Human Behavior Intro 2 Part Question

FIRE PROTECTIONFIRE PROTECTION

T H E O F F I C I A L M A G A Z I N E O F T H E S O C I E T Y O F F I R E P R O T E C T I O N E N G I N E E R S

FALL 2002 Issue No.16

page 0

HUMAN BEHAVIOR

A HISTORICAL REVIEW OF

IN FIRE

COOL UNDER FIRE

LIFE SAFETY, FIRE PROTECTION, AND MOBILITY-IMPAIRED PERSONS

TOOLS FOR THE SIMULATION OF HUMAN BEHAVIOR

33

37

28

BUILDING EXPECTATIONS BY MANAGING FIRE SAFETY

16

ALSO:

page 6

HUMAN BEHAVIOR

A HISTORICAL REVIEW OF

IN FIRE

www.sfpe.org 1

3 VIEWPOINT

4 FLASHPOINTS

16 BUILDING EXPECTATIONS: MANAGING FIRE SAFETY BY PUTTING PEOPLE BACK IN THE PICTURE Considering human behavior more during the design phase may help reduce the occurrence of fire and lessen the negative consequences when they do break out. Mark Chubb

23 SPEECH INTELLIGIBILITY Information to help ensure occupants will be able to hear and understand voice directives in the case of a building emergency. A supplement by the National Electrical Manufacturers Association

28 TOOLS FOR THE SIMULATION OF HUMAN BEHAVIOR The author explains how to factor in human behavior when working with evacuation models, weighing the pros and cons of various methods available. Rita F. Fahy, Ph.D.

33 COOL UNDER FIRE Making decisions during a fire is vastly different than making every-day deci- sions. What goes on in a person’s mind during a fire? How can we increase public awareness of proper evacuation procedures? Guylène Proulx, Ph.D.

37 LIFE SAFETY, FIRE PROTECTION, AND MOBILITY-IMPAIRED PERSONS An exploration of provisions currently in place along with a review of issues that still must be addressed. Brian D. Black

44 PRODUCTS/LITERATURE

47 SFPE RESOURCES

50 BRAINTEASER/CORPORATE 100/AD INDEX

52 FROM THE TECHNICAL DIRECTOR New specification for the Fire Protection Engineering PE Exam. Morgan J. Hurley, P.E.

Cover illustration by Thomas Dannenberg/Masterfile

Online versions of all articles can be accessed at www.sfpe.org. Invitation to Submit Articles: For information on article submission to Fire Protection Engineering, go to http://www.sfpe.org/publications/invitation.html.

contents F A L L 2 0 0 2 6 COVER STORY A SELECTED HISTORICAL REVIEW OF HUMAN BEHAVIOR IN FIRE How do building occupants react during a fire? What is the relationship between performance-based codes and human behavior and how will it affect evacuation modeling in the future? Dr. John L. Bryan

Subscription and address change correspondence should be sent to: Fire Protection Engineering, Penton Media, Inc., 1300 East 9th Street, Cleveland, OH 44114 USA. Tel: 216.931.9180. Fax: 216.931.9969. E-mail: [email protected].

Copyright © 2002, Society of Fire Protection Engineers. All rights reserved.

FIRE PROTECTIONFIRE 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 information to fire protection engineers and allied professionals. The opinions and positions stated are the authors’ and do not necessarily reflect those of SFPE.

Editorial Advisory Board Carl F. Baldassarra, P.E., Schirmer Engineering Corporation

Don Bathurst, P.E.

Russell P. Fleming, P.E., National Fire Sprinkler Association

Morgan J. Hurley, P.E., Society of Fire Protection Engineers

William E. Koffel, P.E., Koffel Associates

Jane I. Lataille, P.E., Los Alamos National Laboratory

Margaret Law, M.B.E., Arup Fire

Ronald K. Mengel, Honeywell, Inc.

Edward Prendergast, P.E., Chicago Fire Dept. (Ret.)

Warren G. Stocker, Jr., Safeway, Inc.

Beth Tubbs, P.E., International Conference of Building Officials

Regional Editors U.S. HEARTLAND John W. McCormick, P.E., Code Consultants, Inc.

U.S. MID-ATLANTIC Robert F. Gagnon, P.E., Gagnon Engineering, Inc.

U.S. NEW ENGLAND Thomas L. Caisse, P.E., C.S.P., Robert M. Currey & Associates, Inc.

U.S. SOUTHEAST Jeffrey Harrington, P.E., The Harrington Group, Inc.

U.S. WEST COAST Michael J. Madden, P.E., Gage-Babcock & Associates, Inc.

ASIA Peter Bressington, P.Eng., Arup Fire

AUSTRALIA Brian Ashe, Australian Building Codes Board

CANADA J. Kenneth Richardson, P.Eng., Ken Richardson Fire Technologies, Inc.

NEW ZEALAND Carol Caldwell, P.E., Caldwell Consulting

UNITED KINGDOM Dr. Louise Jackman, Loss Prevention Council

Personnel

EXECUTIVE DIRECTOR, SFPE Kathleen H. Almand, P.E.

