Figure 1. Fire Safety Concepts Tree alarm subevents necessary for success.

• the flame spread and smoke poten- automatic suppression, fire detection The relationships and interdependen-
tial of the interior finish and alarm system improvements by cies among these various parts of bal-
• the expected/permitted occupant themselves are not likely to permit anced fire protection are complex.
load changes in interior finish, particularly in Codes typically contain one or two sim-
• occupant mobility egress paths. However, by improving ple, reliable, proven combinations of
• the maximum permitted travel dis- fire detection and several other facets of systems to achieve a fire safety objec-
tance protection, such as decreased travel dis- tive. Other possible solutions may also
• the degree of compartmentation tance, reduced occupant load, more be possible, but may incorporate more
• the presence or lack of automatic than two egress paths from a space, and complex combinations and relation-
suppression systems improved containment of fire and ships. NFPA 550, Guide to the Fire
smoke, it may be possible to use some Safety Concepts Tree, is a useful tool for
Building and fire codes specify certain combustible finishes. For instance, it examining these relationships and their
combinations of the various systems may be acceptable to use wood panel- weighted impact on fire safety.2 The Fire
necessary to meet the objective of the ing as a wainscoting in limited horizon- Safety Concepts Tree is an event tree us-
code. For example, in the 2000 IBC, As- tal exit access corridors. Or in rooms ing logical AND and OR gates to relate
sembly occupancies are permitted to (not part of the egress system) that various combinations of subevents that
have Class B interior finish in egress might normally require Class B or better lead to the top level successful event.
components when sprinkler protection interior finish, supplementary fire detec- Figure 1 is the top level of the Fire
is provided. When there is no sprinkler tion that closes fire and smoke doors Safety Concepts Tree.
protection, the interior finish is limited and initiates smoke control may allow Note that the top-level Fire Safety
to Class A. The code does not list any the use of Class C finishes. Objective is connected by an OR gate
similar tradeoffs for interior finish (circle with a plus sign in it) to the
when an automatic fire detection and subevents Prevent Fire Ignition and
alarm system is incorporated. A per- Manage Fire Impact. If probabilities
formance based analysis/design may are calculated or designated for the
permit greater latitude in combining subevents, then the OR gate dic-
various degrees of each protection or tates that the probabilities be added
prevention system. Unlike complete together to determine the probabil-

FALL 2004 www.sfpe.org 35

 ■ A Balanced Approach

Manage exposed

Limit amount exposed Safeguard exposed

Defend exposed in place Move exposed

Restrict Defend Maintain Cause Provide Provide

movement of the essential movement movement safe
exposed place environment of exposed means destination

Go to
A

Defend Provide Provide Provide Provide

Detect Signal Provide Provide
against fire structural route protected route
need need instructions capacity
produst(s) stability completeness path access

Go to
Figure 2. Manage Exposed Branch of the Fire Safety Concepts Tree A

ity of the parent event. That is why the that success is also dependent on other events that might contain detection
OR gate symbol is a circle with a plus events taking place. For instance, fire subevents are shown in green.
sign in it. In other parts of the Fire Safety protection engineers regularly Manage
Concepts Tree AND gates (circle with an Fire Impact, by Moving the Exposed. In PERFORMANCE EFFECTS
X or a dot in it) require all subevents to addition to detecting the fire, it must be
occur. Thus, the probabilities are multi- signaled to occupants and emergency Interior finish has several direct effects
plied to determine the probability of the forces, adequate egress means must be on the design and performance of fire
top-level event. provided, and a safe destination is also alarm systems. The most obvious is on
The entire Fire Safety Concepts Tree is needed. the selection and performance of audi-
too large to reproduce in this short arti- Fire detection is also a part of several ble signals. Building materials such as
cle. However, examination of the re- other Fire Safety Concepts Tree events, glass, carpet, and acoustical tiles are
mainder of the tree shows that paths though not specifically listed. For ex- tested to determine their sound absorp-
containing events related to interior fin- ample, instead of Moving the Exposed, a tion coefficients at different frequencies.3
ish fall under both the Prevent Fire Igni- design might include the event Defend A particular drop ceiling panel may have
tion event and the Manage Fire event. Exposed in Place. (See Figure 2.) One a relatively flat absorption curve. That is,
Fire detection events are found only un- element required to accomplish this is its absorption coefficients are about the
der paths leading to Manage Fire Im- to maintain a tenable environment. In same for low, middle, and high frequen-
pact. In Figure 1, the red event boxes the Fire Safety Concepts Tree, this cies. Sound that is not absorbed is re-
contain paths that eventually lead down event is titled Maintain Essential Envi- flected back into the room or space. Ma-
to a detection event box. The yellow ronment. The tree does not show terials such as glass and gypsum board
events contain paths that relate to inte- subevents required for that event. In tend to have higher absorption coeffi-
rior finish. The purple box leads to both some cases, it is useful to add addi- cients at the lower frequencies and
fire detection and interior finish events. tional subevents to understand what is lower absorption at the higher frequen-
The Fire Safety Concepts Tree explic- required for an event to be successful. cies. Thus, they tend to reflect more of
itly lists three Detect Fire events that lead One subpath might include fire detec- the high-frequency sound. Carpeting
upwards in the tree to the Manage Fire tion AND closing dampers. Another and some acoustical tiles absorb high
Impact event box. However, they all subpath might be fire detection AND frequencies much more efficiently than
connect through AND gates. This means pressurization of a space. In Figure 2, lower frequencies.

