THE OFFICIAL MAGAZINE OF THE SOCIETY OF FIRE PROTECTION ENGINEERS

FIRE PROTECTION

WINTER 2004 Issue No. 21

ALSO:

15 NFPA 921’s IMPACT ON FIRE

SCENE RECONSTRUCTION
THE FIRE
25 FROM THE LAWYER’S
PERSPECTIVE: THE ROLE OF A
INVESTIGATION
FIRE PROTECTION ENGINEER
PROFESSIONAL
page 6
31 FAILURE OF A SPRINKLER
SYSTEM: A CASE STUDY

38 OPPORTUNITIES TO LEARN
FROM 9/11
 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 2004

information to fire protection engineers and allied

professionals. The opinions and positions stated are
the authors’ and do not necessarily reflect those of SFPE. 4
Editorial Advisory Board
Fire Investigation and The Fire Engineer
This article examines the changing role in regard to the analysis of a fire scene, fire
Carl F. Baldassarra, P.E., Schirmer Engineering Corporation
scene management methods (which acknowledge the needs of all parties partici-
Don Bathurst, P.E.
pating in the investigation), and the impact of recent court rulings that directly
Russell P. Fleming, P.E., National Fire Sprinkler Association impact the FPE’s ability to testify as an expert.
Morgan J. Hurley, P.E., Society of Fire Protection Engineers Robert A. Schroeder, Ph.D.
William E. Koffel, P.E., Koffel Associates
Jane I. Lataille, P.E., Los Alamos National Laboratory
Margaret Law, M.B.E., Arup Fire
10 NFPA 921’s Impact on Fire Scene Reconstruction
This article provides an overview of the scientific methods covered in
Ronald K. Mengel, Honeywell, Inc.
NFPA 921, Guide for Fire and Explosion Investigations, along with step-by-
Edward Prendergast, P.E., Chicago Fire Dept. (Ret.)
step procedures for working and testing hypotheses. The authors encourage
Warren G. Stocker, Jr., Safeway, Inc. fire investigators to become familiar with NFPA 921, as well as other authori-
Beth Tubbs, P.E., International Code Council tative references and sources of data and information.
Regional Editors David J. Icove, Ph.D., P.E., and John D. DeHaan, Ph.D.
U.S. H EARTLAND
John W. McCormick, P.E., Code Consultants, Inc. 17 From the Lawyer’s Perspective: The Role of a Fire Protection
U.S. M ID -ATLANTIC
Engineer in the Investigation of Fire Related Losses and
Robert F. Gagnon, P.E., Gagnon Engineering, Inc. Subsequent Litigation
U.S. N EW E NGLAND This article discusses the role of a fire protection engineer who is called as
Thomas L. Caisse, P.E., C.S.P., Robert M. Currey & an expert witness and what is expected of him or her. Topics covered
Associates, Inc. include conducting a proper investigation, proving the case, and working
U.S. S OUTHEAST with attorneys.
Jeffrey Harrington, P.E., The Harrington Group, Inc. Paul R. Bartolacci, Esq. and Georgia S. Foerstner, Esq.
U.S. W EST C OAST
Michael J. Madden, P.E., Gage-Babcock & Associates, Inc.
21 Failure of a Sprinkler System: A Case Study
This article briefly summarizes an unexpected failure of a sprinkler system:
A SIA
the catastrophic separation of a mechanical coupling that had been installed
Peter Bressington, P.Eng., Arup Fire
for nearly three years. The author discusses the investigation, testing
A USTRALIA
procedures, observations, and conclusions drawn.
Brian Ashe, Australian Building Codes Board
Daniel Arnold, P.E.
C ANADA
J. Kenneth Richardson, P.Eng., Ken Richardson Fire 25 Fire Alarm Testing Strategies Can Increase Occupant Response
Technologies, Inc. and Reduce the “Cry Wolf” Syndrome
N EW Z EALAND The previous article in this series identified how and why fire alarm testing
Carol Caldwell, P.E., Caldwell Consulting can contribute to the “cry wolf syndrome.” This article explores alternative
U NITED K INGDOM testing strategies that can improve occupant response and reduce this
Dr. Louise Jackman, Loss Prevention Council syndrome.
Personnel National Electrical Manufacturer’s Association
EXECUTIVE DIRECTOR, SFPE
Kathleen H. Almand, P.E. 27 Opportunities to Learn from 9/11
This article examines some of the opportunities to learn from the September
T ECHNICAL E DITOR
Morgan J. Hurley, P.E., Technical Director, SFPE 11, 2001 terrorist attacks on the World Trade Center and The Pentagon, so
that improvements can be made in the future through fire protection
P UBLISHER
Terry Tanker, Penton Media, Inc. engineering.
James G. Quintiere, Ph.D.
A SSOCIATE P UBLISHER
Joe Pulizzi, Custom Media Group, Penton Media, Inc.
Cover: Photographer ©Chandra Clarke/CORBIS
M ANAGING E DITOR
Joe Ulrich, Custom Media Group, Penton Media, Inc. Online versions of all articles can be accessed at www.sfpe.org.
A RT D IRECTOR Invitation to Submit Articles: For information on article submission to Fire
Pat Lang, Custom Media Group, Penton Media, Inc.
Protection Engineering, go to http://www.sfpe.org/sfpe/fpemagsubmit.htm
M EDIA S ERVICES M ANAGER
Lynn Cole, Custom Media Group, 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: 216.931.9180. Fax: 216.931.9969.
E-mail: asanchez@penton.com.
Copyright © 2004, Society of Fire Protection Engineers. All rights reserved.

www.sfpe.org 1
viewpoint

presentations are made in the second sec-

Book Review tion (Chapters 6-10). This information is

very useful not only for the practitioner us-
ing this book but also for those who seek to
Ignition Handbook clarify the limitations of the data presented.
Chapter 6 deals with the ignition of solids
by Vytenis Babrauskas, Ph.D. and 7 with that of liquids. These chapters
open with a revision of such concepts as
Auto Ignition Temperature, flashpoint, and
fire point. Practical relations between the

T he foreword of this book describes

very well the nature of this work:
“the Ignition Handbook leaves no
stone unturned.” This is a comprehensive
compilation of ignition data together with
previous three sections. The book has a fi-
nal section of two chapters, 14 and 15.
Chapter 14 corresponds to an encyclopedia
where diverse subjects are listed in alpha-
different values coexist with fundamental
explanations on the different physics behind
each parameter. Historical accounts of the
development of assessment methodologies
betical order covering every condition that
selected theoretical descriptions that, in and parameters are presented together with
might lead to ignition. Chapter 15 presents a
the words of the author, “could lead to dispelling comments related to common
great amount of data in summary tables.
conceptual understanding or provide a statements on ignition parameters such as
This section is very useful for those readers
framework for making useful practical cal- the ignition temperature. Consistently, ques-
seeking specific information. A minor criti-
culations.” This book consists of 15 chap- tions are raised on the utility of some corre-
cism is the lack of appropriate references to
ters that cover a multiplicity of areas relat- lations and analysis that link fundamental
authors within the figure captions, which
ed to ignition. A CD-ROM that contains properties to practical parameters. Discus-
makes it difficult to identify the source of
five Excel® files describing various proper- sions on limitations of different techniques
the data.
ties directly related to ignition accompa- and standard test methods are done with
The first two sections deserve more de-
nies the book. The tables provide infor- great thoroughness. A comprehensive treat-
tailed attention, not only because of the
mation for many materials compiled in ment of the theories used for interpretation
large amount of information, but also be-
four different categories: pure chemicals, of piloted ignition data allows a clear under-
cause they truly represent the conceptual
dusts, commercial natural products, and standing of the limitations and advantages
core of this book.
solids. The CD-ROM might be of practical of each theoretical approach. It would have
The introductory section provides a de-
use to those searching for specific infor- been nice if the author could have synthe-
tailed presentation of the subject that covers
mation, but it adds little information not sized the theory and converged to a conclu-
some basic combustion principles as well as
available in the text. sion of what is “good practice” in the inter-
very important statistical facts that put the
This book can be divided in five sections pretation of these results. Chapter 8
reader in context with the problem to be
characterized by a combination of theoreti- represents a brief addition to Chapter 7 that
studied. No detail or derivations are present,
cal material, anecdotal information, and an emphasizes the different treatments neces-
emphasizing the physical description of the
exhaustive compilation of data. An intro- sary for metals and carbon particles. Chapter
processes. Throughout this section the au-
ductory section comprised of Chapters 1 to 9 takes a different path and discusses self-
thor shows a clear preference for simplified
5 opens the book. A second section (Chap- heating. The Frank Kamenetskii theory is re-
models. Practical issues related to piloted ig-
ters 6-10) that addresses in a classical man- viewed again in the context of the FRS test
nition of gases such as spark ignition, Auto
ner all known forms of ignition follows. For method. This is done thoroughly, and the fi-
Ignition Temperature (AIT), Minimum Igni-
organizational purposes, the author divides nal example is very useful. Explosives, pyro-
tion Energy (MIE) concept, and quenching
the material in a classical form – liquids, technics, and reactive substances are treated
are explored. Chapter 5 extends the pre-
solids, self-heating – and adds a novel sec- in Chapter 10 in a manner that is consistent
mixed flame approach to the ignition of
tion on explosives, pyrotechnics, and reac- with Chapters 8 and 9.
dust clouds. This section is equivalent to the
tive substances generally not treated in fire In summary, the author should be com-
introductory chapters of any good combus-
textbooks. The third section looks into ex- mended for being able to compile such a
tion book but has a practical side that will
ternal factors directly related to ignition, fantastic amount of information into a single
make it appealing, especially to fire investi-
such as ignition sources and preventive text. This book will become a broadly used
gators. A detailed definition of terms and
measures. The chapter on preventive mea- reference within Fire Safety Engineering and
acronyms used in the book is also pre-
sures focuses on design and lacks informa- Fire Investigation.
sented. This is very useful because the fire
tion on maintenance-related aspects. Chap- literature is filled with terms that are not de- Jose L. Torero, Ph. D.
ter 13 includes a number of important fined with precision. The University of Edinburgh
miscellaneous items and can be included Chapters 4 and 5 analyze standard test
within this section. The fourth section is a methods, including mostly U.S.-based tests,
magnificent compilation of color plates that but a brief description of German and
illustrate many of the issues discussed in all British methods is also provided. Similar

W INTER 2004 www.sfpe.org 2

 flashpoints
fire protection industry news

FEMA Offers Multihazard Building University of Edinburgh Fire Safety Council Aims to End
Design Summer Institute for College Announces New Degree Programs Deaths from Residential Fires
and University Faculty The School of Engineering and Elec- Taking a stand during National Fire Pre-
The Department of Homeland Security’s tronics at the University of Edinburgh an- vention Month, the Fire Safety Council an-
Federal Emergency Management Agency nounces new BEng/MEng degree nounces the first in a series of steps to
(FEMA) has announced four seminars on haz- courses in Structural and Fire Safety Engi- eliminate deaths from residential fires by
ard mitigation in building design to be held neering, which commenced in October 2020. The Council also unveiled its Web
during a two-week-long institute in July 2004. 2003. The programs cover a complete site, www.FireSafety.gov, which provides
The institute covers protective design for range of Fire Safety Engineering topics up-to-date information about programs
earthquakes, fires, flood, and wind. including Fire Dynamics, the Design of and prevention information, as well as an
The seminars, part of the Multihazard Build- Fire Safety Systems, Numerical Fire Mod- interactive page for children.
ing Design Summer Institute (MBDSI), are de- eling, Human Evacuation, and emphasiz- The Council has initiated a pilot pro-
signed for active college or university faculty ing the understanding of the behavior of gram to combine their smoke alarm instal-
teaching undergraduate or graduate-level ar- structures when exposed to a fire. lation programs to reach the 5 million
chitecture and engineering courses in U.S.- Students with undergraduate degrees American homes without smoke alarms.
based institutions. The seminars were devel- in any area of engineering from non-UK The plan builds on research and programs
oped by staff at the Emergency Management institutions are welcome to apply for the that demonstrate the effectiveness of in-
Institute (EMI) in conjunction with experi- MEng program. An evaluation of the stalling long-lasting, lithium-powered
enced architects and engineers. The sessions, student’s background is made, and a smoke alarms in homes in high-risk com-
free of charge to qualified applicants, are con- customized course of study is proposed. munities.
ducted by EMI, located on the campus of the In support of this new endeavor, a “Each residential fire death is preventable.
National Emergency Training Center in Em- new state-of-the-art laboratory facility, Our primary injury-prevention goal is end-
mitsburg, MD. Qualified applicants are eligible The Rushbrook Fire Safety Engineering ing these unnecessary deaths,” says Julie L.
for no-cost housing at the training center and Laboratory, has been developed. Gerberding, MD, director of the Centers
are reimbursed for travel expenses. For more information, visit for Disease Control and Prevention.
During the summer institute, two seminars www.see.ed.ac.uk/undergraduate/Cours For more information, visit www.Fire-
are offered each week. Flood Protection De- es/struct_fire.shtml. Safety.gov.
sign and Wind Engineering are offered the
week of July 19, 2004. Earthquake Protective
Design and Fire Design are offered during the Corrections
week of July 26, 2004. On page 23 of the Fall, 2003 issue, article “Balancing Safety and Security in the School
For more information, visit Environment,” the reference to endnote #11 should have referred to endnote #10.
http://training.fema.gov/emiweb/MBDSI/ In the last issue the name of Kevin Biando, P.E. was not included in the table of contents
entry for the article entitled “How the GSA Ensures Safety and Security.”

MEMBERS SMALL BUSINESS MEMBERS

The SFPE Corporate 100 Program was founded in 1976 to Altronix Corporation Bourgeois & Associates, Inc.
strengthen the relationship between industry and the fire pro- Arup Fire The Code Consortium, Inc.
tection engineering community. Membership in the program Cybor Fire Protection Company Davidson & Associates
recognizes those who support the objectives of SFPE and have FM Global Corporation Demers Associates, Inc.
a genuine concern for the safety of life and property from fire. GE Global Asset Protection Services Fire Consulting Associates, Inc.
Harrington Group, Inc. Fire Suppression Systems Association
BENEFACTORS HSB Professional Loss Control Futrell Fire Consult and Design Inc.
Koffel Associates, Inc. James W. Nolan Company (Emeritus) Gagnon Engineering, Inc.
Rolf Jensen & Associates, Inc. Marsh Risk Consulting Grainger Consulting, Inc.
Specified Technologies, Inc. National Fire Sprinkler Association J.M. Cholin Consultants, Inc.
Nuclear Energy Institute MountainStar Enterprises
PATRONS The Protectowire Co., Inc. Poole Fire Protection Engineering, Inc.
Reliable Fire Equipment Company Risk Logic, Inc.
Ansul, Inc.
Risk Technologies LLC Slicer & Associates, LLC
Code Consultants, Inc.
TVA Fire and Lifesafety, Inc. S.S. Dannaway & Associates, Inc.
Edwards Systems Technology
Underwriters Laboratories, Inc. Van Rickley & Associates
Gage-Babcock & Associates, Inc.
Wheelock, Inc.
Hughes Associates, Inc.
Williams Fire and Hazard Control
National Fire Protection Association
The Reliable Automatic Sprinkler Company
Schirmer Engineering Corporation
SimplexGrinnell

3 Fire Protection Engineering N UMBER 21

 Fire Investigation
and
The Fire
Engineer
By Robert A. Schroeder, Ph.D.