TECHNICAL EDITOR Morgan J. Hurley, P.E., Technical Director, SFPE

PUBLISHER Terry Tanker, Penton Media, Inc.

ASSOCIATE PUBLISHER Joe Pulizzi, Custom Media Group, Penton Media, Inc.

MANAGING EDITOR Joe Ulrich, Custom Media Group, Penton Media, Inc.

ART DIRECTOR Pat Lang, Custom Media Group, Penton Media, Inc.

MEDIA SERVICES MANAGER Lynn Cole, Custom Media Group, Penton Media, Inc.

COVER DESIGN Dave Bosak, Custom Media Group, Penton Media, Inc.

By Norman E. Grone, Ph.D.

R eaders of this issue will learn that we already know a good deal about human behavior and

fire. But how can fire protection engi- neers apply this knowledge in their everyday work of analysis and design? Finding answers to this question is central to advancing fire protection engineering. Fire protection engineer- ing needs to advance beyond present tendencies to ignore human perfor- mance or use naïve and unvalidated safety factors and assumptions.

Given the best of all worlds, many engineers would prefer to plug data into calculation methods that predict the time people need to safely evacuate buildings. They could set aside consid- erations of human behavior and com- fortably focus their efforts on physical design problems. In my view, this design strategy will not yield optimal designs wherever people in fires need to understand situations and make deci- sions accordingly. To make this point, I use a metaphor that relates the behav- iors of people and fires. In both in- stances, valid predictions require a good understanding of the contexts in which the behaviors occur. Without such con- textual information, levels of uncertainty are simply too great to make acceptable predictions. In the case of fire, the engi- neer needs some knowledge of ventila- tion, nearby fuels, room geometry, and interior finishes before any reasonable predictions are possible. Imagine being asked to predict fire behavior without information about the above-listed fac- tors. In the case of human behavior, some knowledge about the information available to people from the social and physical environment is needed before reasonable predictions are possible. To predict human behavior, knowledge of the social and physical contexts is required.

Where will engineers acquire such contextual knowledge? From the design process itself. The job of the fire protec- tion engineer is to constrain the uncer- tainty of fire behavior by designing the environment. The challenge for the en- gineer will be to reduce the uncertainty of human behavior by supplying people with accurate information by designing the social and physical environment such that they can make timely and rea- sonable decisions. Only when design constrains behavior will acceptable pre- dictions be feasible.

To accomplish this task, fire protec- tion engineers will need analysis and design tools that enable them to predict the value of their design decisions. In the physical realm, FPEs have models that allow them to predict the contextual effects of suppression, fire-resistant bar- rier, and smoke control systems. Engi- neers will need similar validated tools that enable them to predict the contex- tual effects of alarm systems design, communications systems, emergency response capabilities, and occupant training programs. With such tools, FPEs can potential specify requirements that will effectively support people in their efforts to adaptively respond to situa- tions far better than they can at present.

Where can FPEs find such tools? While fields such as risk analysis and systems safety are relevant, I believe that the field of human factors and ergonom- ics (HFE) holds the greatest potential. Many HFE methods concern the analysis and design of technological systems that support humans in their interactions with products, environments, and equip- ment in performing tasks and activities. However, these methods were mostly developed to analyze and design sys- tems that support well-trained operators interacting with tightly coupled systems using well-defined interfaces. Adapting such methods will require clear thinking

and testing on real-world fire safety problems.

To use these tools, fire protection en- gineers will need to take a more expan- sive view of their analysis and design tasks. Please read the following excerpt from testimony that Beverly Eckert pro- vided at the June 24, 2002, NIST public meeting for input about its plans for in- vestigating the World Trade Center building disaster. Imagine a future where your design helps to avert such tragedy by providing information to people so that they can adaptively re- spond even to scenarios you were not asked to consider.

“I was on the phone with Sean for the last half hour of his life, beginning at 9:30 AM. He described the situation, what escape routes he had tried, and asked me for information based on what I was seeing on TV. He was calmly and rationally trying to assess his options. I reached 911 on another phone, but a full half-hour after the planes had struck they had no information to pass along... So despite advanced technology and a multitude of potential ways to transmit information to those whose lives de- pended on it there was no useful infor- mation being relayed, even though many of us were in contact with those who were trapped... Imagine the num- ber of lives that would have been saved had those few who had found an es- cape route from the upper floors of Tower Two been able to communicate to authorities which stairwell was open. It could have been passed along to those in the tower who were in contact with the world outside.”

Norman E. Groner is an independent human factors consultant in Santa Cruz, California.

FALL 2002 www.sfpe.org 2

A Future Where Engineering Supports People Adapting to Survive Fires

viewpoint

UL Investigates Firematic Nickel-plated Sprinklers

NORTHBROOK, IL, – Underwriters Labora- tories, Inc. (UL) is asking building owners and fire sprinkler contractors to submit sam- ples of Firematic model S & A 160°F sprin- klers for testing. UL recently tested field sam- ples taken from a single Connecticut location, revealing that some of these sprinklers did not operate as intended. UL is seeking addi- tional samples from throughout the country to determine if the problem is more wide- spread.