36 Fire Protection Engineering N UMBER 24

 With tone-only signals, materials that cepts Tree, Copyright© 2002, National
are acoustically hard with respect to Fire Protection Association, Quincy, MA
higher frequencies cause a large amount 02269. This reprinted material is not the
of the higher-frequency sound energy to complete and official position of the
bounce off of the surfaces and fill a NFPA on the referenced subject, which is
space. This helps distribute the audible represented only by the standard in its
signal in a space, particularly when us- a variety of conditions, including light entirety.
ing today’s higher-frequency appliances and dark surfaces with high and low
that operate at approximately 3000 Hz. ambient lighting. Thus, NFPA 72 require- REFERENCES
High-frequency signals, such as those ments should cover most situations in
generated by the piezo electric transduc- small spaces. However, the effect of 1 International Building Code, International
ers used in many modern fire alarm au- high ambient light with bright interior Code Council, Falls Church, VA, 2000.
dible appliances, tend to be more direc- finishes in large spaces has not been
2 NFPA 550, Guide to the Fire Safety
tional than lower frequency signals.4, 5 studied and may be beyond reasonable Concepts Tree, 2002 edition, National Fire
Thus, they are not as loud at one side of extension of the UL test results. Protection Association, Quincy, MA.
the device as compared to in front of the Interior finish can also directly impact
3 Everest, F.A., The Master Handbook of
device. In acoustically hard environ- the success of a fire detection system. To
Acoustics, 3rd ed., TAB Books, 1994.
ments, directionality is less important be successful, a fire detection and alarm
because of reverberation. On the other system must do its job before an attack 4 Russell,D., Radiation from a Baffled
hand, acoustically soft surfaces absorb by fire causes the system to fail. If cir- Piston, Kettering University Applied
more of the fire alarm sound energy and cuits pass through spaces with com- Physics, 2002, www.kettering.edu/
~drussell/Demos/BaffledPiston/BaffledPist
help prevent it from reverberating and bustible interior finish or combustible
on.html.
filling a space. Therefore, in acoustically occupant-related goods but without fire
soft spaces, directionality may affect detection, they can be attacked, and 5 Directivity, Huisinga and Olsen
loudness in some locations. So, building they may fail before successful fire de- Publishing, Willmar, MN 56201, 2004,
www.soundinstitute.com/article_detail.cfm
materials and interior finish that absorb tection takes place. Similarly, even
/ID/138
high frequencies may require the use of where detectors are present, if the sys-
a larger number of audible appliances. tem does not remain operational long 6 NEMA Supplement, “Speech
They are spaced to ensure that listeners enough to complete its mission, failure Intelligibility,” Fire Protection Engineering,
are always located in the direct field of occurs. Thus notification appliance cir- Society of Fire Protection Engineers, Issue
No. 16, Fall 2002.
the sounder. cuits may need to be protected and ad-
When voice is used as the fire alarm ditional fire detection may be warranted. 7 NFPA 72, National Fire Alarm Code,
signal, acoustically hard surfaces help Circuits installed on or directly behind National Fire Protection Association,
maintain audibility of the voice but combustible finishes may not survive Quincy, MA, 2002.
cause reverberation, decreasing the in- long enough to do their job. 8 Devoss, F.,“Report of Research on
telligibility of the voice message.6 That is Fire prevention and fire protection Emergency Signaling Devices for Use by
one reason why it is necessary to mea- strategies require that many different the Hearing Impaired,” Underwriters
sure the intelligibility of voice systems, systems work together (balance), and Laboratories, Northbrook, IL, March 20,
not just the audibility. In acoustically in some cases, they must be coordi- 1991.
soft spaces, any reverberation is gener- nated so that they do not interfere with 9 NEMA Supplement, “Engineering Failure
ally at a much lower energy level (not as each other (performance).9 While the or Failure to Engineer?,” Fire Protection
loud). Therefore, persons generally relationship between interior finish and Engineering, Society of Fire Protection
need to be in the direct field of a fire detection and alarm systems is a Engineers, Issue No. 17, Winter 2003.
speaker to receive an intelligible mes- relatively weak one, they still impact
sage since the volume will be too low at each other. Stronger relationships exist
other locations. between interior finish and egress, and
Editor’s Note – About This Article
The color of interior finish and the between fire detection and smoke con-
ambient lighting of a space affect the trol, for example. This is a continuing series of articles
signal-to-noise ratio of fire alarm strobe NFPA’s Fire Safety Concepts Tree pro-
that is supported by the National
lights. The current requirements of NFPA vides a logical graphical format to un-
72 (soon to be adopted directly as part derstand the many facets of fire safety Electrical Manufacturer’s Association
of the Americans with Disabilities Act and the relative impact each may have (NEMA), Signaling Protection and
Accessibility Guidelines) are based on on a stated goal. In some cases, it is Communications Section, and is
research by Underwriters Laboratories useful to expand the tree to provide ad- intended to provide fire alarm industry-
(UL) using indirect signaling in relatively ditional detail and analysis capability. related information to members of the
small spaces such as classrooms and of-
fire protection engineering profession.
fices.7, 8 The tests and resulting guide- Figures reprinted with permission
lines for strobe signaling were based on from NFPA 550-2002, Fire Safety Con-

