P R O V I D I N G P R A C T I C E - O R I E N T E D I N F O R M AT I O N T O F P E s A N D A L L I E D P R O F E S S I O N A L S

FIRE PROTECTION

WINTER 2002 Issue No. 13

Fire
Modeling page 9
ALSO:

19 EVALUATING COMPUTER
FIRE MODELS

24 THE FUTURE OF FIRE

SIMULATION AT NIST

39 THE RIGHT TOOL FOR

THE JOB

46 BRIEFING ON THE
WORLD TRADE CENTER
ATTACKS
 No one in the life safety consulting profession offers more
diverse and exciting career challenges than The RJA Group.
With our reputation as a leader and project involvement all
over the world, we’ll challenge your abilities every day while
we give you the opportunity to grow and develop.

Open your Our disciplines range from fire engineering (Rolf Jensen &
Associates, Inc.) and security (Sako & Associates, Inc.) to

mind to a construction management (Fire Protection Management,

Inc.) and training (Protection Knowledge Concepts, Inc.). So
your career path can take any direction you desire, in any of

world of career our prime urban locations throughout the country.

If you have an open mind about how far you’d like to take

possibilities. your career, contact us at 888-831-4RJA or visit our

web site: www.rjagroup.com.

Offices in major cities throughout the

United States and strategic business
alliances around the world.
 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 WINTER 2002

information to fire protection engineers and allied 9

professionals. The opinions and positions stated are COVER STORY
the authors’ and do not necessarily reflect those of SFPE. FROM PHLOGISTON TO COMPUTATIONAL FLUID DYNAMICS
All of combustion science rests at least in part on Lavoisier’s discovery in the late
Editorial Advisory Board
18th century that combustion involves reaction with the element oxygen. This article
Carl F. Baldassarra, P.E., Schirmer Engineering Corporation
reviews the steps that brought fire dynamics from there to the models of today.
Don Bathurst, P.E.
Harold E. Nelson, P.E.
Russell P. Fleming, P.E., National Fire Sprinkler Association
Douglas P. Forsman, Firescope Mid-America
Morgan J. Hurley, P.E., Society of Fire Protection Engineers 3 VIEWPOINT
William E. Koffel, P.E., Koffel Associates 4 LETTERS TO THE EDITOR
Jane I. Lataille, P.E., Los Alamos National Laboratory
Margaret Law, M.B.E., Arup Fire 6 FLASHPOINTS
Ronald K. Mengel, Honeywell, Inc.
Warren G. Stocker, Jr., Safeway, Inc. 19 EVALUATING COMPUTER FIRE MODELS
Beth Tubbs, P.E., International Conference of Building There are a growing number of computer-based models being used in support of
Officials fire safety engineering design and analysis. The author examines various types,
Regional Editors summarizes guides that help assess uncertainties, and discusses applications.
U.S. H EARTLAND Marc L. Janssens, Ph.D.
John W. McCormick, P.E., Code Consultants, Inc.
24 THE FUTURE OF FIRE SIMULATION AT NIST
U.S. M ID -ATLANTIC
How far will fire modeling have advanced in 10 years? NIST researchers forecast
Robert F. Gagnon, P.E., Gagnon Engineering, Inc.
particular challenges developers will face over the next decade and beyond.
U.S. N EW E NGLAND
Thomas L. Caisse, P.E., C.S.P., Robert M. Currey & Kevin McGrattan, Ph.D., Howard Baum, Ph.D., Ron Rehm, Ph.D., Glenn Forney,
Associates, Inc. Ph.D., and Kuldeep Prasad, Ph.D.
U.S. S OUTHEAST 39 THE RIGHT TOOL FOR THE JOB
Jeffrey Harrington, P.E., The Harrington Group, Inc.
Sometimes it makes sense to use sophisticated calculation tools to perform quan-
U.S. W EST C OAST
titative fire hazard analyses, and sometimes it doesn’t. What should you consider?
Marsha Savin, Gage-Babcock & Associates, Inc.
Frederick W. Mowrer, Ph.D., P.E.
A SIA
Peter Bressington, P.Eng., Arup Fire 46 BRIEFING ON THE WORLD TRADE CENTER ATTACKS
A USTRALIA Based on a briefing paper that was prepared two weeks after the World Trade
Richard Custer, Arup Fire Center attacks, this article describes the many factors that need to be considered
C ANADA when evaluating how a building responds to such an extreme event. Prevention
J. Kenneth Richardson, P.Eng., Ken Richardson Fire
measures are also highlighted.
Technologies, Inc.
Extreme Events Mitigation Task Force, Arup
N EW Z EALAND
Carol Caldwell, P.E., Caldwell Consulting 54 PRODUCTS/LITERATURE
U NITED K INGDOM
Dr. Louise Jackman, Loss Prevention Council 56 SFPE RESOURCES
Personnel 58 BRAINTEASER/CORPORATE 100/AD INDEX
EXECUTIVE DIRECTOR, SFPE
Kathleen H. Almand, P.E., 60 FROM THE TECHNICAL DIRECTOR
T ECHNICAL E DITOR Learning from Tragedy
Morgan J. Hurley, P.E., Technical Director, SFPE Morgan Hurley, P.E.
P UBLISHER
Terry Tanker, Penton Media, Inc.
A SSOCIATE P UBLISHER
Joe Pulizzi, Custom Media Group, Penton Media, Inc.
Cover illustration by Rick Fischer/Masterfile
M ANAGING E DITOR Online versions of all articles can be accessed at www.sfpe.org.
Angela Vannucci, Custom Media Group, Penton Media, Inc.
A RT D IRECTOR Invitation to Submit Articles: For information on article submission to Fire
Pat Lang, Custom Media Group, Protection Engineering, go to http://www.sfpe.org/publications/invitation.html.
Penton Media, Inc.
C OVER D ESIGN
Dave Bosak, Custom Media Group, Subscription and address change correspondence should be sent to: Fire Protection Engineering,
Penton Media, Inc. Penton Media, Inc., 1300 East 9th Street, Cleveland, OH 44114 USA. Tel: 216.931.9159. Fax: 216.696.7668.
e-Mail: [email protected].
Copyright © 2002, Society of Fire Protection Engineers. All rights reserved.

www.sfpe.org 1
 Naturally...
Nontoxic. Noncorrosive. Unregulated.
Unrestricted. Chemical-free.
Safe to breathe.

Call for your FREE brochure:

800-346-3626
715-735-7411

www.ansul.com
viewpoint

THE LACK OF PRESCRIPTIVE INPUTS PROBLEM IN

PERFORMANCE-
the ‘Life Safety Evaluation,’ there should
be a factor of safety, especially in view
of the incomplete technical grasp of
both egress and fire issues at the pre-
sent. For example, in a conservative

BASED DESIGNS
approach, the ‘time available’ should be
at least twice as long as the ‘time
required.’” Despite this well-documented
recommendation, one of the engineers
who participated in a design we
reviewed stated, “Jake Paul’s method
(SFPE Handbook) of doubling the occu-
By Joseph M. Fleming I have reviewed many “deterministic” pant egress time is not commonly
timed egress analyses. A key component accepted or used for fire engineering
of these analyses is the selection of ten- analysis. For almost any engineering

P erformance-based designs appear to

rely on three sources of information:
1. A Set of Objectives. (This could be
ability criteria. This selection could
increase or decrease the time available
for egress by several minutes depending
analysis, you could find someone with
an opposite analysis or result.” Do I
have the right as a code official to reject
a Performance-Based Code.) on the fire scenario. Unfortunately, many this design based on this comment? The
2. A Design Guide. (These are gener- guidelines, such as the SFPE Handbook, engineer felt that I did not since his
al rules on documentation and provide multiple choices for acceptable “engineering judgment” justified the
methodology.) tenability criteria or “factors to consider” opinion. But isn’t this type of thinking
3. Reference Material. (For example, when deciding what tenability limit to circular reasoning? An engineer’s opin-
an engineering handbook.) utilize. Even the SFPE Engineering Guide ion regarding the proper choice of
The problem with the use of this to Performance-Based Fire Protection safety factors has to be justified by
material is that it allows too much flexi- Analysis and Design of Buildings only more than just the engineer’s opinion.
bility in the selection of critical “input” states, “Visibility through smoke might
items such as performance criteria and affect the occupants’ ability to safely exit SOLUTION
assumptions about occupant behavior. A a building. The factors that affect visibili-
code official trying to use these docu- ty include the amount of particulate in It seems reasonable to give designers
ments to ensure the safe design of the path of vision and the physiological the freedom to choose different
buildings is analogous to a police officer effect on the eyes. Low light levels might designs, i.e., different outputs, to
trying to enforce a safe society by using also affect occupants’ ability to egress.” achieve the same goal. At the same
books on philosophy and theology. Utilizing these references would provide time, it does not seem reasonable to
These books may contain valuable the designer has a lot of leeway in this allow designers the freedom to pick
information as to how one should con- matter. If a code official wants 0.25 any criteria and make any assumption
duct the affairs but are almost useless as OD/m to be used, as the criteria for visi- that they can find a reference for in a
a set of enforceable rules. bility and a designer wants to use 0.5 technical handbook or peer-reviewed
These input items could substantially OD/m, how do we “referee” the dis- journal. Until a set of prescriptive inputs
alter the output of computer models agreement? can be published, it seems prudent to
and algorithms. How is a code official The proper use of safety factors is always require the following:
supposed to ensure that the output of a another area where there seems to be 1. The logic justifying the selection of
design process is safe when they have only general guidelines as opposed to criteria or assumptions should be
no method to measure the validity of rules. It does not appear that any algo- explicit. Whenever possible it
the inputs? This is not a problem with rithm or model that is currently utilized should be more than “engineering
prescriptive codes because the output is is 100% accurate. As a consequence, it judgment.”
“prescribed.” With performance-based would seem prudent to utilize safety fac- 2. An independent peer review
designs, the output is not “prescribed.” tors to offset the uncertainty. In actuali- should be provided to the code
To ensure the safety of future occu- ty, many designs that the Boston Fire official.
pants in performance-base-designed Department has seen have not utilized 3. All references used to justify
buildings, I believe that the inputs must any Safety Factor. Others have used a assumptions or criteria should be
be “prescribed.” Failure to prescribe the Safety Factor of 1.5. All of these selec- made available to the code official.
inputs could lead to either unsafe or tions were based on “engineering judg-
overly expensive designs. ment.” In the SFPE Handbook, Jake Joseph M. Fleming is with the Boston
As Fire Marshal of the City of Boston, Pauls recommends that “...in relation to Fire Department.

W INTER 2002 www.sfpe.org 3

 letters to the editor

initiate alarm notification appliances. References:

Yet the question remains: With the 1 NFPA 12, Standard on Carbon Dioxide
incorporation of these changes, are Extinguishing Systems, 2000 Edition. National
Fire Protection Association, Quincy, MA,
the standards adequate to permit the 2000.
use of carbon dioxide in areas sub-
2 NFPA 72, National Fire Alarm Code, 1999
ject to occupancy? If they are not, Edition. National Fire Protection Association,
then more evaluation is required with Quincy, MA, 1999.
the intention of revising the standard 3 Lockheed Martin, 1999. Identification of the
or restricting the use of CO2 in lethal Specific Mechanism by which the CO2
concentrations to areas not subject to System in Building TRA-648 Accidentally
Discharged.
occupancy.
3. The specific mechanism that caused
the CO2 system to discharge without
warning has been attributed to a
design defect in the manufacturer’s
I think the article “Lessons Learned
from a Carbon Dioxide System
Accident” (Issue No. 12) could have
fire panel [Lockheed Martin, 1999].3 gone even further and stressed the

I n Response to “Lessons Learned When significant defects are identi- importance of simplicity over complex-
from a Carbon Dioxide System fied, many manufacturers issue recall ity and of mechanical over electronic
Accident,” Issue No. 12. notices to ensure that not only is the schemes where robustness is of top
consumer made aware of the defect,
priority. Designers of water-based
Having been at the INEEL for about but also that the defect is “correct-
extinguishment systems have known
two years, talked and worked with per- ed.” Standards dealing with fire sup-
the value of indicating valves for over
sonnel involved or familiar with the pression agent design concentra-
a century, and it is still the most
incident, and toured the area where the tions, which, if exceeded, could be
tragic event occurred, I have learned a hazardous to occupants, should robust way of shutting off a system
few “lessons” that were not focused on include provisions that address prod- and having its status be unambiguous-
in the article. Here they are: uct defects. These provisions should ly indicated. Designers, OSHA, and the
1. Carbon dioxide (CO2) at minimum require manufacturers to immediate- NFPA 12 committee should consider
design concentrations required to ly notify users of the defective prod- the wisdom that a heat-seeking missile
suppress fires is lethal to humans. uct and initiate recalls when it is may prove not to be more reliable
Given its inherent hazard, it should determined that the defect is suffi- than a fly swatter in swatting flies. Get
not be used in areas subject to occu- ciently hazardous to endanger mechanical indicating valves in there
pancy – EXCEPT when the risk of human life. now, and don’t accept non-robust
fire is documented to be greater than 4. Standards dealing with fire suppres- “modern” substitutions!
the risk to personnel AND there are sion agent design concentrations,
no viable suppression alternatives. which, if exceeded, could be haz- Vytenis Babrauskas, Ph.D.
When considering the recent ardous to occupants, should include President,
advances our industry has made in provisions that address system config- Fire Science and Technology, Inc.
fire suppression agents (notably urations. System configurations where
water mist and Halon alternatives),
no situations which meet the excep-
tion condition come to mind.
the failure of a single component
could result in hazardous concentra-
tions to occupants should be prohib-
I n the Fall 2001 issue (No. 12) of
Fire Protection Engineering, the arti-
cle “UL 2360, A New Test for Wet
2. When it is evident that adherence to ited (some “approved” gaseous agent Bench Plastics”1 by Jane Lataille con-
a national consensus standard is not reserve manifolds would not meet tains a number of inaccuracies and
sufficient to preclude a fatal accident, this provision). omissions that could misinform read-
that standard must be evaluated to A final thought. As fire protection ers about the state of flammability
determine its adequacy. Immediately engineers, we need to be continually testing for plastics used in semicon-
following the tragic event, NFPA 12, scrutinizing our engineering approach ductor clean rooms.
Standard on Carbon Dioxide to the fire problem. To quote one of my Ms. Lataille notes in the article that
Extinguishing Systems (2000 edition)1, engineers, just because “we can” do Factory Mutual Research introduced
was changed to require the addition something doesn’t mean “we should.” fire tests to determine the acceptability
of an isolation valve to enable the of plastics for use in the semiconduc-
system to be physically isolated. Stephen Thorne, P.E. tor industry. These tests, known as the
Additionally, the appendix guidance INEEL Fire Marshal Factory Mutual Research Test Standard
in NFPA 72 (1999 edition)2 was modi- The opinions expressed above are my 49102 (FM4910), were released in 1997
fied to suggest the use of devices own and do not necessarily reflect and included both small-scale tests
which detect the flow of agent and those of my employer. and an intermediate-scale (parallel

4 Fire Protection Engineering N UMBER 13

panel) test adopted in concept several between the small-scale test method- article, without explanation, was
years later by Underwriters ologies. developed by Factory Mutual Research
Laboratories (UL 2360)3 as discussed in The FM49102 small-scale test uses a as an intermediate-scale test, which
the article. However, Ms. Lataille inac- “vertical” sample of plastic material to would provide an appropriate simula-
curately characterizes Factory Mutual characterize the heat release rate of a tion of the fire hazard in semiconduc-
Research’s small-scale test as one material under conditions in which fire tor clean rooms. This was demonstrat-
which plastics manufacturers are not propagation is supported by the heat ed in full-scale wet bench tests at
able to run “because the nonstandard flux generated by the material’s own Factory Mutual Research.10
test apparatus is not readily available flame. Forty-percent oxygen is used to
and the test results were not repro- simulate the high flame heat fluxes Robert G. Bill, Jr., Ph.D.
ducible.” She adds, “In addition, the that are developed in large-scale tests.8 Assistant Vice President and Director,
test is complex and expensive.” In contrast, in the UL 23603 small-scale Materials Research
As I will explain below, in sharp con- test, the heat release rate is obtained in Factory Mutual Research
trast, FM4910 provides a reproducible, a test with a “horizontal” sample of
standardized, technically sound, and plastic material in normal air with a References:
documented method of selecting non- fixed imposed external heat flux. This 1. Lataillle, J., “UL 2360, A New Test for Wet
fire propagating materials for use in fixed heat flux is not necessarily char- Bench Plastics,” Fire Protection Engineering,
No. 12, 30-33, Fall 2001
semiconductor clean rooms. acteristic of the heat flux that would
The small-scale test apparatus gen- be generated by the material’s own 2. Test Standard, “FMRC Clean Room Materials
Flammability Test Protocol,” Class Number
erating the results alleged by Ms. flame in a large-scale fire. Thus, the FMRC 4910, Factory Mutual Research
Lataille to be nonreproducible is the generated heat release rate may not be Corporation, Norwood, MA, 1997.
apparatus known to many as the appropriate. 3. Standard for Safety, “Test Methods for
FMRC Flammability Apparatus. This This may be seen in the data present- Determining the Combustibility of Plastics
apparatus has been used to generate ed by Ms. Lataille. Table 11 indicates Used in Semi-Conductor Tool Construction,”
UL 2360, 1st Edition, Underwriters
the data in the SFPE Handbook of Fire that clear PVC, using the UL classifica- Laboratories Inc., Northbrook, IL, May 2000.
Protection Engineering, Section tion scheme of Table 4,1 is a limited
4. Tewarson, A.,”Generation of Heat and
3/Chapter 4, “Generation of Heat and propagating material (i.e., flame propa- Chemical Compounds in Fire,” SFPE
Chemical Compounds in Fires.”4 It is gation in the parallel panel is less than Handbook of Fire Protection Engineering,
fair to say that this is one of the most 8 ft.). Indeed, clear PVC just misses by Section 3/Chapter 4, 2nd edition, NFPA
Publications, Quincy, MA, 1995.
widely used sources for flammability 0.2 ft being classified by the UL parallel
properties of materials. Moreover, the panel as a nonpropagating material. 5. ASTM E2058-00, Standard Methods of Test
for Measurement of Synthetic Polymeric
apparatus is considered of sufficient However, the UL 2360 small-scale test Material Flammability Using a Fire
technical soundness to be recognized classifies this material as a propagating Propagation Apparatus (FPA). The
as the ASTM E-2058 Fire Propagation material, with the UL index being more American Soc. For Testing and Materials,
West Conshohocken, PA, 2000.
Apparatus5 and recognized in the than double the requirements to be
NFPA 287, Standard Test Methods for considered as nonpropagating. 6. NFPA 287, Standard Test Methods for
Measurement of Flammability of Materials in
Measurement of Flammability of Testing of this material at Factory Cleanrooms Using a Fire Propagation
Materials in Cleanrooms Using a Fire Mutual Research has indicated that this Apparatus (FPA), National Fire Protection
Propagation Apparatus (FPA).6 The inconsistency would not occur under Association, Quincy, MA, 2001.
apparatus also is available in a com- FM4910, as both the parallel panel test 7. ASTM E-1354, Standard Test Method for
mercial version from Fire Testing and the FM4910 small-scale test would Heat and Visible Smoke Release Rates for
Materials and Products Using an Oxygen
Technology Ltd. (FTT) and is used by classify the material as nonpropagat- Consumption Calorimeter, The American
Factory Mutual Research’s Approvals ing. Similar results have been observed Soc. For Testing and Materials, West
division for its certification testing. for other materials and reported by Conshohocken, PA, 1999.
Although the article asserts to the Factory Mutual Research.9 8. Tewarson, A., Khan, M., Wu, P., and Bill, R.,
contrary, it is also important to point Another correction to note in the “Flammability Evaluation of Clean Room
Polymeric Materials for the Semiconductor
out that test complexity and expense article is the explanation for the equa- Industry,” Fire & Materials, 25, 31-42, 2001.
are not an issue compared to UL 2360.3 tion for the Thermal Response
9. Bill, R., Wu, P., Tewarson, A., and Khan, M.,
The parallel panel test in UL 2360,3 Parameter (TRP). It is not derived from “Characterizing the Vertical Fire Propagation
adapted from the Factory Mutual equations for the flame height and Propensity of Materials Through Small-Scale
Research test, is virtually the same as flame propagation rate. Rather the TRP Flammability Testing,” Proc. Fire & Materials
that in FM4910,2 and the tests conduct- is derived from the temperature 2001, 209-219, 7th International Conf.,
Interscience, London, January 22-24, 2001.
ed with the cone calorimeter used in response of a material under thermally
10.Wu, P., “Parallel Panel Fire Tests for
the UL 23601 are similar in many thick heating conditions. For further Flammability Assessment,” Proc. 8th
respects to the small-scale ASTM E- details, please refer to reference 4. International Fire Science & Engineering
20585 tests in FM4910.2 There are, how- Of final clarification, the parallel Conf., Edinburgh, Scotland, 605-614,
ever, important technical differences panel test mentioned frequently in the Interscience, London, June 29 – July 1, 1999.