Firematic brand sprinklers are equipped with a heat-responsive element (HRE) that uses solder which melts at a specific tempera- ture. Once the solder melts, the HRE assem- bly should separate from the sprinkler body so that water is free to flow from the sprin- kler. The nickel-plating on the HRE for these particular Firematic brand sprinklers may pre- vent the solder from melting properly. This results in the HRE remaining intact, which may prevent these sprinklers from operating in a fire condition.

Firematic Sprinkler Devices manufactured the sprinklers from 1976 through 1979.

Testing on samples received from the one location revealed that the HRE assembly did not operate as intended in approximately 14 percent of the samples tested. UL tested nickel-plated sprinklers from another location and found all samples operating satisfactorily. More test samples are needed before UL can reach a final conclusion.

UL is requesting additional field samples of these sprinklers for operational testing. Although only samples of the model S & A rated 160°F have shown unacceptable results, the following additional models manufactured between 1967 and 1993 with ratings of 160°F, 212°F, 286°F, and 360°F should also be tested: “D,” TU-57, TP-57, TU-80, TU-29, TP-29, TU-39, and TP-39. These models use the same HRE as the S & A 160°F sprinklers.

Building owners or fire sprinkler contrac- tors desiring to have installed sprinklers tested should remove representative sprinkler sam- ples from the installation and send them to UL for testing. Building owners may wish to con- sult with their sprinkler contractor or Firematic for information regarding removal and replacement of the test samples before sub- mitting sprinklers to UL for testing. Once properly removed and packaged according to the instructions, sprinkler samples can be sent directly to Mr. Stephen Angeliu at Underwrit- ers Laboratories, Inc. (UL), 333 Pfingsten Rd., Northbrook, IL, 60062; Phone: 847-664-3687; e-mail: [email protected] for operational testing.

In keeping with UL’s not-for-profit, testing for public safety mission, UL will conduct these operational tests at no cost to the sub- mitter during the course of UL’s on-going in- vestigation, with the exception of expenses related to sprinkler removal, replacement, shipping and handling.

Firematic Sprinkler Devices can be reached at 900 Boston Turnpike, Shrewsbury, MA, 01545; Phone: 800-225-7288; Fax: 508-842-3523.

Federal Investigation Into World Trade Center Collapse

NEW YORK, NY – August 28, 2002 – The National Institute of Standards and Technol- ogy (NIST), a non-regulatory agency of the Department of Commerce, recently an- nounced the details of its $16 million, two- year federal building and fire safety investi- gation into the structural failure and subsequent collapse of WTC Buildings 1 and 2 (“The Twin Towers”) and WTC Building 7.

The investigation, which has been supported by the recent passage of an emer- gency supplemental appropriations bill, will focus on the building construction, the materials used and all of the technical condi- tions that contributed to the outcome of the WTC disaster. The study will involve the participation of experts from industry, acad- emia and other laboratories, complementing NIST’s own in-house capabilities.

NIST will also draw on the expertise of a private-sector coalition that includes profes- sionals from the Structural Engineering Institute of the American Society of Civil Engineers (SEI/ACSE), the National Fire Protection Association (NFPA), and the American Institute of Steel Construction (AISC). NIST will also maintain liaison with the Soci- ety of Fire Protection Engineers (SFPE), the Council on Tall Buildings and Urban Habitat (CTBUH), and the Structural Engineers Asso- ciation of New York (SEAoNY).

NIST expects to complete its investigation and issue a final report within an estimated twenty-four months from the start of the pro- gram. The investigation is part of a broader NIST response plan to the WTC disaster.

3 Fire Protection Engineering NUMBER 16

flashpoints fire protection industry news

HEADQUARTERS TERRY TANKER Publisher 1300 East 9th Street Cleveland, OH 44114-1503 216.696.7000, ext. 9721 fax 216.696.3432 [email protected]

NORTHEAST MARK D. FORKER District Manager P.O. Box 880 Oaks, PA 19456 610.676.9170 fax 610.676.9172 [email protected]

NORTH CENTRAL JOE DAHLHEIMER District Manager 1300 East 9th Street Cleveland, OH 44114-1503 216.696.7000, ext. 9279 fax 216.696.3432 [email protected]

CENTRAL AMY COLLINS District Manager 3240 Shadyview Lane North Plymouth, MN 55447 763.404.3829 fax 763.404.3830 [email protected]

SOUTHEAST DEBBIE ISGRO District Manager 707 Whitlock Avenue SW, Suite B-24 Marietta, GA 30064 770.218.9958 fax 770.218.8966 [email protected]

WEST COAST LORA ATHERTON ALBAINY National Sales Manager 1300 East 9th Street Cleveland, OH 44114-1503 216.696.7000, ext. 9282 fax 216.696.3432 [email protected]

Sales OfficesSales Offices FIRE PROTECTION

4 Fire Protection Engineering NUMBER 16

By Dr. John L. Bryan

INTRODUCTION

T his article reviews selected historical literature relevant

to the behavior of the occupants during a fire occur-

rence. Although the primary emphasis is on the devel-

opments in this study area within the United States from the

early 1900s until 2002, the literature cited is of worldwide ori-

gin. This article also presents an analysis of the recent influ-

ence of the performance-based code concept on the research

on human behavior in fire with the resulting emphasis on

evacuation models.