FALL 2004 www.sfpe.org 37

 Products/Literature
Prepiped Mist Protection Beam Smoke Detectors
The new MCC (Mist Control Center) Single-ended, reflected-type Beam
AquaMist® Delivery System is a com- Smoke Detectors, featuring a four-
pletely prepiped and prewired pump wire design, protect open areas with
skid assembly that includes the pump, high ceilings. Alignment is accom-
controller, control valves, deluge plished via an optical sight and a
valve, control panel, and system built-in, two-digit signal strength
strainer. Compact enough to fit meter. Available with an integral sen-
through a standard doorway, it is designed to install easily in a single sitivity test module, the detectors provide 16- to 328-foot protection
step. The patented fine mist delivery system provides an alternative to coverage (in temperatures ranges from -22°F to 131°F) to give early
gaseous, foam, and heavy density sprinkler systems. warning in environments where temperatures reach extremes.
www.tyco-fire.com www.systemsensor.com
—Tyco Fire & Building Products —System Sensor

“Stand-Off” Hanger Fire Alarm Power Supplies

TOLCO™ introduces the Figure 28, a Two new fire alarm power supply products
“stand-off” hanger and restrainer for CPVC – the HPF602ULADA and the HPF902ULA-
plastic pipe that has a new one-piece DA – connect to virtually any fire alarm
design for fast and easy installation. control panel to provide the additional
Designed to reduce installation time by power needed for notification appliance cir-
eliminating the need for wood blocking cuit expansion. They cost-effectively deliver
extensions. Flared edges protect plastic significant power and built-in synchroniza-
pipe from scratches and abrasions. UL-list- tion protocols for many common strobes.
ed for installations into 3/8-in. thick composite wood. Available in sizes www.honeywellpower.com
from 3/4 in. through 2 in. —Honeywell Power Products
www.tolco.com
—TOLCO

38 Fire Protection Engineering N UMBER 24

 FIRE PROTECTION

B R A I N T E A S E R
Sales
Offices
HEADQUARTERS

Yahtzee® is a game played with five six-sided dice. Players take turns TERRY TANKER Publisher
rolling the dice, trying to get certain combinations of 1s, 2s, 3s, etc. Players 1300 East 9th Street
Cleveland, OH 44114-1503
may roll the dice up to three times during each turn and are permitted to 216.696.7000, ext. 9721
set aside any subset of the five dice after each roll. fax 216.696.3432
ttanker@penton.com
A player rolls the following combination on the first roll: 2, 2, 3, 4, 5. If
NORTHEAST
the player keeps the 2, 3, 4 & 5, what is the probability of obtaining a TOM CORCORAN District Manager
“large straight” (the numbers of all five dice fall in a consecutive sequence) 929 Copes Lane
in the two remaining rolls? West Chestor, PA 19380
610.429.9848
fax 610.429.1120
Solution to last issue’s brainteaser tomcorcoran@penton.com