W INTER 2002 www.sfpe.org 5

 flashpoints
fire protection industry news

Spivak Receives Free, Revised ESFR New Award for

W.T. Cavanaugh Award Sprinkler Datasheet Environmental
Awareness in Fire
W. CONSHOHOCKEN, PA – The JOHNSTON, RI – FM Global, the
2001 William T. Cavanaugh Memorial developer of Early Suppression, Fast- Protection
Award was recently presented to Dr. Response (ESFR) sprinkler technology
Steven M. Spivak, professor emeritus for warehouse fire protection – particu- ASHLAND, MA –
and immediate past-chairperson of the larly high rack storage – has released a Kidde plc, in associ-
Department of Fire Protection newly revised, easier-to-understand ver- ation with the U.S.
Engineering at the University of sion of its Property Loss Prevention Environmental
Maryland in College Park, MD. Data Sheet 2-2 Installation Guidelines Protection Agency
Established in 1987 in memory of the for Early Suppression, Fast-Response (EPA), has created
late William T. Cavanaugh, ASTM chief Sprinklers. FM Global is making this the annual David
executive officer from 1970 until his data sheet, written for fire protection Ball Award for Fire
death in 1985, the award is the highest system designers and installers, avail- and the Environment. The award will
recognition ASTM gives to a person or able for free to address all-too-common recognize the significant contribution
persons of widely recognized eminence misunderstandings about ESFR technol- made by an individual, team, or organi-
in the voluntary standards system. ogy and the increased international use zation to furthering the understanding
Dr. Spivak has been on the faculty of of such sprinklers. of the impact of fire or developing
the University of Maryland since 1970; The revised and reorganized instruc- improved methods of fire control that
he has been with the Department of tions contain more in-depth explana- benefit or protect the environment. The
Fire Protection Engineering since 1992. tions and illustrations for proper instal- first award will be presented at the
The prior 20 years, he worked in textile lation of ESFR sprinklers. The data Fourth Earth Technologies Forum in
fire research, fiber science, and poly- sheet also emphasizes the challenges Washington, DC, March 25-27, 2002.
mer/chemical engineering. He has tak- that can impact the effectiveness of The award is named in honor of Dr.
en three sabbaticals involving standards ESFR sprinklers including roof height, David Ball, a 25-year employee of
work, one at the General Services structural members, lighting, HVAC, Kidde, who died in July 2001. David
Administration and two at the National storage heights, and certain types of was Kidde’s research manager and was
Institute of Standards and Technology. commodities. well known in the global fire protection
For a free copy, visit www.fmglobal. industry as a passionate advocate of
com/esfr, or call FM Global Customer environmental awareness.
Services at 781.255.6681. More more information, contact
Stephen Summerill at 304.728.3489
([email protected]).

Sales Offices FIRE PROTECTION

HEADQUARTERS NORTH CENTRAL SOUTHEAST

TERRY TANKER Publisher JOE DAHLHEIMER District Manager DEBBIE ISGRO District Manager
1300 East 9th Street 1300 East 9th Street 3312 Piedmont Road, Suite 250
Cleveland, OH 44114-1503 Cleveland, OH 44114-1503 Atlanta, GA 30305
216.696.7000, ext. 9721 216.696.7000, ext. 9279 404.237.5731
fax 216.696.3432 fax 216.696-3432 fax 404.237.1372
[email protected] [email protected] [email protected]

NORTHEAST CENTRAL WEST COAST

MARK D. FORKER District Manager AMY COLLINS District Manager LORA ATHERTON ALBAINY Senior Regional Manager
485 Devon Park Drive, Suite 107 11900 Wayzata Blvd., Suite 216K 1300 East 9th Street
Wayne, PA 19087 Minnetonka, MN 55305 Cleveland, OH 44114-1503
610.995.0202 952.545.2054 216.696.7000, ext. 9282
fax 610.995.0208 fax 952.545.2056 fax 216.696.3432
[email protected] [email protected] [email protected]

6 Fire Protection Engineering N UMBER 13

 Thomas Edison,
who invented the
light bulb, was
afraid of the dark.

Custom-Tailored Protection from SFPE: Call 1-800-424-9883 for FREE

information featuring costs,
eligibility, renewability, limitations
for What Keeps YOU Up at Night and exclusions of the following
SFPE-sponsored Plans:

Y ou’re an individual, with your own

hopes, dreams ... and worries: Would
Whether it’s solid Member Assistance
protection for you, Term Life Insurance
Disability Income Plan

Term Life Plan

your spouse be able to afford your for yourself and your spouse, or High-Limit Accident Plan
family’s current lifestyle if something Comprehensive HealthCare coverage for Member Assistance Plan
happened to you? Will you be around to your whole family, the SFPE-sponsored
Program is your one-stop source for Catastrophe Major Medical Plan
put your kids through college? If not,
what then? custom-tailored protection for what Medicare Supplement Plans
matters most in your life. And each plan
Unfortunately, it’s not always is available to members like you at
Comprehensive HealthCare Plan
possible to—like Thomas Edison— SFPE group-negotiated rates that fit Dental Discount Plan
invent your own instant solutions to your budget.
what keeps you up at night. Cancer Expense Plan
Rest easy.... The SFPE-sponsored
SFPE can help. As an SFPE member, Insurance Program looks out for you.
you can choose from nine quality
SFPE-sponsored insurance plans to help
safeguard you and your family.
Sponsored by:

For FREE information about any of the SFPE-sponsored Plans, call toll free 1-800-424-9883.
The Term Life, High-Limit Accident and Comprehensive HealthCare Insurance Plans are underwritten by New York Life Insurance Company, 51 Madison Avenue, New York, NY 10010. The Catastrophe
Major Medical and Dental Discount Plans are underwritten by The United States Life Insurance Company in the City of New York, Member American General Financial Group. The Member Assistance
Plan is underwritten by Unum Life Insurance Company of America. The Medicare Supplement and Cancer Expense Insurance Plans are underwritten by Monumental Life Insurance Company.
677-01
From PHLOGISTON to
COMPUTATIONAL
FLUID
DYNAMICS
By Harold E. Nelson, P.E.

D
uring the last decade, fire
models have had a major
influence on applied fire pro-
tection engineering. Modern compart-
ment fire models solve fire dynamics
equations that have been only very
recently seriously investigated. The
information developed as a result of
these investigations has helped transfer
fire dynamics methods from the
research realm to field application. Fire
protection engineering applications
now range from straightforward alge-
braic equations, such as those for
describing fire plumes, to complex
computer models that solve complex
simultaneous fire dynamics equations
iteratively.
Over the past 50 years, fire dynamics
information was derived from scientific
principles abstracted from other disci-
plines. The basis for the fire dynamics
in compartment fire models lies in the
physics relationships inherent in com-
bustion, such as heat balance and con-
servation equations.
All of combustion science rests at
least in part on Lavoisier’s discovery in
the late 18th century that combustion
involves reaction with the element oxy-
gen. This article reviews the steps that
brought fire dynamics from there to the
models of today.

THE DISCOVERY OF OXYGEN –

THE FALL OF THE PHLOGISTON
THEORY

In ancient times, fire was considered

one of the four elements, the other
three being earth, air, and water. From

W INTER 2002 www.sfpe.org 9

 the time of the early Greek philoso- 1940s. One of Britain’s leading physi- In the mid 1970s, Edward Zukoski,
phers until the late 18th century, it was cists, Sir Geoffrey Taylor of Cambridge along with his colleague T. Kubota and
held that the essence of fire is a sub- University, used his knowledge of fluid his doctoral student B. Cetegen at the
stance called phlogiston. The phlogis- dynamics to develop a system for clear- California Institute of Technology, pur-
ton theory explained most combustion ing fog at wartime airfields. Gasoline sued the entrainment phenomena fur-
reactions, at least to the degree of mea- fire plumes rising out of ditches beside ther and developed one of the earliest
surement available to the experimenters runways were used to entrain and plume equations. Their plume equation
of that day. remove fog sufficiently for returning air- was formally published in 1981 and has
At the time of the French Revolution, craft to find the fields and land. been used in fire models such as ASET,
Antoine Lavoisier, French philosopher Critical to quantifying the fire plume FIRE SIMULATOR, and CCFM.
and chemist, demonstrated the phlogis- is an analytical understanding of the Gunner Heskestad of Factory Mutual
ton theory as inaccurate. By careful amount of air (or other surrounding Research Corporation developed the
experiment and measurement, Lavoisier gases) entrained into the rising fire concept of virtual source to account for
showed for the phlogiston theory to be plume. The initial work of the defini- the fact that most accidental fires are
correct, phlogiston had to have a nega- tion of fire plumes, including the not point source in nature. Heskestad
tive mass, an untenable position. At the entrainment coefficient, appears to have then developed a coordinated set of
same time, the English scientist Joseph been published by W. Schmidt in algebraic equations on fire plumes for
Priestly isolated what he called Germany in 1941. Taylor analyzed engineering use. His work, first pub-
“dephlogisticaded air.” Lavoisier Schmidt’s work and a description of a lished in 1984, is accessible to engi-
demonstrated that Priestly’s finding was thermally driven plume was prepared neers as the chapter on fire plumes in
a substance, which he labeled oxygen. for the U.S. Government under the title the SFPE Handbook.
Lavoisier further demonstrated that oxy- of “Dynamics of a Hot Gas Rising in Others, including Philip Thomas at
gen was part of a combustion reaction Air” in 1945. The first major publication the UK Fire Research Station and Kunio
and added weight to the product. of plume theory impacting the analysis Kawagoe at the Building Research
Also in the last decade of the 18th of accidental fires, however, was the Institute of Japan, also examined and
century, another English scientist, paper titled “Turbulent Gravitational improved the fire plume equations.
Benjamin Thompson, demonstrated by Convection from Maintained and Craig Beyler, then a visiting scientist at
studying experiments that no mass Instantaneous Sources.” by B. R. the UK Fire Research Station, examined
changes are associated with tempera- Morton, Taylor, and J. S. Turner, all of 14 fire plume and ceiling jet equations
ture change and that heat is a manifes- Cambridge University, in 1956. This in 1986.
tation of motion. These discoveries described the first heat-driven plume
constituted the final criticism and equation that included entrainment, IMPACT OF RESEARCH
destruction of the phlogiston theory. plume velocity, and plume temperature. INSTITUTIONS
While most current fire models and
FARADAY’S CHRISTMAS calculations do not use the early The development of the understand-
LECTURES Morton, Taylor, and Turner equations, ing of the dynamics of fire has strong
all of the current methodologies for cal- ties to a number of institutionalized
In the 1840s, the famous scientist culating plume properties are, at least programs. It’s worthy to briefly discuss
Michael Faraday in his Christmas lec- to some extent, descendants of the some of the major participants.
tures at the Royal Institution of Great Morton, Taylor, and Turner work. The
Britain presented one of the earliest work of Morton, Taylor, and Turner The UK Fire Research Station. In
demonstrations of the interaction of ele- was a turning point that has become the UK, both the fire impact of
ments in fire dynamics. Faraday used the basis for the modern zone model. wartime attacks and the use of alterna-
the burning of a candle to demonstrate A progression of investigators carried tive materials, some of which were
many of the phenomena involved in the work forward increasing the level very combustible (such as low-density
combustion. The candle lecture includ- of understanding. In 1960, Sizuo Yokoi fiberboard linings), produced fires that
ed demonstrations of radiant energy, of the Building Research Institute of exceeded expectations. The British
liquidation of the candle wax, and cap- Japan examined both flame heights and came to realize that there was a wide
illary action. the hot gas currents of a plume flowing range of speeds of involvement and
from a window, exposing the floor extent of fire development, spread of
THE START OF MODERN FIRE above. fire between buildings, and other
DYNAMICS Bernard McCaffery of the Center for aspects of fire danger permitted by the
Fire Research, in his 1979 report on then-current bylaws and regulations.
Early Investigations of Fire experimental results, examined entrain- During the war, the British science
Plumes. Modern fire dynamics, howev- ment in the area of the flame as well as community had worked diligently on
er, finds its initiation in and following the rate of entrainment in the portion of fire prevention and control, as well as
the World War II period in the early the plume above the flame. on using fire as a weapon. In 1947,

10 Fire Protection Engineering N UMBER 13

one of the most important modern fire ings. Yokoi’s plume research has been gram was primarily focused on civil
research and development groups was used to develop the spandrel panel defense, particularly the potential
assembled: The Fire Research Station. requirements for high-rise buildings in impact of nuclear attack. The OCD
Also, the Joint Fire Research Japan. programs involved an expenditure of
Organization, created in 1946, enabled Also in the late 1950s, Kawagoe and approximately 1 million dollars annu-
funding from both the government his colleagues observed that early in ally on fire research, the vast majority
and private interests, particularly those the development of the room fire, a of which was directed at large-scale
in the fire insurance industry. reasonably clean-cut smoke layer – or nuclear attack-initiated fires. However,
The British fire research programs zone – is established. Kawagoe further work relative to fire dynamics was part
have long enjoyed a well-deserved established that in the simple situation of this effort. Extensive work was
leadership position as developers of where there is a compartment with a done at the Naval Radiological
fundamental and applied fire science single opening and no forced air, the Defense Laboratory on understanding
and engineering. Examples include the available air for combustion is limited the ignitability of materials from radi-
analysis of radiation between buildings to that which can be drawn in the ant energy sources. Much of this infor-
as a function of openings, separations, room through the opening to replace mation was published in the book The
and fire condition by Margaret Law; the mass expelled through the same Effects of Nuclear Weapons by Samuel
the venting of hot gases from large opening. The rate of air intake was Glasstone. Extensive work on
space fires by Peter Hinkley and established to be a function of the area flashover was undertaken by the
Thomas; and the study of various heat of the opening times the square root Armor Research Foundation (now
release and heat transfer means by of its height. Kawagoe mathematically Illinois Institute of Technology
Richard Chitty and Geoffrey Cox. Fire expressed these phenomena in what is Research Institute). The principle
dynamics developments at the Fire frequently referred to as Kawagoe’s investigator was Thomas Waterman.
Research Station in recent years have equation. Kawagoe’s observations and Professor Willis Labes of the IIT
concentrated on the advancement of calculations are key elements in zone Department of Fire Protection
CFD modeling. modeling. Engineering performed investigations
There is an anecdote of Kawagoe of radiant fire spread as part of the
Building Research Institute of visiting Zukoski at Cal Tech and telling OCD effort.
Japan – Japanese Fire Research him about the layer interface phenom-
Institute. In Japan, the immediate ena, which was then contrary to many OCD Sponsors National Academy
problem in the first years following the commonly held beliefs. Prior to of Science Committee on Fire
war was the reusability of structures Kawagoe’s work, many researchers Research. In the early 1950s, OCD
that had been damaged during the air believed that the turbulence in the sponsored the establishment of the
campaign. The program initially room would cause a mixing of all the National Science Foundation
addressed structural problems, both gases into one zone referred to as a Committee on Fire Research. This
with and without earthquake consider- well-mixed reactor. According to the committee was initially headed by
ations, but soon extended to the ques- anecdote, Zukoski challenged Hoyt Hottel of Massachusetts Institute
tion of what fire would do in these Kawagoe to prove his point and they of Technology and subsequently by
buildings, and which buildings would went to a nearby mechanics laboratory Howard Emmons of Harvard. In the
be safe for continued use. The where they initiated the phenomena in early 1970s, the committee sponsored
Building Research Institute of Japan a liquid tank. They pumped a light the publication of Fire Research –
created a fire research section, and density liquid into a higher density liq- Abstracts and Review, essentially the
about the same time, a separate Fire uid as an analogue to fire and clearly first U.S. science-based fire journal.
Research Institute was established in saw the layer development. The Fire Research – Abstracts and
The Ministry of Home Affairs. Understanding these two zone phe- Review was published 3 times a year
There are many contributions to fire nomena was critical to the concept of for approximately 12 years from the
safety developed in the Japanese fire zone models. early 1960s to the 1970s. Robert
research programs. In terms of fire Fristrom of Johns Hopkins Applied
dynamics and modeling, examples of RESEARCH ON FIRE DYNAMICS Physics Laboratory was editor. The
important work at the Building AND MODELING IN THE UNITED exchange of information provided by
Research Institute includes that of STATES this journal was very important to the
Yokoi and Kawagoe. spread of knowledge among the vari-
Yokoi was the first director of fire Office of Civil Defense Funds ous groups interested in fire research.
research at the Building Research Major Fire Research Program. In
Institute. Under his direction in the the 1950s and 1960s in the United Woods Hole Conference
1950s, important work was developed States, the impetus for the OCD fire Appraises Application of Science to
on flame height and fire plumes, par- research program, directed by James Fire Problem. In 1962, the NAS
ticularly those projected from build- Kerr, was the Cold War. The OCD pro- Committee on Fire Research autho-

W INTER 2002 www.sfpe.org 11

Screen shown actual size
 s
Fire Safety

Introducing the world’s first

BIG-SCREEN fire detection system.
FireFinder, the fastest intelligent fire detection system
with the biggest, easiest display screen.

It makes all the difference in an emergency. The big 6” display

on the FireFinder panel is the easiest for firefighters to under-
stand when trying to locate a fire. Because it has an unrivaled
capacity for hundreds of easy-to-read, large-font text characters,
Hazmat icons, NFPA fire safety symbols and graphic maps. All of
these features work to help fire personnel identify fire location,
hazards and occupants faster than ever. Unlike small screens with
room only for vague, abbreviated messages which can delay and
confuse fire personnel.

The FireFinder system also responds with unrivaled speed,

detecting fire and notifying occupants in under 3 seconds. And it’s the simplest intelligent system
to install and maintain. The display lets operators view the entire system and locate specific alarms
with a simple “user prompting” sequence of lighted buttons. Field programmable from a touch
screen, it makes system control and changes simple for all users. For installers, revolutionary wiring
technology can reduce wiring errors and cut installation time by up to 25%*.

What’s more, the FireFinder offers a very-low life-cycle cost. Learn how this big-screen system can
bring you big savings and big peace-of-mind. Call Siemens Building Technologies, Fire Safety Division
at 973-593-2600 or visit cerbpyro.com

* Depending on type and size of facility. © 2002 SIEMENS

 rized the Woods Hole Study in Woods efforts by eminent scientists, Long trust fund established in the consent
Hole, MA, which brought together a encouraged and eventually sponsored order of 1974 between the Federal
representation of a cross-section of sci- centers of excellence. Four such cen- Trade Commission and 25 responders
ences that potentially could contribute ters were established at Harvard; involved in the manufacture of cellular
to solutions for fire safety problems. University of California, Berkley; Johns plastics. The work of the Product
Leaders in the applied fire protection Hopkins Applied Physics Laboratory; Research Committee was undertaken in
arena were also included. The confer- and the University of Utah. Of these, the form of grants totaling $5,000,000
ence met for four weeks with the first the work at Harvard most directly over a five-year period.
week’s agenda consisting of external applies to fire dynamics and fire mod-
speakers on the sciences as they could eling. OTHER CONTRIBUTORS
be applied to fire problems. The major
recommendation of the conference Harvard Program. The Harvard It is impossible to review the contri-
was the need to establish a national program, under the direction of butions by all the other organizations
fire research program within the Howard Emmons, was undertaken in and individual researchers to the under-
Federal government. This concept, cooperation with the basic research standing, quantification, and engineer-
however, was not widely accepted program at Factory Mutual Research ing modeling of the individual fire phe-
outside of the science community. Corporation. This center concentrated nomena and their interactions that
Leaders of fire protection and preven- on the fire environment in a room and make up fire dynamics. A sampling of
tion organizations, such as the NFPA the various phenomena needing exam- these contributors include:
and the fire insurance industry, strong- ination to accurately predict the room • Factory Mutual Research
ly opposed a Federal research pro- fire conditions. Their project was titled Corporation
gram. Their position was that there “The Home Fire Project.” The Harvard • National Research Council of
was no need for new information, fire model came from this program. Canada
rather a need for applying the knowl- In the late 1970s, the National Fire • University of Lund, Sweden
edge we currently had and that the Foundation transferred the RANN pro- • University of Edinburgh, Scotland
Federal government should not gram to NBS. The RANN projects were • University of Maryland
impose itself in this arena. Fire merged with the established grant pro-
research and safety bills proposed in gram of the Center for Fire Research. FIRE MODELS
1963 and 1964 were never enacted.
The Woods Hole Conference, how- FEDERAL FIRE PREVENTION AND Impact of Livonia and Jaguar Fires
ever, had an important impact in that CONTROL ACT OF 1974 SETS UP on Modeling. In 1953, there was a
it generated a consensus in the scien- CENTER FOR FIRE RESEARCH major fire at the General Motors
tific community that there was a hydraulic transmission plant at Livonia,
potential to mathematically describe, The National Bureau of Standards MI. A few years later there was a fire,
i.e., model, fire phenomena and the (NBS), now known as the National which got out of control and exceeded
impact of the fire on the environment Institute of Standards and Technology the ability of hose streams, at the
within a building. (NIST), has had a fire research program Jaguar Automobile Production Plant in
since 1914. Prior to the creation of the the UK. In both of these cases, there
THE RANN PROGRAM TAKES Center for Fire Research (CFR), the pro- was an interest in determining if roof
PLACE OF OCD FIRE RESEARCH gram emphasis was generally not venting could have made a difference
PROGRAM directed at the dynamics of fire in a and provide protection in a manner
space. It is currently an international that would eliminate the need to erect
In the mid-1960s, the U.S. govern- leader in research and development of firewalls in these spaces.
ment established an administrative pol- the understanding of fire dynamics and Thomas and Hinkley conducted
icy that the military should shed itself the creation of fire models and other extensive experiments at the UK Fire
of research that did not directly impact fire protection tools and systems. Research Station simulating conditions
military objectives. A research program in these fires. While conducting these
entitled “Research Applied to National PLASTICS INDUSTRY SUPPORTS experiments, they noted a large-scale
Needs” (RANN) was established in the FIRE RESEARCH THROUGH version of the same type of layer
National Science Foundation, and the PRODUCT RESEARCH reported by Kawagoe. A plume devel-
fire safety aspects undertaken by the COMMITTEE oped feeding smoke into that layer fol-
Office of Civil Defense were trans- lowing the general descriptions of
ferred to the RANN organization. Dr. At approximately the same time as Morton, Taylor, and Turner, and
Ralph Long was placed in charge of the enactment of the Federal Fire Kawagoe. They then devised means of
the fire aspects. The RANN program Prevention and Control Act of 1974, the determining roof vent requirements,
was focused on research at academic Products Research Committee was based on an equilibrium between the
institutions. In addition to individual formed. This committee administered a mass flow into the upper smoke layer,