A Selected Historical Review of

Human Behavior in

FIRE

FALL 2002 www.sfpe.org 5

This review is an attempt to identify the development with both the acade- mic and professional recognition of the study area consisting of the human be- havior variables that appear to occur in fire incidents. The review of literature has attempted to recognize the sources of the significant and historical publica- tions relative to these issues. Any omis- sions of significant and historical stud- ies is truly unintentional. The emphasis throughout this presentation is on the history and development of this study area within the United States. For the purposes of continuity, this article has been organized into three chronologi- cal time periods: The “Prerecognition Years” period was arbitrarily selected as the period from the evacuation studies

of the early 1900s to the 1970s. The time period for the “Productive Years” of the study area of research on human behavior in fire was selected to include the 1970s and the 1980s. The “Perfor- mance Code Incentive Years” were identified as the 1990s into the 2000s. There has also been an attempt to delineate the need for the use of fire- related validation data in the develop- ment of the human behavior models.

THE PRERECOGNITION YEARS FROM THE 1900s TO THE 1970s

The earliest documented human behavior studies in the United States involved capacity counts of pedestrian velocity for the New York City design

of the Hudson Terminal Building in 1909.1 The first (1927) edition of the National Fire Protection Association’s Building Exits Code was developed from evacuation studies conducted dur- ing 1917-1927.2 Classical evacuation studies involving railway terminals, subway stations, theaters, department stores, and federal government office buildings with both “normal” exiting flows and “fire-drill” exiting flows were conducted in the early 1930s and pub- lished in 1935.1

The London Transit Board and other evacuation studies were conducted in Great Britain;3, 4 however, there was a lack of interest in the United States dur- ing the 1940s and the 1950s. Even in large loss of life fires such as the Coconut Grove, dedicated human be- havior studies of the activities of the occupants were not conducted.5

The academic study of the behavior of people in buildings at the time of a fire occurrence was initiated in the United States in the 1950s. Interviews with selected occupants of the Arundel Park fire incident in 1956 verified a reentry behavior by members of family groups.6 The prevailing rationale at this time was that fire protection engineers developed building features to enhance the fire safety of the occupants, to con- trol the ignition of fires, and to effec- tively suppress the fires that did occur. However, it was recognized by some that a difference between a minor fire incident and a major fire incident often involved the human behavior of the personnel immediately prior to the fire incident or during the fire incident. The recognition of occupants’ behavior often identified in engineering investi- gations of major fire incidents had long been documented, but little study and

6 Fire Protection Engineering NUMBER 16

analysis had been conducted to iden- tify and determine the casual and intervening variables involved.

Two politically significant fires occurred in 1967 that created and ac- celerated a change in fire protection engineering. The first fire was the Apollo spacecraft fire on January 27, at Cape Kennedy, Fla., with the loss of the three astronauts.7 The second fire occurred on February 7 in Dale’s Pent- house Restaurant in Montgomery, Ala., with 25 fatalities and 12 injuries.8 The U. S. House of Representatives Com- mittee on Science and Astronautics conducted hearings in May and June of 1967 on the Fire Research and Safety Act of 1967. This act was signed by President Johnson in 1968 establishing the National Commission on Fire Pre- vention and Control, which resulted in the “America Burning” report of 1973.9

Thus, a clear mandate was given to the federal government to conduct research in areas of the fire problem previously not considered. The Center for Fire Research at the National Bu- reau of Standards in The Department of Commerce was formed in 1974 and became vitally involved with many questions of the occupants’ behavior in fire incidents. Two individuals at the bureau were primarily involved in guiding and fostering this human be- havior research, Harold E. “Bud” Nelson and Irwin Benjamin.

THE PRODUCTIVE YEARS OF THE 1970s AND 1980s

The most productive time span for research and publications on human behavior in fire in the United States was from 1970 through the mid-1980s. The National Fire Prevention and Con- trol Administration provided a federal government focus on national fire problems. This organization guided the new and enhanced federal financial support for all facets of fire research, including human behavior. This agency envisioned the primary role of human behavior research being applicable pri- marily to the educational aspects of fire prevention.

During the 1970s and 1980s, the Na- tional Bureau of Standards through the Center for Fire Research, was the pri- mary source for funding human behav- ior in fire studies in the United States.