INTRODUCTION
Since the 1960s, FPEs have been

F
called upon to participate in the investi-
ire Protection Engineers (FPEs) are, with increased
gation of large-loss fires, including fires
frequency, included in the initial investigation of in high-rise structures, fires resulting in
fires. The role of the FPE in fire investigation has the multiple losses of lives, and inci-
dents of failed fire protection systems.
evolved from an examiner of fire protection-related sys- Originally, the primary role of an FPE
tems and life safety code analysis to that of principal was to support the cause-and-origin in-
vestigators, answer the “whys” of fire
investigator. With this transformation comes new protection system performance, and as-
responsibilities and skills not commonly associated with sess the adequacy of life safety-related
the discipline. This article reviews the changing role of design issues such as exiting capacity. In
the 1970s, the number of FPEs conduct-
an FPE in regard to the analysis of a fire scene, fire ing cause-and-origin investigations and
scene management methods, the needs of all parties assessing building and life safety sys-
tems was very limited. Often the inves-
participating in the investigation, and the effect of recent tigative teams were composed of “spe-
court rulings that directly impact the FPE’s ability to tes- cialists,” engineers or technicians with
specific areas of expertise – alarm sys-
tify as an expert.
tems, sprinkler design, construction, life
safety – all employed by the same firm.
This practice continues today and has

4 Fire Protection Engineering N UMBER 21

expanded into fire cause and origin, of analysis. This evolution has been fur- about the fire scenario; specifically,
computer modeling, and fire testing. ther supported by the creation of for- who is involved. Employees of multi-
By the early 1980s, the number of malized college-level coursework that service providers must seek as much
tragic and complex fires revealed the addresses the elements of the fire scene information as possible regarding the
chasm between the analysis (princi- examination, resulting in an increase of interface of any segment of their com-
pally cause-and-origin based) of fire practitioners with engineering and sci- pany with affected parties, including
scenes and the failure of the current de- entific backgrounds. The judiciary has the potential involvement of company
sign of buildings. In the United States, also been driving this technical transfor- clients in the cause or spread of the
the magnitude of litigated cases involv- mation. The U.S. Judicial Courts speak fire. Asking questions early and often
ing huge losses led to a comprehensive volumes each time expert testimony is can ensure an FPE’s good faith efforts
re-examination of fire protection engi- excluded from trial due to lack of pro- in exploring conflicts of interest,
neering-related issues, such as sprinkler fessional education or comprehensive thereby reducing the likelihood that a
performance, life safety requirements, experience. The unspoken, but strongly problem will arise at a later date.
fire detection and alarm effectiveness, influencing, trend that is emerging from
and building construction and its rela- the U.S. courts is to exclude expert tes- MANAGING THE SCENE
tionship to smoke and fire migration. timony by those who have not met a
At the same time, a burgeoning amount level of professional education or expe- Responsible fire officials should be
of fire research was taking place at the rience. The days of qualifying as an ex- contacted before an FPE responds to the
National Bureau of Standards (NBS) pert solely on years of fire investigative scene. Local practices will begin to sur-
and other significant institutions as a work are waning. face as officials describe their investiga-
result of the fire-loss statistics. The The presiding judge determines the tive process, conclusions, and means to
study of fire was being funded. Re- acceptable benchmark. An investigator an FPE and other interested parties
searchers and professors began to ad- trained in fire scene examination will seeking access or control of the scene.
dress post-fire analysis in their papers most likely be allowed to testify as to Policies of releasing reports, statements,
and classrooms. The disturbing rise in where the fire originated. As to why the photographs, communication tapes, and
life and property loss led people to re- fire started may not be permitted, and the interviewing of first-responding
alize the need for fire scenes to be ana- most certainly will be challenged. emergency workers will vary. Strong in-
lyzed for more than merely cause and terpersonal communication skills are
origin. THE INVESTIGATION useful. Patience and sensitivity may be
Two major steps to increase profes- advantageous when various interests are
sionalism in the fire investigation field An FPE’s job begins when he or she vying for attention, and are even more
also took hold in the early 1980s: the is contacted by an insurance adjuster or important if lives have been lost, fire-
establishment of an NFPA Committee to an attorney and is hired to undertake or fighters injured, or other complications
develop a guide to fire investigation1 participate in the investigation and have occurred which change the fire
and the International Association of Ar- analysis of a fire loss. The hiring party scene from a property loss to a tragedy.
son Investigators (IAAI) Certified Fire may have called an FPE for several rea- However, most information will eventu-
Investigator credential. Both efforts sons: the magnitude of the loss; the ally be revealed.
sought to increase the knowledge base presence of sprinkler or fire detection An initial meeting with local fire offi-
of the fire investigator as well as rein- systems; the potential opposition has cials or investigators at the fire scene
force sound practices and procedures. hired a FPE; or the client recognizes the will normally involve a walk-through
The influences of changing attitudes need for analysis beyond basic cause and concise briefing of suppression and
in the 1980s that prompted advanced and origin. Once an FPE determines investigative activities. If the incident is
knowledge in the industry stimulated that a conflict of interest does not exist, of an incendiary origin (arson), there
the transition of fire investigation from the investigation is underway. can be reluctance to release particular
a perceived art into an applied science. information; such release may be de-
Fire investigators who once received CONFLICT OF INTEREST pendent on the arson immunity laws of
their primary training and guidance individual states.
from seasoned investigators, inheriting Although the identities of all the po-
their habits and practices, are today ed- tential parties may not be known when THE SCENE
ucated in the scientifically based con- the initial call is received, before ac-
cepts of combustion, fire dynamics and cepting the assignment, an FPE will The overriding urge is to get on the
heat transfer, and the scientific method need to know as much as possible scene and start digging it out. This urge

W INTER 2004 www.sfpe.org 5

 ■ Fire Investigation and The Fire Engineer

must be resisted. Once the local author- underway will not be well received. Be-
ities have completed their site investi- cause of this, there will be delays in get-
gation and have turned control of the ting the plans, meeting with the mainte-
scene back to the property owner, the The first phase nance personnel, locating pre-fire
FPE representing the owner(s) or their of a fire investigation photographs.
insurance carrier must take preliminary directed by an FPE The second reason for waiting before
steps before the dig begins. Regardless initiating the scene dig-out is “spolia-
of the condition of the scene, after the 1. Check for conflicts. tion.” In fire-scene vernacular, spolia-
Authorities Having Jurisdiction (AHJs) tion means knowingly destroying or
have completed their on-site activities, 2. Contact public officials (AHJs) changing the scene (or suspect objects)
the scene must be treated as if it were responsible for the fire scene and without providing other “interested par-
undisturbed and sterile. The condition the public investigation. ties,” such as potential defendants or
of the scene at this juncture must be plaintiffs in a legal action borne from
3. Initiate the gathering of pre-fire
documented. Artifacts examined, the fire, an opportunity to also see it in
information including plans, pho-
opened, and/or removed during the an unmolested state. For example, the
tographs, etc. Arrange to interview
AHJs’ site analysis need to be pho- disassembly, modification, or alteration
individuals who understand the
tographed in situ, and those who al- of potential ignition sources, suspected
construction layout, the site’s use,
tered or moved them must tell of their equipment, or switches which are
as well as the first witnesses of the
location and condition when they were found within the involved area could
fire. The Principal Investigator
found. A concerted, documented effort cause spoliation. Even the appearance
(FPE) should participate in these
must be made to accomplish this. This of impropriety may cause the opposi-
interviews.
effort will give insight into how the tion to cry “spoliation.” This accusation
AHJs’ examination was conducted what 4. Perform a preliminary site survey will not be conveyed to the engineer
they focused on, which will provide a and walk-through. If the fire’s area digging out the scene but rather to the
defensible record for the FPEs if spolia- of origin contains a suspect piece judge if the engineer’s client decides to
tion arises later on. The scene should of equipment, if it was recently litigate. Spoliation, including the dig-
be approached as if there will be only serviced, or if a construction fea- ging out of a fire scene without giving
one opportunity to process it. The FPE ture appears to be contributory to the potential interested parties an op-
should research the construction fea- the loss, put those parties “on portunity to participate in its process-
tures of the building and locations of its notice.” ing, has resulted in dismissal of subro-
contents before the first shovel is gation lawsuits. It may also provide an
turned. 5. Set a date in the reasonable future avenue of financial recovery against an
The technical reason for delaying the to begin the dig-out and disassembly. FPE’s professional liability coverage if
destructive examination of the scene lies 6. Resist efforts to short-cut these the decision to proceed is solely that of
in the understanding of the scene prior steps, especially when the identity the engineer’s.
to the fire. This knowledge can only be of the suspect fire cause or contri-
gained from interviews with those per- bution is known. THE GATHERING OF THE CLAN
sons knowledgeable about the site and
the fire area, the last occupants, the first 7. Identify special equipment or per- After written notices to interested par-
witnesses of the fire, and, if applicable, sonal protective equipment (PPE) ties have been delivered and the manag-
the job site superintendent. Construction needs. Arrange for the equipment ing FPE has performed a site safety as-
plans and details, pre-fire photographs, ahead of time, and convey the PPE sessment, the processing of the scene
and actual fire footage are also valuable needs to all interested parties. begins. The site safety assessment in-
sources. cludes identification of physical hazards,
such as unstable walls and roofs; the
INTERVIEWING WITNESSES – directing the investigation need to be in- distribution of loads and, in particular,
WHAT THEY DIDN’T TEACH YOU IN volved in the interviewing process. loads in the fire-affected areas; biohaz-
ENGINEERING SCHOOL Once again, patience and self-awareness ards; the potential energy and electrical
are necessary. A person who has just hazards; the recommended or required
More often than not, witnesses deter- lost a business, a home, or – worse yet – PPE; as well as an analysis of the envi-
mine the life and death of a fire investi- a loved one is in no frame of mind to be ronment.
gation. The sooner the key witnesses are interviewed. Being overly assertive in A number of steps should be taken
interviewed, the more likely the infor- trying to get all of this “needed-now” in- once the group of interested parties has
mation received will be reliable. Those formation because an investigation is assembled. After everyone has signed

6 Fire Protection Engineering N UMBER 21

 ■ Fire Investigation and The Fire Engineer

in and identified whom they represent, constraints on how much information sufficient to prevent the destruction of
a briefing and group discussion should can be shared. artifacts by the twist of a knob or the
be held. Preferably, the briefing will in- The managing FPE should facilitate an opening of a panel. Sometimes, not-so-
clude public information about the fire, open discussion about the examination subtle reminders are required.
witness information (if it helps identify protocol including distribution of re-
the area of origin), fire alarm data (in- sponsibilities such as shoveling, procur- ATTORNEYS AT FIRE SCENES
cluding report printouts), copies of ac- ing equipment, and conducting nonde-
curate floor plans, and generally any structive measurements. The group Attorneys will often be at the scene
non-prejudicial information which will should discuss how the scene will be during the group investigation. Their
assist in the group investigation. The documented (mapped); if this documen- principal roles are to protect their
client or counsel must preapprove the tation will be “pooled” and made avail- clients’ interests, to become familiar
distribution of information, including able to all; how the evidence will be with both the scene and recovered evi-
nonprejudicial information, before it is identified, removed, tagged, and stored; dence, and to ascertain what the poten-
shared. Generally, other investigators and, finally, site safety. tially litigated issues may be. They also
are not given open access to the FPE’s The group should also plan to for- provide cover for their respective engi-
clients or their employees. The FPE mally meet at a designated time to re- neers and investigators when control
should prearrange for first-in firefight- view what has been accomplished, up- and decision issues arise. Consider, for
ers to provide a group briefing and date any safety concerns, identify example, this scenario: At a fire investi-
walk-through, and videotape the meet- which artifacts need to be secured, and gation scene, an interested party wants
ing, if possible, to memorialize what assign new responsibilities. The group to partially disassemble the battery
transpired. investigation works well if everyone is charger of a forklift to see if it still has a
allowed to physically participate. specific component. This action con-
DIRECTION OF THE Things usually become skewed if peo- tradicts an earlier consensus of the
INVESTIGATION ple are allowed to remain idle and group to not open equipment at the
watch the site-controlling team do the scene. Those who want to open it,
When multiple “interested parties” work. however, have established an alliance
have been invited to participate in the As the scene is being processed (dug with other involved parties. They want
scene examination, it can trigger a po- out), it is essential to strive for contin- to deviate from the agreed-upon exami-
tentially burdensome avalanche of in- ued consensus among the group. Of- nation protocol. Rather than straining
quiries for local authorities. Requests to tentimes, seemingly small actions can the dynamics of the investigating team
interview the first-arriving firefighters are challenge this dynamic, especially if with a potentially deteriorating postur-
expected and normally accommodated, they involve opening up a piece of ing discussion, the group can turn to
but this can become time-consuming. In equipment. The forced opening of a the attorneys. It is the responsibility of
anticipation of these types of needs, an thermally damaged widget can alter a legal counsel representing the involved
FPE can prearrange a group briefing or potentially important condition. It is far parties to discuss the various positions
interview. These gatherings may be better to conservatively approach these with each other and to make the deci-
more productive when conducted in a “let’s just open it up” desires. “X-ray sion whether to open it or to wait, thus
meeting room as though they were before disturbing or distorting” is the allowing the esprit de corps of the in-
briefings: information is given out, and mantra of forensic electrical engineers, vestigating group to continue.
questions and answers follow. This is a conservative position also held by
also the proper setting to distribute most fire investigators. There are those NOT YOUR SCENE
plans, documents, pre-fire photos, and whose natural curiosity can cloud their
fire photos, if such are available. This ef- under- standing of “destructive exami- If an FPE represents an equipment
ficient and considerate vehicle for dis- nation.” Frequently, these individuals manufacturer or service provider, it
tributing information will help expedite have little or no fire scene experience, may be very beneficial to the clients’
the actual scene examination and pre- and thus are not as disciplined in the position to provide some technical in-
serve the group’s relationship with the art of “looking without altering.” The formation to the group. Consider the
AHJs and witnesses. usual suspects include manufacturer’s manufacturer of an electrical switch
This group approach can also be used field engineers, maintenance employ- panel who is “put on notice” and
when important witnesses are employ- ees or owners of the fire-affected busi- chooses to send an investigator to par-
ees of an FPE’s clients, with a caveat that ness, curious insurance adjusters, or ticipate in the dig-out. The switch panel
the responsible attorneys should decide anyone who carries on in a poke-and- is opened (nondestructively), and a ser-
if and how this can be accomplished. prod, look-see manner. Preliminary re- ial number and cryptic date codes are
The attorneys should also define the minders during the briefing may not be located. The investigator is then asked

7 Fire Protection Engineering N UMBER 21

 ■ Fire Investigation and The Fire Engineer

to identify when the panel was made. REPORTS DISCLOSING THE the Court’s position. According to Rule
The motivation to reveal this informa- FPE’S OPINIONS 702, U.S. Federal Rules of Evidence:
tion is clear: Many states have “statutes Discovery is the process by which in- “If scientific, technical, or other spe-
of repose” – a defined period of time formation regarding all aspects of the lit- cialized knowledge will assist the trier
for which a manufacturer may be held igation is developed. This includes the of fact to understand the evidence or to
liable for design and fabrication de- posing of questions, document produc- determine a fact in issue, a witness
fects. If the switch panel in the scenario tion, depositions, and expert reports. qualified as an expert by knowledge,
above is older than the defined period When discovery is completed and an skill, experience, training, or education
of time, the appeal of the manufacturer FPE has been identified as a testifying may testify thereto in the form of an
as a potential defendant is greatly di- expert, when all of the fact witnesses opinion or otherwise, if (1) the testi-
minished. If the information as to the have been deposed, when the attorneys mony is based upon sufficient facts or
date of manufacture is not shared, and have conducted their information-gath- data, (2) the testimony is the product of
the switch maker is sued, the time and ering investigation throughout the dis- reliable principles and methods, and
economic costs to all parties are greatly covery process, when the recovered arti- (3) the witness has applied the princi-
increased. If sharing technical informa- facts have been examined, and the ples and methods reliably to the facts
tion can help extract an FPE’s client expert disclosure deadline looms, it can of the case.” 2
from potentially lengthy and costly liti- only mean one thing: expert report time. This component of Rule 702 is
gation, then the information should be Once again, an FPE should keep close known as “Daubert” and from it came
shared – after permission from the contact with the responsible attorney to the Daubert Tests or Factors. Simply
client to do so has been obtained. determine what, if any, formal report stated, they are as follows:
prepared by the FPE will be required 1. Has the theory or technique been
REPORTS and, if one is required, what needs to be tested?
included in it. The report requirements 2. What is the rate of error or reliabil-
The premature preparation of re- span a vast range, which is a function of ity of the theory or technique?
ports, which declare or suggest the the jurisdiction in which the litigation is 3. Has the theory or technique been
opinions of an FPE or others in the being conducted. Some states simply re- subjected to peer review and pub-
team, can be detrimental. All docu- quire that the expert’s opinions and the lication?
ments prepared by an FPE will be dis- basis for them be included in the “An- 4. Does the theory or technique en-
closed to the opposing parties if the swers to Interrogatories,” which is pre- joy general acceptance within the
matter is litigated. The FPE will have to pared by the attorneys. For the expert, scientific community?
answer for each opinion and its basis. these disclosures are the least invasive, These factors are applied in U.S. Fed-
If changes are later made, the modified for the submitting attorney, not the ex- eral Court and in many state courts as
opinion will be a red flag and invite ad- pert, signs the document. If this is the well. Theories or opinions held by an
ditional scrutiny by the opposition. Lac- means of opinion disclosure, the FPE FPE will be subjected to the Daubert
ing opinions with caveats such as should review what the attorney has tests in these courts; in states that have
“based upon the information as identi- written and attributed to the FPE before not adapted the Daubert standard, the
fied in appendix E” or “if additional in- documents are sent to the opposing par- expert disclosure and qualifying re-
formation is received, I reserve the ties. The way in which the position is quirements may be less stringent. How-
right to change or alter my opinions” presented must be critically scrutinized. ever, expect opposing counsel to use
helps, but not much. The expert should not expect counsel to Daubert as an examination tool in at-
The volley of questions from oppos- be familiar with the nuances of the fire tempts to discredit witnesses during
ing counsel in a deposition or, worse, safety engineering language, but know cross-examination.
at trial can be avoided by not prema- that the experts engaged by the oppos- The impact of Daubert is immense.
turely preparing opinionated reports. ing side will be. An FPE may first face Daubert’s effect
Clients will often require status reports when opposing counsel files a motion
on the state of the investigation. Simply EXPERT REPORTS AND THE to have the expert’s opinions and testi-
providing information detailing the FEDERAL COURTS mony excluded because they did not
process and what was observed, with- meet the standard set forth by Rule 702.
out expressing opinions, may fulfill this Within the last decade, the U.S. This pre-emptive motion may be based
requirement. Preparing reports which Supreme Court has gone to great solely on the expert report provided
publish an FPE’s opinions prior to the lengths to clarify and promulgate rules during discovery. If FPEs fail to ade-
completion of all litigation-driven dis- regarding the admissibility of scientific- quately present their opinions and the
covery should be resisted, if not com- based expert testimony. Tomes have basis for them, they are vulnerable to
pletely avoided. been written on the cause and effect of exclusion. The basis must also formally