14 Fire Protection Engineering N UMBER 13

and the mass flow out the roof vents. by either David Evans of NBS or ships that would define the individual
As this work was done before the Leonard Copper, then with NBS, to phenomena involved and the interac-
common availability of computers to reflect the impact of the hot upper tion between phenomena. In 1976, it
practicing engineers, Hinkley and zone on the hot ceiling jet. reached to the point where Professor
Thomas expressed their results in the Emmons brought in help, particularly
form of nomographs. The nomographs Quintiere’s Paper “Growth of Fire Henry Mitler, to write a predictive sin-
provided the engineer with a reason- in Building Compartments.” In 1976, gle-compartment fire model. From this,
able approximation of the venting James Quintiere, then at the Center for in 1978, emerged the Harvard Model
requirements under a range of build- Fire Research, presented a paper at an Mark I. The Mark I model was crude
ing dimensions and exposure fire size. ASTM symposium marking the 75th by current standards; however, it
In a way, Hinkley and Thomas devel- anniversary of the National Bureau of demonstrated the potential of zone
oped the first compartment zone fire Standards. In this paper, he discusses compartment fire models. With the
model. the possibility of creating a fire model object of making the Harvard Model as
While both the US and UK were and outlines the variables the model comprehensive as possible, the model
familiar with and have undertaken pro- would have to solve and the algorithms was expanded and enlarged through a
jects related to both zone models and needed to be solved. He subsequently series of improved versions. The last
Computational Fluid Dynamics (CFD) generated a simple model, which he version developed at Harvard was
models, a tendency developed in the called RUNF. That model has never Mark V issued in 1983, the year of
US to use zone models while activity in been in common use, but it and Emmons’ retirement. Emmons had
the UK was more concentrated on CFD Quintiere’s paper influenced thinking hoped that the Harvard Model could
models. The rest of the world has been on the part of other researchers includ- become so complete and accurate that
similarly divided. ing Ronald Pape at IIT Research it could be used to judge the validity
Institute and Zukoski at Cal Tech. of simpler and special purpose mod-
ZONE MODELS els. This objective was not reached.
First Published Model, the IITRI After Emmons retired, the Harvard
A zone model usually divides each Model. Pape developed the IITRI Model was turned over to the National
room into two spaces or zones: an Model, the first published compart- Bureau of Standards and renamed
upper zone containing the hot gases ment fire model. It was a simple sin- FIRST. A small cadre of scientists still
produced by the fire and a lower zone gle-room compartment fire model use FIRST, but it is not generally used
containing all space beneath the upper based on the approaches proposed by in engineering applications at this
zone. The lower zone is a source of Quintiere which used Zukoski’s plume time. Jonathan Barnett and his stu-
air for combustion and is usually the equations. Pape also used correlations dents at WPI have modified FIRST into
location of the fire source. During the of burning rate data from prior work an extensively expanded model they
course of the fire, the upper zone can at IITRI for the OCD fire threat pro- call the WPI Model.
expand to occupy virtually all of the gram. The IITRI Model was published
space in the room. slightly ahead of the Harvard model. Ad Hoc Committee on Fire
In a zone model, the upper zone is Its importance lies in being the first to Modeling Passes Information
considered a control volume that demonstrate the potential of a mathe- among Modelers. During most of the
receives both mass and energy from matical model as a fire protection engi- 1980s, scientists and engineers devel-
the fire and loses energy to the sur- neering design tool. oping zone fire models met and
faces in contact with the upper zone exchanged modeling knowledge and
by conduction and radiation, by radia- First (The Harvard Model). One of discussed related algorithms at a series
tion to the floor, and by convection or the main elements of the Home Fire of open meetings called by the Ad
mass movement of gases through Project at Harvard and Factory Mutual Hoc Committee on Fire Modeling. The
openings. Some models evaluate con- Research Corporation was a detailed committee meetings were called and
ditions in the lower layer, others study scientifically measuring and ana- provided with secretarial support by
assume that the lower layer remains at lyzing the development of a fire from Robert Levine of the Center for Fire
ambient conditions. Mass is conserved, its early ignition through flashover in a Research. This committee had no offi-
accounting for mass entering or vented bedroom situation. For three years, the cial standing but met several times a
from the control volume. project conducted one well-instrument- year and circulated minutes with such
Many zone models superimpose the ed and carefully directed test each attachments as the attendees submit-
correlations developed initially by year. The ensuing year was used to ted. Unfortunately neither the minutes
Ronald Alpert at Factory Mutual analyze the data, often inviting nor any other permanent record of this
Research Corporation to describe the researchers from other fire research committee was ever published. Even
hot ceiling jet. In most of these mod- programs to join the analysis. Much so, it was an important factor in the
els, the correlation by Alpert had been attention was directed in establishing exchange of knowledge and the
adjusted using techniques developed the algorithms, equations, and relation- advancement of fire modeling.

W INTER 2002 www.sfpe.org 15

 Zukoski’s Equations Underpin neering public in 1990. The FPETOOL zone model that includes procedures
the Model ASET. Also in the time collection was eventually expanded to for predicting fire growth and smoke
period of the late 1970s into the 1980s, include a model of smoke flow in a transport from a compartment of fire
Zukoski produced a set of equations corridor, and was later extended to origin to connecting compartments. It
suitable for modeling impact of fire in add a model related to smoke filling of maintains conservation of energy,
a room. These equations were then a room remote from the room of fire mass, momentum, and species. The
used by Cooper at the Center for Fire origin as a result of smoke transferred basic conditions predicted by CFAST
Research to produce the compartment to that room from smoke in the corri- include the depth of smoke layer in
fire model we now know as ASET, dor. each compartment, smoke layer tem-
which stands for “Available Safe Egress The FPETOOL package gained great perature, and species concentration
Time.” The ASET model originally popularity in the fire protection engi- (oxygen, carbon dioxide, carbon
produced by Cooper was designed for neering field. Similar packages fol- monoxide) in the smoke layers. CFAST
mainframe computer and had, by com- lowed in other places in the world has been enhanced over its life to
parison to the ASETB approach, a such as, ASK FRS, published by the include additional capabilities such as
tedious input program. An important UK Fire Research Station, and FIRE- the impact of forced airflow or venting
event in the transfer of modeling tech- CAL, published in Australia. A survey of smoke and the prediction of other
nology from the research community in 1997 indicated that 60% of the prac- toxic species concentrations such as
to the practicing fire protection engi- ticing fire protection engineers used hydrogen chloride or hydrogen
neer was the development by W. D. the FPETOOL package as their princi- cyanide.
(Doug) Walton of ASETB. Walton took pal computerized fire dynamics calcu-
the basic equations and reduced them lator. SFPE Computer Committee
to 200 lines of simple BASIC program- Spread the Word on Models. For
ming that could be run at very high FAST, CFAST, CCFM, AND BRI2. In about 10 years starting in 1984, the
speed on the relatively slow portable 1983, Takeyoshi Tanaka of the SFPE had a committee on computers.
computers of that day, improving Building Reserch Institute of Japan This committee served as an open
accessibility to the practicing engineer. published his paper “A Model of Multi- exchange, usually meeting at NFPA
ASETB was released to the public in Room Fire Spread.” He spent the two meetings. The committee chairman
1985. preceding years as a visiting scientist was Jack Watts, with major contribu-
at the Center for Fire Research tions by Walton. This committee holds
DETACT Opens Up the Ability to expanding on his previous fire model- an important position in the technolo-
Predict Sprinkler and Detector ing work in Japan. After he returned to gy transfer of modeling and related
Response. Within the same period of Japan, work continued on his model in fire science to the practicing fire pro-
the mid-1980s, Evans combined the both Japan and the US. His efforts pro- tection engineer. These meetings had
equations for ceiling jet temperature duced the model BRI-2. the distinct value of presentations by
and velocity developed by Alpert in A Center for Fire Research team lead practicing fire protection engineers
the late ’60s with the response time by Walter Jones extensively revised who were using models relating their
index and related heat transfer equa- and extended Tanaka’s initial program experiences, including problems as
tions from Heskestad’s work to devel- to a point of development of a distinct, well as successes, to their colleagues
op the DETACT models. These models separate model. The model was titled considering their use.
are used to predict the response of Fire and Smoke Transfer (FAST) and
sprinklers and heat detectors. used the core model of Hazard Mark I. CFD MODELS
The first version of FAST was released
FPETOOL Is Developed by Nelson. in 1985. Cooper concurrently devel- Computational Fluid Dynamics
Harold Nelson gathered a series of oped the competing model, (CFD) models divide the space being
hand equations to which he added the Consolidated Compartment Fire Model modeled into many small cells (on the
ASETB Model and the DETACT Model. (CCFM). In a consolidation action, the order of hundreds of thousands to mil-
He assembled this collection in a com- two models were compared function lions). The basic laws of mass,
puter program title FIREFORM, an by function, and the approach judged momentum, and energy conservation
abbreviation for “fire formulas.” the best and most suitable was adopt- are applied in each cell and balanced
The FIREFORM program was subse- ed. To a large extent, the physics solu- with all adjacent cells. The computa-
quently expanded by Nelson to tions from FAST and the source codes tional modeling is a complex fluid
include the room compartment model structure and solver procedures from mechanics solution of both turbulent
FIRE SIMULATOR, which added meth- CCFM were adopted into a single and laminar flow derived from classic
ods for estimating the burning rate his- model given the name Consolidated fluid dynamics theory. The governing
tory of exposure fires. This collection Compartment Fire and Smoke equations are the Navier-Stokes equa-
was titled FPETOOL. The FPETOOL Transport Model (CFAST). First tions. These equations involve a set of
package was first released to the engi- released in 1990, CFAST is a multiroom three-dimensional, nonlinear partial

16 Fire Protection Engineering N UMBER 13

differential equations expressing con- escalators bent over to a point where the Fire Research Station including the
servation of mass, momentum, and the flame was virtually parallel with work by Professor Spalding were
energy. Important sub-models related the escalator steps, igniting the wood under the direction of Geoff Cox. A
to individual cells address turbulence, rails, resulting in a fierce and fast fire special fire model known as JASMINE
radiation, soot, pyrolysis, flame spread, spread up the escalators to the ticket- was developed based on PHOENIX.
and combustion. Approaches to the ing space near street level. JASMINE is now being replaced with
modeling differ among the various The development of faster comput- an improved version called JOSEPINE.
CFD models. ers has made it possible to simulate
CFD models can examine the fire complex flows using other higher-level SUMMARY
environment in much greater detail turbulence models. An alternative
than zone models. In general, CFD model to the time averaged k- epsilon The application of fire dynamics to
models are significantly more expen- turbulence model is known as Large the solution of fire problems is quite
sive to obtain and use. Important engi- Eddy Simulation (LES), which allows recent. Twenty years ago, Nelson gave
neering decisions are required in set- for the prediction of the instantaneous a lecture at the SFPE Technical
ting up the problem and interpreting values of turbulence as opposed to Sessions at an NFPA meeting and at
the output produced by the model. time-averaged valves. Drs. Howard several SFPE chapter meetings. The
However, the use of CFD models in Baum, Ronald Rehm, and Kevin lecture consisted of a series of simple
fire protection problems is increasing. McGratten have developed and algebraic equations designed to quan-
CFD models are particularly well suit- released an LES-based model for use tify fire dynamics applications. For
ed for situations where the space is on fire problems. This model, titled example one equation was Thomas’
irregular, turbulence is a critical ele- “Fire Dynamics Simulator” (FDS), has equation for determining the size of
ment, or very fine details are sought. demonstrated its ability to identify con- fire required to flashover a space. At
CFD models usually require large- ditions and phenomena not recognized the time, there was virtually no interest
capacity computer workstations or by other types of analysis, for exam- in these types of computations by
mainframe computers. Advancements ple, how fire overcame three firefight- practicing fire protection engineers. In
in personal computers and the ers in a simple row house fire in the 1985, Randy Lawson and Quintiere
improvement in the solver routines in District of Columbia. published the report “Slide Rule
the CFD models are, however, allow- The fact that FDS is free and avail- Estimates of Fire Growth.” The reac-
ing some cases to be run on high-end able on the NIST Web site while most tion was about the same as that to
personal computers. of the other CFD models have annual Nelson’s presentation of equations.
The most popular approach for sim- fee often in the tens of thousands of Then came the proliferation of the
ulation of turbulent flows in most CFD dollars will result in widespread use of portable computer and the generation
codes has been some variant of the FDS as a tool in actual fire protection of user-friendly fire models. Now fire
k-epsilon turbulence model. In this design and analysis problems. protection engineering demands the
approach, additional transport equa- Fortunately, it is a good, soundly application of fire modeling to the
tions are solved for time-averaged tur- based model with good support. problem. The trend to performance-
bulence variables (i.e., turbulent based design expects the engineer to
kinetic energy, k, and its dissipation CFD MODELING HAS BEEN LONG model a spectrum of fire scenarios.
rate, epsilon). Examples of this ESTABLISHED IN THE UK The SFPE has established a committee
approach are the Fire Research Station examining the validity of models. To
model, JASMINE, and most commer- During the early 1970s Professor D. date, the only model they have been
cially available CFD codes. While this Bryan Spalding and his team at the able to appraise is DETACT. They are
model is computationally economical, Imperial College of Science and currently examining ASET, by far the
it is only moderately accurate due to Technology in London entered into a least complex of all compartment fire
some simplifying assumptions. close relationship with the Fire models. This does not mean that it is
CFD models are by no means Research Station with the object of inherently improper to use other mod-
restricted to fire problems. They are developing a new generation of math- els; indeed, fire protection engineering
widely used in the fluid dynamics and ematical models with the treatment of practice has progressed to almost com-
combustion fields. One of the most fires in enclosures. mon use of either much more sophisti-
famous cases involving CFD models The model developed, titled cated zone models or CFD models.
was the analysis of the Kings Cross PHOENIX, was the first to use compu- The using engineer is, however, as
underground station fire in 1989. In tational fluid dynamics as a compart- always responsible for the quality of
this particular case, the model used ment fire model. As noted above, this his or her product and the appropri-
was the CFD model FLOW 3D devel- approach to modeling divides the ateness of the solution means chosen.
oped by the Harwell unit of the British space into mass numbers of cells and
atomic energy authority. The details of relates the transportation of fluids and Harold Nelson is with Hughes
the model showed how the fire in the energy between cells. The efforts at Associates, Inc.

W INTER 2002 www.sfpe.org 17

 The K- 22-Magnum
ESFR Sprinkler
Minimum Water Flows
for Maximum
Impact...
This Sprinkler has been optimized for 45
feet high buildings. The minimum system
design savings in gpm required is 252 gpm —
below the K-25!
FM Approved
Approved!
Performance Characteristics for Factury Mutual Requirements:
Single Sprinkler Pressure & Flow Requirements

Bldg.
K-14 K-25 K-22
Height (Ft.) Pressure Flow Pressure Flow Pressure Flow
30' 50 psi 99 gpm 20 psi 113 gpm 25 psi 112 gpm
35' 75psi 121 gpm 30 psi 138 gpm 35 psi 133 gpm
40' 75 psi 121 gpm 40 psi 159 gpm 45 psi 150 gpm
45' w/ 35' stor. 90 psi* 133 gpm* 50 psi 178 gpm 50 psi 158 gpm
45' w/ 40' stor. 90 psi* 133 gpm* 50 psi 178 gpm 50 psi 158 gpm
* K.14 Must Include one row of in-rack sprinklers.

Lower flows offers opportunities to reduce:

• Interior Piping • Underground Pipe
• Fire Pumps Sizes • Tank Sizes

Maximum deflector distance from ceiling (roof) = 18 inches.

Available in 155°F and 200°F

The Reliable Automatic Sprinkler Co., Inc.
Manufacturer & Distributor of Fire Protection Equipment

1.800.431.1588
www.reliablesprinkler.com
Corporate Offices United Kingdom Office China Office
Tel: 9.14.668.3470 Tel: 441.372.728899 Tel: 010.6489.3998
 E V A L U AT I N G

By Marc L. Janssens, Ph.D. COMPUTER FIRE MODELS

E volutions in fire science and technology and computing What Is a Computer Fire Model?
have resulted in a growing number of powerful mathe- A fire model is a physical or mathematical representation
matical models that are used in support of fire safety engi- of burning or other processes associated with fires.1
neering design and analysis. Good engineering practice requires Mathematical models range from relatively simple formulae
that a statement of uncertainty accompany all computer fire that can be solved analytically to extensive hybrid sets of dif-
model calculations. This article summarizes a series of standard ferential and algebraic equations that must be solved numeri-
guides that help assess uncertainties in computer fire models. It cally on a computer. Software to accomplish the latter is
also discusses application of the guides by SFPE and others. referred to as a computer fire model.

W INTER 2002 www.sfpe.org 19

 Types of Computer Fire Models necessary to determine whether the
The most commonly used computer model is suitable for the intended use,
fire models simulate the consequences i.e., the simulation of fire scenarios of
of a fire in an enclosure. Zone models interest. Model documentation pre-
as well as field or CFD models are pared according to the guidelines in
used for this purpose. Enclosure fire ASTM E 1472 contains all the elements
models have been extended to simu- needed for a proper evaluation.
late the spread of fire and smoke The second step consists of a
through multiroom structures. detailed review of the theoretical basis
A second category of computer fire of the model, and an assessment of
models predicts how materials, sys- the correctness of the assumptions that
tems, or people respond when are made and the approaches that are
exposed to specific fire conditions. used. An independent expert who has
Sprinkler and detection activation not been associated with the develop-
models fall in this category. Other ment of the model should perform this
examples are calculation methods to task. In practice, often only the model
assess the load-bearing capacity of developer has enough incentive to
structural elements and assemblies conduct such a tedious and time-con-
exposed to fire, and models that simu- suming task.
late human behavior and egress in the A model is then verified by assess-
event of fire. ing its mathematical and numerical
robustness. Verification can be per-
Uses of Computer Fire Models formed by comparing model output to
Computer fire models are used pri- analytical solutions of simple problems
marily for two purposes: reconstruc- for which such solutions exist (e.g.,
tion and analysis of a fire, and fire-safe steady-state problems), by checking
design of a structure. The former is the computer source code for irregu-
usually an easier task, because there is larities and inconsistencies, and/or by
always other information available investigating the accuracy and conver-
such as forensic evidence, eyewitness gence of the numerical solutions of
accounts, fire department reports, etc. the guides are applicable to other the model equations.
A computer fire model in this case is types of fire models, the focus is on Step four is usually based on a com-
most often used to supplement the compartment zone models. parison between model output and
other information in demonstrating experimental data, and provides an
that a particular hypothesis is or is not Evaluating the Predictive Capability indirect method for validation (step
plausible. of Fire Models two) and verification (step three) of a
ASTM E 1355 was first approved in model for the scenarios of interest
ASTM STANDARD GUIDES 1990 and slightly revised in 1992. A (described in step one). It is generally
major revision based on work per- assumed that the model equations are
ASTM subcommittee E05.39 on Fire formed at the National Institute of solved correctly, and the terms valida-
Modeling (later merged into Standards and Technology (NIST) tion and evaluation are therefore often
Subcommittee E05.33 on Fire Safety resulted in the current edition, which used interchangeably. Experimental
Engineering in 1996) initiated the was approved in 1997. The model data for model evaluation can be
development of a set of standard evaluation process, according to ASTM obtained from standard fire tests, ad-
guides in the late 1980s. Four guides E 1355, consists of the following four hoc fire tests conducted as part of the
are now available, which cover specif- steps: model development and evaluation
ic issues pertinent to computer fire 1. Define the scenarios for which process, the literature, and/or experi-
modeling: the evaluation is to be conducted. ence.
• ASTM E 1355 addresses evaluating 2. Validate the theoretical basis and Three types of uncertainties con-
the predictive capability of fire assumptions used in the model. tribute toward the accuracy of fire
models; 3. Verify the mathematical and models when quantified by comparing
• ASTM E 1472 provides guidelines numerical robustness of the model predictions with experimental
for documenting fire models; model. data. The input uncertainty is primarily
• ASTM E 1591 describes proce- 4. Evaluate the model, i.e., quantify due to the errors and assumptions for
dures to obtain input data for fire its uncertainty and accuracy. the input data. Sensitivity analyses are
models; and used to identify the critical input para-
• ASTM E 1895 addresses uses and The first step of the process consists meters, which must be specified with
limitations of computer fire mod- of a description of the fire scenarios much greater care than the parameters
els. for which the evaluation is to be con- to which the model is relatively insen-
Although many of the provisions in ducted. Sufficient documentation is sitive. A sensitivity analysis of a com-