Some of the early and distinguished re- searchers supported by the National Bureau of Standards were John “Jack” Keating and Elizabeth Loftus from the University of Washington. These re- searchers established the parameters of the voice alarm system for the Seattle Federal Office Building.10 Norm Groner and Bud Levin were also involved with some of the early studies sponsored by the National Bureau of Standards in the Center for Fire Re- search. They both have continued their involvement with the research on hu- man behavior in fire, specifically with their studies involving areas of refuge, evacuation, and elevator usage.11, 12

It was during this period that there were two international seminars on hu- man behavior in fire. The first seminar was conducted at the University of Surrey in March of 1977, having been organized by David Canter and the members of the Fire Research Unit at the university. Most of the papers pre- sented at this seminar, with additional invited papers, became the first com- plete book on human behavior in fire.13

The second international seminar was conducted in October of 1978 at the National Bureau of Standards. The re- searchers at both of these conferences were primarily involved with the exam- ination and development of the meth- ods for the investigation of the behavior of the occupants in fire situations in both the United States and Great Brit- ain.14, 15, 16, 17 Funding was also provided in the early 1970s by the National Bureau of Standards for the formation with Japan of the U.J.N.R. (United States and Japan Natural Resources) Panel on Fire Research and Safety in which the panel meetings included

the study area of research on human behavior in fire. The emphasis of the studies in human behavior during this later portion of the “Productive Years” was in defining the behavioral actions of the occupants in fire situations, the examination of the then-popular con- cept of “panic behavior,” and a new emphasis on the study of the evacua- tion process as it occurred in high-rise buildings. As a result of this decade of research in human behavior in fire, the study area attracted additional re- searchers and interest throughout the world.18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33

In the 1970s and early 1980s, two critical studies were conducted at the University of Maryland. The first study replicated the methodology of Peter Wood’s study in Great Britain, in which fire service personnel interviewed 2,193 occupants from 952 fires from a struc- tured questionnaire.34 Peter Wood assisted in the development of the questionnaire for the University of Maryland study to assure the compati- bility of the study data for comparative purposes. This study, which was con- ducted in jurisdictions in the Washing- ton, D.C., and Baltimore, Md., area, involved 584 participants in 335 fire incidents and confirmed the reentry be- havior where members of the primary group were involved. The study also established the tendencies of occupants to move through smoke and the fire- fighting behavior of occupants in resi- dential occupancies.35 This study was also one of the few human behavior in fire studies comparing the cross-cultural samples of occupant behavior from Great Britain and the United States.34, 35

The second study primarily involved healthcare occupancies and selected

■ Human Behavior in Fire

FALL 2002 www.sfpe.org 7

significant fire incidents involving re- ports of human behavior activity.36 This study was conducted by university per- sonnel applying questionnaires and interviews of staff, patients, and first- arriving firefighters. A fire incident studied in a high-rise apartment build- ing provided the first indication of the formation by the occupants of conver- gence clusters in selected apartments, which was later confirmed to a much greater degree in the fire incident at the MGM Grand Hotel.37, 38 Two fire inci- dents in university residence halls involved jumping behavior by occu- pants previously injured while using the means of egress system in evacua- tion attempts.39, 40

In the United States, federal support for fire research started to emphasize computer modeling of fire dynamics in the early 1980s. Thus, the financial sup- port relative to human behavior in fire since the early 1980s has been primar- ily directed to computer models con- cerned with the human behavior in building evacuation.41, 42, 43, 44, 45, 46

The International Association for Fire Safety Science from 1985 to 1996 had become the primary organization for the publication of research studies on human behavior in fire. At the first symposium in 1985, there were six pa- pers, and at the 1996 symposium, there were nine papers, with two of the pa- pers involving evacuation models.47, 48

The 1994 symposium had a record number of human behavior in fire pa- pers with a total of twelve papers, 75 percent of which involved evacuation models.49

The concern for the evacuation of occupants with mobility impairments and other physical constraints was ini-

tially established in the 1970s as a result of studies of the evacuation of high-rise buildings.16, 32 This concern in the United States resulted in a two-session national conference on Fire and Life Safety for the Handicapped in 1980.50, 51

This area of human behavior in fire has continued to receive emphasis through both interview and experimental stud- ies, the examination of the use of ref- uge areas, and the use of elevators for evacuation.2, 50, 51, 52, 53, 54, 55, 56

The basis for most evacuation mod- els which predict the velocity of move- ment and the flow of occupants in the egress system have been developed from evacuation studies by Jake Pauls, who conducted his initial study in 1969.26 Since the mid- and late 1980s the emphasis of studies on human be- havior in fire in the United States has been on the documentation of evacua- tions and the development of evacua- tion models, with consideration of the human behavior factors identified in the fire incident studies of the 1970s, ’80s, and ’90s. One of the most inten- sively studied evacuations was the bombing of the World Trade Center on February 26, 1993. This incident was studied with one interview study and two questionnaire studies relative to the behavior of selected occupants.57, 58, 59

This fire incident evacuation was also the source of one of the few detailed and documented interview studies of the evacuation of disabled occupants.60

Fahy utilized the data collected in a questionnaire study of the evacuation behavior of 406 fire wardens to im- prove her evacuation model.58, 61

Human behavior in fire in the United States has been neglected since the mid-1980s. The leaders in this area of

critical fire research are now in Austra- lia, Canada, Great Britain, Japan, New Zealand, Northern Ireland, Norway, and Sweden. The application of experi- mental studies with human subjects relative to wayfinding and smoke effects have been conducted in both Japan and Norway.62, 63, 64

THE PERFORMANCE CODE INCENTIVE YEARS OF THE 1990s INTO THE 2000s

The worldwide movement toward performance codes has created a de- mand for computer evacuation models that will provide an estimate of the evacuation time for a building.44, 45, 61, 65, 66