8 Fire Protection Engineering N UMBER 21

cite the treatises relied upon by an FPE Transcripts of trial testimony and expert REFERENCES
that support a theory and confirm its depositions are saved, databased, and
general acceptance by the scientific traded as though they were baseball 1 NFPA 921 – Guide for Fire and explosion
community. cards by attorneys. A number of fee- Investigations, 1992 edition; National Fire
The second wave of challenge be- based testimony and deposition reposi- Protection Association, Boston, MA., 1992.
comes personal for the FPE. He or she tories also exist. ▲ 2 Federal Rules of Evidence, Federal
will be evaluated on knowledge, skill, ex- Judicial Center, Washington, 2000.
perience, training, and education. Oppos- Robert Schroeder is with Schroeder Fire.
ing counsel will conduct an examination
and attack on an FPE’s background and
ability to render the opinions that have
been presented. Does this person have
formal training with the information used
to support the opinions? Does he or she
possess the experience to adequately
qualify them in this particular area? Is the
FPE simply parroting the words of the at-
torney or client for whom he or she is
working? Was this supposed expertise
gained through a nontested short course
or seminar? Has the FPE ever conducted
any research or performed the actions be-
ing commented on? This is just a sampling
of the scrutiny that the expert should ex-
pect. If the inquisitor finds that the FPE
has overstated his or her qualifications,
experience, or understanding of the sup-
porting basis of their opinions, results can
be humiliating or even disastrous.
If an FPE has not received some for-
mal fire investigation training combined
with supervised site examination expe-
rience, the likelihood of successfully
qualifying as a cause-and-origin expert
is grim, as it should be. If opposing
counsel is successful in disqualifying an
FPE in one area such as cause and ori-
gin, the remaining opinions and areas
of expertise will also be challenged. At-
tacks on the credibility of an FPE will
not come early in the life of the litiga-
tion. They are often held back until the
eve of trial or, worse yet, when an FPE
is on the stand.
It should also be noted that if the op-
position is successful in getting the
FPE’s opinions disqualified or discred-
ited, the discomfort does not end in the
courtroom. There is now a public
record. Although similar records of
each deposition taken of an FPE exist,
ill-supported theories or overstated
qualifications made by a discredited
FPE become part of the public domain.

W INTER 2004 www.sfpe.org 9

 NFPA 921’S IMPACT ON

FIRE SCENE
RECONSTRUCTION
By David J. Icove, Ph.D., P.E., and ment, and it explains aspects of the fire In order to be effective, these expert
John D. DeHaan, Ph.D. and smoke development, the role of fu- opinions must be able to pass the even-
els, effects of ventilation, the impact of tual scrutiny of cross-examination during
INTRODUCTION manual and automatic extinguishment, sworn depositions, peer review, and
the performance of the building, life courtroom testimony. Recent court deci-

C ourts now prefer science-based

methods when describing and
providing testimony in fire
scene reconstruction. These science-
based methods include a reliance on
safety features, and manner of injuries or
death.
Forensic fire scene reconstruction re-
lies upon a comprehensive review of the
sions place more weight on expert
forensic testimony based upon scientific,
rather than merely experience-based,
knowledge.1
fire pattern damage, sound fire protec- When determining the origin and
The National Fire Protection tion engineering principles, human fac- cause of a fire, a comprehensive recon-
Association’s NFPA 921 – Guide for tors, physical evidence, and an appropri- struction often involves a fire engineering
Fire and Explosion Investigations, as ate application of the scientific method. analysis that tests various scenarios. This
well as on other expert treatises. These factors often form the basis for an analysis may use fire modeling to com-
Fire scene reconstruction is the expert opinion as to the most probable pare actual events with predicted out-
process of determining the most likely origin and cause of the fire or explosion. comes by varying causes and growth sce-
development of a fire using a scientifi- The expert opinion may be part of a narios. This engineering analysis adds
cally based rationale. Reconstruction fol- written report or the basis for oral testi- value, understanding, and clarity to an al-
lows the fire from ignition to extinguish- mony in depositions or courtrooms. ready complex fire scene investigation.

W INTER 2004 www.sfpe.org 10

 ■ Fire Scene Reconstruction

THE NEED FOR SCIENCE IN FIRE and testing, is the best approach for The use of the scientific method is
SCENE RECONSTRUCTION conducting fire scene analysis and re- well received in both the technical and
construction. NFPA 921 – Guide for Fire research communities. An investigation
A federal conference in November and Explosion Investigations defines conducted using this approach is more
1997 assessed the current state of the the scientific method as the systematic likely to be embraced by those who
art and identified technical gaps in fire pursuit of knowledge involving the rec- would tend to doubt a less vigorously
investigation.2 The International Con- ognition and formulation of a problem, conducted probe. Second, the scientific
ference on Fire Research for Fire Inves- the collection of data through observa- method is an accepted protocol of
tigation concluded that many scientific tion and experiment, and the formula- practice for NFPA 921.
gaps existed in the methodology and tion and testing of a hypothesis.3 Those persons ignoring or deviating
principles used to reconstruct fire from NFPA 921 practices would bring
scenes. Benefits of the Scientific Method closer scrutiny to their reports and
opinions. Another guide that repeatedly
The Scientific Method There are numerous benefits for us- cites NFPA 921 is the National Institute
ing the scientific method to examine of Justice’s Fire and Arson Scene Evi-
The underpinning of forensic fire fire and explosion cases, including: dence: A Guide for Public Safety Per-
scene reconstruction is the use and ap- sonnel.4 The NIJ guide notes that ac-
plication of relevant scientific principles • Acceptance of the methodology in tions taken at the scene of a fire or
and research in conjunction with a sys- the scientific community, arson investigation can play a pivotal
tematic examination of the scene. This • Use of a uniform, peer-reviewed role in the resolution of a case.
is particularly true in cases that later re- protocol of practice, such as NFPA A thorough investigation is key to en-
quire expert witness opinions. The sci- 921, and suring that potential physical evidence is
entific method, which embraces sound • Improved reliability of testimony neither tainted nor destroyed, nor po-
fire protection engineering principles from opinions formed using the tential witnesses overlooked. More im-
combined with peer-reviewed research scientific method. portantly, expert testimony in fire and

11 Fire Protection Engineering N UMBER 21

explosion cases will rely more heavily • Identifies important future research knowledge and time was required to
on opinions formed using the scientific areas in fire investigation. collect and analyze the information on
method. Recent U.S. Supreme Court de- the development of the fire.
cisions underscore these principles, with Historically, the use of engineering Engineering analysis and modeling
many state courts following the trend. analysis and modeling in fire scene re- are becoming more commonplace. De-
construction was conducted on a case- spite the complexity of the processes,
Engineering Analysis as a Scientific by-case basis, due mainly to the com- fire engineering analysis and modeling
Method plexity of the process. A vast amount of should be applied to cases involving

Fire engineering analysis, which can

range from a basic “back-of-the-enve-
lope” first-level calculation to a sophis-
ticated fire model, is an important step
in fire scene reconstruction that pro-
vides numerous benefits. Historical in-
vestigations using fire engineering
analysis have accurately assessed fire
development, measured the perfor-
mance of fire protection features and
systems, and predicted the survivability
and behavior of people during the
incident.
Several significant fire engineering
analysis studies have been undertaken
by the NIST over the years. These de-
tailed reports consist of the Dupont
Plaza fire,5 First Interstate Bank Build-
ing fire,6 Pulaski Building fire,7 Hill-
haven Nursing Home fire,8 Happyland
Social Club Fire,9 62 Watts Street fire,10
and Cherry Road fire.11 Some form of
fire engineering analysis is now consid-
ered a necessary step in comprehensive
fire scene analysis and reconstruction.

Benefits of Engineering Analysis

Fire engineering analysis, combined

with modeling, offers the following
benefits:
• Establishes the basis for the collec-
tion of data needed to construct
event time lines;
• Provides ignition sequences, and
failure modes and effects analysis
(FMEA);
• Invites application of the scientific
method when testing hypotheses
and validating fire scene recon-
struction;
• Provides a viable alternative to full-
scale fire testing and extends full-
scale test results to differing ranges
of conditions (sensitivity analysis);
• Provides answers to many impor-
tant questions raised on human
factors, ignition sequences, equip-
ment failure, and fire protection
systems (detectors, alarms, and
sprinklers); and

W INTER 2004 www.sfpe.org 12

 ■ Fire Scene Reconstruction

multiple deaths, or cases where code knowledge and validate the methods sistent with on-scene observations,
deficiencies contributed to the fire, or shared in common by both groups. physical evidence, and testimony
in the anticipation of extensive civil or from witnesses.
criminal litigation. THE SCIENTIFIC METHOD AND
A fire engineering analysis often uses NFPA 921 6. Test the Working Hypothesis (De-
fire modeling to compare actual events ductive Reasoning). The working
with predicted outcomes using varying Listed below is an annotated seven- hypothesis should be compared with
fire causes and growth scenarios. Sur- step systematic process, based upon all other known facts, the incidence
face temperatures on the walls can be NFPA 921, for applying the scientific of prior loss histories, authoritative
calculated by the fire model and dis- method to fire investigation and recon- fire test data, sound published trea-
played using a color-coded gradient struction. tises, and experiments. The working
scale. Results from the surface tempera- hypothesis should be used to elimi-
tures can be compared with burn pat- 1. Recognize the Need. First respon- nate all other reasonable origins and
terns to confirm the sequence of the ders must recognize that a scene causes for the fire or explosion. The
fire. This analysis often adds value, un- needs to be protected until a full in- investigator should recommend the
derstanding, and clarity to complex fire vestigation can be started. This in- collection and analysis of additional
scene investigation. Fire modeling can cludes not only the origin and cause data, seek new information from wit-
also serve as an alternative to full-scale of the event, but also recognizes the nesses, and develop or modify the
fire testing to explain burn patterns and responsibility to determine if future working hypothesis. This may in-
the fire dynamics involved. fires, explosions, or loss of life can volve reviewing the analysis with
Fire engineering analysis and model- be prevented through new designs, other investigators with relevant ex-
ing may also provide answers to impor- codes, or enforcement strategies. perience and training. Interactively
tant questions raised during fire investi- repeat Steps 4, 5, and 6 until there
gations. These questions might include 2. Define the Problem. A tentative in- are no discrepancies between the
the following: vestigative plan should be devised to working hypotheses. By testing vari-
preserve and protect the scene, de- ous working hypotheses rigorously
• What was the most probable cause termine the cause and nature of the against the data, those that cannot be
of the fire (i.e., can several possi- loss, conduct a needs assessment, conclusively eliminated should still
ble causes be eliminated)? formulate and implement a strategic be considered viable.
• How long did it take to activate fire plan, and prepare a report.
or smoke detectors and sprinklers? 7. Select Final Hypothesis (Conclu-
• What were the smoke and carbon 3. Collect Data. Facts and information sion or Opinion). When the work-
monoxide levels in each room after about the incident should be col- ing hypothesis is thoroughly consis-
10 min? lected through direct observations, tent with evidence and research, it
• Why didn’t the occupants of the measurements, photography, evi- becomes a final hypothesis and can
building survive the fire? dence collection, testing, experimen- be authoritatively presented as a
• Was an accelerant used in the fire, tation, historical case histories, and conclusion or opinion of the investi-
and if so, what type? witness interviews. All collected data gation. If a final hypothesis cannot
• How much time elapsed from should be subject to verification of be determined, the cause should be
when the owner of the structure how it was legally obtained, its chain reported as “undetermined.”
left the building until the fire of custody, and notation as to its reli-
reached flashover? ability and authoritative nature. SCIENCE-BASED EVIDENCE
• Did a negligent building design or
failed fire detection and suppres- 4. Analyze the Data. The investigator Recently, the U.S. Supreme Court has
sion system contribute to the relies upon his or her knowledge, ruled on the admissibility of expert sci-
growth of the fire? training, and experience in evaluat- entific and technical opinions, particu-
• What changes to policy, building, ing the totality of the data. This sub- larly as they relate to fire scene investi-
or fire codes are necessary to en- jective approach to the analysis may gations. These decisions impact how
sure that similar fires will not occur include knowledge of similar loss expert testimony is accepted and inter-
in the future? histories (observed or obtained from preted.
references), training and understand- Although much controversy exists
Discussion and cooperative research ing of fire dynamics, fire testing ex- over these Court decisions, fire and
between fire investigators and fire pro- perience, and experimental data. explosion investigation is emerging
tection engineers underscore the need more as a “science” and less of an
to include fire modeling as an integral 5. Develop a Working Hypothesis. “art.” This is particularly true when
and required step in fire scene recon- Based upon the data analysis, a ten- combining the use of the scientific
struction. Fire investigators and engi- tative working hypothesis should be method with relevant engineering prin-
neering professionals need to collabo- developed to explain the fire’s origin, ciples and research in providing expert
rate further in order to expand their cause, and development that is con- witness testimony.12, 13