20 Fire Protection Engineering N UMBER 13

plex model might involve a very large grammer’s guide. The technical docu- applied the ASTM guides to any of its
number of runs to assess the effect of mentation describes the theoretical and models, the contents of the ASTM
all input parameters individually and mathematical foundations of the guides are largely based on the pio-
of possible interactions between differ- model. The user’s manual provides neering work done by NIST. For
ent parameters. Special mathematical instructions for installing and operating example, the documentation for the
techniques can be used to drastically the software. Sample runs should be CFAST and FDS models covers most, if
reduce the number of computer model included to allow the user to verify not all, of the elements that are
runs without losing much information.2 correct operation of the program. The described in ASTM E 1472.9, 10 Several
The model uncertainty is primarily due programmer’s guide includes the extensive studies have been performed
to the assumptions made by the source code and instructions for users to evaluate the predictive capability of
model, and can be quantified as a who want to customize the program. CFAST and its predecessors.11
result of the validation process (step Model documentation prepared
two of the evaluation). Full-scale fire according to ASTM E 1472 contains all FIRM
test data are subject to experimental the elements that are needed for an In 2000, the author of this article
uncertainty. Therefore, discrepancies evaluation according to ASTM E 1355. published the second edition of an
between model predictions and exper- introductory text on mathematical fire
imental data might be, at least partly, Data for Computer Fire Models modeling.12 The book describes the
due to measurement errors. There are Computer fire models typically development of a zone fire model that
procedures to determine the precision require physical, chemical, and flam- predicts the consequences of a user-
of standard test methods on the basis mability properties of materials specified fire in a single compartment.
of interlaboratory trials or round robins involved in the fire. ASTM E 1591 The model is an extension of ASET
(e.g., see ASTM E 691 or ISO 5725). describes procedures to measure many developed by Cooper at NIST. It
Custom, nonstandard, full-scale fire of these properties, and includes includes algorithms to determine flows
experiments are usually not repeated numerous references to the open liter- through a ventilation opening in one
for cost reasons. However, the uncer- ature where property values can be of the vertical walls. The model is
tainty of custom test data is probably found. ASTM E 1591 was first referred to as FIRM (Fire Investigation
comparable to that of standard full- approved in 1994, and was slightly and Reconstruction Model). The revi-
scale fire tests. Round robins of stan- revised in 2000. A major revision is in sion presented a unique opportunity
dard full-scale fire test methods have progress to extend the scope to CFD to use the ASTM guides. This book is
shown that the uncertainty of some codes and other types of computer fire the first formal application of the
measurements may be as high as models. ASTM guides to a fire model.13
± 30 %.3
There are many problems in com- Use and Limitations of Computer SFPE Task Group on Model
paring the results from fire model sim- Fire Models Evaluation
ulations to data from full-scale fire ASTM E 1895 was approved and In June of 1995, SFPE formed a Task
experiments. Some of the problems published in 1997. Several surveys Group to evaluate the scope, applica-
are due to the differences between have been published and should be tions, and limitations of computer fire
the form of the recorded experimental consulted to determine which models models. The approach used by the
data and the form needed for compar- are available.6, 7, 8 ASTM E 1895 pro- SFPE Task Group is largely based on
ison with model predictions. For vides guidance to model users on uses the ASTM guides, in particular E 1355.
example, contrary to the assumption and limitations of computer fire mod- The DETACT-QS model was selected
of pre-flashover compartment fire els, and thus facilitates the selection of as a test case. DETACT-QS is a model
zone models, there often is not a clear the model that is most suitable for a for predicting the response of detec-
and sharp change distinguishing the particular task. The document also tors and sprinklers to a user specified
lower and upper gas layers.4 provides guidance to model develop- fire. Despite its simplicity, it took five
Quantifying the agreement between a ers and to authorities having jurisdic- years to complete a draft and initiate a
calculated and measured curve of tion that review designs based on ballot. Efforts are currently under way
variable expressed as a function of model calculations. to add more comparative data and to
time also presents a major challenge. enhance or clarify some of the docu-
Researchers at NIST have made some APPLICATIONS OF THE ASTM mentation. This illustrates how difficult
recommendations, but more work is GUIDES and time-consuming it is to properly
needed to address this problem.5 evaluate the predictive capability of a
NIST Models fire model. The SFPE Task Group is
Documenting Computer Fire NIST has been at the forefront of currently also evaluating the ASET
Models computer fire model development for model.
ASTM E 1472 was first published in more than two decades. Many different
1992, and reapproved in 1998. This fire models can presently be down- RECOMMENDATIONS
guide requires that computer fire mod- loaded freely from the NIST BFRL Web
els be documented with a technical site (http://www.fire.nist.gov). Over the past decade, ASTM pub-
reference, a user’s manual, and a pro- Although NIST has not formally lished a series of standard guides that

W INTER 2002 www.sfpe.org 21

 facilitate the assessment of computer Marc Janssens is with the University Complex Fire Models,” Fire Safety
fire model uncertainties. So far, the of North Carolina, Charlotte. Journal, Vol. 30, pp. 103-136, 1998.
application of these guides has been 3. Beitel, J., “International Fire Standards
very limited. Model users in the fire Project Report, Interlaboratory Test
protection engineering community are REFERENCES Program, Proposed ASTM Standard
encouraged to apply the guides and Method for Room Fire Test of Wall and
share their experience with the SFPE Ceiling Materials and Assemblies,” PCN
1. ASTM E 176, Standard Terminology of
Fire Model Evaluation Task Group and 33-000012-31, ASTM, Institute for
Fire Standards, Annual Book of
Standards Research, Philadelphia, PA,
the ASTM E05.33 Subcommittee. This Standards, Vol. 04.07, ASTM, West
1994.
feedback is essential to improve the Conshohocken, PA.
guides and increase the acceptance of 4. Janssens, M., and Tran, H., “Data
2. Peacock, R., Reneke, P., Forney, L., and
computer fire modeling. Reduction of Room Tests for Zone
Kostreva, M., “Issues in Evaluation of
Model Validation,” Journal of Fire
Sciences, Vol. 10, 1992, pp. 528-555.
5. Peacock, R., Reneke, P., Davis, W., and
Jones, W., “Quantifying Fire Model
Evaluation Using Functional Analysis,”
Fire Safety Journal, Vol. 33, pp. 167-184,
1999.
6. Friedman, R., “An International Survey
of Computer Fire Models for Fire and
Smoke,” Journal of Fire Protection
Engineering, Vol. 4, 1992, pp. 81-92.
7. Janssens, M., “Room Fire Models, Part A:
General,” Chapter 6 in Heat Release in Fires,
Elsevier, New York, 1992, pp. 113-157.
8. Sullivan, P., Terro, M., and Morris, W.,
“Critical Review of Fire-Dedicated
Thermal and Structural Computer
Programs,” Journal of Applied Fire
Science, Vol. 3, 1993-94, pp. 113-135.
9. Peacock, R., Reneke, P., Jones, W.,
Bukowski, R., and Forney, G., User’s
Guide for FAST: Engineering Tools for
Estimating Fire Growth and Smoke
Transport, NIST SP 921, National
Institute of Standards and Technology,
Gaithersburg, MD; March 2000.
10. McGrattan, K., Baum, H., Rehm, R.,
Hamins, A., and Forney, G., Fire
Dynamics Simulator: Technical
Reference Guide, NISTIR 6467, National
Institute of Standards and Technology,
Gaithersburg, MD, January 2000.
11. Peacock, R., Jones, W., and Bukowski,
R., “Verification of a Model of Fire and
Smoke Transport,” Fire Safety Journal,
Vol. 21, pp. 89-129, 1993.
12. Birk, D., An Introduction to
Mathematical Fire Modeling, Technomic
Publishing Co., Lancaster, PA, 1990.
13. Janssens, M., An Introduction to
Mathematical Fire Modeling (Second
Edition), Technomic Publishing Co.,
Lancaster, PA, 2000.

*The SFPE Task Group and ASTM E05.33

Subcommittee are chaired by Mr. Dan
Madrzykowski of NIST (daniel.madrzykowski
@nist.gov) and Dr. Ron Alpert of FM Global
([email protected]), respectively.

22 Fire Protection Engineering N UMBER 13

 Your job’s complicated enough. Fortunately,
specifying the right candela doesn’t have to be.
Specifying the right Fire Alarm Notification Appliance has never been easier with
the introduction of Wheelock’s revolutionary Multi-Candela product line.
In fact, whatever the candela intensity required, there’s now a single product to
fulfill all requirements. Each audible and visual appliance now includes strobe settings
of 15, 30, 75 or 110 candela intensities.
It’s called SPEC1FIRE™, as in ONE spec for all your Fire Alarm Notification
Appliance needs. It will make your job less complicated, and far more cost effective in
the process.

Introducing

SPEC 1 FIRE
A Family of MultiCandela AppliancesSM
™

FEATURES
• Field Selectable Candela Settings of: 15, 30, 75 or 110cd • Satisfies Virtually ALL
Requirements for Indoor, Wall Mount Applications • Both Sync and Non Sync In ONE
Appliance • Approvals: UL 1971; New York City (MEA), California State Fire Marshall (CSFM),
Factory Mutual (FM), Chicago (BFP) and ULC • ADA / NFPA / ANSI Compliant • Synchronize
with Wheelock SM, DSM Sync Modules or Wheelock’s PS-12/24-8 Power Supply with Built-In
Sync Protocol • Backward Compatible With Other Wheelock Appliances

273 Branchport Ave. • Long Branch, NJ 07740

TEL: (732) 222-6880 • FAX: (732) 222-2588
http://www.wheelockinc.com INC.
E-Mail: [email protected] Helping People Take Action™
 The Future of
Fire
Simulation
By Kevin McGrattan, Ph.D., lenges facing fire modelers: useful to fire protection engineers and
Howard Baum, Ph.D., First, there are an enormous number researchers alike.
Ron Rehm, Ph.D., of possible fire scenarios to consider.
Glenn Forney, Ph.D., and Second, we do not have either the BLOWING SMOKE
Kuldeep Prasad, Ph.D. physical insight or the computing
power (even if we had the insight) to Even as we develop more sophisticat-
INTRODUCTION perform all the necessary calculations ed numerical algorithms to describe the
for most fire scenarios. Finally, since growth and suppression of fires, the
Scientists and engineers are often the “fuel” in most fires was never majority of design calculations will con-
asked to make predictions of the state intended as such, the data needed to tinue to address a subject for which
of technology in the future and are characterize both the fuel and the fire zone and field fire models were first
usually laughably wrong. The best environment may not be available. developed – smoke movement. Because
prognosticators get the trends right, Ten years from now, these issues smoke inhalation and carbon monoxide
but cannot possibly fill in the details. will remain. Certainly the wide range poisoning are and will remain the most
Think of Jules Verne predicting a trip of fire scenarios will persist, even wid- dangerous actors in a fire, code officials
to the moon, albeit in a projectile ening due to the constant emergence will continue to enforce regulations
decked out in lavish red velvet, of new materials and new architectural designed to ensure safe evacuation of a
manned by champagne-sipping adven- forms. Computing power will certainly building. Originally, two-zone fire mod-
turers, and shot out of a giant cannon. increase, but not to the point of allow- els were developed to predict the
Unlike Jules Verne, we dare only look ing for direct numerical solutions of descent of the smoke layer in a fairly
10 years into the future, rather than the governing equations. As models simple building, but as building geome-
100. Also, we focus on the work going focus in on the small-scale combustion tries become more complex, fire protec-
on in the Building and Fire Research processes in a fire, ever-more complex tion engineers are turning to field mod-
Lab at NIST because, even though challenges will emerge that are, for els to track the smoke in open-plan,
most agree that modeling will play an now, neglected. Fortunately, there is multilevel buildings.
increasing role in fire research, the hope. The reason is that models based Ten years from now, engineers will
nature of the models is a subject of on fundamental principles will still be interested in smoke movement
intense debate. We do not presume to improve automatically as computers from fires whose size and growth rate
speak for the entire community, and get faster and the temporal and spatial will be predefined. Current field models
we welcome the opinions of other resolution improves. In looking can handle these problems in theory,
researchers as to the direction of mod- towards the future, we need to adopt but computation times are often too
eling in the future. fundamentally sound physical mecha- long or the grid resolution is too coarse
In the Spring 2000 issue of Fire nisms that will retain an elegance and to capture important features of the flow.
Protection Engineering,1 Howard Baum simplicity over time, that will shift us The solution to this problem is faster
wrote a brief history of fire simulation from empirical to deterministic descrip- computers, better allocation of grid
in which he listed three major chal- tions of fire behavior, and that will be cells and parallelization, all of which

24 Fire Protection Engineering N UMBER 13

at NIST
are subjects of active research by com- ing. The design engineer must demon- heat transfer to pyrolyzing surfaces,
puter scientists because the application strate to the official what is being cal- and radiation from gaseous combus-
of these ideas goes way beyond fire. culated with something more than stat- tion products. All of these phenomena
What we can do now is adopt ic images or time-temperature plots. occur in both laboratory-scale experi-
models that improve automatically as Animations of smoke flow provide a ments and material test apparatus.
numerical grids become more refined. visual check of the building geometry, These processes will serve as the start-
The best example of this idea is Large grid resolution, and other features of
Eddy Simulation (LES). We have found the calculation that are difficult to con-
over roughly twenty years that good vey any other way.
simulations result from solving the
Navier-Stokes equations with as few WHAT ABOUT THE FIRE?
empirical parameters as possible on
grids with as many cells as possible. Skeptics of fire models have com-
While we always want more, we have plained from the outset that the fire is
found that very good results are not really modeled in a fire model. To
obtainable with modest calculations, a large extent, this criticism is valid,
allowing engineers running our mod- and here are two reasons why. First,
els to investigate a wide variety of most engineers are usually interested
problems without having to worry in smoke movement, so there’s no rea-
about numerical parameters for which son to model the fire other than as a
they have little training. point source of smoke and heat.
It is our belief that the LES concept Second, the combustion processes
will emerge as the prevailing method- occur at length and time scales below
ology for smoke and heat transport the resolution limits of most practical
because of its ability to render realis- calculations, so much so that informa-
tic, time-resolved animations of the tion obtained from the resolvable
flow of gases throughout a building. It scale, like an average cell temperature,
is inevitable that as computers get is useless in even the most simplistic
faster, users of CFD models will of combustion models.
demand more lifelike simulations Much of the future research in fire
rather than time-averaged or steady- modeling will focus on improvements
state images. This is particularly true of in the way small “subgrid-scale” physi- FIGURE 1. Simulation of a sample of
the fire community since the audience cal processes are modeled. Examples wood burning in the cone calorimeter
for many of the simulations are the of such processes include (but are not performed with FDS. Courtesy Simo
authorities having jurisdiction who limited to) soot formation and growth, Hostikka, VTT Building and Transport,
often have little training in fire model- combustion in vitiated atmospheres, Finland.

W INTER 2002 www.sfpe.org 25

 ing point for developing better fire
submodels because they are well-con-
trolled, relatively simple, and, most
importantly, small. Because they are
small, the calculations can be per-
formed at sufficiently high resolution
to capture the important phenomena
directly, and then the same calcula-
tions can be performed at lower reso-
lution to see how well the new algo-
rithms perform on larger-scale simula-
tions. The objective of this effort is not
to produce the most detailed descrip-
tion of the phenomena. Very detailed
submodels of most fire phenomena
exist now; the challenge is to design
an overall fire model that balances the
accuracy of each submodel. Balance
means that the level of detail incorpo-
rated into each is roughly the same.
Everyone learns in high school that FIGURE 2. Simulation of a rack storage commodity fire. The simulation predicts
adding a measurement accurate to the the growth and suppression of a fire that originates at the floor. These types of
nearest millimeter and one accurate to simulations are by far the most challenging attempted to date, and it remains to
the nearest centimeter yields a result be seen how much the relatively simple solid phase pyrolysis and suppression
that is only accurate to the nearest algorithms can be improved.
centimeter. Similarly, a fire model will
only be as accurate as the least accu- work with this description of the com- time, they were not nearly as primitive
rate of its components. bustion while the solid phase mecha- as the sprinkler spray and suppression
A good starting point for a better nisms are brought up to par. models. With the exception of the
fire model is a well-controlled test There are two advantages to this thermal activation equation, which by
apparatus, like the cone calorimeter evolutionary strategy. First, the various that time had become widely adopted,
(Figure 1). A set of solid and gas submodels, even in their primitive the water droplets emerging from the
phase models should be developed states, have been useful to FPEs for pipe, landing on the commodity, and
that would hopefully provide a reason- smoke movement and simple heat eventually interacting with the fire
able, balanced description of the inter- transfer calculations, and to introduce were by far the greatest source of
action of the fire with the test appara- the next generation to the technology. uncertainty in the simulation.
tus. In essence, this is the procedure Second, all aspects of the simulation A series of bench-scale experiments
that was followed in the development improve at the same pace – sprinklers, was conducted at NIST to develop
of the Fire Dynamics Simulator (FDS), radiation, burning objects – so that no necessary input data for the model.
a general-purpose fire field model part of the calculation looks out of These experiments generated data
released into the public domain in the place. A good analogy is classical describing the burning rate and flame
year 2000. The approach had been to sculpture. The artist transforms a block spread behavior of the cartoned plastic
model the large-scale gas phase trans- of marble into a human form by commodity, thermal response parame-
port as faithfully as possible for a painstakingly chipping away stone first ters and spray pattern of the sprinkler,
given numerical grid, and then intro- to reveal the gross outlines of the and the effect of the water spray on
duce extra features that were consis- head, arms, torso, etc., and only then the commodity selected for the tests.
tent with the detail (or lack thereof) focuses in on finer details. Consider The missing link in the analysis was
afforded by the smoke transport algo- that the most beautifully sculpted hand the spray characterization of the sprin-
rithm. In some sense, the fire itself would look ridiculous if one arm were kler itself; that is, the water was
was just another one of these extra longer than the other. assumed to leave the sprinkler in a
features. At first, the fire was a simple umbrella pattern quantified
Gaussian-distributed blob of heat INDUSTRIAL-STRENGTH FIRES only by visual observation. What made
superimposed on the numerical grid, the model work reasonably well was
then the fire was a set of hot particles A few years ago, in parallel with the fact that the spray parameters were
ejected from the burning object, and large-scale tests, the development of tweaked until a match between com-
for the time being the fire is a surface FDS turned towards the problem of puted and observed water density pat-
on which fuel and oxygen meet and fire suppression in large warehouses terns on the floor was obtained.
react infinitely fast. The emerging fire and warehouse retail stores2 (Figure 2). Hundreds of hours were needed to
model may move beyond this, but at As simplistic as the combustion and roughly characterize one fuel and one
the moment it should be possible to heat transfer algorithms were at the sprinkler because the characterization

26 Fire Protection Engineering N UMBER 13

 One Company. One Call.
Total protection. SimplexGrinnell.
We are the number one choice
in comprehensive single-source
protection for fire and life safety.

e are the one company in the

industry that takes total responsibility
TOTAL for meeting all of your needs. From
SERVICES portable fire extinguishers and sprin-
klers to fire alarm and suppression
Total systems, we’ll inspect, test, maintain
Excellence and repair any and all of your fire and
security equipment. Think of it this
Inspection way:Anything. Anytime. Anywhere .
We’ll also give you an ironclad guar-
Testing antee from SimplexGrinnell: We won’t
stop until you’re satisfied that you’ve
Preventive received the best possible service.
Maintenance
It’s all part of our total services, total
excellence approach to protecting
Central
your people and property. With
Monitoring
SimplexGrinnell as your protection
partner, you can rest assured that in
Ongoing an emergency situation, your safety
Support systems will operate properly. And
when all is said and done, what
24/7 could be more important than that?
Emergency
Service Call us today for your free services
evaluation.800-746-7539 in the U.S.
or 800-565-5400 in Canada. Or visit
our web site for more information.
www.simplexgrinnell.com/safe

© Copyright 2001, SimplexGrinnell LP Protecting People, Property and Peace of Mind

 With mounting flexibility, and a NEMA 6P rating, our outside switches can take
the heat (and cold) and still perform heroically. Day in and day out, even in the
P ® harshest conditions, underwater or above ground. For over a century, our strength
The Symbol of Protection
has been our proven record of performance, protecting businesses just like yours.
1.800.325.3936 w w w. p o t t e r s i g n a l . c o m

Remember, if it doesn’t have P The Symbol of Protection, It isn’t Potter.

 was almost all empirical – little of it there ought to be more investment in
was based on fundamental physical the measurement techniques required
models because the phenomena was for input data. The Catch-22 is that it’s
so very complex. As a result, users of hard to show benefits with little data.
the FDS model were not able to apply Understanding how various standard
it easily to other commodities and commodities burn and how they
sprinklers; a problem that persists to respond to water ought to be less
this day. empirically based than sprinkler
Sprinkler spray characterization will sprays, assuming the necessary solid
remain largely empirically based phase models are developed that
because each sprinkler has its own retain enough of the fundamental
unique design that makes predicting physics to accommodate a better
which way the water will go difficult. description of suppression, yet simple
To simulate the sprinkler spray, we enough to be used in large-scale simu-
need to know the initial distribution of lations. We discussed above the need
the droplet size and velocity. for more fundamentally based models
Measuring these quantities has proven of pyrolysis, starting with relatively
FIGURE 3. Image of a sprinkler spray to be very difficult and still very small-scale calculations of standard test
created with Particle Image expensive. The most promising tech- apparatus and eventually moving to
Velocimetry (PIV). The green arrows nique for measuring droplet size is large-scale. It is unclear how to
represent the velocity vectors of water through Phase Doppler Interferometry describe the burning of real commodi-
droplets leaving the sprinkler orifice. The (PDI) and droplet velocity through ties, which are mixtures of cardboard,
technique involves taking two Particle Image Velocimetry (PIV) (see plastics, woods, etc., other than with
photographs of the spray in rapid Figure 3). Both are nonintrusive, laser- the simple lumped parameter models
succession and backing out the velocity based techniques that require very developed to date. It is hoped that at a
from the displacement of the droplets. expensive equipment and skilled tech- minimum, we will have a way of relat-
Courtesy Dave Sheppard, Northwestern nicians with a high level of training in ing the burning rate of the fuel to the
University and Underwriters laser diagnostics. This is worrisome heat feedback to the surface based on
Laboratories. because calculations should be cheap- the thermophysical properties of the
er than experiments, or else what’s the fuel rather than simply an exhaustive
point? If high-level modeling of chal- series of experiments that are often
lenging industrial fire scenarios too expensive to perform given the
becomes more routine and starts to wide variety of fuels burning in a sin-
show potential benefits to sprinkler gle fire. This is possible now with a
manufacturers and building owners, limited number of pure fuels, liquids
especially, but hopefully this list can
be extended in the future.