The basis for most computer evacua- tion models has been the data provided from normal and practice evacuations in earlier studies.67, 68

Regardless of the perceived advan- tages or disadvantages of performance- based fire codes69, 70, 71, 72, the utilization and adoption of these codes will en- hance all of the research areas relative to the understanding of human behav- ior in fire. Thus, with performance- based fire codes, the fire protection en- gineer and the building designer can no longer adopt a policy of benign ne- glect of the results of the research on human behavior in fire. Meacham has included in his ten needs for the suc- cessful application of a performance code approach “to understand how people react in a fire situation.”71

Computer models of occupant evac- uation have continued to be refined and to become more valid in their sim- ulation of occupant activities in normal evacuations and evacuation drills.65, 66 It is of interest to note that the only hu- man behavior in fire study supported by the National Institute for Standards and Technology in 1996 or 1997 was EXIT89 evacuation model Fahy’s.61

Shields and Dunlop have emphasized the areas of refinement and improve- ment needed in the computer models as: the accommodation of mixed-abili- ties populations, accepting manage- ment fire safety profiles, recognition of the phenomenon of contra flows in egress ways, and most importantly, to eliminate invalid assumptions.55

The future of research on human be- havior in fire is indeed more promising today than at any time in the past, with

8 Fire Protection Engineering NUMBER 16

the recognition of the need to deter- mine valid evacuation times in the design of the building under the perfor- mance code concept.

However, with this increased appli- cation of computer evacuation models there is an increased need for model designers to effectively communicate to potential users the limitations of the model, e.g., if human behavior is not simulated by the model and the popu- lations or occupancies for which the model has not been validated. In addi- tion, the model designers should pro- vide a detailed inventory and descrip- tion of the validation procedures and studies to which they have subjected the model. Critical variables in human behavior have been identified as fol- lows:

Proulx has found that the means of alerting occupants directly impacts the time delay before initiation of evacua- tion.73 Proulx, Sime, and Fahy stress the need to accurately assess the time delay before evacuation when predicting evacuation times.74, 75, 76, 79 They point out that this delay is influenced by building occupancy and population characteris- tics, and that it includes investigation, information-seeking, alerting others, assisting others, and firefighting.

In incorporating data from studies, developers of evacuation models must also recognize that evacuation drills are less stressful than evacuations from fires, since there is less ambiguity and physio- logical or psychological exposure to the effects of the smoke or heat in drills than in fires.35, 36, 37 Drills are also different in that occupants do not seek out areas of refuge for communication, relief from contaminants, or psychological support as has been seen in fires in high-rise res- idential occupancies.38, 72 The ambiguity

of alerting cues in the fire-induced evac- uations can prolong the time delay prior to evacuation.6, 38, 75, 76, 77, 78, 79 Fire-induced wayfinding complications also reduce occupant evacuation velocity.35, 37, 38, 77, 79

In 1999, the Society of Fire Protec- tion Engineers created a task group on human behavior. This task group is developing a guide for application by fire protection engineers to identify and develop the critical human behav- ior data in their utilization of the per- formance-based fire protection analysis and design of buildings.80

The Fire SERT Centre at the Univer- sity of Ulster has produced significant studies relative to disabled individuals in the late 1990s. These studies have developed data on the mobility of the population in corridors, on ramps, through doors, and the perception of exit signs.81, 82, 83, 84

The University of Ulster also created and conducted the First International Symposium on Human Behaviour in Fire in Belfast in 1998.85 This was the first symposium dedicated to human behavior in fire research in almost twenty years.13, 14 The symposium was a landmark event, with 81 presentations on the following human behavior in fire topic areas: Fire-Related Human Be- haviour; Fire Regulations and Fire Codes; Learning from Human Behav- iour in Real Fire Situations; The Assimi- lation, Evaluation, and Use of Informa- tion by People Exposed to Fire; Characterization of Building Occupan- cies and Complex Environments; Evac- uation of Disabled People from Fire; Assessing the Risk to People Exposed to Fire; Determining Evacuation Time and Evacuation Performance Measures; Occupant Characteristics and Human Behavior Transport and Ancillary Facili- ties; Wayfinding Performance in Com- plex Environments; Occupant Percep- tion and Escape through Smoke; Evacuation Simulation Modeling and Approaches to Validation.85

The University of Ulster also formal- ized the Second International Sympo- sium on Human Behaviour in Fire held at the Massachusetts Institute of Technology in 2001. This symposium included 56 presentations in the follow- ing topic areas of human behavior in fire: Human Behavior Theory; Evacua- tion Modeling and Functionality; Occu- pancy Characterization; Occupant

Response; Toxicity and Smoke Effects; Non-Engineering Solutions for Reducing the Incidence and Impact of Fire; Visual Accessibility and Spatial Analysis; Human Performance Criteria for Inclu- sion in Regulations and Codes.86

CLOSING REMARKS

It is recommended that all interested fire protection engineers review the presentation on the development of the study area of human behavior in fire by Jake Pauls at the First International Sym- posium on Human Behaviour in Fire.87

The increased interest and research activity generated by the performance code concepts and the Fire SERT Centre symposiums have created a promising potential for the utilization of research on human behavior in fire. The fulfill- ment of this potential is the responsibil- ity of human behavior researchers, building designers, and most impor- tantly, fire protection engineers. ▲

John Bryan is Professor Emeritus in the Department of Fire Protection Engi- neering, University of Maryland.