Substitute a unique integer from 1 to 9 for each different letter in the NORTH CENTRAL
subtraction problem below. JOE DAHLHEIMER District Manager
1300 East 9th Street
FI RE Cleveland, OH 44114-1503
-HE A T 216.696.7000, ext. 9279
OUT fax 216.696.3432
jdahlheimer@penton.com
There are at least three solutions:
CENTRAL / WEST
2598 AMY COLLINS District Manager
-1834 3240 Shadyview Lane North
Plymouth, MN 55447
764 763.404.3829
fax 763.404.3830
6198 acollins@penton.com
-5824
SOUTHEAST
374
TOM GRAVES District Manager
9126 1180 Pin Oak Court
-8673 Cumming, GA 30041
770.205.1870
453 fax 770.205.1872
tgraves@penton.com

39 Fire Protection Engineering N UMBER 24

 from the technical director

The Continuing Need for Fire Research

broadly applied in engineering practice. improved understanding of the science

This differs from “testing,” which is fre- of fire, engineers could safely reduce ex-
quently applied to solve specific prob- cess conservatism while still providing
lems and which is typically not readily an appropriate level of safety. Addition-
applied in a general sense. ally, continued research would expand
Underlying the decline in research the types of problems that fire protection
productivity has been a decline in fund- engineers could solve.
ing of national government laboratories. The Society of Fire Protection Engi-
Several governmental fire laboratories neers has focused some efforts in coun-
have been privatized, with the result that tering this trend. SFPE held a workshop
funding for research must be sought in 1999 to develop a research agenda for
from the private marketplace, where in- the fire protection engineering profes-
terest tends to favor testing over funda- sion (available from www.sfpe.org/
mental research. Public fire laboratories sfpe30/pdfsanddocs/pbdfr.pdf). While
that were not privatized have suffered a broad in scope, this research agenda in-
diminishing level of funding from their dicates the types of research that engi-
Morgan J. Hurley, P.E. national governments. For example, neers could use to benefit society.
Technical Director government-appropriated fire funding at One of the conclusions reached dur-
Society of Fire Protection Engineers the U.S. National Institute of Standards ing the development of the research
and Technology (NIST) rose from ap- agenda was that the public sector alone
proximately four million dollars in 1974 will likely not return to the state of re-
to about seven million dollars in 1999. search funding during the mid- to late
While this may seem like an increase, it 20th century. Public/private partnerships
actually represents a decrease in pur- will be necessary to increase the amount
chasing power of approximately 50% of research funding that is available.
due to a decline in the value of the dol- While not identified in the research
lar. While NIST has received additional agenda, the first step is the development

B
eginning with the publication funding to analyze the building failures of a sound business plan to attract fund-
of the SFPE Handbook of Fire that occurred on September 11, 2001, ing to support fire research.
Protection Engineering and this funding increase may be only SFPE has also directly supported re-
continuing with the publication of per- temporary. search through its Educational and Sci-
formance-based codes and several Fire protection engineering is the entific Foundation. The Educational and
engineering design guides, the fire bridge between fire research and the Scientific Foundation has historically
protection engineering profession has built environment. A fundamental tenet supported a number of fire research pro-
matured tremendously over the past of engineering is to do the best job pos- jects, typically conducted at academic in-
decade-and-a-half. Underpinning this sible with the information that is avail- stitutions. Funding for this support has
advancement is a foundation of fire able, and despite declining research pro- come from contributions from SFPE
research. However, a great deal of this ductivity, fire protection engineers will members and chapters. Additionally, the
research was conducted in the 1950s continue to apply the knowledge avail- Foundation is currently exploring mech-
through the mid-1980s. While quality able to protect people and property anisms to expand its support.
research continues, significantly less from fire. When faced with a less-than- While relatively modest in magnitude,
funding is available to support this total understanding in an area of prac- the Educational and Scientific Founda-
research, and hence, much less is tice, engineers typically compensate by tion made valuable contributions since its
being conducted now than before. building in conservatism. This excess 1979 inception, and this support has the
Before proceeding further, it is useful conservatism translates into higher de- potential to grow. A sound foundation of
to define the term “research” as used sign costs, which are ultimately passed fire research allows fire protection engi-
here. “Research” refers to a scientific in- on to the public through higher overall neers to provide the best possible service
vestigation which has results that can be costs of products and services. With an to the public, clients, and employers.

40 Fire Protection Engineering N UMBER 24