THE FIRE HAS LEFT THE BUILDING

The fire models with which we are

familiar were developed to describe
residential and commercial building
fires. However, there is a different
class of models developed by the for-
est management and agricultural com-
munities designed to predict the
spread of wildland and forest fires.
These models are semiempirical and
are built upon very different assump-

FIGURE 4. Simulation of a brush fire

advancing on a house. Preliminary
calculations such as these are now being
performed to assess the feasibility of
simulating community-scale fire spread.
Here, the domain is a few hundred
meters on a side, the grid cells about 1
m. The trees serve as a drag on the
oncoming wind.

W INTER 2002 www.sfpe.org 29

 EST3 SYNERGY-ENABLED LIFE
SAFETY SYSTEM.THE BACKBONE
OFA SECURE ENVIRONMENT.
EST 1125 Bui l di ng Spr ead 9/ 5/ 01 2:
Page
13 PM
2

For over 130 years, Edwards Systems Technology

has been designing and manufacturing leading EST3 Life Safety Platform supports up to
160,000 points, making it the most flexible
edge fire alarm systems for Fortune 500 compa-
life safety controller in the business.
nies around the world. Now, we’re proud to bring
the superiority of world class fire alarm technology
The 3-SAC Security Access Control
to the security industry.
Module resides on the local rail of the EST3
Unlike fire alarm systems, security and access panel, providing synergy between fire, secu-
rity and access control equipment.
control systems are not mandated by code to
meet stringent agency life safety requirements. This large LCD display and telephone-
style keypad allows for easy management
But with a Synergy Enabled EST3 platform as the of security events.
backbone, all of your building’s systems – from fire
to access control to security and CCTV—work at Fully compatible with Genesis‘and Signature
Series devices. Shown here are: Genesis
a higher level of fault-tolerant interoperability. So Horn Strobe, Signature intelligent multiple
not only can the systems work together to fend off sensor detector, SIGA-MD passive infrared
motion detection module and SIGA-SEC2
any type of intrusion, EST3 survivability insures Security Module.
they won’t die trying.
The End Of The Spineless Solution. Adding CCTV surveillance capability
enhances the synergy of the overall system.
The EST3 System supports all types of security
devices by integrating them into our system through
Each Card Reader Controller monitors and
Signature Series Security Modules. One backbone. controls a door with both entry and exit
readers, door strikes or magnetic locks.
One source of command. One great system.
Lower Your Costs
The Access Control Database program
While Raising Your Standards. uses familiar Windows-based commands
You may expect to pay more, but the fact is, you to grant or deny cardholders access to
any combination of doors.
pay less. Less for both installation and mainte-
nance costs. Less to manage it. So don’t settle for
a spineless security, access control and fire solu- The FireWorks family of software and hard-
tion. Insist on the Synergy Enabled Life Safety ware options utilizes a four-quadrant
approach, presenting prioritized informa-
System from Edwards Systems Technology. tion in an easy-to-manage, graphical and
text format.
Just call 1-888-EST-2DAY or visit our website.

TM enabled
E D WAR D S S Y S T E M S T E C H N O L O G Y

World leader in life safety technology.

www.est.net/SecurityAccessFire
 tions than building fire models. They fuels and structures. As with any useful time-consuming the simulation. The
are closer to zone models in philoso- model, these descriptions must be vali- payoff for investing in faster comput-
phy since they are designed to run dated using experimental data and ers and running bigger calculations is
faster than real time so that they can must then be integrated into a CFD the proportional gain in realism mani-
be used the day of the actual fire. It flow solver generalized to account for fested by the images.
would be hard to believe that field an atmospheric boundary layer flow As the community at large becomes
models for community-scale fire conditioned by natural topography, accustomed to looking at various pic-
spread could be developed in ten upwind structures, and trees. tures and animations, model develop-
years’ time to be run in real time, but In addition to the numerical chal- ers will find new ways to dazzle. Up
field models are being developed lenges posed by community-scale fire to now, most visualization techniques
now, both at NIST and elsewhere, to prediction, it is often difficult or have provided useful ways of analyz-
study the behavior of large outdoor impossible to obtain meteorological ing the output of a calculation, like
fires to aid in planning efforts. and topological information in a form contour and streamline plots, without
Last year, wildland fires cost an esti- that can be used in the calculation. much concern for realism. A rainbow-
mated one billion dollars for just the The meteorological conditions driving colored contour map slicing down
firefighting. However, fires in the built the fire have to be postulated or through the middle of a room is fine
environment are generally even more derived from a mesoscale weather for researchers, but for those who are
costly. The famous 1991 Oakland Hills model with a minimum resolution only accustomed to looking at real
fire in Berkeley, CA, alone did an esti- measured in kilometers and the terrain smoke-filled rooms, it may not have as
mated $1.5 billion in damage while features obtained from a database that much meaning. Visualization in the
killing 25 people and injuring 150 oth- may or may not exist at the required next 10 years will turn towards provid-
ers. While large-scale models of fire resolution for that particular patch of ing as much information as the rain-
propagation are useful in wildland set- the earth. Fortunately, there are now bow contour map but in a way that
tings, corresponding models for com- efforts within the meteorological and speaks to modelers and nonmodelers
munity-scale (rural, suburban, or geographical research communities to alike. Take, for example, Figure 5.
urban) fire spread, i.e., fires spreading develop and maintain models and Presented are two ways of visualizing
between structures and natural fuel, databases that would be useful to the the same calculation, each figure made
are still in their infancy. Development fire community. For example, digital for a different audience. The trend in
of such models suffer from the follow- elevation data from LIDAR overflight scientific visualization is to combine
ing Catch-22 – validated models of measurements is being made increas- the features of each into one to reach
community-scale fire spread are need- ingly available and cost-effective. both groups of people.
ed because experiments on that scale A good example is smoke visibility.
are almost impossible to carry out; but PROOF BY PRETTY PICTURE Unlike temperature or species concen-
without experiments, how do we vali- tration, smoke visibility is not a local
date the models? Modelers are looked upon with quantity but rather depends on the
The main numerical problem to skepticism by the rest of the fire com- viewpoint of the eye and the depth of
community-scale fire prediction is grid munity because of the perception that field. Advanced simulators and games
resolution. Consider a square kilome- they all too often hide behind an eye- create the illusion of smoke or fog in
ter of terrain containing both structures catching image or animation without ways that are not unlike the tech-
and dry vegetation. Any field model understanding the physics underpin- niques employed by fire models to
tracking the progression of a brush fire ning the model. In fact, some have handle thermal radiation (Figure 6). It
through the area would require several started to refer to CFD as “colorful is envisioned that eventually graphics
million grid cells, which, even if clev- fluid dynamics.” This is often a fair hardware and software will play a role
erly distributed, would provide spatial assessment, but it is short-sighted. in actually computing results rather
resolution of at best a meter. Existing While the rapid improvement in visual- than just drawing pretty pictures. AHJs
large-scale models of wildland fires ization techniques has been a boon to often ask whether or not building
regard the fuel (vegetation) as continu- many in the field who use field mod- occupants will be able to see exit
ous and assume the fire to propagate els on a regular basis, within the next signs at various stages of a fire. The
as a line. Resolvable-length scales for 10 years what is now gee-whiz will fire model can predict the amount of
these models are tens to hundreds of become ho-hum. This is good because soot in each grid cell, but that doesn’t
meters. The technical challenge for the as field modeling becomes routine, the answer the question. The harder task
community-scale fire model is to discussion will be raised beyond the is to compute on the fly within the
develop a mathematical description for superficial level we are at now to a visualization program what the occu-
the ignition and burning of individual point where the quality of a simulation pant would see and not see.
trees and shrubs, and to determine fire will be judged by the spatial and tem-
spread between wildland elements and poral fidelity of its images and anima- CAN’T YOU MAKE IT GO FASTER?
structures. Such a mathematical tions. With any field model, the user
description must include fire spread by chooses a numerical grid on which to Computational Fluid Dynamics was
brand generation, transport, and sub- discretize the governing equations. The built around weather prediction and
sequent ignitions for both wildland more grid cells, the better but more aerospace design. A quick browse

32 Fire Protection Engineering N UMBER 13

 FIGURE 5. Two different ways of
visualizing a fire simulation. On the
left, contours of gas temperature are
shown superimposed on the numer-
ical grid. On the right, the fire and
smoke are shown as an orange-
colored sheet and black dots.

through any CFD

textbook will make
this point clear. Although
fire modeling borrows
many of the same physical
assumptions from weather modelers and
numerical algorithms from the aerody-
namics community, it is different in one
important way – it is practiced by rela-
tively small organizations whose engi-
neers have limited backgrounds in
numerical methods and computing.
Although many small FPE firms have
been absorbed by larger, more diversified
design and architecture firms, the typical
fire modeling effort within one of these
organizations is modest – a few engi-
neers working for a few weeks on a
given design problem with computers
not much more powerful than those
found in the home. The reason for this is
that fire protection is but one of many
features of an overall building design,
and one which is typically squeezed
when other items in the budget run in
the red.
Because of how it is practiced, fire
modeling has always emphasized sim-
plicity and efficiency. One of the first
questions that we are asked whenever
we demonstrate the latest simulation is
how long did it take to set up the case
and how long did it take to run. The
answer to both of these questions needs
to be on the order of a day or less (and
don’t tell me I can’t run it on my lap-
top!). If it’s more, then we’ve lost 90% of
our audience. This presents us with a bit
of a dilemma – how do we stay at the
forefront of CFD but still serve the com-
munity of practitioners? One way is to
design the model so that one can easily FIGURE 6. This figure is an example of how fog is used to bring more realism into
start doing simple calculations with sim- the scene. Shown is a simulated smoke plume made with the ALOFT (A Large Outdoor
ple geometries and then systematically Fire plume Trajectory) model. The plume is embedded within a fog-shrouded oil field
work up towards more elaborate appli- visualized with the commercial software package SGI Performer.

W INTER 2002 www.sfpe.org 33

Clas• sic (klas’ik) adj. 1a. Belonging to the highest rank or class; b. Serving as the established
model or standard; c. Having lasting significance or worth. 2. Formal, refined, and restrained in style.

AnaLASER® II from Fenwal

High sensitivity smoke detection for

telecommunications, data processing and
other mission critical facilities where business
interruptions are simply not tolerated.

w w w. A n a L A S E R . c o m
Active detection technology that’s thousands of times more sensitive than
traditional smoke detectors. Selectable sensitivity for installation flexibility.
Networking for up to 127 devices in one or multiple locations. Engineered
for reliability from Fenwal, mission-critical specialists for more than 30 years.

FENWAL PROTECTION SYSTEMS 400 Main Street • Ashland, MA 01721 • (508) 881-2000 Protection Systems
cations. The trouble with many types of
engineering software is that it is packed
with so many features that the learning
curve for even the simplest of problems
is too steep. Fire modeling, especially
field modeling, will advance only if
there is a large enough core of users to
justify the time and expense of develop-
ing and maintaining a very complex
computer code. Sustaining that core of
users means making the software acces-
sible to a wide audience.
Not only must the software be easy to
use, but the calculations must run as fast
as possible. Veteran CFD practitioners do
not find week-long calculations unusual,
but fire protection engineers who only
have experience with zone models find
it intolerable. Faster computers have
soothed some, but the demands for
more-detailed calculations often negate
gains made in computer speed. To keep
up with demand, the fire models will
need to exploit advances in computer
science and numerical methods that go
beyond just faster chips. Parallel process-
ing is becoming more of a reality in cer-
tain fields, but still is a few years away
for those using the current generation of
personal computers. However, in the
not-too-distant future, relatively inexpen-
sive desktop computers will come with
2, 4, or 8 processors, plus the necessary
hardware and software to make these
chips work together effectively. Also,
techniques to better distribute the grid
cells will allow for greater flexibility in
the design of simulations. One technique
that is used by many CFD packages (but
not yet FDS) is called multiblocking. An
example of how this would work is a
house in which every room has its own
numerical grid. Those rooms requiring
more spatial resolution could have finer
grids, those that don’t need it could
remain coarsely gridded. The numerical
algorithms presently used in single-block
codes will not change except there
needs to be extra logic built into the
algorithms so that information is proper-
ly communicated across block interfaces.
Such a technique is perfect for fire mod-
els because most simulations investigate
buildings with relatively simple, rectan-
gular geometries. Contrast this with the
aerospace industry where simulations are
performed on very complicated body
shapes. These models utilize numerical FIGURE 7. Several snapshots of a fire spreading through a townhouse. The fire
grids that are far more sophisticated and originates in the kitchen area (lower left), and eventually spreads throughout the house.
difficult to construct than those needed The front door (right) is assumed to be open, as are the windows on the second level.
for fire models.

W INTER 2002 www.sfpe.org 35

 WHO’S GOING TO USE FIRE MODELS? REFERENCES
The discussion thus far has focussed mainly on the applica- 1 Baum, H.R., “Large Eddy Simulations of Fire.” Fire Protection
tion of fire models to design problems. This is not surprising, Engineering, 6:36-42, Spring 2000.
since fire models have traditionally worked best when the fire 2 McGrattan, K.B., Hamins, A., and Stroup, D., “Sprinkler, Smoke & Heat
itself was considered merely a model input rather than a model
Vent, Draft Curtain Interaction – Large Scale Experiments and Model
output. However, fire models have been used as forensic tools
Development.” Technical Report NISTIR 6196-1, National Institute of
in the past, and their use as such will accelerate in the future.
Standards and Technology, Gaithersburg, MD, September 1998.
In fact, much of the work to improve fire models past the point
of just smoke movement has benefited the fire investigators 3 Madrzykowski, D., and Vettori, R.L., “Simulation of the Dynamics
moreso than the designers who are most often content to dial in of the Fire at 3146 Cherry Road NE, Washington, D.C., May 30,
a design fire rather than try to predict its growth and suppression. 1999.” Technical Report NISTIR 6510, National Institute of
The fire service in particular has traditionally been skeptical of Standards and Technology, Gaithersburg, MD, April 2000.
any type of model, usually preferring full-scale experimental data
over computer simulations. However, recent work3 with fire mod-
els to reconstruct several fire losses has moved some in the fire CITY OF LAS VEGAS
service to consider the use of fire models as training tools for fire-
fighters. If the present interest in simulation by the fire service ASSISTANT FIRE PROTECTION ENGINEER
continues, a great deal of effort will be placed on understanding
APPROX. MONTHLY SALARY: $3,769 – $5,878
the spread of fire through an entire house, not just a single room.
FILING DEADLINE: Feb. 4, 2002 @ 4:30 P.M.
Reconstructing a raging house fire goes way beyond simple
ORAL TEST: Feb. 19, 2002
smoke and heat transport because, as the fire spreads from its
point of ignition to envelop entire rooms in flame, the response QUALIFICATIONS: Bachelor’s degree in engineering, preferably in fire
of the wall materials becomes tightly coupled to the progression protection engineering. Two years of experience in plans review, including one
of the fire in a way that up to now fire models have largely year in fire protection engineering work, is desirable. Valid driver’s license on
neglected. Presently simulations of entire house fires are being the date of application. Uniform Fire Code certification and Uniform Building
performed to roughly scope out the grid resolution and physical Code Plans Examiner certification within 18 months of the date of hire.
mechanisms necessary to at least capture qualitatively the To download an application, visit our Web site at www.ci.las-vegas.nv.us
sequence of events from primary ignition to second-object igni- and click on employment opportunities. Or call us at 702-229-6315 to request
tion to room flashover to room-to-room fire spread (Figure 7). It an application through the mail. The application can be faxed to us at 702-
is difficult to validate such calculations, but at least we are starting 385-1259 or mailed to the address on the application.
to understand what we’re up against. Validation will come from EOE/AA M/F/D
more controlled single-room experiments, like the ISO 9705 room TDD# 386-9108
corner test, and from simulations of test apparatus.
Ultimately, the users of fire models, whether they be
researchers, design engineers, firefighters, or litigators, will drive
the development of new features. The challenge to developers Engineering Manager/
is to create fire models that can be used and understood by all Principal-in-Charge
of these groups. Even if some do not need the entire set of
Interested in career advancement with one of the most respected fire protec-
model features, their use of the models will speed the develop- tion and security engineering firms in the United States? If so, Gage-
ment and acceptance of them because much of the skepticism Babcock & Associates may have the opportunity for you. We’re currently
associated with modeling will diminish as more people grow looking for a dynamic engineering professional to be the Principal-in-
comfortable with the capabilities and limitations. Charge of our San Francisco regional office. This Principal-level manage-
ment position would have complete oversight to all business and engineer-
ing operations of the office, covering all of the Pacific Northwest. As a
ACKNOWLEDGMENTS Principal within the GBA management team, you will report directly to the
President of the firm.
The work described in this article is the contribution of Only candidates with a proven track record of effective management of busi-
many people at NIST and beyond. We are indebted to a num- ness operations and strong business development skills will be considered.
ber of guest researchers and post-doctoral associates who have If you get satisfaction from developing new business opportunities and
contributed to the modeling effort over the past 10 years, building a client network, as well as from the challenge of managing and
developing clients and building an office, send your résumé. Position
including Dike Ezekoye, Ruddy Mell, Javier Trelles, Francine
requires engineering registration and at least 7-10 years of proven manage-
Battaglia, Jason Floyd, and Simo Hostikka; in addition, our col- ment experience in the engineering field.
leagues at NIST for input data, validation experiments, and
Please submit your résumé and
software testing: Jason Averill, Dave Evans, Paul Fuss, Bill
salary history in confidence to:
Grosshandler, Anthony Hamins, Dan Madrzykowski, Rick
Peacock, Bob Vettori, Doug Walton, and John Widmann. Thomas W. Jaeger, P.E., President
Finally, thanks to the many users of our various fire models, Gage-Babcock & Associates, Inc.
5175 Parkstone Drive; Suite 130
for useful feedback, suggestions, and critiques. Chantilly, VA 20151-3816
Fax: 703-263-1549
The authors are with the National Institute of Standards
e-Mail: [email protected]
and Technology.

36 Fire Protection Engineering N UMBER 13

 ALARM
LEVEL

SENSITIVITY
AMBIENT
ACCLIMATE™ reduces nuisance alarms. Faced with an increase in “noise,” NOISE
such as dust or fibers, ACCLIMATE will increase its alarm level. When the
disturbance subsides, the unit becomes more sensitive. TIME

What Fire Sensor Has the Intelligence to

Make Sense of Your Changing Environment?
ACCLIMATE,™ by System Sensor, is an advanced photoelectric-thermal Quality System Sensor quality is an integral part of every fire
fire sensor with unique capabilities. ACCLIMATE continuously monitors sensor and accessory. Our fire sensors are built to stringent requirements,
its environment and automatically adjusts its sensitivity accordingly. exceeding those of regulatory agencies throughout the world.
Whether a location is an office environment or a manufacturing facility,
occupied, vacant or in transition, ACCLIMATE will always provide Service System Sensor’s dedication to meeting your needs
maximum protection. Time saving features, quality and innovation, doesn’t end at the point of sale. We have application engineers ready
coupled with System Sensor's experience and service, are the heart to help…a customer support department that is on call to support
of our field-proven designs. your needs…and, technical documentation available 24/7 through
automated FAX or CD-ROM.
Time ACCLIMATE intelligent fire sensors are your best choice
to save you time. By automatically adjusting its sensitivity, ACCLIMATE Find out why ACCLIMATE makes sense for your
eliminates the need for technicians to return to a site when property application. Call for a free E•DOCSTM CD-ROM,
use changes. In addition, on-board drift compensation offsets the effect a comprehensive resource of technical
of dust build up in the sensing chamber and reduces maintenance. information 800/927-6676.