13 Fire Protection Engineering N UMBER 21

 ■ Fire Scene Reconstruction

A judge has the discretion to exclude Appeals applied Daubert and excluded THE IMPACT ON TESTIMONY IN
testimony that is speculative or based the testimony of a fire investigator in COURT CASES
upon unreliable information. In the case the Benfield case.19 The Court held that
Daubert v. Merrell Dow Pharmaceuticals, the investigation of fires is science- The importance of NFPA 921 has
509 U.S. 579 1993,14 the Court placed the based and that the Daubert criteria also been cited along with Daubert as
responsibility upon a trial judge to en- apply. an interlinking element of expert testi-
sure that expert testimony was not only NFPA 921 specifically cautions that mony since it establishes guidelines for
relevant but also reliable. The judge’s “the elimination of all accidental causes the reliable and systematic investigation
role is to serve as a “gatekeeper” to de- to reach a conclusion that a fire was in- or analysis of fire and explosion inci-
termine the reliability of a particular sci- cendiary is a finding that can rarely be dents.21
entific theory or technology. The Court justified scientifically.” If pursued vigor- Several clusters of recent federal
defined four criteria to be used by the ously, however, the scientific method court opinions and rules fall into the
“gatekeeper” to determine whether the can be used to demonstrate success- following areas.
expert’s theory or underlying technol- fully that the only mechanism for igni- • Use of investigative protocols,
ogy should be admitted. tion had to be deliberate by demon- guidelines, and peer-reviewed cita-
The four-guideline Daubert criteria strating that all relevant accidental tions,
consist of testing, peer review and pub- mechanisms had been specifically eval- • Methodological explanations for
lication, error rates and professional uated, tested, and eliminated. burn patterns, and
standards, and general acceptability. In • Qualifications to testify.
short, expert testimony must rely on a Error Rates, Professional Standards, Professional education is paramount
balance of valid peer-reviewed litera- and Acceptability in any profession. It is incumbent on all
ture, testing, and acceptable practices if professional fire investigators continu-
it is to be considered credible by the Under Daubert, the Court considers ously to read and keep abreast of all
courts. the known or potential rate of error and relevant fire, engineering, and legal
In a more recent decision,15 Kumho the existence of acceptable professional publications and critically evaluate their
Tire Co. Ltd. v. Carmichael, the Court standards on the techniques used by conclusions with this ever-changing
applied the four-guideline criteria to ex- the expert. Error rates from repeated knowledge. Today’s knowledge may
pert testimony to determine whether it tests form the basis for many equations, change with new developments, truly
was based upon science or experience. relationships, and models used to de- affecting an established hypothesis.
scribe fire and explosion dynamics.
Peer Review and Publications These error rates are produced during Use of Guidelines and Peer-Re-
data analysis from fire testing. viewed Citations
A credible, reliable theory must take A particular practice that has a broad
into account the body of research that impact on investigators is the Standard In a case involving a residence fire, a
has been compiled, verified, and pub- Practice for Examining and Testing number of independent investigators
lished by experts in the field. The ne- Items That Are or May Become Involved attempted to determine the exact origin
cessity for credibility underscores ad- in Litigation.20 This practice covers evi- and cause of the fire. One investigator,
vantages of using the scientific method. dence (actual items or systems) that an electrical engineer, offered the opin-
Recently, there has been a tendency for may have future potential for testing or ion that a television set located in the
courts to hold experts to the same stan- disassembly and are involved in litiga- basement family room caused the fire.
dards that scientists use in evaluating tion. The plaintiffs sued the television’s man-
each other’s work, sometimes referred This practice sets forth the following ufacturer, claiming product liability,
to as peer review.16 guidance: negligence, and breach of warranties.22
Investigators can also participate in • Documentation of evidence prior A federal judge hearing this case held
peer review when their cases are sub- to removal and/or disassembly, a Daubert hearing and concluded that
mitted to supervisors for review. In law testing, or alteration, the one investigator’s causation testi-
enforcement, this is the primary func- • Notification of all parties involved, mony was inadmissible. The judge cited
tion of supervisory review – to assure and that the issue in this case was the sec-
that all questions, logical investigative • Proper preservation of evidence ond Rule 702 factor, that an expert’s
leads, laboratory examinations, and after testing. opinion should be based upon a reliable
plausible theories are addressed. Some methodology. The judge noted that the
experts develop opinions based solely This practice also stresses the impor- investigator did not use a fixed set of
on the results of tests conducted specif- tance of safety concerns associated with guidelines in determining the cause of
ically to support expert testimony. testing and disassembly of evidence. the fire. Notably, the investigator did
The Daubert II 17 Court places greater This is particularly important when confirm that, even though he was aware
weight on testimony based upon pre- dealing with energized equipment or of the existence of established guide-
existing research that uses the scientific evidence containing potentially haz- lines in NFPA 921 and Kirk’s Fire Inves-
method, as it is considered more reli- ardous chemicals. tigation,23 he relied upon his own expe-
able.18 The Eleventh Circuit Court of rience and knowledge.

14 Fire Protection Engineering N UMBER 21

 ■ Fire Scene Reconstruction

During the hearing, the plaintiffs did tion to bar opinion testimony as to the This article is based in part upon ma-
not submit any books, articles, or wit- origin and cause of the fire at the time of terial from the authors’ textbook, Foren-
nesses on proper fire causation tech- trial.26 In this case, both a municipal fire sic Fire Scene Reconstruction, published
niques other than the one investigator. department investigator and a private in- by Prentice Hall. ISBN: 0-13-094205-7,
The judge specifically commented on surance investigator concluded from ©2004.
the fact that the one investigator made their examinations of the scene that the
reference to his reliance on a burn pat- fire was incendiary in origin. No physi-
REFERENCES
tern inside the television set. No cita- cal evidence was taken for laboratory
tions were made to peer-reviewed examination.
1 Lentini, J.J., Standards Impact: The
sources to support this opinion. The plaintiff argued that these inves-
Forensic Sciences, ASTM Standardization
tigators were not reliable and their testi-
News, American Society for Testing and
Methodology Needed monies were inadmissible under Rule
Materials, February 2001. pp. 17-19.
702, because they did not use the scien-
A 1996 residential fire started in the tific method as outlined in NFPA 921 2 Nelson, H., and Tontarski, R., “Fire
corner of a kitchen containing a dish- and relied upon only physical evidence Research for Fire Investigation,”
washer, toaster oven, and microwave observed at the scene. The case also International Conference. Proceedings.
oven. The municipal fire marshal con- contained parallels with the Benfield ATF Inaugural Conference. HAI Report
cluded that the fire had been caused by case.19 In denying the plaintiff’s motion, 98-5157-001. November 11-14, 1997,
the microwave oven, while in a federal the federal magistrate noted in his deci- Baltimore, MD, 353 pp, 1998.
civil case, the plaintiff’s expert asserted sion that the investigators were able to 3 NFPA 921 – Guide for Fire and
that a defective toaster oven caused the provide an adequate methodological Explosion Investigations, National Fire
fire.24 explanation for the analysis of burn Protection Association, Quincy, MA,
A senior federal judge held a Daubert patterns that led to how they reached 2001 Edition.
hearing in which evidentiary and testi- their conclusion as to the fire’s incendi-
monial records were reviewed and ary origin. 4 Fire and Arson Scene Evidence: A
ruled that the opinion of the plaintiff’s Guide for Public Safety Personnel,
investigator was not admissible under Methodology and Qualifications National Institute of Justice,
Rule 702. The judge concluded that the Washington, DC., June 2000.
plaintiff’s investigator did not provide In the case of a residential fire, 5 Nelson, H., “Engineering Analysis of the
sufficient reliable evidence to support which resulted in a federal lawsuit for Early Stages of Fire Development – The
the methodology for investigating the products liability, the judge granted the Fire at the Dupont Plaza Hotel and
cause of the fire. The judge also noted defendant’s motion to exclude the tes- Casino – December 31, 1986,” NBSIR
that the comprehensive nature of NFPA timony of the electrical engineer citing 87-3560, National Institute of Standards
921 contained a methodology that Daubert, Rule 702 and Rule 704.27 In and Technology, Gaithersburg, MD, 1987.
could have supported the opinion and this case, an insurance company inves-
6 Nelson, H., “Engineering View of the
tested the hypothesis for the fire’s tigator called upon an electrical engi-
Fire of May 4, 1988, in the First
cause. neer to remove and examine the
Interstate Bank Building, Los Angeles,
In a federal civil case, an insurance charred remains of a bathroom exhaust
California,” NISTIR 89-4061, National
company sued to recover subrogation fan, clock, lamp, timer, compact disc
Institute of Standards and Technology,
paid by the company for fire damage to player, computer with printer and
Gaithersburg, MD, 1989.
the property of its subrogors. The judge monitor, ceiling fan, power receptacle,
denied the defendant’s motion to ex- and power strip. The engineer came to 7 Nelson, H.E., “Fire Growth Analysis of
clude the trial testimony of the plaintiff’s an opinion that a defect in the compact the Fire of March 20, 1990, Pulaski
origin-and-cause expert witness. The disc player caused the fire. His obser- Building, 20 Massachusetts Avenue,
judge noted that the expert’s testimony vations were based on burn patterns in N.W., Washington, DC,” NISTR 4489.
is the product of reliable principles the room, on appliance remains, and National Institute of Standards and
based upon the scientific method as on his experience, education, and Technology, Gaithersburg, MD, 1990.
outlined in NFPA 921 and that the ex- training. 8 Nelson, H., “Engineering Analysis of
pert applied these principles and meth- In his ruling, the judge noted that the the Fire Development in the Hillhaven
ods in a reliable and relevant manner to training and experience of the engineer Nursing Home Fire, October 5, 1989,”
satisfy the admissibility requirements of did not qualify him to offer an analysis NISTIR 4665, National Institute of
Daubert and the Federal Rules of Evi- of burn patterns and theory of fire ori- Standards and Technology,
dence 702.25 gin. Furthermore, the judge noted that Gaithersburg, MD, 1991.
the engineer did not use the scientific
Methodological Explanations for methodology recommended by NFPA 9 Bukowski, R.W., and Spetzler, R.C.,
Burn Patterns 921 to form a hypothesis from the “Analysis of the Happyland Social Club
analysis of the data, nor did he satisfy Fire With HAZARD I,” Journal of Fire
In the case of a building fire, a federal the requirements for expert testimony Protection Engineering, Vol. 4, No. 4,
U.S. magistrate denied the plaintiff’s mo- under Daubert. ▲ 117-131, 1992.

15 Fire Protection Engineering N UMBER 21

10 Bukowski, R.W., “Modeling a Backdraft: 25 Royal v. Daniel Construction. (Royal 26 Eid Abu-Hashish and Sheam Abu-
The Fire at 62 Watts Street,” NFPA Insurance Company of America as Hashish v. Scottsdale Insurance
Journal, Vol. 89, No. 6, 85-89, Subrogee of Patrick and Linda Magee v. Company, 2000.
November/December 1995. Joseph Daniel Construction, Inc.), Civil 27 American Family Insurance Group v.
Action No. 0 Civ. 8706 (CM), 208 JVC Americans Corp., Civil Action No.
11 Madrzykowski, D., and Vettori, R.L.,
F.Supp. 2d 423, 2002 U.S. Dist. LEXIS 00-27 (DSD-JMM), 2001 U.S. Dist. LEXIS
“Simulation of the Dynamics of the Fire
12397. Decided July 10, 2002. 8001. Decided April 30, 2001.
at 3146 Cherry Road NE, Washington,
DC, May 30, 1999,” NISTIR 6510,
National Institute of Standards and
Technology, Gaithersburg, MD, 2000.
12 Chesbro, K.J., “Taking Daubert’s focus
seriously: The methodology conclusion
distinction.” Cardozo Law Review, Yeshiva
University, 15 (6-7): 1994. pp. 1745-1753.
13 Ogle, R.A., “The need for scientific fire
investigations.” Fire Protection
Engineering, Fall 2000, pp. 4-8.
14 Daubert v. Merrell Dow Pharmaceuticals,
Inc., (509 U.S. 579; 113 S. Ct. 2756; 215
L.Ed.2d 469 (1993).
15 Kumho Tire Co. Ltd. v. Carmichael, 119
S. Ct. 1167 (1999), U.S. LEXIS 2199
(March 23, 1999).
16 Ayala, F.J., and Black, B., “Science and
the Courts.” American Scientist, 81
1993, pp. 230-239.
17 Daubert v. Merrell Dow Pharmaceuticals,
Inc., 43 F 3d 1311, 1317 (9th Cir. 1995).
18 Clifford, R.C., Qualifying and Attacking
Expert Witnesses. James, Costa Mesa,
CA, 2000.
19 Michigan Miller’s Mutual Insurance
Company v. Benfield, 140 F.3rd 915
(11th Cir. 1998).
20 ASTM E860, “Standard Practice for
Examining and Testing Items That Are
or May Become Involved in Products
Liability Litigation,” American Society
of Testing and Materials, West
Conshohocken, PA, 1997.
21 Campagnolo, T., “The Fourth
Amendment at Fire Scenes.” Arizona
Law Review, Fall 1999, p.601.
22 Andronic Pappas, et al. v. Sony
Electronics, Inc., et al. 2000.
23 DeHaan, J.D., Kirk’s Fire Investigation,
Fifth Edition, ISBN 0-8359-5056-5,
Prentice-Hall, Inc., Upper Saddle River,
NJ, 1997.
24 Booth, Jacob J., and Booth, Kathleen v.
Black and Decker, Inc., Civil Action No.
98-6352, 2001 U.S. Dist. LEXIS 4495.
Decided April 12, 2001.

W INTER 2004 www.sfpe.org 16

 FROM THE LAWYER’S PERSPECTIVE:

The Role of a Fire Protection Engineer

By Paul R. Bartolacci, Esq., and
Georgia S. Foerstner, Esq.

A ny discussion surrounding the

role of an expert witness, e.g.,
a qualified fire protection engi-
neer, in loss site investigations and liti-
gation must begin by defining the pur-
poses for which the fire protection
engineer is retained. From the lawyer’s
perspective, the purpose of an expert
in a fire investigation is two-fold: he or
she is an investigator of the loss and
an advocate for his client during litiga-
tion. Too often, experts consider them-
selves only as investigators and disre-
gard or minimize the importance of
being an advocate for their client dur-
ing later litigation. Both roles are
equally important, and an expert
should be mindful of these dual roles
throughout the assignment as they go
hand in hand.
Not only do clients and attorneys rely
on experts to investigate losses and for-
mulate opinions as to their causes, but
they rely on experts to convincingly tes-
tify as to their opinions during litigation
or trial. The best, most accurate opinion
as to the cause of a loss can be severely
in the Investigation of
Fire-Related Losses and
undermined if not testified to or pre-
sented with confidence and conviction.
A way in which to develop confidence
in one’s opinions is to conduct a thor-
ough investigation to ensure that prop-
erly supported conclusions are reached.
This, in turn, will foster conviction in
one’s opinions and assist in communi-
Subsequent Litigation
cating that conviction to an opposing
party, judge, or jury.
In order to provide effective testi-
mony and advocate strongly on behalf fact-finder as weak and not believable. From an attorney’s prospective, select-
of a client, the fire protection engineer Finding the facts to support one’s opin- ing and retaining the proper expert is
must be comfortable with the facts that ions begins at the investigation stage. essential to a successful resolution of a
support his or her opinions and be able claim or defense. The use of expert wit-
to strongly present his or her opinions CONDUCTING A PROPER nesses in litigation has blossomed, in
and conclusions. Opinions and conclu- INVESTIGATION part, because of the specialization and
sions that are not supported by the facts, concentration of practices that all profes-
or the standards of care applicable to The first and most critical step in any sions are currently experiencing. The at-
the industry, will surely come across to a investigation is to have the right expert. torney’s task, on behalf of his or her

W INTER 2004 www.sfpe.org 17

 ■ From the Lawyer’s Perspective

client, is to initially identify the need for a particular forensic The exchange of information, facts, and ideas will typically push
analysis and then ask the proper questions of particular consul- an investigation into areas and directions not immediately evi-
tants to make sure that the consultant has the appropriate exper- dent when the project originates.
tise for the situation. Fire protection engineers should not feel Perhaps the most obvious situations calling for the expertise of
offended if an attorney vigorously “cross-examines” him or her a fire protection engineer arise when there are issues surround-
during the engagement process. It is the attorney’s job to know ing the design, installation, and operation of automatic sprinkler
the law and the standards for admissibility of testimony and systems and smoke/fire detection devices. In order to provide
opinions. Background questioning and interviewing of expert value for a client, a fire protection engineer must understand the
witnesses is designed to allow the attorney to make an assess- design criteria that existed at the time that devices were de-
ment as to whether the particular witness possesses the back- signed, installed, modified, or upgraded. Similarly, a fire protec-
ground, experience, and knowledge to be able to meet the stan- tion engineer should be retained in situations involving compli-
dards for admissibility. Some points are obvious, but warrant ance with building codes or standards, such as fire separations,
repeating. As with any witness, honesty is always the best qual- fire areas, seals around openings and conduits for utilities, and
ity. Fire protection engineers should never stretch their qualifica- flammability standards for particular materials, all of which are
tions or experience in order to “get the project.” extremely important in any fire investigation. Typically, fire cases
After confirming that the right expert has been retained, the involve two critical areas of analysis. The first is why the fire
next step is to define the tasks and duties of the fire protection started, and the second is whether there is some particular event
engineer. Almost universally, the earlier that an expert can be- or design or construction defect that permitted the fire to spread
come involved in a conflict, the more useful and beneficial the into areas that it should not have otherwise spread under the cir-
role of that expert will be. Personally inspecting a fire scene cumstances. Usually, the fire protection engineer’s role is more
and the related equipment, material, burn patterns, and building critical in the second stage of the analysis.
construction, as well as interviewing eyewitnesses, will prove ul- Simply because a property, building, job, or project was de-
timately to be invaluable when testimony is offered. The fire signed, built, or installed does not necessarily mean that all as-
protection engineer should not only address the specific issues pects of the work were performed in compliance with applicable
he or she has been asked to review, but also guide counsel in codes and standards. All too often, fire or other catastrophic
spotting additional issues and educate counsel along the way. events occur in municipalities where code officials or building in-
spectors have simply relied upon certifications of others, such as
sprinkler inspection companies, design professionals, or contrac-
tors, rather than inspecting the items themselves. This is not to
suggest that there is anything unusual or inappropriate about this
situation; in fact, this is probably the norm in most jurisdictions.
Many jurisdictions simply do not have the manpower, time, or re-
sources to determine compliance with every code, standard, or
design practice. Rather, they rely on the certifications of other
professionals who are especially retained to conduct such de-
tailed and in-depth inspections. However, many times, these
companies fail to properly perform their duties. Fire protection
engineers can be particularly useful in identifying design and
construction errors that are overlooked by parties merely because
the work has received prior “approval” or certification.
In addition to determining factors and circumstances that pro-
mote fire, smoke, and water spread, nonoperation of fire-suppres-
sion and fire-detection devices, and design and installation issues,
fire protection engineers are necessary to establish the “standard
of care” with which a particular party is alleged to have failed to
comply. A fire protection engineer can expect to research the ap-
propriate standard of care for design issues, such as hydraulic
analysis, proper sizing of pipes, sprinkler ratings, and the protec-
tion of sprinkler devices from environmental effects. Fire protec-
tion engineers should be able to establish the standard of care
with respect to construction and installation errors involving auto-
matic sprinkler systems and comment upon practices relating to
inspection, testing, and maintenance of sprinkler systems. The
bottom line is that without a fire protection engineer, or a similarly
qualified witness, a party to a lawsuit will likely face a difficult task
establishing what should have been done in a particular set of cir-
cumstances and why the act or failure to act fell below the stan-
dard of care within the governing industry.
The overall task of the fire protection engineering forensic pro-
ject should be to show that a different design, construction, tech-