REFERENCES

1 National Bureau of Standards. Design and

Construction of Building Exits, Washington,

D.C., 1935.

2 National Fire Protection Association, Quincy,

MA: Life Safety Code. NFPA 101, 2000, 1.

3 London Transit Board. Second Report of the

Operational Research Team on the Capacity of

Footways, London, 1958.

4 Melinek, S. J. and S. Booth. An Analysis of

Evacuation Times and the Movement of Crowds

in Buildings, Borehamwood: Fire Research

Station, 1975.

5 Lyons, Paul R., Fire in America! Boston:

National Fire Protection Association, 1976,

180-183.

6 Bryan, John L., A Study of the Survivors

Reports on the Panic in the Fire at The

Arundel Park Hall in Brooklyn, Maryland, on

January 29, 1956, College Park, MD: University

of Maryland, 1957.

7 Fire Onboard.” NFPA Journal, 90, 1, 1996,

68-89.

8 Juillert, Ernest E., “Fire at Dale’s Penthouse

Restaurant,” Fire Journal, 71, 3, 1967, 5-9, 13.

9 National Commission on Fire Prevention and

Control. America Burning, U.S. Government

Printing Office, Washington, D.C., 1973.

Continued on page 13

■ Human Behavior in Fire

FALL 2002 www.sfpe.org 9

10 Keating, John P,. & Loftus, Elizabeth F., People

Care in Fire Emergencies – Psychological

Aspects 1975, Society of Fire Protection

Engineers, Technical Report 75-4, 1975.

11 Groner, Norman E., & Levin, Bernard M.,

Human Factors Consideration in the Potential

for Using Elevators in Building Emergency

Evacuation Plans, NIST, Building and Fire

Research Laboratory, NIST-GCR-92-615, 1992.

12 Levin, Bernard M., & Groner, Norman E.,

Human Behavior Aspects of Staging Areas for

Fire Safety in GSA Buildings, NIST, Building

and Fire Research Laboratory, NIST-GCR-92-

606, 1992.

13 Canter, David (Ed.) Fires and Human

Behavior, New York: John Wiley & sons, 1980.

14 Levin, B. M., & Paulson, R. L., (Eds.) Second

International Seminar on Human Behavior in

Fire Emergencies, October 29-November 1,

1978, Proceedings of Seminar, National Bureau

of Standards, Center for Fire Research, NBS-

GCR-80-2070, 1980.

15 Canter, D., & Matthews, R.,The Behaviour of

People in Fire Situations: Possibilities for Fire

Research, Borehamwood: Fire Research

Station, 1976.

16 Keating, John P.; Loftus, Elizabeth F.; & Groner,

Norman E., An Evaluation of the Federal High-

Rise Emergency Evacuation Procedures, Seattle:

University of Washington, 1978.

17 Lerup, Lars, People in Fires: A Manual for

Mapping, Berkel.; & Hokugo, A., “A Case Study

of Fire and Evacuation in a Multi-Purpose

Office Building, Osaka, Japan,” Fire Safety

Science – Proceedings of The First

International symposium, New York:

Hemisphere Publishing Corp., 1986, 523-532.

18 Stahl, Fred I.; Crosson, James J.; & Margulis,

Stephen T., Time-Based Capabilities of

Occupants to Escape Fires in Public Buildings:

A Review of Code Provisions and Technical

Literature, Washington, D.C.: National Bureau

of Standards, 1982.

19 Bryan, John L., Implications for Codes and

Behavior Models from the Analysis of Behavior

Response Patterns in Fire Situations as Selected

from the Project People and Project People II

Study Reports, Gaithersburg, MD: National

Bureau of Standards, 1983.

20 Canter, D., Studies of Human Behavior in Fire.

Empirical Results and Their Implications for

Education and Design, Borehamwood: Fire

Research Station, 1985.

21 Horiuchi, S.; Murozaki, Y.; & Hokugo, A., “A

Case Study of Fire and Evacuation in a Multi-

Purpose Office Building, Osaka, Japan,” Fire

Safety Science – Proceedings of The First

International symposium, New York:

Hemisphere Publishing Corp., 1986, 523-532.

22 Jin, Tadahisa, “Visibility Through Fire Smoke,”

Journal of Fire and Flammability, 9,1978,

135-157.

23 Jones, B. K., & Hewitt, J. Ann, “Leadership and

Group Formation in High-Rise Building

Evacuations,” Fire Safety Science – Proceedings

of The First International Symposium, New

York: Hemisphere Publishing Corp., 1986,

513-522.