Experience System Sensor’s intelligent fire sensor products

are based on proven expertise from millions of installations around the
world. System Sensor makes research and development a priority, with
focused engineering and well-equipped test facilities. System Sensor’s
years of experience ensure your peace of mind whether you are
specifying or installing.
By Frederick W. Mowrer, Ph. D., P.E.

INTRODUCTION

A master builder wouldn’t use a fram-

ing hammer for finish work or a tack
hammer for framing. The master

The
builder knows what tools are available

RIGHT TOOL
for different tasks, knows how to
properly use each tool, and under-
stands which tool is appropriate for
the task at hand. If not qualified for a
particular task, the master builder
either develops the skills to perform
the necessary tasks to complete the

for the
project, or, more likely, hires another
professional with the required skills
and expertise.
The practice of virtually any trade or
profession demands a similar
approach. We all have a limited set of

JOB
skills and expertise. Consequently, we
must all understand our limitations as
well as our capabilities. This demands
that we understand the appropriate
uses and limitations of the tools that
are available to us as well as our own
competencies in applying the available
tools. Where our competencies are
lacking, we need to either enhance
our skills or obtain the services of One of the first things the fire protec- • The calculation clearly demon-
people with the required skills. tion engineer must do before perform- strates achievement of the objec-
In the field of fire protection engi- ing a quantitative fire hazard analysis is tive performance criteria;
neering, the “toolbox” of models and decide what is the right tool for the job. • The calculation does not clearly
calculations available for quantitative demonstrate either achievement of
fire hazard analysis has expanded WHY FIRE MODELING? or failure to achieve the objective
greatly over the past two decades. We performance criteria;
have gone from “slide rule estimates of All of the tools discussed in this article • The calculation clearly demon-
fire growth”1 and other hand calcula- permit the calculation of one or more strates failure to achieve the objec-
tions and correlations embodied in aspects of enclosure fire dynamics. It is tive performance criteria.
many chapters of the SFPE Handbook important to keep sight of the objectives It is for this reason that calculation
of Fire Protection Engineering,2 through of such analyses. We perform calcula- methods ranging from “back-of-the-
computer-based single- and multiroom tions and other analyses to evaluate the envelope” hand calculations through
two-zone fire models,3, 4 to emerging adequacy of a proposed or existing the most sophisticated computer-based
highly detailed computational fluid component, system, or process to fulfill CFD models all have a place in the fire
dynamics (CFD) models.5, 6 its functions. In other words, we per- protection engineering toolbox. A rela-
Spurred by the ever-increasing form calculations to support design deci- tively simple hand calculation may
processor speeds and memory capaci- sions. As noted in the SFPE Guide to clearly demonstrate the achievement of,
ties of desktop computers, it is now Performance-Based Fire Protection or the failure to achieve, a specified
practical to perform detailed fire Analysis and Design of Buildings,7 the performance criterion, in which case
dynamics calculations and visualize first steps in the process involve the use of a more powerful tool would be
results in ways that were virtually establishment of fire safety goals, objec- unnecessary as well as uneconomical.
unimaginable only a few years ago. tives, and performance criteria. Ideally, we would like to think that the
While these ever-increasing computa- Once the performance criteria are more powerful the tool used for an
tional capabilities are valuable, they are established, different calculations can analysis, the smaller would be the
not needed for every job. Given the be performed to determine whether the uncertain middle area of the possible
enormous computational needs of specified performance criteria will be outcomes. The perfect analytical tool
these sophisticated calculations, it achieved by a proposed design. Some would be one that eliminates the
makes sense to make use of computa- evaluations may involve only a few uncertain middle area, such that the
tionally less-intensive tools where pos- simple calculations, while others, such results of a calculation would either
sible and save the more-detailed, com- as CFD analyses, may require millions demonstrate achievement of the objec-
putationally intensive calculations for or even billions of calculations. In gen- tive performance criteria or the failure
those tasks where the simpler tools are eral, there are three possible outcomes to achieve it. This concept is illustrated
inadequate. for a calculation within this context: in Figure 1. Unfortunately, the perfect

W INTER 2002 www.sfpe.org 39

 analytical tool does not exist, so there • The lower layer Some of the simpler models calculate
will always be uncertainty that must be • Vent flows and mechanical air entrainment into a fire plume based
considered. ventilation on a particular correlation while more
• Boundary heat transfer sophisticated models calculate entrain-
ELEMENTS OF ENCLOSURE FIRE • Target heating and response ment based on the fundamental
MODELS dynamics of the fluid flow. Similarly, in
These are the phenomena that are simple models the ceiling jet is calculat-
Before considering the analytical addressed with varying degrees of ed in terms of an available correlation,
tools available for quantitative fire haz- sophistication and detail by different while in more complex models the ceil-
ard analysis, it is useful to consider the fire models. To a large extent, the level ing jet develops as a natural part of the
elements of enclosure fires. Based on of detail with which a model treats fluid flow calculations. The simpler
observations and measurements during these elements distinguishes one model models can only be used reliably for
real and laboratory enclosure fires, it is from another. In some of the simpler conditions and scenarios where the cor-
possible to subdivide the complex and models, the user must specify some or relation is reasonably accurate, while
interrelated phenomena occurring dur- all of these elements, while in more the more sophisticated models can be
ing an enclosure fire into a number of complex models, the model calculates used with more confidence for a wider
discrete elements. As illustrated in some of these same phenomena based range of scenarios.
Figure 2, these phenomena include: on fundamental principles.
• The fire source For example, most models require TOOLS OF THE TRADE
• The fire plume specification of a fire source; only a
• The ceiling jet few attempt to calculate fire growth Over the past 25 years, considerable
• The smoke layer based on fundamental principles. progress has been made with respect to
understanding the dynamics of fires in
buildings. While models to simulate
conditions in post-flashover fires have
been around for even longer than this,
efforts to understand and model the
Increasing power Performance criteria Performance criteria elements and interactions of pre-
of analytical tools clearly achieved clearly not achieved flashover fires began in earnest during
the mid-1970s.
Not clear if Models for evaluating the dynamics
performance of enclosure fires can be classified in
criteria clearly three categories, ranging from the sim-
achieved plest to the most complex:
• Hand calculations and correlations
• Zone models
• Field models
Simple tools More powerful tools may be Simple tools Hand calculations generally include
adequate useful to reduce uncertainties adequate “closed-form” equations that can be
solved directly without the need for iter-
ation. For example, a number of corre-
lations have been developed that permit
Figure 1. Ideal representation of calculation results as a function of the increasing the calculation of flame heights, air
power of the analytical tools employed. entrainment into fire plumes, fire plume
and ceiling jet temperatures and gas
velocities, and smoke layer descent
rates in closed rooms for fires with
Ceiling jet
known energy release rates. While
called hand calculations because of
their closed form, these calculations
Mechanical ventilation

typically involve the evaluation of expo-

Fire Smoke nents, so they are usually performed
Boundary plume layer
heating
with the aid of a calculator, a spread-
sheet template, or a basic computer
program to make the process easier.
The best-known compilation of hand
Vent
flows
calculations in the United States is the
Fire FIREFORM (for Fire Formula) suite
source Lower developed at the Building and Fire
Target layer Research Laboratory (BFRL) at the
heating National Institute of Standards and
Technology (NIST). While now more
than ten years old, this suite of calcula-
Figure 2. Elements of enclosure fire models. tions is still widely used, either within

40 Fire Protection Engineering N UMBER 13

the FPETool program,3 as part of recent
versions of the FAST program,4 or sepa-
rately. Similar suites have been devel-
oped in the United Kingdom, where
ASKFRS8 was developed at the Fire
Research Station of the BRE, and in
Australia, where FIRECALC9 was devel-
oped at CSIRO.
The term “zone model” is generally
applied to the many two-zone models
that have been developed over the past
25 years. In a two-zone model, a room
is divided into at least two thermody-
namic control volumes or zones, a
smoke layer of buoyant gases and
products of combustion that develops
beneath the ceiling, and a lower layer
of relatively fresh air that remains near
the floor, as illustrated in Figure 2.
Conditions within each zone are
assumed to be uniform, and the zones
are separated by a distinct interface.
Mass, species, and energy conservation
equations are applied to each zone to Figure 3. Results of an example SOFIE calculation.
determine the average temperature and
smoke composition within each zone.
conditions and boundary conditions for sortium of European fire research labo-
While clearly an idealization of what
the calculation domain. ratories. SOFIE is a field model of the
actually occurs in room fires, the two-
Field models generally demand much type known as k-epsilon based on the
zone model has proven to be an effec-
more in the way of computational way that turbulence is addressed by the
tive tool for estimating room fire condi-
resources than zone models do. While model. Results of an example SOFIE
tions.
once relegated to advanced scientific calculation are illustrated in Figure 3.
A large number of two-zone fire
workstations, field model calculations More information about the SOFIE pro-
models with varying levels of sophisti-
can now be run on many personal gram is available at the SOFIE Web site
cation have been developed interna-
computers. Some practical applications (www.cranfield.ac.uk/sme/sofie/).
tionally. The first two-zone model to
may take days or even weeks to run, The second model being developed
enjoy widespread use among practicing
but with the multitasking capabilities of specifically for fire dynamics calcula-
fire protection engineers was the ASET
modern computers, these calculations tions is known as the FDS (for Fire
(for Available Safe Egress Time)
can be performed in the background Dynamics Simulator) model. The FDS
model,10 which calculates the descent of
while a computer is used for other tasks. model is being developed at the
a smoke layer within a single closed
They can also be performed over eve- Building and Fire Research Laboratory
room as a result of gas expansion and
nings and weekends, providing full em- at NIST. FDS is based on the concept of
air entrainment into an axisymmetric
ployment for otherwise idle computers. Large Eddy Simulation (LES). A recent
fire plume. Today, the most widely
A number of general-purpose com- paper by Baum in Fire Protection
used two-zone model is probably the
mercial CFD models have been applied Engineering11 provides an overview of
FAST (for Fire And Smoke Transport)
to fire calculations. These models gener- the FDS model. The FDS model is
model,4 which calculates smoke layer
ally require some modification to deal rapidly gaining popularity among prac-
conditions in a multiroom building.
with some of the unique aspects of fire ticing fire protection engineers. One
Both of these models were developed
problems, including the large tempera- reason for the growing popularity of
at NIST and can be downloaded from
ture and density variations and gradients FDS is inclusion of the SMOKEVIEW
the BFRL Web site (www.fire.nist.gov).
in the fire plume, and the significant program that permits the results of an
The term “field model” is generally
role that thermal radiation plays at high FDS simulation to be visualized and
applied to the types of models more
temperatures. These commercial CFD animated. These free programs provide
generally known as computational fluid
models typically include preprocessors capabilities normally associated with
dynamics (CFD) models in other disci-
and postprocessors to simplify data expensive CFD and visualization soft-
plines. In a field model, a space is sub-
input and visualize computed results, ware packages. Results of an example
divided into a very large number of
but the cost for these commercial pack- FDS calculation are illustrated in Figure
cells, which may be in the hundreds,
ages can add up to tens of thousands of 4. Version 1 of the FDS model is cur-
thousands, or even tens of thousands,
dollars per year for a site license. rently available for download at the
depending on the desired resolution of
At least two CFD models are being BFRL Web site (www.fire.nist.gov/fds/).
the calculation and the available com-
developed specifically for fire dynamics Version 2 is currently in beta testing.
putational resources. The conservation
calculations. The first, known as SOFIE Field fire models, including SOFIE
equations for mass, species, energy,
(for Simulation Of Fire In Enclosures), and FDS, require considerable knowl-
and momentum are then applied to
is being developed primarily by a con- edge and substantial computer power
each cell along with appropriate initial

W INTER 2002 www.sfpe.org 41

 released by the fire affect the results. One of the benefits of
and the amount of fire modeling is the ability to vary input
oxygen consumed parameters over expected ranges of
from the atmosphere. conditions in order to evaluate the sen-
Many fires become sitivities of the results to variations in
“ventilation limited,” input parameters. An exception to this
which means that their would be for situations where an analy-
energy release rate sis is being performed for a “reasonable
becomes governed by worst case” scenario. In this case, a sin-
the rate of airflow to gle calculation might suffice provided
the fire enclosure that the input parameters are truly rep-
rather than by the fuel resentative of reasonable worst case
Figure 4. Results of an example FDS calculation. release rate. Models conditions.
that do not account for
this ventilation limita- GARBAGE IN – GOSPEL OUT?
to apply properly. While casual users tion on combustion can produce inac-
may have some success running these curate results. Most of us have heard the phrase
models, they require a high level of associated with computers: “garbage in
expertise to use them properly. Despite MODELING UNCERTAINTIES – garbage out.” While this phrase has
these caveats, CFD models are the pro- become trite from overuse, it is
grams of choice for situations where Models, by definition, are incomplete overused because it remains so true.
adequate correlations do not exist to representations of the component, sys- While it may be appropriate to trust
make use of the simpler tools. tem, or process being modeled. that a computer calculation will be per-
Consequently, there will always be formed accurately once a program has
FIRE MODELING OR FIRE uncertainties associated with the results been debugged and verified, this is no
CONSEQUENCE ANALYSIS? of a calculation. In general, these guarantee that the calculation result will
uncertainties will be associated either accurately reflect reality. If the scenario
Many, if not most, applications of fire with the model itself or with the input being modeled is not specified accu-
modeling should more accurately be parameters. Modeling uncertainty is pri- rately, then the model will not produce
called fire consequence analyses marily epistemic because it relates to meaningful results. On the contrary,
because the fire itself is not being mod- the level of knowledge related to the results of a sophisticated computer-
eled, only the expected consequences phenomena being modeled. Parameter based calculation may provide the per-
of a specified fire are being calculated. uncertainty is primarily aleatory as it ception of a high level of accuracy
Even the most sophisticated fire models relates to random variations in the despite producing erroneous results.
have only rudimentary ability to model properties or attributes of the input This has sometimes been referred to as
the actual development of a fire in an parameters. “garbage in – gospel out.” For this rea-
enclosure at the present time. More Enclosure fire models are a type of son, it is critical that the assumptions
typically, the user specifies a fire ener- deterministic model because they pro- and uncertainties associated with the
gy release rate history, which may duce the same specific set of output input parameters be clearly described
include an incipient period of low results for the same input parameters. and adequately justified.
power output, a growth period of Only a few enclosure fire models have Consider, for example, the practice of
accelerating power output, a fairly attempted to address uncertainty using a design fire heat release rate of
steady period of peak power output, through specification of probability dis- 5 MW for the analysis and design of
and a decay period of decreasing tribution functions for the input para- smoke management systems in large
power output, as shown in Figure 5. meters and in some cases for the mod- spaces such as atria. Such a specifica-
This approach has an obvious poten- eling equations. The COMPBURN tion is widely used in building codes
tial pitfall. A user can specify an incor- model12 that is widely used for nuclear (e.g., 1997 Uniform Building Code and
rect fire history, resulting in an inaccu- power plant fire hazard analyses is 2000 International Building Code) and
rate analysis regardless of the accuracy probably the best known of these mod- referenced in design standards (e.g.,
of the actual calculation method els. Such models produce probability 1995 NFPA 92B). While such a specifi-
employed. While there is not a simple distribution functions for the output cation may be reasonably conservative
solution to this problem, it is important parameters by repeating the same cal- for the vast majority of applications, it
for both fire modelers and for those culations hundreds or thousands of is inadequate for those applications
reviewing fire modeling results to times with the different input parame- where larger fires might be expected to
understand and accept the bases for the ters in a process known as Monte Carlo occur. Appropriate administrative con-
specified fire histories. simulation. trols are needed to limit fire sizes to
While most current fire models may Whether uncertainty is addressed those used for analysis. Otherwise,
not be able to accurately calculate the quantitatively or qualitatively, it is “garbage in – garbage out.”
growth and spread of fire, many mod- important that uncertainty is addressed.
els can and do address the influence of Perhaps the worst mistake a fire model- WHAT IS THE RIGHT TOOL FOR
oxygen on burning rates within an er can make is to run a model only THE JOB?
enclosure. For most typical com- once and then rely on those results as
bustibles that burn in fires, there is a conclusive without considering how It would be nice if there were a sim-
direct relationship between the energy variations in input parameters might ple and universal answer to the ques-

42 Fire Protection Engineering N UMBER 13

 power plant fire safety analysis that
uses such a screening step. As illustrat-
ed in Figure 1, simple screening tools
may be adequate to clearly demonstrate
Incipient Growth Fully developed Decay
that performance criteria clearly are, or
are not achieved, for many scenarios.
While the ultimate objectives of an
Heat release rate (kW)

analysis may require the use of more

sophisticated tools for scenarios that do
not screen out with a simple tool,
Total
screening steps frequently permit many
energy
scenarios to be screened so that the
released
more sophisticated tools can be applied
(kJ)
only when needed.
This step should also include consid-
eration of how the tools being evaluat-
ed have been validated and verified.
Ideally, the modeling tools being con-
Ignition Time sidered for use will have compared
favorably with experimental data for
Figure 5. The stages of fire energy release rate. scenarios similar to those being consid-
ered. If not, some validation studies
tion: What is the right tool for the job? purposes are documented along with may be needed to develop confidence
Unfortunately, there isn’t one. The right the performance criteria that will be in the accuracy of the modeling tools
tool depends on the task at hand. It is used to evaluate the acceptability of a for the scenarios being considered.
the job of the engineer to understand proposed design. Sometimes fire pro-
the uses and limitations of the available tection engineers take for granted the 3. Determine what input parame-
tools and to select a tool that is suitable purpose of an analysis because they ters are needed for the models being
for the task. It is equally important that routinely perform such analyses. considered, and identify and docu-
the engineer understands and docu- However, the purpose of an analysis ment the data sources to be used as
ments the bases and limitations for a may not be apparent to other stake- bases for the input parameters.
particular analysis. holders, so this documentation is Most of the input parameters need-
It is easier to develop a procedure important. ed for different fire models are basi-
that can be followed to help the engi- cally the same. While they may be
neer in the selection and application of 2. Identify modeling tools that can specified in different terms, all enclo-
the right tool than it would be to pro- support the level of detail demanded sure fire models require specification
vide specific guidance on which specif- by the analysis. of the enclosure and ventilation open-
ic tools apply to which scenarios. This In order to identify appropriate ing dimensions, the boundary thermal
procedure would be: modeling tools for a particular task, it properties, and the fire properties,
is important for the fire protection either in terms of a specified energy
1. Identify and document the pur- engineer to understand the uses and release rate or in terms of flammability
pose of the analysis in as much limitations of the models that are avail- properties. While results of a simula-
detail as possible. able for a task. For example, the wide- tion may be relatively insensitive to
Fire models are generally used to ly used model to predict fire detector many of these input parameters, they
demonstrate that fire mitigation strate- activation, DETACT,13 is based on will be relatively sensitive to others,
gies will be effective before unaccept- unobstructed axisymmetric plumes in notably the fire history.
ably hazardous conditions develop. large spaces with smooth horizontal For this step, it is important for the
For example, they can be used to help ceilings. The engineer needs to recog- data sources that are used to justify the
demonstrate that a large space can be nize what the effects of obstructed selection of different input parameters
evacuated before occupants become plumes, small spaces, and obstructed to be identified and documented. This
immersed in the descending smoke or sloped ceilings will be before will provide the justification for the val-
layer. Or they can be used to evaluate applying this model to such plumes or ues selected as well as provide an audit
when a fire detection device or sprin- spaces. Where it is determined that the trail so that others can confirm and ver-
kler would be expected to activate in DETACT model does not apply due to ify that appropriate values have been
different fire scenarios. Ideally, they complicated geometries, other more selected. This documentation of input
could be used to demonstrate the sophisticated tools may be needed, parameter selection may also provide a
effectiveness of a sprinkler system such as a CFD model, that can address basis for an uncertainty analysis.
design, although such capabilities are these aspects.
generally beyond the current state of Many analyses include a screening 4. Consider how model results
the art. step during which simple tools are used will be used to support design deci-
For this first step, it is important that to perform conservative analyses. sions.
all the different purposes for which an Dungan14 recently provided an Finally, before a model is selected for
analysis is being conducted are identi- overview of the FIVE Methodology, a use, consideration should be given to
fied. It is equally important that the risk-based methodology used in nuclear how the model results will be used. At

W INTER 2002 www.sfpe.org 43

 this step, model results must be communicated to other, gen- 6 Rubini, P.A., “SOFIE – Simulation of Fires in Enclosures,”
erally nontechnical people. While such people may not be Proceedings of 5th International Symposium on Fire Safety Science,
interested in the details of an analysis, it is important for them International Association for Fire Safety Science, Melbourne,
to understand the limitations of the analysis so that they can Australia, March 1997.
make better risk-informed decisions. 7 SFPE Engineering Guide to Performance-Based Fire Protection
Analysis and Design of Buildings, Society of Fire Protection
Fredrick Mowrer is with the Department of Fire Protection Engineers, Bethesda, MD, 2000.
Engineering at the University of Maryland.
8 Chitty, R., and Cox, G., “ASKFRS: An Interactive Computer Program
for Conducting Engineering Estimations,” HMSO, London, 1988.
REFERENCES
9 “FIRECALC Computer Software for the Fire Engineering
1 Lawson, J. R., and Quintiere, J. G., “Slide-Rule Estimates of Fire Professional,” CSIRO Division of Building, Construction, and
Growth,” Fire Technology, Vol. 21, No. 4, 267-292, November 1985. Engineering, North Ryde, NSW, Australia, 1991.