18 Fire Protection Engineering N UMBER 21

nique, product, or material would have made a difference in the forensic matters in order to overcome the Daubert challenges.
outcome of the event. There must be a nexus or causal connec- For example, today, in a fire spread case, it is risky for a claimant
tion between the event and the issue being reviewed. Code vio- to rely solely on a basic cause-and-origin fire investigator, with
lations or installation errors are factually irrelevant if they play no no fire protection engineering background, to give opinions on
role in leading to the “bad” result or if the same result would issues involving fire spread, automatic sprinkler operation, fire
have nevertheless occurred, even if the design, installation, test- detection devices, or system designs. Retaining the proper fire
ing, and maintenance of a particular system or building compo- protection engineer is essential to complement the other mem-
nent was originally performed correctly. Of course, identifying a bers of the forensic investigative team. The fire protection engi-
series of code violations and other deficiencies in conjunction neer can assist counsel in a manner that will allow for the presen-
with a particular violation that caused a fire or fire spread might tation of facts and opinions that will prove a client’s case.
be extremely beneficial in showing a pattern of sloppy work on Second, Daubert means that the fire protection engineer must
the part of a particular party. ensure that his or her opinions are factually and scientifically sup-
ported. An expert should be especially mindful of this require-
PROVING YOUR CASE ment when preparing his report and testifying during deposition,
as these two things generally define the scope of the opinions and
After the fire protection engineer has inspected the fire scene, conclusions that an expert can offer at the time of trial. Although
reviewed all of the appropriate codes and standards, reviewed all an expert will prepare a report in preparation of litigation, the re-
of the documents exchanged during the course of litigation, and port itself is not admissible at trial. Rather, the expert must testify
prepared his or her report, the next step is proving the case. This as to his or her opinions at trial. However, this does not mean
is where the advocacy role of the expert comes in. Before ad- that an expert’s reports, letters, and other communications are not
dressing how to best go about proving one’s case, it is important discoverable by the other side. To the contrary, experts should be
to have an understanding of the potential pitfalls that attorneys mindful of what documents they place in their files as everything
and experts face in having an expert disqualified or having ex- that an expert reviews and relies upon will most likely be discov-
pert opinions excluded because they are not properly supported. erable by the opposing party and can be used to impeach an ex-
In today’s courtrooms, there is a harsh legal climate restrict- pert’s opinions and credibility during a deposition or at trial.
ing the ability of parties in lawsuits to present “opinion” evi- Experiments and tests that a fire protection engineer under-
dence that supports their respective positions. Obviously, the
goal in any lawsuit brought by a plaintiff, or resisted by a de-
fendant, is to develop facts that will convince a fact-finder
(typically a jury, but sometimes a judge) that a particular the-
ory, scenario, or position is correct. But equally important, and
sometimes overlooked, is the need to build the proper founda-
tion that enables those parties to present that evidence, i.e.,
the story, the documents, the physical evidence, through ap-
propriate witnesses whom the court will permit to provide ex-
pert testimony at the time of trial.
Now, more than ever, this potential stumbling block domi-
nates trial and evidentiary rulings as they relate to expert or
opinion testimony. The strict standard for admissibility has its
genesis in the 1993 United States Supreme Court landmark deci-
sion in Daubert v. Merrell Dow Pharmaceuticals, Inc. There, the
Court directed trial judges to be “gatekeepers” of evidence and
responsible for evaluating and guarding against improper evi-
dence making its way to a jury for consideration. The Court im-
plored trial judges to keep “junk science” out of the courtroom.
To identify “junk science,” the courts generally look at the over-
all “reliability” of an expert’s opinions as the cornerstone of ad-
missibility. Courts consider factors such as the background and
experience of the proposed expert witness; the existence of
standards against which the acts in question can be judged;
whether there have been peer review processes involving the
expert’s opinions on the same issue; whether the opinions, the-
ories; and scenarios have been tested; and whether the expert
has utilized the appropriate methodology in reaching his or her
conclusions.
How then does this preliminary evaluation by a judge impact
the role of a fire protection engineer? First, it potentially translates
into more business opportunities for the expert. Because of the
heightened standard of admissibility for expert testimony, more
and more fire protection engineers will need to be involved in

W INTER 2004 www.sfpe.org 19

 ■ From the Lawyer’s Perspective

takes also become part of his or her which attention spans are very limited. system would or would not have been
work product. While it is certainly en- Thus, there is an increasing use of com- overwhelmed by an intentional fire,
couraged that the expert and the attorney puter-generated fire modeling and anima- whether an alarm system should have
exchange ideas and discuss particular is- tions that should be within the capabilities alerted a central station or fire depart-
sues in the case, an expert witness, of a fire protection engineer. These types ment at an earlier point in time, or
whether a fire protection engineer or not, of endeavors are particularly helpful whether particular building materials
should never undertake testing or signifi- when there are issues involving storage should have restricted or contained a fire
cant work on a particular project without practices of combustible materials, allega- in particular areas. Here again, the ulti-
discussing it in detail with counsel and tions that a fire was intentionally set and mate admissibility of the work is the pri-
having the project approved. with multiple points of origin, and theo- mary concern. The greatest computer-
The final question then is what is the ries including deficiencies with regard to generated program does an attorney or
best way to communicate the expert’s system design and building construction. his or her client no good if a court ulti-
opinions convincingly so as to prove the Most importantly, in order for these mately rules that a jury cannot see the
client’s case? As with most things, prepa- types of computer programs to be admis- fruits of the fire protection engineer’s
ration is the key and cannot be under- sible in court, they must be properly sup- work. Generally, if the work is presented
valued. Experts should insist upon meet- ported by the facts of the situation. The as a “depiction” of the ultimate opinions
ings with counsel in order to have suffi- proponent of the visual aids and models and conclusions of the fire protection en-
cient time to prepare for anticipated must be able to establish the reliability of gineer who will testify to the content of
cross-examination questions during depo- the program. This requires working the program and modeling, it will be ad-
sitions and at trial. In addition to being closely with counsel on a regular basis missible. On the other hand, if the work
well-prepared for one’s testimony, there and an ongoing exchange of information is intended to be a “re-creation” of an ac-
are other things that an expert can do to and facts that are incorporated into the tual event, courts have typically frowned
convincingly set forth his or her opinions. finished product. These projects are ef- upon that type of potential evidence. ▲
It is now well known that jurors like to fective in terms of establishing what a
see examples of what they are listening sprinkler system would have done had it Paul R. Bartolacci and Georgia S. Foer-
to. We live in a visual society, one in operated properly, whether a sprinkler stner are with Cozen O’Connor.

20 Fire Protection Engineering N UMBER 21

 Failure of a Sprinkler System:

A CASE STUDY
By Daniel Arnold, P.E. pipes. On-duty hospital personnel re- A material and dimensional analysis
sponded immediately, investigated the of the connecting pipe and fitting deter-
INTRODUCTION separation, and shut down the flow of mined that their wall thicknesses and
water. In the meantime, the water groove dimensions were within toler-

F ire sprinkler systems have a rapidly flooded the upper floor of the ances specified by the manufacturer. No
deserved reputation of reliability hospital and cascaded to all floors be- material defects were observed in either
and effectiveness. When properly low, causing extensive damage to the component. However, upon inspection,
designed, installed, inspected, and building, equipment, and contents. striations were evident on the separated
maintained, they are expected to lie In response to the massive flow of end of the pipe section. (See Figure 1.)
dormant for decades, poised ready to water, the hospital instituted their formal
respond to a potentially damaging fire. Internal Disaster Procedures to assure
As a result, fire protection engineers patient care and safety, and to mitigate
have confidence that sprinkler systems the impact of the water. Off-duty per-
will reliably function when called sonnel were called in. Patients were re-
upon to do so. However, on rare occa- located, and efforts were made to stem
sions, sprinkler systems may fail to the flow of water. There were no injuries
perform as intended or expected. reported.
When this occurs, fire protection engi-
neers may be called upon to deter- SEPARATED COUPLING
mine the cause of the system failure.
This article summarizes a portion of an After the water was shut off, it was Figure 1. Separated pipe section
investigation and analysis of one such observed that the pipe separation had
failure: the unexpected separation of a occurred at a mechanical coupling used
mechanical coupling that had been to join a piece of feed main to a fitting AUTOMATIC SPRINKLER SYSTEM
installed for nearly three years. on a wet pipe sprinkler system.
A mechanical coupling connects pipe When originally constructed in the
INCIDENT SUMMARY ends and fittings together using a spe- early 1970s, the hospital was not pro-
cially designed clamping collar and gas- vided with automatic sprinklers through-
At approximately 11:00 p.m., a 2-inch ket. The coupling used two bolts that, out the upper floors. As a result, many
(50 mm) diameter wet pipe sprinkler line when tightened, clamp the pipe ends sprinkler systems were retrofitted into
installed in a large metropolitan hospital and fittings together. Hospital staff took areas of the hospital after the original
suddenly separated and began to imme- photographs of the coupling still in construction. A review of available
diately discharge water. The separated place on the pipe end as it was found records showed that the system that ex-
pipe was above the suspended ceiling of after the water to the sprinkler system perienced the coupling separation had
the high-rise hospital’s eighth floor. The was shut off. only been installed for approximately
hospital building housed in-patient The separated coupling was a flexible, three years at the time of the separation.
sleeping areas, diagnostic and treatment rubber-gasketed fitting, which was listed The automatic fire pump supplying
suites, operating rooms, support, and re- and approved for fire protection service. the sprinkler system and other areas of
lated patient care areas. The joining pipe was manufactured of the hospital campus was rated for 1,000
It was estimated that water discharged thin-wall steel with rolled-groove treated gpm (0.63 m3/s) at 165 psi (1.13 MPa).
from the separated pipes at a flow rate ends with the coupling still attached on The pump developed a maximum churn
in excess of 1,000 gallons per minute one end to a rolled-groove coupling. discharge pressure of more than 230 psi
(0.63 m3/s). In total, more than 10,000 The coupling was compatible with and (1.6 MPa). The fire pump was pre-exist-
gallons (40 m3) of water flowed directly listed for use with the installed fittings ing when the subject sprinkler system
into the hospital from the separated and pipe. was designed and installed.

W INTER 2004 www.sfpe.org 21

 ■ Failure of a Sprinkler System

INITIAL INVESTIGATION From an initial inductive analysis of tection Engineering at the University of
the collected data, the following factors Maryland. A substantially similar replica
Due to the extensive damage and dis- were identified as potential contributing of the fire sprinkler system was con-
ruption caused to the hospital and to factors to the coupling separation. structed using exemplar fittings, cou-
guard against a possible reoccurrence, it • Improperly installed coupling – plings, and piping. The layout of the
was critically important to hospital rep- Photographs of the separated coupling system mirrored detailed measurements
resentatives that the cause of the cou- showed a gap between the mating sur- and technical data from the failed sys-
pling separation be determined. As a re- faces of the coupling’s bolt shoulders. tem. The replicated system addressed
sult, an engineering investigation was (See Figure 2.) A coupling installed per factors such as downstream piping vol-
undertaken to determine the underlying the manufacturer’s installation instruc- ume, specific hanger configuration, and
cause. tions for the coupling exhibited no such fire/jockey pump arrangement and set-
The initial phase of the investigation gap. (See Figure 3.) tings.
included gathering and documenting all • Lack of pressure reducing valve A series of initial tests was conducted
available factual data surrounding the (PRV) – Hydraulic calculation errors to validate the testing concept for the
circumstances of the coupling separa- were made during the design of the sub- replica system and to investigate the
tion. Data was collected through a de- ject sprinkler system that resulted in the general relationship between the identi-
tailed review and documentation of the omission of a PRV for the subject sprin- fied potential factors and their tendency
existing sprinkler system installation, kler system. A PRV was required by ap- to cause or contribute to a coupling sep-
study of available construction records, plicable standards based on the operat- aration. Factors including coupling bolt
nondestructive examination and testing ing pressures of the existing fire pump torque, the presence of a pressure re-
of system components, and eyewitness serving the hospital campus, the sprin- ducing valve, jockey pump cycling, fire
interviews. kler system’s relative location to the fire pump pressure surges and air entrain-
pump, and the rating of ment in the piping system were varied
sprinkler system compo- to determine which, if any, may have
nents. led to the coupling separation (alone or
• Improper hangers in combination with other potential fac-
– Certain pipe hangers tors).
and supports that were
required by applicable JOCKEY PUMP CYCLING TESTS
standards were omitted
near the location of the One of the initial test series focused
coupling separation. on the potential impact of jockey pump
• Improper pump cycling on the integrity of the installed
start pressures – An hy- coupling. The purpose of the testing
draulic calculation error was to investigate if the cycling of the
resulted in the fire pump jockey pump observed by the hospital
start pressure being estab- staff on the afternoon of the separation
lished well below the set- would generate enough pressure, pipe
ting recommended by rel- movement, etc., to loosen an installed
evant standards (NFPA coupling and/or contribute in any way
20). to the coupling separation.
Figure 2. Separated coupling • Jockey pump cycling – On the af- Jockey pump cycling tests were run
ternoon of the separation, the hospital’s on both a properly installed coupling
jockey pump was noted by hospital staff (tightened to the manufacturer’s speci-
to be cycling on and off very frequently. fied bolt torque) as well as on exemplar
• Air in the piping system – The couplings with loosened and finger-
subject sprinkler system piping that sep- tightened bolts. Conservative pump
arated served the uppermost level of the pressure ranges and cycle frequencies
hospital, providing a greater opportunity (i.e., start/stop duration) were used. The
for the accumulation of trapped air. cycling tests were conducted for dura-
tions as long as 15 hours. In each test,
COLLEGE PARK TESTING the sprinkler system piping simply
swayed gently with each start/stop cy-
To assess the contribution of the cle. No separations occurred, and no
above potential factors, tests were con- leaks were observed anywhere in the
ducted at the Maryland Fire and Rescue system including during tests with loos-
Institute (MFRI) in College Park, Mary- ened and finger-tight couplings.
land, under the direction of Dr. Fred Further jockey pump cycling tests
Figure 3. Properly installed coupling Mowrer, Associate Professor of Fire Pro- were conducted using the actual jockey

22 Fire Protection Engineering N UMBER 21

 ■ Failure of a Sprinkler System

pump that was removed from the hospi- was not accomplished. The “finger-tight” tal’s fire pump.
tal. The regenerative turbine jockey and loosened bolts resulted in a gap be- The following identified potential fac-
pump was physically replaced and rein- tween the mating surfaces of the bolt tors were considered in the preliminary
stalled in a test set up to evaluate a the- shoulders similar to that observed in the surge testing. Each potential factor was
ory developed that the repetitive jockey photographs taken of the actual sepa- tested independently and in combina-
pump cycling observed by the hospital rated coupling. tion.
staff would result in a cumulative in- In each hydrostatic test, the replicated • Coupling installation (i.e., varied
crease in downstream system pressure system successfully passed the accep- bolt tightness)
due to the specific configuration of the tance requirements of NFPA 13. No • Installation of pressure reducing
hospital system’s check valves and back- leaks developed or occurred during any valve (set per NFPA 13 requirements)
flow preventer. Reportedly, an indepen- hydrostatic test. Surprisingly, the hydro- • Air in the piping system
dent surge analysis/computer model static tests required by NFPA 13 did not During each of the preliminary surge
predicted that extraordinary pressures uncover the improperly installed cou- tests, the piping system moved dramati-
well in excess of the rated system pres- plings, including those with their bolts cally in response to the rapid introduc-
sure could develop due to the cumula- installed only “finger-tight.” tion of water flow and increase in pres-
tive effects of jockey pump cycling. sure. However, as expected, the
While the model was not reviewed, it SURGE TESTING magnitude of the piping system’s re-
was theorized that excessive forces from sponse to the induced surge was clearly
these predicted excessive pressures Following the initial exploratory test- dampened for tests that included a pres-
could be related to the separation. ing, a series of preliminary surge tests sure reducing valve and those when the
The additional testing clearly demon- was conducted on the replicated system. piping system was “water solid” (i.e.,
strated that excessive pressures pre- For each preliminary surge test, water with no introduced air).
dicted by the computer model would was introduced into the piping system For all but one of the preliminary
not develop from the observed jockey by rapidly opening a valve installed in surge test sequences, the piping system
pump cycling. As expected, the maxi- the sprinkler system’s supply. The char- (including with loosened couplings)
mum pressure achieved by the hospital’s acter and magnitude of the induced maintained the system pressure bound-
jockey pump system was equivalent to surge was established from a study of ary and did not separate or even leak.
the jockey pump shut-off pressure plus the flow and pressure potential of the However, during one of the surge tests,
the maximum available static pressure hospital’s fire pump and the pump “start the test coupling catastrophically sepa-
from the pump source of supply. The pressure” settings. and recorded using rated. (See Figure 4.) It was noted that
maximum potential pressure from the pressure transducers and data recording striations similar to those observed on
hospital’s jockey pump system was well equipment. A similar capacity mobile the actual separated pipe were also cre-
within the working pressure of the pump was used to simulate the hospi- ated on the pipe from the test.
sprinkler system and coupling.

HYDROSTATIC PRESSURE TESTS

Another important portion of the in-

vestigation was to conduct hydrostatic
tests on the replica system to the re-
quirements of NFPA 13, Standard for In-
stallation of Sprinkler Systems. The test-
ing was performed to assess the
hydrostatic performance of the sepa-
rated coupling under varying bolt tight-
ness conditions to the acceptance testing
criteria contained in NFPA 13. To ac-
complish this, the replicated sprinkler
system was hydrostatically tested several
times under two configurations:
• Each coupling on the system was
installed in strict accordance with the
manufacturer’s installation instructions.
Specifically, all coupling bolts were
tightened to the required torque.
• The exemplar coupling at the loca-
tion of the separation was installed with
its bolts loosened and only “finger-tight.”
That is, the final tightening step speci-
fied by the manufacturer’s instructions Figure 4. Separated coupling

23 Fire Protection Engineering N UMBER 21

 To system

PT

Jockey
Jockey pump
PRV pump controller
(Tests 5-7 only)

From Drain From

pump hydrant

CATASTROPHIC PIPE SEPARATION this particular occurrence, these factors pling configuration and separation, and
would have been inconsequential but for the related information should not be
The factors present at the time of the the loose coupling bolts. considered a general review of coupling
coupling separation during the prelimi- The use of a torque wrench by sprin- performance. However, it is somewhat
nary surge testing included: kler system installers is unusual, and instructive on a practical level that there
1. “finger-tight” coupling bolts, traditional rolled-groove couplings do were more than 350 opportunities for a
2. no pressure reducing valve, and not have a visual tell-tale, such as coupling to separate during the surge
3. air introduced into the piping system. break-away bolt heads, to assure mini- testing, yet the only separations or leaks
When only one or two of these factors mum torque is reached. Fire protection that occurred were with couplings that
were present, the coupling did not sepa- engineers should recognize that the were “finger-tight.” This outstanding
rate. absence of a gap between mating cou- performance, as well as the rarity of
Based on the results of the preliminary pling surfaces is not a reliable indicator coupling separations generally in fire
surge test series, subsequent surge test- that the specified bolt tightness has protection systems, is clearly due to the
ing was performed varying coupling bolt been achieved. (For example, the gap robust design of the coupling and dili-
torque, installing a compliant pressure on this coupling tested completely gent installers. Nevertheless, installation
reducing valve, and introducing air in closed when less than 10 percent of the oversights can occur. Fire protection en-
the sprinkler system piping. Again, the required bolt torque was applied.) Fur- gineers in conjunction with the installing
coupling separated only when all of the ther, while counterintuitive, the hydro- contractor should give appropriate con-
above three conditions were present. static testing requirements of NFPA 13 sideration to specific methods to assure
may not reveal improperly installed the tightness of coupling bolts in accor-
GENERAL OBSERVATIONS couplings, including those that are only dance with manufacturer’s requirements.
“finger-tight.”
Sprinkler systems have a deserved Notwithstanding the above, rolled- ACKNOWLEDGEMENTS
history of reliable and durable perfor- groove mechanical couplings are well
mance. This investigation supports this accepted and reliably used in the fire Mr. Duane Johnson, Fire Protection
reputation. However, when failures do protection industry. Overall, the testing Engineering student at the University of
occur, the lessons learned from related conducted in this investigation supports Maryland, is specially acknowledged for
investigations can provide useful knowl- this comfortable reliance. Even when the his assistance in constructing the repli-
edge to fire protection professionals. coupling bolts were installed with signif- cated system and conducting of the test-
The fundamental cause of this cou- icantly less torque than specified, the ing. Also, thanks are extended to the
pling separation was the improper instal- couplings performed well and did not Maryland Fire & Rescue Institute and
lation of a mechanical coupling. Specifi- separate or leak when subjected to this College Park Volunteer Fire Department
cally, the bolts of the coupling were left surge test protocol. for use of equipment and facilities. ▲
finger tight when originally installed. The sole purpose of this investigation
While the lack of a PRV and accumu- was to review a discrete and specific set Daniel Arnold is with Seneca Fire Engi-
lated air in the system also contributed to of potential factors on a particular cou- neering, LLC.

W INTER 2004 www.sfpe.org 24

 Strategies for Reducing the

“CRY WOLF” SYNDROME Part 2

ALTERNATIVE TESTING areas that would fail. Because this parts of the building where the signals
STRATEGIES method does not meet the intent of the are not expected to be marginally close
code, it should only be used where to the required minimum or maximum.

I n some occupancies, such as office specifically permitted by the AHJ and For example, for this batch of meters,
buildings, it is easy to separate the approved by the owner and engineer. one might use an error bar of +/- 10 dB
occupants from fire alarm testing. Testing time can be further decreased based on +/- 6 dB in the tests and a fac-
Background noise measurements can by the use of more meters and more tor of safety of 4 dB. If a measurement
be taken during occupied times, and personnel. It takes very little training to indicates a level of 100 dBA in a space
the audible system tests can be con- use a sound level meter – particularly where the goals is 85 dBA, it can be
ducted at unoccupied times. Similarly, one with a peak-hold function. One concluded that the system meets the
strobes can be inspected during occu- problem is that very few contractors, dis- minimum and does not exceed the max-
pied times and tested during unoccu- tributors, AHJs, and even manufacturers imum: 100 - 10 = 90 > 85 dBA minimum
pied times. Unfortunately, this is not of fire alarm systems show up for a test and 100 + 10 = 110 < 120 dBA maxi-
possible in many occupancies, such as with a certified and calibrated ANSI mum.
apartment buildings, multiple-shift fac- Type 2 meter. Most use a nonrated me- A good testing strategy is to designate
tories, and hospitals. ter available at many electronic stores short, exact times when the system will
Can the number of tests and the total for about $50. be tested. The test plan should be com-
testing time be reduced? For audible An informal test of five nonrated me- municated to the occupants one week
testing, some authorities, owners, engi- ters against a calibrated Type 2 meter and again one day prior to the test using
neers, and service companies use an en- showed that the nonrated meters had an mail, e-mail, bulletin boards, posters,
gineered, statistical approach and target error range of approximately +/- 6 dB.1 etc. On the day of testing, traffic cones
specific locations where the system is Two of the analog-type meters were con- can be placed with signs at all entrances
most likely to produce the lowest SPLs. sistently within 2 dB of the ANSI Type 2 before occupants arrive. If the building
They then check all other areas subjec- meter. A digital version ranged from -4 has an intercom, public address system,
tively and verify that all appliances are dB to +6 dB. A third analog-type ranged or if the fire alarm system has voice ca-
at least operating. This is not a simple from -6 dB to 0 dB difference from the pability, it should be used (in addition to
statistical sampling that has, by defini- rated meter. The largest range for any sin- the other means of communication) on
tion, some probability of missing areas gle test was from -2 dB to +6 dB. the day of testing to inform the building
likely to fail. A careful engineering Because many system designs are try- occupants.
analysis and a thorough site inspection ing to meet a +5 dB signal-to-noise ratio The test should be planned on the
are needed to develop a successful test- (S/N), these nonrated meters would hour, say 10:00 a.m., for exactly three
ing plan. An engineered statistical test clearly not be suitable for accurate test- minutes. One or two more test times can
plan reduces the probability of missing ing. However, they may be useful in be scheduled with at least one or two

25 Fire Protection Engineering N UMBER 21

hours between. If a test time has is a failure of the appliances
been designated, the test team themselves since they are not
should test at the designated time, monitored. For example, the
even if testing is completed in the flash tube and lens of a strobe
first or second time slot. Expecta- light can be broken without
tions should be set, and the occu- generating a trouble signal.
pants should be trained that man- Similarly a pillow or duct tape
agement, the test team, and the fire can be placed over audible
alarm system are reliable. jective evaluations, they may uncover ar- and visible appliances, causing failure of
A good test plan includes provisions eas that should be investigated by the their mission. These types of failures are
for what to do if a real fire occurs during professional testing team. Sufficient time not discoverable by the electrical super-
the testing. This includes how occupants should be allotted before the second vision of the circuits. However, these
should report emergencies and how a scheduled test time of the day to do a vi- failures can often be uncovered as part
real alarm will be communicated to the sual inspection of the “problem” areas of the required visual inspection man-
occupants and to the fire service. The and determine if measurements are war- dated by the fire alarm code. By testing
comic piece “One Morning at the Of- ranted. the occupant notification portion of a
fice” is a story of an alarm technician’s Often, occupants report many areas system only once per year, the “Cry
attempt to come up with such a plan af- where they feel the system is too loud Wolf” ratio is reduced.
ter testing has already been announced.2 and some areas where they feel it is not Several formats or methods should be
To be effective, the plan needs to be loud enough. The test team can be “cali- used to provide information to staff and
thought out in advance and communi- brated” with a little bit of informal train- occupants, and use different sources.3
cated in writing to the occupants. In of- ing and education. The test team should For example, the owner may send a
fices, hospitals, and other buildings be shown the code requirements and al- newsletter, management may send e-
where many of the occupants are em- lowed to take a few readings with a me- mails or place notices on video bulletin
ployees, the emergency procedures ter using a test horn in an area where it boards, the alarm contractor may pro-
should be a part of employee training. won’t disturb others. Systems that are vide posters and signs, and the fire de-
The regular occupants of the space too loud and too soft should be demon- partment may post notices at entrances.
can be solicited to assist in subjectively strated, although they should not be ex- Professionals need to be cognizant
evaluating the fire alarm signal in their posed to noise levels that could damage that fire safety is a system that includes
spaces. In facilities such as hospitals, of- their hearing. Instead, the system’s de- the occupants. It is necessary to weigh
fices, and factories, designated employ- sign maximum should be used. Occu- the effects of one part of the system –
ees can be charged with walking pants should be invited to approach the testing – on the other parts, such as oc-
through specific areas to subjectively sounder from a distance and allowed to cupant response and behavior. By using
evaluate the loudness and effectiveness stop when they feel uncomfortable. Oc- carefully planned testing strategies, the
of the system. Occupants can be given cupants can be given hearing protection success of the system as a whole is
forms to record the room numbers or lo- to approach closer and make measure- increased. ▲
cations and their perceptions of the sys- ments.
tem. Properly done, the subjective part Newsletters, posters, mail, and e-mail REFERENCES
of an engineered statistical approach should be used to inform occupants
serves several functions. Using the regu- when testing has been completed. Oc- 1 These tests were not exhaustive scientific
lar occupants of the building, including cupants should be told that they will be tests: The meters were not arranged to
residents in apartment buildings, makes given notice before the next scheduled provide data logging; analog meters were
them a part of the test, not an outside test of the system, and reminded that read by eye; the meters being compared
varied in age from 2 years to 10 years
observer that is annoyed by the testing. any alarms they hear should be assumed
old; new meters were not included; and
They take on some ownership and de- to be real and that it is critical that they tests were done using only two different
velop some pride in being part of a respond and evacuate quickly. Occu- noise samples, one broadband and one
safety team. They can be trained to re- pants should also be told that the next dominant at approximately 3 kHz.
mind others several minutes before a test won’t be for X months, or better, for
2 Shay, K., “One Morning at the Office,”
test starts. This sets up the expectation one year.
Timothy McSweeney’s Ultimate Makeover,
that any other time the system sounds, it Although statistically it’s better to test http://www.mcsweeneys.net/2001/08/20m
must be real since they were not noti- some part of the occupant notification orning.html, August 2001.
fied in advance and asked to participate. system at each regular test, it causes fail-
3 Sorensen, J., “Risk Communication And
This helps to condition their response. ure in the occupant response part of the
Disaster Warning: Lessons For Counter-
Immediately after each test, occupants protection scheme by increasing the Terrorism,” Oak Ridge National
that assist the test team should be asked “Cry Wolf” syndrome. The wires of the Laboratory, National Academy of Sciences
to relay their recorded perceptions to a system are monitored for integrity so Forum, February 28, 2002,
central location – either by hand, fax, that faults will be known. Between tests, http://www7.nationalacademies.org/ndr/1
phone, or e-mail. Although they are sub- the failure most likely to be undetected Sorenson_Presentation.pdf

W INTER 2004 www.sfpe.org 26

 away, but the ultimate question is what
damage was rendered inside. Some said

Opportunities
the steel columns in the core were dam-
aged, some said the insulation was
blown away, some said the floor truss
systems were severely damaged. State-
ments on each of these issues require
more than opinions; they require scien-
tific analysis. Even simple analyses can

to Learn from
be sufficient, and such are necessary be-
fore anything more complex is even
considered.
The Federal Emergency Management
Agency (FEMA) report1 did not find any
substandard features in the design of the
World Trade Center buildings, “in fact,
many structural and fire protection fea-

9/11
tures... were found to be superior to the
minimum code requirements.” The
FEMA report further pointed out that
prior to 9/11, high-rise buildings have
not fallen in fire. However, the collapse
of the high-rise buildings at the World
Trade Center complex should be ana-
lyzed to determine if the fire resistance
requirements that are presently used are
adequate. Indeed, the process presently
used is not based on engineering at all
in the U.S., but rather a prescription of
good intentions. A Russian text3 recom-
mends a resistance rating for structures
based on the fire duration multiplied by
a factor of safety, which is a function of
the building type. What is the correct
approach? These are some of the fire
protection issues stemming from 9/11.