2 DiNenno, P., Editor-in-Chief, SFPE Handbook of Fire Protection 10 Cooper, L. Y., “Mathematical Model for Estimating Available Safe
Engineering, 2nd edition, National Fire Protection Association, Egress Time in Fires,” Fire and Materials, Vol. 6, No. 3-4, 135-144,
1995. September/December 1982.

3 Nelson, H. E., “FPETOOL: Fire Protection Engineering Tools for 11 Baum, H. “Large Eddy Simulations of Fires,” Fire Protection
Hazard Estimation,” NISTIR 4380, National Institute of Standards Engineering, Society of Fire Protection Engineers, No. 6, Spring 2000.
and Technology, Gaithersburg, MD, October 1990. 12 Ho, V., Siu, N., Apostolakis, G., “COMPBRN III—A Fire Hazard
4 Peacock, R. D., Forney, G. P., Reneke, P. A., Portier, R. W., Jones, Model for Risk Analysis,” Fire Safety Journal, Vol. 13, No. 2-3,
W. W., “CFAST, The Consolidated Model of Fire Growth and 137-154, 1988.
Smoke Transport,” NIST TN 1299, National Institute of Standards 13 Stroup, D. W., Evans, D. D., “Use of Computer Fire Models for
and Technology, Gaithersburg, MD, February 1993. Analyzing Thermal Detector Spacing,” Fire Safety Journal, Vol. 14,
5 McGrattan, K. B., Baum, H. R., Rehm, R. G., Hamins, A., Forney, 33-45, 1988.
G. P., “Fire Dynamics Simulator: Technical Reference Guide,” 14 Dungan, K. W., “Practical Applications of Risk-Based Technologies,”
NISTIR 6467, National Institute of Standards and Technology, Fire Protection Engineering, Society of Fire Protection Engineers,
Gaithersburg, MD, January 2000. No. 10, Spring 2001.

44 Fire Protection Engineering N UMBER 13

 F O R E WA R D
Arup has established an Extreme Events Mitigation
Task Force to help understand the events of
September 11 in New York and Washington, and
their consequences for the buildings industry.
Results from this work will be made available to the
industry.
The following is based on a “briefing paper” that
was prepared two weeks after the World Trade Center
disaster by this task force. The briefing paper is post-
ed on the Arup Web site and is being periodically
updated as more information becomes known.
Members of the task force include:

Tony Fitzpatrick – Chairman, Arup Americas Board

Ray Crane – Team Leader, New York
Ashok Rajii – Mechanical, New York
Bob Cather – Materials, London
A
A Briefing
Briefing on
on the
the
Brian Meacham – Risk/Fire, Boston

WORLD
Craig Gibbons – Structures, Hong Kong
Dick Custer – Fire and Forensics, Boston
Faith Wainwright – Structures, London
Jim Pinzari – Risk/Business Continuity, Boston
Jim Quiter – Fire, San Francisco
John MacArthur – Structures, New York
Leo Argiris – Structures, New York
Michael Willford – Structures, London
Richard Hough – Structures, Sydney
This brief was prepared for a broad audience
inside and outside of Arup. Thus, some material will
be familiar to readers of this magazine. However, the
brief described many of the factors that need to be
TRADE
considered when evaluating how a building responds
to such an extreme event. Accordingly, it forms one
part of a broader risk assessment designed to advise
clients on individual risk, whether in a new building
or an existing one. Prevention measures designed to
reduce the probability of a terrorist attack will be
CENTER
taken into account in such an assessment.
Attacks
Attacks
INTRODUCTION

Towers 1 and 2 of the World Trade

Center in New York were both hit by
commercial aircraft on Tuesday,
September 11, 2001. The aircraft had
been hijacked, and the incidents caused
both towers to collapse. Neighboring
buildings also subsequently collapsed,
and there was damage over a wide area.
This was a deliberate act of terrorism,
planned to create devastation and
destruction. Such events are not antici-
pated, and current codes and design
practice do not address damage on this
scale being inflicted on buildings.
Nevertheless, looking at the total col-
lapse of both towers and its impact on
society, we must see if we can do
something in the way we design build-

46 Fire Protection Engineering N UMBER 13

 ings to prevent this happening again,
and we must explore the options and
offer the choices to society. These were
manmade structures – society makes
choices about how we design buildings
and how they should perform, but we
are the professionals with the needed
expertise to inform the decisions.
The consequences of the September
11 tragedy will be far-reaching, and as
governments look at security and inter-
national political actions, we need to
think how we will respond on our pro-
jects. What are the options, and how
might they be pursued?
This report sets out some facts,
issues, and questions, and suggests
how we might begin to address these.
It was written shortly after the event
and will be further developed and
updated by the Extreme Events
Mitigation Team Arup has established.

THE EVENT: WHAT HAPPENED?

Events on the morning of

September 11, 2001, until the col-
lapse of the WTC Towers 1 and 2
At 8:45 am, American Airlines Flight
11 carrying 92 people was flown by
hijackers into the north tower, called
One WTC of the World Trade Center
complex, hitting the tower on the north
face around the 90th floor. At 9:06 am,
United Airlines Flight 175 carrying 65
people and also hijacked, hit the south
face of Two WTC (the South Tower) at
around the 60th floor.
Both aircraft penetrated the buildings
and burst into flames. Large and wide-
spread fires were observed following
impact, involving both combustion of
the fuel from the aircraft as well as
combustible materials typically found in
office buildings.
At 10:00 am, 54 minutes after the
impact, Two WTC collapsed complete-
ly. At 10:29 am, one hour, 44 minutes
after the impact on One WTC, it, too,
collapsed to the ground. Many people
escaped in the time that the buildings
remained standing, but 2,442 people
are currently unaccounted for. More
than 550 people have been confirmed
dead as of December 20, 2001.
The planes were Boeing 767 200-
Pictures courtesy of Paul Doherty and Henry Gifford.
ERs, which have a cruising speed of
830 km/h (530 mph) at 10 km (35,000')
and a fuel capacity of 90,000 l (23,980

W INTER 2002 www.sfpe.org 47

 US gallons). At this point in time, the
impact speed is not known. At an alti-
tude of less than 600 m (2000'), it is
likely that the air speed was less than
500 km/h (300 mph).
It is reported that normal occupancy
of the buildings would be 40,000 to
50,000 total in the two towers. It is not
yet known how many actually were in
the buildings; but significantly fewer
were likely there at that time in the
morning.
On the same morning, the Pentagon
was also struck, causing nearly two
hundred fatalities. Analysis of this event
is outside of the scope of this article.

Neighboring buildings
The twin towers were part of the combination of the impact loads from Preliminary analysis of what led to
World Trade Center complex of seven the collapse of One and Two WTC, the total collapse of One and Two
buildings. the tremendous fire that ensued, and WTC
In addition to these high-rise build- foundations being affected by the col- The initial impact destroyed many
ings, there was a 47-story high-rise lapsing towers. Specifics of the col- columns and more than one floor. How
building (Seven WTC), a 22-story high- lapse mechanism are being analyzed much was damaged may never be
rise building (Marriott hotel), two by many parties. known, but the impact itself did not
9-story buildings, and one 8-story build- Falling debris also damaged build- cause the towers to collapse. The build-
ing. Excluding the hotel, most of the ings beyond the World Trade Center ings remained standing long enough to
occupied space within the buildings complex. An assessment of these allow many people to escape.
was dedicated for office use. All the buildings was undertaken during the The initial impact probably removed
buildings, except for Seven WTC, were week of September 17, by SEAoNY some columns and significantly weak-
constructed over a plaza area that con- (Structural Engineers Association of ened many others. Deformed, and
tained a shopping mall, four under- NY) and is available on the SEAoNY probably unbraced over two or more
ground levels of public parking, and Web site at www.seaony.org. The floors, these columns would also have
two utility levels. assessment was based on an ATC-20 been subjected to additional load from
The entire center contained approxi- Post-Seismic Building Assessment. The damaged floors and debris. The spread-
mately 1.25 million square meters (13.5 affected perimeter buildings contain ing fire would have continued to weak-
million square feet) of rentable space approximately 1.5 million square en the structure, most likely leading to
including .18 million square meters (2 meters (16 million square feet) of a collapse under vertical load.
million square feet) in Seven WTC. space. In all, more than 30% of down- The aircraft were heavily laden with
The collapse of these neighboring town Manhattan office space has been jet fuel, having just taken off from
buildings was probably due to the affected. Boston en route to Los Angeles. The

48 Fire Protection Engineering N UMBER 13

resulting hydrocarbon-fueled fire would
have been much hotter and more rapid
than the type of fire against which
buildings are normally designed.

THE BUILDINGS – DESCRIPTION

OF THE DESIGN

Structure
The Towers were steel framed, 110
stories high, and square on plan. The
buildings were designed by Skilling,
Helle, Christianson, Robertson, which
was the structural engineering firm of
record for the World Trade Center com-
plex, and completed in phases starting
in 1970. One WTC was 417.0 meters THE WORLD TRADE CENTER
(1368') tall, and Two WTC was 415.1
meters (1362') tall. Each 110-story tower
had a floor plate 637 x 637 meters (209' (BOCA) when developing designs. The
by 209'). The central core in each steel frame had been clad in fire spray.
building was 26.2 x 42.3 meters (86' x Asbestos-based spray had been used ini-
139'), constructed with steel columns tially in the building, but later changed
and lightweight drywall for infill. to a nonasbestos material. Further specif-
Around the perimeter of the build- ics, including which structural members
ings, 356 mm (14") steel box columns were treated and to what level of fire
were spaced at 1 meter (3'-3") on cen- resistance, are still being determined.
ter, with 1.2 meters (48") deep plate
girder spandrels at each floor. At the Blast resistance
third level, the columns transitioned in One WTC had already survived an
an arch-like formation to a 3.05 meter explosion in February 1993 when a
(10'-0") spacing for the lower story. rental truck packed with explosives was TYPICAL FLOOR PLATE
Floors were supported by steel trusses detonated in basement level B2. That
spanning 18.3 meters (60'-0") from the bomb blew out one section of a north
core to the perimeter wall on each side tower cross-brace between two of the
of the building. perimeter columns. The blast ripped buildings, incorporating specific mea-
The building perimeter structure was out sections of three structural slabs in sures intended to address lessons
the key element in the performance of the basement levels, but did little dam- learned. Because building and fire
the building. In addition to taking verti- age to the columns. codes do not typically consider deliber-
cal gravity loads, it also resisted all hor- According to one of the original ate, willful attacks, such as arson or ter-
izontal loads by framing action designers, the towers were originally rorism, these events have not played a
between the close-centered columns designed to take the impact of a significant role in code development.
and the spandrel beams, such that the Boeing 707. A member of the original The current code-based design
perimeter structure acted as a pierced design team has clarified recently that approach on its own is inadequate for
tube in resisting loads. this addressed impact only and not the addressing the question “how would
Vertical structure inside the building consequences of a jet-fuel fire. buildings perform under such an
supported gravity loads only. attack?” However, it is possible to look
THE QUESTIONS at any building on a case-by-case basis.
Fire protection To inform future design, perhaps it
The World Trade Center was devel- How would our current buildings may be better to turn the question
oped and constructed by the Port fare under such an attack? around – “What are the implications for
Authority of New York and New Jersey, The answer is not simple. Structures design if the building were required to
a self-supporting agency of the two are designed to building and fire codes, remain standing for longer after such
states. Although the Port Authority has and these documents reflect our soci- an extreme event?” This is addressed
not adopted any specific fire safety ety’s tolerance of risk from fire, earth- below.
code, it considers the requirements quake, floods, and a host of other haz-
developed by the National Fire ards, and establish the minimum levels Is this a question just for tall
Protection Association (NFPA), New of safety. Historically, building and fire buildings?
York City, and the Building Officials and codes have evolved in response to fire The World Trade Center’s promi-
Code Administrators International and other events that have affected nence and symbolism must have been

W INTER 2002 www.sfpe.org 49

 factors in why it was chosen as a tar- the design for safety in such an sent time, any engineered or perfor-
get. The Pentagon was also a target. extreme event. mance-based designs are typically used
As we face the question “will this affect only under the “alternate methods and
the future of tall buildings?”, we have CURRENT BUILDING DESIGN materials” clauses found in most US
to recognize that tall buildings will PRACTICE codes. It is not unusual, for example, to
always be prominent structures on the undertake egress analyses for large or
cityscape. However, avoiding symbol- Performance required unusual buildings under the “alternate
ism and expression in buildings can Current building codes in the United methods and materials” clauses, taking
only lead to a bunker mentality – and States are predominantly prescriptive in into consideration such factors as fire
even if this were pursued, would we nature. In other words, the code pre- resistance ratings, sprinklers, compart-
be safe from determined terrorist acts? scribes specific minimum requirements mentation, travel distances, and various
Wherever large quantities of people for parameters such as structural loads, fire scenarios.
gather in one building or complex, be fire resistance ratings, and egress It is worth noting that to operate in a
it for work or for play watching some requirements. In some other countries, a performance-based environment, per-
sporting event, then the actions of peo- performance- or functional-based system formance criteria and tools and meth-
ple determined to kill and themselves has been adopted, wherein the overall ods for assessing performance must be
to die in the process are a complete performance or function of the building available.
new paradigm for design. The focus and its systems are defined, but the
may be on tall buildings initially, but it design solutions are not mandated. Evacuation and refuge
is not a question only for tall buildings. Frequently these performance codes The US does not have a single,
contain acceptable or deemed-to-satisfy nationally enforced building code.
How will this affect the approach to solutions that serve as a prescriptive Rather, there is a system of model build-
building design? option to the performance approach. ing codes that are adopted by states
To answer this, we first have to look A performance-based approach is and local jurisdictions, sometimes with
at the current approach to design and currently being introduced into the modifications. In some jurisdictions,
then consider how we would address United States as well, but for the pre- such as New York City, there are also

DecoShield Systems, Inc.

Whatever You Need in Fire Protection
The Creative Cover Support System
Automatic Control Valves...
COVERS IT ALL!!!
CONCEALING sprinkler piping,
plumbing, electrical wire and cabling
FOR MORE has never been easier, thanks to the
UL-Listed DecoShield® Cover and
INFORMATION Support Systems.

&
SAMPLES Watts ACV Has It Now!
CALL Whether you need UL-Listed Pressure Reducing
Valves, Factory Mutual System approved and
800-873-0894 UL-Listed Fire Pump Relief Valves, Factory Mutual
System approved Fire Pump Suction Control Valves
OR VISIT OUR
or UL-Listed Pneumatic, Hydraulic, or Electronically
WEBSITE @ Actuated Deluge valves, Watts ACV has them in
www.decoshield.com We have been serving the fire sprin- sizes from 3” through 10”, in globe or angle pattern.
kler retrofit industry for more than ten For more detailed information, call us at:1-713-943-0688
years. Our system is clean, attractive, or visit our website at www.wattsacv.com.
paintable, durable, tamper-resistant,
economical and quick to install. That
is why we are the #1 choice for retro-
fitting all types of buildings.

50 Fire Protection Engineering N UMBER 13

locally developed codes. As a result, water supply volume is determined device, which, when a defined temper-
comments about “standard” US proce- according to the occupancy type of ature is reached, allows water to flow.
dures are necessarily generalizations. the building. The sprinklers would not The aim is that a sprinkler will actuate
Having said that, standard practice is be designed to cope with a wide- when a fire is relatively small, and the
to evacuate automatically only those spread aviation fuel fire. available water will control or suppress
floors in the vicinity of the event (usu- The sprinklers themselves are the fire.
ally floor of origin, floor above, and designed such that each sprinkler head In the WTC towers, the initial colli-
floor below, although some jurisdic- actuates individually. In typical office sion and explosion likely damaged a
tions require more). The responding occupancy, each individual sprinkler considerable part of the sprinkler sys-
emergency professionals can then has a fusible link, or other reaction tem and fire protection water supply
make a decision whether further evac-
uation is necessary. The voice alarm
system, which is much like a special-
ized public address system, can be
used to notify people on a selective or
general basis.
There are several reasons to do this:
a. By evacuating only those in the
vicinity of the event, it is likely
that they can more rapidly enter
the stairs, rather than queuing at
the entrance while other nonaffect-
ed people (from other floors) fill
the stairwell. The Soffi-Steel™ system is today’s
b. Crowd management is more effec- premier modular steel pipe covering
tive if only those directly at risk system, designed to provide a durable,
are moved. maintenance-free facing that is attractive
c. In tall buildings, people can be and easily installed.
relocated to alternate floors, rather The Soffi-Steel™ system accommodates
than out of the building. all types of sprinkler devices including
This approach allows those directly institutional, recessed, and concealed
affected to leave their floors and enter sprinklers. In addition, it also functions as
areas of safety, such as pressurized fire- a shield for utility lines and HVAC.
resistive stairs or other floors. People Soffi-Steel’s clean design and durability
remote from the event are assumed to make it perfect for the use in commercial/
be safe, at least for some time. residential properties where the aesthetics
Using this scenario, responding per- of the shield are as important as its
function.
sonnel will generally go to the floor
Soffi-Steel™ is secure-it is virtually
below (or some number of floors
irremovable once assembled (Patent
below) the fire, either via elevator or
#5,526,617 and others pending).
stair, depending on the Fire Installation service, training, and technical
Department. support available at competitive pricing.
This approach clearly implies that cat-
astrophic failure within a short period of
time cannot be allowed to happen.

Sprinklers • Secure
Fire protection by automatic sprin- • Attractive
klers assumes that there will be water
available at a sufficient pressure to
121 E. Burbank Ave. • Versatile
Janesville, WI 53546
deliver a spray of water to the fire area. 800/800-3213; Fax: 608/757-1452 • Economical
Sprinkler systems are not typically
designed to support water flow from
E-mail: [email protected] • Durable
www.soffisteel.com
all sprinklers simultaneously, as the
water demand could be enormous and
is typically not needed. The intent is
that the water supply system provides
only enough water for a specific
design area, with a certain number of
sprinklers at the most hydraulically
remote location. The density and

W INTER 2002 www.sfpe.org 51

 system on the floors of initial impact. If Design for robustness under such fighters or emergency personnel trying
so, the sprinkler system would not have an extreme event to enter the building. For example,
been able to operate as intended. For The challenge here is that an event with a moderate fire at low level then
those parts of the system that remained such as the WTC collision involves total evacuation would bring many
intact, the rapid growth and spread of impact, explosion, and fire all at once. people through the danger zone and
the fire resulting from the jet fuel would The “design scenario” cannot be one of into the area of the firefighting effort.
likely have actuated a large number of a set of loads, but of a level of damage These are difficult decisions under dif-
sprinklers, taxing the available water inflicted by whatever cause. One might ficult circumstances requiring accurate
supply. In addition, if any fire protec- develop a set of “what ifs?” to try to information.
tion water supply risers were taken out envisage a “worst possible” or “worst
on a lower floor, the water supply to credible” event. Then by posing possi- Way forward
floors above may have been impacted. ble design options, we can ask, “If xxx Design measures to mitigate the
had been incorporated, would this pre- effects of this and other extreme events
Progressive collapse vent a collapse?” are possible and, therefore, must be
There are no explicit requirements For example: researched and evaluated as a matter of
for design to prevent progressive col- • If the steelwork at Towers 1 and 2 urgency. This requires an holistic
lapse of buildings in the US model had not succumbed to the fire, the assessment of all the issues discussed
building codes. There is an under- buildings might still be standing. What above. The results can then be part of
standing that requirements in the codes if we incorporated additional fire resis- the decision process on the level of
develop a limited resistance to progres- tance into the structure? How much is robustness to which a building should
sive collapse although this is not an enough? Would traditional fire resis- be constructed.
explicit design requirement. For zones tance have withstood the fire regardless In the United States, the next genera-
of high seismicity, US model building of the rating? Would fire resistance have tion regulatory approach, a perfor-
codes, including the UBC, NBC, SBC, adhered to the steel given the impact of mance-based one, is about to be intro-
and IBC, all have special detailing pro- the aircraft? duced. This will be an important venue
visions that have the goal of increasing • If the floors had not continued to facilitate the overall discussion of
ductility and toughness in structures unchecked in a progressive collapse to risk tolerance and the related social,
that reduce the possibilities of progres- the ground, then the buildings might technical, and financial impacts. This
sive collapse during seismic events. still be standing. What if we incorporat- terrible event emphasized our vulnera-
The buildings were constructed by ed “collapse” stories every ten floors – bility in dramatic fashion, but policy-
the Port Authority of New York and a level of structure designed to carry makers should avoid a knee-jerk reac-
New Jersey, and compliance with the the debris load from the nine collapsing tion in seeking to prevent a recurrence.
New York City Building Code was vol- floors above? Obviously, it is not just high-rise build-
untary. There was no seismic detailing • If the towers had had a reinforced ings that are vulnerable to terrorist
provision in the code at the time the concrete core able to continue to pro- attack; any change in codes will impact
WTC was designed and constructed. vide stability after the loss of the floors on many other occupancy types.
There are no explicit requirements in and perimeter frame, then a shaft pro-
the US model building codes or New viding a safe refuge might have
York City Building Codes for debris
loading.
remained – should we consider dual-
stability structures, where one stands
RESOURCES
after the other has failed? • www.sfpe.org
WHAT ARE THE IMPLICATIONS • www.asce.org
FOR DESIGN? Phased or total evacuation • www.seaony.org
Information is slowly being pieced • www.ncsea.com
• www.aisc.org
From a code perspective, the WTC together about escape from the towers,
• www.arup.com
disaster would have to be considered principally from people who escaped. • www.m-yamasaki.com
an unexpected event at the time it was This will ultimately help in the discus- • www.newscientist.com
designed and constructed. The US sion of phased or total evacuation. • www.greatbuildings.com
has not previously been subjected to However, regardless of the intended • www.skyscrapers.com
such a deliberate act of terrorism of evacuation approach, some phasing • www.nytimes.com
this magnitude, and therefore the will occur naturally based on people’s • www.usatoday.com
building and fire Codes did not have a perception of the threat, the presence • www.cnn.com
direct past history to draw upon. In the of loved ones in the building, and the • FEMA – Federal Emergency
aftermath, it will be a societal decision limitations for the flow of people Management Agency
whether we want to tolerate the risk of • SEAoNY – Structural Engineers
through exit routes.
Association of New York
another similar event or whether we Total evacuation is not necessarily • NCSEA – National Council of
want to accept the restraints associated always the best answer and needs to Structural Engineers Associations:
with minimizing this type of event. be considered at each event assessing • AISC – American Institute of Steel
How do we begin to consider the who is most at risk, how safe they are Construction.
basis on which to make such decisions? without evacuation, and access for fire-

52 Fire Protection Engineering N UMBER 13

 Products/Literature
Internet-Based Vehicle Management IBC Analysis Guide
With these Internet- Rolf Jensen & Associates, Inc. (RJA) intro-
based Information duces From Model Codes to the IBC: A
Management solutions, Transitional Guide, a publication aimed at
Emergency Service helping architects, engineers, and builders
Providers can communicate vehicle status throughout their department interpret the new International Building
and simplify paperwork for vehicle and facilities management. Using a Code (IBC). The book, written by RJA con-
Web browser, personnel can create, review, approve, and distribute sultants and produced and marketed by
information on vehicle or equipment repairs. Firefighters can update R.S. Means, presents a convenient way to
the status of any vehicle at any time. compare IBC requirements to other model
www.manifolddata.com building codes.
—Manifold Data Systems www.rjagroup.com
—The RJA Group, Inc.