THE EVENT

By James G. Quintiere, Ph.D. tated the collapse; the steel melted or WTC 1, the first hit, fell in about 104
softened; the collision by the aircraft was minutes, and the South tower (WTC 2)
INTRODUCTION too much for the 110-story twin towers fell in 56 minutes.1 It is noteworthy to
despite the fact that the buildings had consider that the FEMA report1 states

O n September 11, 2001, terror-

ists crashed 767s into the
World Trade Center twin tow-
ers, ultimately causing their collapse.
The Pentagon was similarly attacked.
been designed to withstand a Boeing
707 (traveling at 180 mph [290 km/h]
with a weight of 230,000 pounds
[100,000 kg]) compared to the actual im-
that the impact zone (floors 94-98) of
WTC 1 had an upgraded 1-1/2 inch (38
mm) thickness of insulation on the floor
truss members, while the South tower
pact of the 767 (traveling at approxi- had only 3/4 inches (19 mm). The insu-
The investigation of these catastrophes mately 530 mph [850 km/h] weighing lation thickness specification for the
has been varied. This article examines about 274,000 pounds [124,000 kg]).1 truss system was upgraded to 1-1/2
some of the opportunities to learn The building weight was about 750,000 inches (38 mm) in 1996. Only about a
from these tragic events, so that tons (680 Gg)2 or about 5,500 times that third of the trusses received this upgrade
improvements can be made in the of the 767. The energy expended in by September 11, 2001.4
future through fire protection engi- bringing these aircrafts to rest is mass x Initially, the collapse of WTC 1 & 2
neering. velocity2/2, or about .4 GJ. The 59 14- was attributed to the airplanes and the
Initial statements following 9/11 of- inch (350 mm) square box columns huge amounts of jet fuel. Some simple
fered a variety of explanations as to the could have absorbed this, in part or en- analyses can illuminate these claims.
reason for the collapses of the World tirely, with a plate thickness of about 1/4 Each plane was estimated to have had
Trade Center buildings: the jet fuel fires inch (6 mm) on the face of each tower.1 10,000 gallons (37 m3) of jet fuel on
were significant and therefore precipi- The columns could have been sheared board at impact.1 Based on the asymp-

27 Fire Protection Engineering N UMBER 21

 10

9 Theory

[mF(AoHo1/2)][W1/W2]1/2 kg(min-m5/2)
8

7
Figure 1. WTC 1 & 2 floor and exterior
6
construction
5
totic burning rate for jet fuel (JP-4) of 60
W1W2H
g/m2-s,5 a total burning rate of 242 g/s 4
would result if the fuel burns homoge- Symbol Shape
neously over the entire floor of 63 x 63 m. 121
3
10,000 gallons (37 m3) of jet fuel has a 221
Figure 2. Hit of WTC 2
mass of approximately 28,500 kg. About 211
2
four fireballs of roughly 60 m in diame- With these values, the quantity 441
ter (D) resulted from a jet impact. The AA0-1H-1/2 can be estimated as between 1
mass (M) burned by these fireballs can 15 and 33 m-1/2. For wood crib-like fires,
be estimated as:5 the results of a CIB study6 shown in Fig- 0
ures 3 and 4 correspond to temperatures 10 20 30 40 50
D ( )
1
A/(AoHo1/2) m-1/2
M=
0.314 of 800°C-1000 °C and burning rates of
 5.25  14.5-18.7 kg/s. For typical floor loadings Figure 4. CIB correlation for the
of 30 to 40 kg/m2, the fire duration per burning rate of wood7
or about 9,400 kg per impact. The burn- floor would be in the range of approxi-
ing duration of the remaining fuel over mately two to three hours.
one floor is It is interesting to contrast these re- gate the Pentagon fire. They subse-

D ( )
1 sults with the Standard Time-Tempera- quently spent several days at the Penta-
M=
0.314 ture Curve8 (Figure 5). The burning of
= 79 s. gon site and left when the Federal Bu-
 5.25  fuel types other than wood can increase reau of Investigation no longer needed
the temperatures and the burning rate as them. They were also told that no team
Hence, the jet fuel fire was short-lived. the heat of combustion to heat of gasifi- was needed in New York since the
The jet fuel acted merely as the arsonist cation ratio increases. Thus, for fuel con- cause was clearly arson. ATF has the au-
torch in igniting large quantities of the trolled burning, lighter-weight furnish- thority to investigate federal arson and
building contents. Indeed, all of the jet ings that are more volatile can give would have done more had they be-
fuel did not burn on the floor since ac- shorter fire durations with higher tem- come involved. The National Transpor-
counts have described spillage into the peratures than shown here. tation Safety Board was relegated to
elevator shafts, burning people in the searching for aircraft debris, and the
lobby. THE OFFICIAL RESPONSE WTC and Pentagon events were investi-
Therefore, the majority of the heat gated as terrorist actions, not fire events.
produced in the fires in the buildings Following the attacks on the morning A report was written by The National In-
was due to the burning of the contents – of 9/11, the Bureau of Alcohol, Tobacco stitute of Standards and Technology
the same contents that the fire resistance and Firearms (ATF) convened one of (NIST) on the Pentagon, but that is clas-
design is based upon. This raises a ques- their National Response teams to investi- sified. FEMA did render an investigative
tion that must be answered: Should
buildings be protected from fires similar
to those which occurred in the World 1200
Trade Center buildings? An appropriate
fire protection design should consider
1000
the duration of the fire that is likely un-
der the plausible fire scenarios for the
T(˚C)

building. 800
The temperatures that occurred in
WTC 1 & 2 can be estimated. The area
of openings, Ao, has been estimated as 600
ranging from 124 and 267 m2 for WTC 11
with an opening height, Ho , equal to 400
3 m and a floor area, A, based on a floor 0 10 20 30 40 50
of 63.5 m by 63.5 m and a utility core of A/AoHo1/2(m-1/2)
26.5 m by 41.8 m that is about 4100 m2. Figure 3. Compartment temperatures from the CIB study6

W INTER 2004 www.sfpe.org 28

 ■ Opportunities to Learn from 9/11

1400 effort by contracting to the American Society of Civil Engineers

(ASCE).
The response to the WTC event fell mainly to NYC. The
1200
clean-up was conducted by the Department of Design and
ISO 834 Construction.10 They decided to remove the steelwork and sell
1000 it. Metallurgical analysis of the steel could have detected
changes in the phases of the steel from temperature and there-
Temperature, ˚C

fore an estimate of its temperature, e.g., 1000°C from an analy-

800 sis by Barnett, et al.11 Moreover, each piece of steel used in the
WTC buildings was stamp-labeled to indicate its location in the
building. A limited amount of the steel was, however, recov-
600 Maximum estimated ered by ASCE from the dumpsites.
for the WTC FEMA assumed the federal government role to provide an in-
based on Wood vestigation and enlisted the support of ASCE, which had previ-
400 Collapse and 10 C/min decay ously established a voluntary team. The resulting investigation
North was intended to collect data, make initial observations and rec-
South ommendations, including areas of further research. Its report
200
was presented in May 2002.1 While this study is packed with
much information, it does not go into the detail that an investi-
0 gation of an incident as significant as this warrants. In August
0 50 100 150 200 250 300 350 400 2002, an official federal investigation was begun by NIST,
Time, min funded at a level of $16 million of $60 million that had been re-
Figure 5. An illustration of the WTC fires with the Standard quested. The NIST investigation is planned to take two years.
Time-Temperature Curve9
THE INVESTIGATION

Meanwhile investigative efforts from litigation claims have

been underway and can shed light on the events, but are al-
ways fraught with special interests and restrictions. The Silver-
stein Group (the lease holder on the WTC) commissioned a
study to estimate the damage wrought by the aircraft to the
buildings. Their results, reported by The New York Times,12, 13
suggest that the severe damage to the columns, including the
major load-bearing columns of the core, were responsible for
the collapse. The fire played a role a weakening the columns
according to the study. The The New York Times further reports
that these investigators felt “there was no need to invoke a fail-
ure of the floors to produce collapse.”
Alternatively, the floor trusses have been considered as the
weak links. A conduction analysis of the truss rods, 1.09 inch
(27.7 mm) diameter A36 steel, with properties for the steel and
the insulation obtained from the literature, gave results for the
rod temperatures as shown in Figure 6. These results shown
here are slightly different from the results presented by Quin-
tiere, diMarzo and Becker.14 The results in Figure 6 were devel-
oped by taking the variation of thermal conductivity and spe-
cific heat with temperature into account. While these were
rendered by numerical computations, approximate hand-calcu-
lations gave similar results. The critical temperatures, for buck-
ling of the truss diagonal rod, can range from 630 to 770 °C.
Since the rate of change of temperature with respect to time is
low in this range, the buckling times can vary. A 700°C mean
critical temperature gives failure times of 60 and 100 minutes for
the south and north towers, respectively. These times are con-
sistent with the actual building failure times of 56 and 104
minutes.
The correct answer to the cause of the collapse can have a
profound impact on the future design and construction of
buildings, and must be determined. Other buildings of the WTC

29 Fire Protection Engineering N UMBER 21

 ■ Opportunities to Learn from 9/11

800 REFERENCES
700 1 McAllister, T., ed. “World Trade Center
Building Performance Study: Data Collection,
600
Temperature, ˚C

Preliminary Observations, and Recommenda-

South
500 tions,” Fed. Emer. Man. Agency, FEMA 403,
Washington, DC, May 2002.
North
400 2 Sometime Lofty Towers, A Photographic
Memorial of the World Trade Center,
300 Browntrout Pub., San Francisco, 2001, p. 16.
200 3 Bushev, V.P., Pehelintsev, V.A.,
Fedorenko, V.S., and Yakovlev, A.I., Fire
100 Resistance of Buildings, 2nd ed., V.A.
Pchelintsev ed., Construction Literature
0 Publishers, Moscow, 1970, p.8.
0 20 40
80 60100 120 140 160
4 Graham-Rowe, D., New Scientist, Feb. 8,
Time, min 2003, p.15.
Figure 6. Steel truss rod temperatures due to a fire of 900°C14
5 Mudan, K., and Croce, P., “Calculations
for Large Open Hydrocarbon Fires,” SFPE
also fell, but were not hit by aircraft, and from the federal government to make a Handbook of Fire Protection Engineering,
similarly require a thorough analysis. change in the infrastructure that ad- National Fire Protection Association,
WTC 7, a 47-story building, fell after dresses fire safety and, now, the vulner- Quincy, MA, 1995.
fires were initiated by falling debris. ability to terrorism. The federal govern-
6 Thomas, P.H., and Heselden, A.J.M.,
ment is the only entity that has the “Fully-Developed Fires in Single
THE NEEDS resources and objectivity to develop the Compartment – A Co-operative Research
needed research for standards and the Programme of the Conseil International du
While the collapse of the buildings at education and training for the field. The Batiment” (CIB Report No 20), Fire Res.
the World Trade Center complex was private sector needs to administer the Station, UK, FR Note No. 923, Aug. 1972.
initiated by aircraft collisions into build- process of developing regulations ap- 7 Quintiere, J.G., “Fire Behavior in Building
ings 1 & 2, fires of similar scale could propriate for society, but grounded in Compartments”, Proc. of the Combustion
have occurred without the aircraft colli- science and engineering. Institute, Vol. 29, The Combustion Inst.
sions (the First Interstate Bank and Pittsburgh, PA, 2002.
1 Meridian Plaza fires offer cases in EPILOGUE 8 ISO 834, “Fire-resistance tests – Elements
point). The codes and standards gov- of building construction,” International
erning structural fire protection design Since this article was submitted in Standards Organization, Geneva, 1999.
presently have little science and engi- November 2002, new information has
9 Torero, J.L., Quintiere, J.G., and Steinhaus,
neering supporting them. This lack of emerged. But conclusions on the cause
T., “Fire Safety in High-Rise Buildings.
science makes it impossible to deter- of the WTC collapses remain to be de- Lessons Learned from the WTC,” vfdb-
mine exactly the level of safety compli- termined. NIST is nearing the end of its Jahresfachtagung, Dresden, June 2-5, 2002.
ance provided by the codes and stan- investigation. James Glanz and Eric Lip-
10 Langewiesche, W., American Ground,
dards. While such safety regulations can ton of The New York Times have written
Unbuilding the World Trade Center,
be viewed as a drag on the economy, a book, The City in the Sky, the Rise and North Point Press, 2003.
their ineffective application can produce Fall of the WTC, which sheds some light
high cost effects and may even be defi- on the construction process and deci- 11 Barnett, J.R., Biederman, R.R., and
Session, Jr. R.D., “An Initial Microstructural
cient in the safety they portend. It is sions. The SFPE held a conference in
Analysis of A36 Steel from WTC Building
highly likely that the root cause of the September 2003, “Designing Structures
7,” JOM, Dec. 2001, p. 18.
collapse of the WTC buildings can help for Fire,” that brought together an inter-
illustrate the cost and consequences national perspective on theories of the 12 Glanz, J., and Lipton, E., The New York
these deficiencies cause in the current collapse. Most believe the fire was key. Times, Oct. 22, 2002.
state of the art. Hopefully, these studies will address the 13 Glanz, J., and Lipton, E., The New York
The total cost of fire can approach role of fire protection, the process of as- Times, Oct. 29, 2002.
1% of the GDP.15 The investment of re- signing ratings, and the relevance of the 14 Quintiere, J.G., diMarzo, M., and Becker,
search and education for the develop- standard furnace test method. In a step R., “A Suggested Cause of the Fire-
ment of the science and technology for toward performance-based design, the Induced Collapse of the World Trade
fire safety is about $20 million in the SFPE will soon publish a new guide: Towers” Fire Safety J., 37, 2002, pp. 707-
USA. Based on a total cost of fire of “Engineering Guide to Fire Exposures to 716.
$100 billion, only 0.02 % of this is de- Structural Elements.” ▲ 15 Wilmot, R. (Ed.) “United Nations Fire
voted to research and education. This is Statistics Study,” World Fire Statistics
too little to assure cost-effective and en- James Quintiere is with the University Center, London: 1998.
gineered fire safety. Help is needed of Maryland.

30 Fire Protection Engineering N UMBER 21

 Resources
2004 SFPE Professional
Development Calendar
Monday Tuesday Wednesday Thursday
Spring Professional Development Week March 22 March 23 March 24 March 25

March 22-25 An Enforcer's Guide to Performance-Based Design Review

Tuscany All Suites Hotel Sprinkler Design for Engineers

Las Vegas, Nevada Principles of Fire Protection Engineering

By popular demand, SFPE returns to the west Introduction to Fire Dynamics

with four days of seminars of interest to practic- Simulator and Smokeview
ing engineers and allied professionals. This
Tenability Systems
event will feature 7 seminars including SFPE’s Fire Alarm Systems Design for Smoke
newest in-depth seminar for the AHJ – An En- Management
forcer’s Guide to Performance-Based Design
Welcome Southern Nevada How to Study for the
Review. Reception Chapter Dinner FPE/P.E. Exam

2nd Annual Meeting and Professional 5th International Conference on

Development Conference Performance-Based Codes and
September 20-24 Fire Safety Design Methods
Palmer House Hilton, Chicago, Illinois October 6-8
European Commission Facilities,
Our fall meeting moves to Chicago! Luxembourg
Following in the path of 2003’s
successful event, the 2004 Annual
Performance-based codes, regulations, and de-
Meeting format will include a one-
sign methods are gaining formal and widespread
day professional program addressing acceptance in many countries. For example, in
issues of concern for the practicing the European Union, the recently completed
FPE, as well as the familiar ice cream Benefeu study by the European Commission
social. This will be followed by the found “... it is clear that a majority of Member
traditional Awards and Honors States intend to implement an alternative fire
Banquet, and by four days of educa- safety engineering approach into national build-
tion events, including 10 seminars, a ing fire safety legislation in the future.”
Symposium on Flammable and Combustible Liquids, and an This conference promises to continue the tradi-
expanded Engineering Technology Showcase. tion established by SFPE to present the state of
Seminars being held are: Principles of Fire Protection Engi- the art in performance-based code approaches
and design methods.
neering; An Enforcer’s Guide to Performance-Based Design
Co-sponsored by: Institution of Fire Engineers,
Review; Sprinkler Design for Engineers; How to Study for the International Council for Research and Innova-
FPE/P.E. Exam; Human Behavior in Fire; Tenability Systems for tion in Building and Construction (CIB), and In-
Smoke Management; Dust Explosion – Hazard Recognition, ter-jurisdictional Regulatory Collaboration Com-
Assessment, and Management; and much more. mittee.