Loop Hanger Surge Restraint Halon Replacement Alternative

The CADDY Loop Hanger Surge A new fire extinguishing material
Restraint (LHSR6) restricts the – 3M™ Novec™ 1230 Fire
upward surge movement of activat- Protection Fluid – was devel-
ed fire sprinkler systems to meet oped to be a long-term replace-
NFPA 13 and ensure proper spray ment for halon. It is designed to
patterns. One clip fits 1/2 through balance the need for extinguish-
2-in. loop hangers. The LHSR6 ing performance, human safety,
grips the loop hanger and not the and low environmental impact.
nut, allowing adjustment of the Its chemical structure is low in
hanger on the threaded rod. acute toxicity, making it ideal for use in occupied spaces.
www.erico.com www.3M.com
—ERICO®, Inc. —3M

Analog Addressable MultiSensor Advanced Detection

The latest microprocessor technology in
smoke detection is available in the
Systems
XP95-M Analog Addressable Multisensor, NOTIFIER’s new, eight-page Advanced
which is optimal for most general Detection brochure provides simple solu-
applications. The XP95-M includes both tions for environments with complex fire
photoelectric and thermal sensing ele- detection needs – from sterile cleanrooms to
ments integrated into a single device, harsh industrial environments. Written for
designed to enhance performance, end-users and specifiers, it highlights state-
reduce false alarms, and provide quick, of-the-art detection and control technologies.
accurate response. It also includes application suggestions for
www.gamewell.com the full spectrum of NOTIFIER detection systems.
www.notifier.com
—Gamewell™ Worldwide —NOTIFIER®/Honeywell Home and Building Control

Dry System Trouble Alarm Applications Guide on CD-ROM

The DSTA (Dry System Trouble Alarm) is a System Sensor has released a MAC or
microprocessor-based local supervisory Windows-compatible Applications Guides
annunciator designed to monitor low/high CD-ROM, free to qualified individuals.
air pressure and low room temperature of a The CD-ROM includes information on
Dry Pipe Sprinkler System Riser (the pres- the proper application of various detec-
sure and room temperature switch is sold tion products in life safety and property
separately). Each unit includes an LED protection environments. It provides
display and an internal buzzer and may be comprehensive data on every System
riser- or wall-mounted. Sensor application, navigates easily, and
www.pottersignal.com is interactive.
—Potter Electric Signal Co. www.systemsensor.com
—System Sensor

54 Fire Protection Engineering N UMBER 13

 Fire-Resistive Remote Sensing Fire Alarm
Cable To address the problem of nonworking fire
alarms due to dead or missing batteries, First
Pyrotenax Cables Ltd. manu- Alert has developed model SA302, which uses
factures mineral-insulated (MI) “smart sensing” and remote control test/silence
cable that has no organic com- technology. The SA302 can be tested or
pounds, so nothing burns, releases toxic fumes, or propagates flames. silenced by anyone, using virtually any home
The Pyrotenax System 1850™ two-hour fire-rated power cable is used remote control device. A “smart-sensing”
for fire protection of critical life safety circuits in commercial buildings; microchip also helps the alarm distinguish non-
it arrives with a UL fire-resistive classification, so conduit or additional threatening conditions from real emergencies.
fireproofing is not required. www.firstalert.com
www.pyrotenax.com —First Alert
—Pyrotenax Cables Ltd.

More Commercial Riser Manifold Options SENIOR PLANS EXAMINER

The EasyPac Commercial Riser Manifold
combines a flow switch, drain valve, sight
(Fire Protection)
glass, and pressure gauge into a lightweight, City of Pasadena – Fire Department
compact unit. It is available in five orifice Salary: $59,120 up to $73,900
sizes ranging from 3/8 to 5/8 in., in sizes 2-6
in., with a standard take-out dimension of 13 (effective 3/25/02)
in. There is a choice of end connections,
and installation may be horizontal or vertical. Oversees the plan review and inspection process
www.vikingsupplynet.com for fire protection; reports to the Fire Marshall;
—Viking SupplyNet provides advice and technical assistance to
citizens, staff, and fire personnel.

The Security and Access Qualifications: BS in fire protection or mechanical

engineering or related field, plus two years’ experience
performing plan review or fire protection design. Must possess an
Control Handbook ICBO Plans Examiner and Uniform Fire Code Inspection Certification.
This handbook outlines practical solutions for Apply immediately to: City of Pasadena, (626)744-4366.
security and access control applications. This www.ci.pasadena.ca.us/humanresources/currentopenings.asp
handbook provides more than 40 different
case studies, each including a description,
parts list, diagram, and application tips. The
Security and Access Control Handbook is a
valuable tool for engineers who design or
specify security and access control systems. (Security and Access Control
Handbook, A Practical Guide to Application and System Design, pub-
lished by EST Press, an imprint of Edwards Systems Technology, Inc.) Established in 1939, Schirmer Engineering was the first independent fire
protection engineering firm to assist insurance companies in analyzing and
www.est.net
minimizing risk to life and property. Schirmer continues to be a leader in the
—Edwards Systems Technology (EST) evolution of the industry, using insight from tradition and experiences of our
past. Today, Schirmer Engineering is synonymous with providing high-quality

Fire Protection Reference Guide engineering and technical services to national and international clients.
Career growth opportunities are available for entry-level and senior-level
fire protection engineers, design professionals, and code consultants.
Viking has released a new color product Opportunities available in the Boston, Chicago, Charlotte, Dallas, Denver, Las
brochure entitled, “Your Fire Protection Vegas, Los Angeles, Miami, Phoenix, San Diego, San Francisco, and
Solutions Partner.” A reference guide for Washington, DC, areas. There is also a need for experienced loss control
fire protection specialists, it contains draw- engineers countrywide. We offer a competitive salary/benefits package. EOE
ings of sprinkler systems suitable for train-
ing. The sprinkler section contains a full Send résumé to:
G. Johnson
range of products, with photographs and Schirmer Engineering Corporation
technical reference tables. Other areas cov- 707 Lake Cook Road
ered include foam and foam equipment Deerfield, IL 60015-4997
and other related fire protection products. Fax: 847.272.2365
www.vikingcorp.com e-mail: [email protected]
—Viking Corp.

W INTER 2002 www.sfpe.org 55

 Resources
The 4th International
Conference on
Performance-Based
Codes and Fire Safety
Design Methods
March, 20-22 2002, at the
Melbourne Exhibition & Convention Centre in Melbourne, Australia.
Performance-oriented codes, regulations, and design implementation of these codes and the challenges they
methods are gaining formal and widespread acceptance in present. On the second day, the theme of the conference
many countries. This three-day conference will again bring will turn to design in the performance environment, and
together professionals from around the world to discuss design aspects such as structural fire performance, smoke
the state of the art in performance-based codes and design management, and reliability assessments will be explored.
methods. As in the prior international conferences held in The third day is a workshop dedicated to an international
1996, 1998, and the most recent held in June 2000 in Lund, comparison of a standard building design using perfor-
Sweden, this conference will benchmark the rapidly accel- mance design methods in regulatory environments from
erating codification and implementation of performance- around the world. A mini Trade Fair will be held in
based design. conjunction with the Conference.
The first day of the conference will be dedicated to per-
formance codes and will feature updates from Australia, REGISTRATION: To register, visit www.sfpe.org or
United States, Canada, Japan and the UK on the status of contact SFPE at [email protected]

March 20-22, 2002

UPCOMING EVENTS
4th International Conference on Performance-Based
UPCOMING EVENTS
Codes and Fire Safety Design Methods
Melbourne, Australia
Info: www.sfpe.org
February 5-6, 2002
May 19-23, 2002
Flame Retardants 2002
NFPA World Fire Safety Congress and Exposition
London, England
Minneapolis, MN
Info: www.intercomm.dial.pipex.com/fr2002cfp.htm Info: www.nfpa.org
March 18-19, 2002 June 16-21, 2002
Structures in Fire The 75th International symposium on Fire Safety Science
University of Canterbury, Worcester Polytechnic Institute
Christchurch, New Zealand Worcester, Massachusetts, USA
Info: www.civil.canterbury.ac.nz Info: www.iafss.org

56 Fire Protection Engineering N UMBER 13

 Advancing the Science and Practice of Fire Protection Engineering

Society
of Fire Protection
Engineers
An Invitation to Join
What is The Society of Fire Protection Engineers (SFPE)?
SFPE, established in 1950, is a growing association of professionals involved in advancing the science
and practice of fire protection engineering and fostering fire protection engineering education.
What are the benefits of SFPE membership?
The Society will provide you with many new opportunities for professional advancement, education,
and networking. The specific benefits members receive are:
Free access to SFPE’s periodicals
This includes:
▲ Fire Protection Engineering magazine. ▲ The peer-reviewed Journal
▲ SFPE Today - Our bimonthly Society newsletter. of Fire Protection Engineering.
Substantial discounts on continuing education
This includes:
▲ Technical symposia on current fire protection issues.
▲ International conferences on state-of-the-art applications of fire protection engineering.
▲ Short courses and seminars offering hands-on instruction.
▲ Discounts on fire-related publications.

Other benefits include:

▲ Recognition of your professional qualifications. ▲ Contribute to the profession through technical task
▲ Opportunity to participate in the SFPE Annual Meeting. groups and committees.
▲ Opportunity to network in local chapters. ▲ A periodic profile of the fire protection engineer,
▲ Low cost group life, health, and liability insurance. including salary information.

I’m interested in learning more about joining SFPE. Please send me additional information.

Name Title

Company/Organization

Address City State/Province Zip/Postal Code

Country

Work Phone Fax E-mail

Fax to 301/718-2242 ▲ Visit the SFPE Web Site: www.sfpe.org

For more information, contact The Society of Fire Protection Engineers:

7315 Wisconsin Avenue, Suite 1225 West ▲ Bethesda, MD 20814
Phone: 301/718-2910
 CORPORATE 100
The SFPE Corporate 100 Program was founded

B R A I N T E A S E R in 1976 to strengthen the relationship between

industry and the fire protection engineering
community. Membership in the program
recognizes those who support the objectives
of SFPE and have a genuine concern for the
safety of life and property from fire.

BENEFACTORS
Rolf Jensen & Associates, Inc.

PATRONS
Code Consultants, Inc.
Solve the following equation for x: Edwards Systems Technology
Hughes Associates, Inc.
The Reliable Automatic Sprinkler Company
( x 2 − 9 x + 20 )
( )
Schirmer Engineering Corporation
x 2
− 5 x + 5 =1 MEMBERS
Arup Fire
Automatic Fire Alarm Association
Thanks to Jane Lataille, P.E., for providing this BFPE International
Factory Mutual Research Corporation
Fike Corporation
issue’s brainteaser. Grinnell Fire Protection Systems
Harrington Group, Inc.
HSB Professional Loss Control
Hubbell Industrial Controls
Joslyn Clark Controls, Inc.
James W. Nolan Company (Emeritus)
Industrial Risk Insurers
Koffel Associates, Inc.
Marsh Risk Consulting
National Electrical Manufacturers Association
Solution to last issue’s brainteaser National Fire Protection Association
National Fire Sprinkler Association
Nuclear Energy Institute
The difference between any two numbers in the set The Protectowire Co., Inc.
Reliable Fire Equipment Company
{2, 3, 4} is equal to their greatest common factor. The Risk Technologies, LLC
Siemens Cerberus Division
same is true of any two numbers in the sets {6, 8, 9, 12} SimplexGrinnell
TVA Fire and Lifesafety, Inc.
and {8, 9, 10, 12}. Find a set of five numbers for which Underwriters Laboratories, Inc.
Wheelock, Inc.
this is true. W.R. Grace Company

DONORS
Gage-Babcock & Associates, Inc.
Answer: Two sets of solutions are {900, 912, 915, 918,
SMALL BUSINESS MEMBERS
920} and {1664, 1665, 1666, 1668, 1680} and the multi- Bourgeois & Associates, Inc.
ples of both sets. Demers Associates, Inc.
Fire Consulting Associates, Inc.
MountainStar Enterprises
Performance Technology Consulting Ltd.
Poole Fire Protection Engineering, Inc.
S.S. Dannaway & Associates, Inc.
The Code Consortium, Inc.

Index of • Advanced Fire Technology, Inc..............Page 44 • Marsh Affinity Group Services ..................Page 7
Advertisers • Ansul, Inc. ..................................................Page 2 • NOTIFIER Fire Systems...........Inside Back Cover
• Central Sprinkler ......................................Page 59 • Potter Electric Signal Company ..............Page 28
• Commercial Products Group ..................Page 45 • The Protectowire Co., Inc. ......................Page 22
• DecoShield Systems, Inc. ........................Page 50 • The RJA Group .......................Inside Front Cover
• Edwards Systems Technology ...........Page 30-31 • Reliable Automatic Sprinkler ..................Page 18
• Essex Fluid Controls................................Page 44 • Siemens Fire Safety ............................Page 12-13
• Fenwal Protection Systems .....................Page 34 • SimplexGrinnell .......................................Page 27
• Fike Protection Systems ..........................Page 53 • System Sensor ..........................................Page 37
• Fire Control Instruments, Inc. ...................Page 8 • Tyco Fire Products .............................Back Cover
• Grice Engineering....................................Page 51 • Watts Regulator (ACV).............................Page 50
• Koffel Associates......................................Page 38 • Wheelock, Inc..........................................Page 23

58 Fire Protection Engineering N UMBER 13

Central’s K17 ESFR Pendent Sprinkler
for warehouse and high-piled storage
fire protection. A generation ahead
in ESFR technology.
Central’s new K17 ESFR Pendent Sprinklers usher in a new
generation of ESFR fire protection. The K17 ESFR offers an increased
level of flexibility for designers, while providing the fast, proven and
effective response of ESFR technology for warehouse and high-piled
storage fire protection needs. The K17 ESFR utilizes the same flows as
traditional 14 K-Factor ESFR sprinklers, but at much lower pressures.
The results are significant installation cost savings due to reduced
branch line pipe sizing and the potential elimination of fire pumps.

Max. Storage Ht. Max. Ceiling Ht. K17 ESFR Min. Press. K14 ESFR Min. Press.
Ft. (m) Ft. (m) psi (bar) psi (bar)
40 (12.2)* 45 (13.7) 63 (4.3) 90 (5.2)
35 (10.7) 40 (12.2) 52 (3.6) 75 (5.2)
25 (7.6) 32 (9.8) 42 (2.9) 60 (4.1)
25 (7.6) 30 (9.1) 35 (2.4) 50 (3.4)
451 North Cannon Avenue • Lansdale, PA 19446
*Indicates One Level Of In-Rack Sprinklers Required. Phone: (800) 523-6512 • Fax: (215) 362-5385
See Tech Data Sheet 3-1.5 for other ESFR applications. www.tyco-central.com
 from the technical director

Learning from Tragedy

Also, any changes must be consid-

B y now, we have all read numer-

ous reports regarding the per-
formance of the Twin Towers
and the Pentagon and the occupants
of those buildings on September 11,
ered in a broad context. Prior to
September 11, a scenario involving the
collision of a jumbo jet with a building
was typically used as an example of the
2001. Some of these reports offer type of high-consequence/low-probabil-
explanations of what went wrong, ity scenario that would typically not be
what went right, and the changes that considered in the design of a building.
should be made in the way that build- Since September 11, numerous changes
ings are designed in the future. Some have been made in the aviation indus-
of these reports have a sound basis in try to reduce the likelihood of similar
science and engineering, while others actions in the future. Additionally, we
do not. need to be careful to ensure that
A fundamental tenet of engineering is changes made to mitigate similar events
learning from past events to prevent or do not have negative consequences for
mitigate the effects of similar events in more frequent events.
Morgan J. Hurley, P.E. the future. SFPE has developed a three- Designing to mitigate the effects of
Technical Director phase plan for learning as much as we the types of events that occurred on
Society of Fire Protection Engineers can from the events that occurred on September 11 in the future could bring
September 11, 2001, at the World Trade with them significant costs – costs that
Center and the Pentagon. could dissuade developers from build-
The first phase is to document avail- ing structures of national significance in
able factual information and data. This the future. Therefore, it is important for
includes information such as the struc- the engineering community to inform
tural frame of the buildings, installed decision-makers as they consider
fire protection, fuel loading, number whether any changes in the design and
and distribution of occupants, impact construction of buildings are warranted.
location of the aircraft, etc. While there is only preliminary infor-
The second phase involves filling in mation available at this time, it appears
the gaps in the information gleaned that the structures of both of the twin
during the first phase and developing a towers were damaged by collisions on
better understanding of what actually September 11. The resulting fires further
happened. Such an analysis will involve reduced the load-carrying capability of
the development of hypotheses and the structures, which led to the subse-
testing those hypotheses via analytical quent progressive collapses. If these
methods such as modeling. preliminary findings are correct, it
The third phase is be to develop a shows that the structural behavior and
listing of the costs and benefits of the fire behavior were closely coupled.
types of changes that might be consid- It is too early to consider what, if
ered in how buildings are designed and any, changes should be made to the
constructed in the future. While it is not design and construction of buildings as
exclusively up to the engineering com- a result of the lessons learned from this
munity to decide what, if any changes tragedy. However, one lesson that can
should be made, the engineering com- be learned is the importance of taking a
munity has a responsibility to provide comprehensive, multihazard approach
information on the costs and benefits of when designing buildings in the future.
any changes that might be considered,
so that policymakers can make
informed decisions.

60 Fire Protection Engineering N UMBER 13

The Flexible Onyx 640 Panel ™

Need flexibility in a fire detection panel? NOTIFIER

introduces the Onyx 640 — the first panel in the new
Onyx series of fire detection products.

The new Onyx 640 panel is:

• Expandable — one loop to two, 318 points per loop
• Scalable — can put multiple two-loop units into a
single enclosure
• Networkable — stands alone or can be linked to
other panels
• Backward compatible — works with existing
NOTIFIER devices
• Voice ready — using the new Onyx XPIQ transponder

Plug into the Onyx 640 panel — first in the new Onyx
series of products from NOTIFIER.

Call (203) 484-7161 or visit www.notifier.com

 At Tyco Fire Products, innovations such as
the ESFR-1, K17 Storage Sprinklers, ESFR-17,
ESFR -25 and EC-25 are just a few examples of our
commitment to meet the dynamic challenges of
the 21st century.
These applications will continue to change —
and we’re committed to develop specific products
for them.
For the latest in storage protection, visit your
local Gem or Star Independent Distributor. For the
TFP Distributor nearest you, visit our website at
www.tyco- fire.com or call us at 800-558-5236.