31 Fire Protection Engineering N UMBER 21

Name ________________________________________________________________________________________________________________________

Title ______________________________________________________________Employer ___________________________________________________

Address ______________________________________________________________________________________________________________________

City, State/Province ____________________________________________________________________Zip/Postal Code ____________________________

Phone __________________________________________Fax _________________________________e-Mail ___________________________________

Area(s) of Interest:
❒ Spring Professional Development Week ❒ Contract Seminar
❒ 2nd Annual Meeting and Professional Development Conference ❒ Interested in Being an Instructor
❒ 5th International Conference on Performance-Based Codes and Fire Safety Design Methods

UPCOMING EVENTS
March 2-4, 2004 May 2-7, 2004 July 5-7, 2004
Use of Elevators in Fires and Other Emergencies CIB World Building Congress 2004 Interflam, 2004
Atlanta, Georgia Toronto, Ontario, Canada Edinburgh, UK
Info: www.asme.org/cns/elevators/cfp.shtml Info: www.cibworld.nl Info: www.intercomm.dial.pipex.com

March 17-19, 2004 May 9-14, 2004 September 1-3, 2004

Fire & Safety at Sea 5th International Scientific Conference – Public Fire Safety – Professionals in Partnership
Melbourne, Australia Wood & Fire Safety 3rd International Symposium
Info: conference@rocarm.com
Slovak Republic Belfast, United Kingdom
The High Tatras Info: www.intercomm.dial.pipex.com/html/
March 17-20, 2004 Info: uvt.tuzvo.sk/wfs/english/info/ events/hbif.htm
6th Asia-Oceania Symposium on Fire g_info.htm
Science and Technology
October 6-8, 2004
Paegu, Korea May 23-26, 2004
Info: 203.232.135.200/english/index.htm 5th International Conference on Performance-Based
NFPA World Safety Conference and Expo
Codes and Fire Safety Design Methods
Salt lake City, Utah
March 22-24 2004 Info: www.nfpa.org
Info: www.sfpe.org

International Fire Safety Engineering Conference

Sydney, Australia
Info: www.sfs.au.com

W INTER 2004 www.sfpe.org 32

Products/Literature
Heat and Smoke Detector Code Comparison Book
The F220 Series Smoke Detectors are the perfect
Rolf Jensen & Associates, Inc., has released
smoke detectors for an imperfect world. The series
Designing & Building with the IBC, Second Edition.
of detectors combines revolutionary new detector
This reference on the latest building codes allows
strategies including photoelectric only, photoelec-
architects, engineers, building officials, and contrac-
tric with heat detection, photoelectric with carbon
tors to easily compare the new International Building Code® (IBC)
monoxide detection, and heat only versions. All
2003 to the three model building codes and the 2000 edition of the
the detectors in the F220 Series feature an easy-to-clean chamber that does
IBC.
not require the units to be disassembled, a valuable time-saver. These
www.rjagroup.com
detectors also automatically compensate sensitivity to contamination in the
—Rolf Jensen & Associates, Inc.
chamber, extending the time between cleanings and reducing false alarms.
www.boschsecurity.com
—Bosch Security Systems

Flexible Fire Sprinkler Connections Smoke Seal Compound

FlexHead now offers flexible fire sprinkler Thermafiber® Smoke Seal™ Compound is
connections to the commercial market- an elastomeric acrylic water-based caulking
place. These connections are designed for material that is applied over a fire-resistant
use in suspended ceilings and are fast and forming material. Fire/smokestop systems
simple to install. Flexible connections with this product are available for through-
enable engineers to specify a product that penetrations in walls and floor-ceilings and
provides simplified project management, for perimeter fire-containment systems. The
faster occupancy, and cost-effective code compound is flexible, yet effectively bonds to keep the firestop in
compliance. place.
www.flexhead.com www.usg.com
—FlexHead —United States Gypsum Co.

Fire Alarm Control Panel Video-Based Fire Detection

The IdentiFlex™ (IF) 602 analog-addressable fire alarm Video-based fire detection systems can
control panel is ideally suited for small- to medium- cover a large area and do not require
sized installations of commercial, institutional, and the smoke to reach any specific point
industrial life safety applications. The IF602 uses location. The use of a video-based sys-
SmartLink™ peer-to-peer protocol to connect up to tem does not add additional costs for
250 network nodes for added flexibility for challenging design applica- the testing or maintenance of hardware
tions. It features SmartStart® advanced self-programming and XP95 fully than would already exist for the video
digital communications protocols, which allow for easy set-up and com- surveillance system. The systems may
patibility with any type of new or existing wiring. be designed to provide digital snapshots
www.gamewell.com associated with other events, such as pre-alarms or trouble conditions.
—The Gamewell Co. www.firesentry.com
—Fire Sentry Corp.

Upgraded Software Fire Sprinkler Evaluation Software

Fire Control Instruments has released version 6.1 New software allows Atlantic Mutual
of its Windows®-based Field Configuration loss-control professionals to more
Program (FCP-7100) and System Operating quickly evaluate a customer’s fire
Software (SOS) for the 7100 Series fire alarm con- sprinkler system and water supply, resulting in a more accurate loss-
trol panel. The 7100 Series is a microprocessor- control assessment and better protection of a customer’s property and
based, analog addressable fire control panel sys- business assets. Named SAVE (Sprinkler Analysis Verification
tem. The basic 7100 Series control panel provides two analog-address- Evaluation), the software calculates sprinkler design adequacy, includ-
able signaling line circuits, each supporting up to 99 analog-addressable ing density and area adjustments; pipe schedule capabilities; water sup-
sensors and up to 98 addressable monitor modules and/or control ply adequacy for sprinklers and hose demand; two-inch drain test flow
points. estimates; and in-rack demands.
www.firecontrolinstruments.com www.atlanticmutual.com
—Fire Control Instruments —Atlantic Mutual Insurance Co.

W INTER 2004 www.sfpe.org 33

 Products/Literature
Combo Exit Sign/Emergency Lamp Fixed Monitor
The new mcPhilben® CCHX Series of compact The Copperhead™ Fixed Monitor is ideal
thermoplastic combos combines an exit sign for small or restricted spaces. Just 16 inches
with two internally designed lamp heads into high, it has an all-brass configuration which
a small unit. With contemporary round cor- delivers efficient flows up to 1,250 gpm
ners, the units blend into any decor where with minimal friction loss. Vertical move-
emergency lighting is needed. The CCHX ment of 90 degrees above to 45 degrees
Series uses red or green LED lamps for energy below horizontal, and full and continuous
efficiency and low cost. Engineered for easy installation, units feature a 360-degree rotation combine for safe, easy
snap-together quick-mount design for wall or ceiling mounting, and operation.
have Xtest™ self-diagnostics. www.elkhartbrass.com
www.daybritelighting.com —Elkhart Brass Mfg. Co.
—Day-Brite Lighting

Website Offers Spec Info Portable Detectors

SureSpec.info allows users to review Durable, portable detectors from Crowcon are
and compare multiple fire alarm data available in several versions including the
and specification sheets from one site Gasman II single-gas version for applications
on the Web. In partnership with many where only one hazard is to be monitored.
leading fire alarm manufacturers in the Several sensor types are available for different
U.S., the SureSpec site was developed to help users prepare fire alarm applications including combustible gas, carbon
specifications with information that is updated continuously. Additional monoxide, and hydrogen sulfide. For applica-
site features include instant e-mail updates, blueprint review service, tions involving multiple gas risks, Triple Plus
links, and more. four-channel instruments are available, which
www.SureSpec.info may also be used for sampling gas from vessels and pipework.
—SureSpec www.crowcon.com
—Crowcon Detection Instruments

Firesafe Ball Valves Pressure Relief Valve CD Catalog

Firesafe-tested and -certified to API- Oklahoma Safety Equipment Co., Inc.,
607 4th edition specifications, Triac now offers its product catalog on a
ball valves combine operating confi- free CD, which contains detailed
dence with wide specification choic- information on the company’s custom-
es, direct-mount compactness, and engineered rupture disks, rupture
complete pre-assembled valve/actua- panels, and other pressure relief prod-
tor packaging convenience. Two model lines – 3-piece F88 and flanged ucts. The CD catalog contains links
full-port FD9 – enable application versatility. The valves come in 316 from the table of contents, PDFs containing in-depth information, illus-
stainless steel or carbon body steel. trations, and specifications.
www.a-tcontrols.com www.oseco.com
—A-T Controls, Inc. —Oklahoma Safety Equipment Co., Inc.

Micro Thermal Imaging Camera Smoke Control Panels

The Fire-Warrior, a new micro thermal imag- Using the latest technologies, H.R.
ing camera, enables firefighters to concur- Kirkland manufactures smoke control
rently use both hands, both eyes, and the panels with full color electronically
camera during fire attack, search-and-rescue, printed graphic displays, printed on the
and related operations. It is light enough, rear of UL listed lexan, and laminated
small enough, and advanced enough to be to .125” CNC punched aluminum. The
helmet-mounted without obstructing normal full color feature is beneficial as the air
vision. handling functions are color coded.
www.totalfiregroup.com The panels offer the flexibility of being hardwired or wired for inter-
—Morning Pride face with FA or HVAC software.
www.hrkirkland.com
—H.R. Kirkland Company, Inc.

34 Fire Protection Engineering N UMBER 21

 FIRE PROTECTION

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

HEADQUARTERS
Four people of different ages told each other how old they were. One TERRY TANKER Publisher
of them said “If I multiply my age by any of your ages, the product is a 1300 East 9th Street
Cleveland, OH 44114-1503
permutation of the digits of the two ages.” How old is everyone? 216.696.7000, ext. 9721
fax 216.696.3432
Thanks to Jane Lataille, P.E., for providing this issue’s brainteaser. ttanker@penton.com

NORTHEAST
TOM CORCORAN District Manager
929 Copes Lane
West Chestor, PA 19380
610.429.9848
Solution to last issue’s brainteaser fax 610.429.1120
tomcorcoran@penton.com
A car is driven 20,000 km per year. The car averages 12 km per liter of
NORTH CENTRAL
fuel. Assuming perfect combustion (i.e., no carbon monoxide or unburned
JOE DAHLHEIMER District Manager
hydrocarbons are created) and that the chemical formula for gasoline is 1300 East 9th Street
C8H18, how much carbon dioxide is emitted by the car annually? Cleveland, OH 44114-1503
216.696.7000, ext. 9279
Since the car averages 12 km per liter of fuel, driving 20,000 km in a year fax 216.696.3432
jdahlheimer@penton.com
requires 1,670 liters of fuel. From Table B.3 in the SFPE Handbook of Fire
Protection Engineering, the specific gravity of gasoline is 0.68, which cor- CENTRAL / WEST

responds to a density of 680 kg/m3. Therefore, 1,670 liters of gasoline AMY COLLINS District Manager
3240 Shadyview Lane North
weighs 1,140 kg. Since the reaction for burning C8H18 is Plymouth, MN 55447
763.404.3829
C8 H18 + 25 2 O2 → 8CO2 + 9 H2O fax 763.404.3830
acollins@penton.com

SOUTHEAST
for each 114 grams of gasoline that are burned, 352 grams of carbon DEBBIE ISGRO District Manager
dioxide are produced. Therefore, burning 1,140 kg of gasoline would 707 Whitlock Avenue SW
Suite B-24
create 3,500 kg of carbon dioxide. Marietta, GA 30064
770.218.9958
fax 770.218.8966
disgro@penton.com

AGF Manufacturing, Inc. .................................Page 9 NOTIFIER Fire Systems .................................Page 19

Index of Ansul Incorporated..........................................Page 2 OCV Control Valves.......................................Page 43
Advertisers BlazeMaster® Fire Sprinkler Systems/ Potter Electric Signal Co................................Page 21
NOVEON ..................................................Page 33 Prescolite........................................................Page 14
Chemetron Fire Systems................................Page 30 The RJA Group.............................Inside Front Cover
Chemguard.....................................................Page 49 Siemens Building Technologies, Inc.
Clarke Fire Protection Products, Inc. ...........Page 13 Faraday Division ......................................Page 11
DACS Inc........................................................Page 29 Siemens Building Technologies, Inc.
Detection Systems .........................................Page 41 Fire Safety Division ....................................Page 5
Edwards Systems Technology .................Page 26-27 SimplexGrinnell .............................................Page 24
Fire Control Instruments ...............................Page 51 Tyco Fire & Building Products................Back Cover
Flexhead Industries .......................................Page 23 UNC Charlotte................................................Page 28
Grice Engineering..........................................Page 17 Wheelock, Inc................................................Page 46
Koffel Associates, Inc....................Inside Back Cover Worcester Polytechnic Institute.....................Page 16
National Fire Protection Association ............Page 40

35 Fire Protection Engineering N UMBER 21

 from the technical director

Structural Fire Resistance Design – We Can Do Better

specific building, the required rating is ing basis requires the following steps:
determined by referring to the applica- • Determination of the fire conditions
ble building code. Any needed protec- to which a structure or portion thereof
tion for the structure is determined by could be exposed. These conditions are
considering the type of structure that is a function of the amount and distribu-
desired, the rating that is required by the tion of combustible items, the available
building code, and the results of fire re- ventilation, and the thermal properties of
sistance tests on specific assemblies. If an enclosure.
an assembly proposed for a structure • Based on the fire conditions, estima-
differs from a tested assembly, simple tion of the thermal response of the struc-
engineering calculations are used to vary ture or portion thereof.
the size of members, protection or cover • Based on the thermal response of
thickness, etc.3 the structure or portion thereof, determi-
Designing structural fire resistance in nation of the response of the structure.
this manner has a number of shortcom- Methods exist to perform each of
ings. These include: these tasks. Internationally, structural de-
• The fire exposure used in the stan- sign standards for the fire condition have
dard test does not consider factors that been developed and are presently in
would influence the temperature and use.4 Additionally, the National Institute
Morgan J. Hurley, P.E.
duration of a fire, such as ventilation, of Standards and Technology has em-
Technical Director
fuel load, and thermal properties of ma- barked on a major program to bench-
Society of Fire Protection Engineers
terials of construction. mark and further develop the state of the
• Single elements are tested in isola- art in structural fire safety design.
tion, and true performance of structural The Society of Fire Protection Engi-
systems and frames is not considered. neers is working in collaboration with

F ire resistance is ubiquitous in Moreover, structural fire resistance de- the American Society of Civil Engineers
commercial buildings as a means sign is typically performed by design and representatives from the steel, ma-
to protect people, property, and professionals other than fire protection sonry, concrete, and timber industries to
mission from damage resulting from engineers. In most cases, an architect more formally implement this design in-
fire-induced collapse. In most cases, determines the required fire resistance frastructure in a series of standards and
fire resistance is designed based on rating and applies the results of fire re- guides. Once completed, these standards
single-element tests that are conducted sistance testing for a structural assembly and guides will allow the design of build-
in accordance with standard tests such that is similar to the one that will be ings in which the structural fire resistance
as ASTM E1191 or ISO 834.2 In these used in the building. While these design is designed such that the building perfor-
standard tests, exemplar elements or professionals do a fine job of specifying mance in the event of fire is known.
assemblies, such as beams or columns, fire resistance in this manner, they do
are subjected to a standardized fire not necessarily have a background that REFERENCES
exposure. The pass/fail criteria of enables them to understand how struc-
these standard tests generally relate to tures will perform in a fire, and hence, 1 ASTM E119, “Standard Test Methods for
the ability of the tested element to they may not be able to determine how Fire Tests of Building Construction and
maintain a load (in the case of beams) unique building characteristics could af- Materials,” American Society for Testing
fect structural fire performance. and Materials, West Conshohocken, PA:
or stay below specified temperature
1995.
criteria (in the case of columns). The While structures that are designed in
rating that is given to structural ele- this manner have an excellent record of 2 ISO 834, “Fire-resistance tests – Elements
ments is expressed in units of hours, performance in fire, the result is that the of building construction,” International
which is a measure of the time before actual performance that will result if Standards Organization, Geneva, 1999.
the element reaches the endpoint cri- these structures are exposed to fire con- 3 ASCE/SFPE 29-99, Standard Calculation
teria in the standard test. This hourly ditions is unknown. Fortunately, the en- Methods for Structural Fire Protection,
rating is intended for comparative pur- gineering infrastructure exists today to American Society of Civil Engineers,
poses only and is not intended to rep- design structures based on the expected Reston, VA, 1999.
resent the length of time that a struc- fire loads and the structural fire resis- 4 Eurocode1 – Basis of design and actions
ture would withstand any given fire tance needed to accomplish the desired on Structures Part 2.2 Actions on struc-
without failure. performance in fire. tures – Actions on structures exposed to
When designing fire resistance for a Designing structures on an engineer- fire. ENV 1991-2-2: 1995, CEN.

36 Fire Protection Engineering N UMBER 21