24 Kimura, Michiharu, & Sime, Jonathan D., “Exit

Choice Behavior During The Evacuation of

Two Lecture Theatres,” Fire Safety Science –

Proceedings of The Second International

Symposium, New York: Hemisphere Publishing

Corp., 1989, 541-549.

25 Latane, Bibb, & Darley, John, “Group

Inhibition of Bystander Intervention in

Emergencies,” Journal of Personality and

Social Psychology, 10, 3, 1968, 215-221.

26 Pauls, J. L., Evacuation Drill Held in The B. C.

Hydro Building, 26 June 1969, National;

Research Council of Canada, 1970.

27 Pauls, J. L., & Jones, B. K,. “Research in

Human Behavior,” Fire Journal, 74, 3, 1980,

35-42.

28 Pauls, Jake, “The Movement of People in

Buildings and Design Solutions for Means of

Egress,” Fire Technology, 20, 1, 1984, 27-46.

29 Predtechenskiy, V. M., & Milinskiy, A. I.,

“Standardization of Evacuation,” Roytman, M.

Ya., Principles of Fire Safety Standards for

Building Construction, New Delhi: Amerind

Publishing Co., Pvt. Ltd., 1975.

30 Quarantelli, E. L., Panic Behavior. Some

Empirical Observations, Columbus, OH: Ohio

State University, 1975.

31 Quarantelli, E. L., Panic Behavior in Fire

Situations. Findings and a Model from the

English Language Research Literature, Tokyo:

4th. Joint Panel Meeting of the United States-

Japan Panel on fire Research, 1979.

32 Rubin, Arthur I., & Cohen, Arthur, Occupant

Behavior in Building Fires, National Bureau of

Standards, Washington, D.C., 1974.

33 Sime, Jonathan D., “The Concept of Panic,”

Fires and Human Behavior, (Ed.), David

Canter, New York:. John Wiley & sons, 1980,

63-81.

34 Wood, Peter G., “The Behaviour of People in

Fires,” Borehamwood: Building Research

Establishment, Fire Research Note 953, 1972.

35 Bryan, John L., Smoke as a Determinant of

Human Behavior in Fire Situations (Project

People), Gaithersburg, MD: National Bureau of

Standards, Center for Fire Research, NBS-GCR-

77-94, 1977.

36 Bryan, J. L., & Milke, J. A., The Determination

of Behavior Response Patterns in Fire

Situations, Project People II, Final Report

Health Care, Gaithersburg, MD: National

Bureau of Standards, Center for Fire Research,

NBS-GCR-81 -343, 1981.

37 Bryan, John L., & DiNenno, Philip J., An

Examination and Analysis of the Dynamics of

the Human Behavior in the Fire Incident at

The Georgian Towers on January 9, 1979,

College Park, MD; University of Maryland,

Department of Fire Protection Engineering,

1979.

38 Bryan, John L., “Convergence Clusters: A

Phenomenon of Human Behavior Seen in

Selected High-Rise Building Fires,” Fire

Journal, 79, 6, 1985, 27-30, 86-90.

39 Bryan, J. L., & DiNenno, P. J., An Examination

and Analysis of the Dynamics of the Human

Behavior in the Fire Incident at Thurston Hall

on April 19, 1979, College Park, MD:

University of Maryland, Department of Fire

Protection Engineering, 1979.

40 Bryan, John L., & Milke, James A., An

Examination and Analysis of the Dynamics of

the Human Behavior in the Fire Incident at

Chesapeake Hall on February 3, 1980, College

Park, MD: University of Maryland, Department

of Fire Protection Engineering, 1980.

41 Alvord, D. M. Status Report on the Escape and

Rescue Model and the Fire Emergency

Simulation for Multifamily Buildings, National

Bureau of Standards, Gaithersburg, 1985.

42 Berlin, Geoffrey N.; Dutt, Ashoke; & Gupta,

Surenda M., “Modeling Emergency Evacuation

from Group Homes,” Fire Technology, 18, 1,

1982, 38-48.

43 Fahy, Rita, “Building Fire Simulation Model,”

Fire Journal, 77, 4, 1983, 93-95, 102-105.

44 Kisko, T. M., & Francis, R. L., Network Models

of Building Evacuation. Development of

Software System, National Bureau of Standards,

1985.

45 Levin, Bernard M., “EXITT – A Simulation

Model of Occupant Decisions and Actions in

Residential Fires,” Fire Safety Science –

Proceedings of The Second International

Symposium, New York: Hemisphere Publishing

Corp., 1989, 561-570.

46 MacLennan, Hamish A., “Towards an

Integrated Egress/Evacuation Model Using an

Open Systems Approach,” Fire Safety Science –

Proceedings of The First International

Symposium, New York: Hemisphere Publishing

Corp., 1986, 581-590.

47 Grant, Cecile E., – Pagni, Patrick J., (Eds.) Fire

Safety Science – Proceedings of The First

International Symposium, New York:

Hemisphere Publishing Corp., 1986.

48 Hasemi, Yuji, (Ed.) Fire Safety Science –

Proceedings of The Fifth International

Symposium, International Association of Fire

Safety Science, 1997.

49 Kashiwagi, Takashi, (Ed.) Fire Safety Science –

Proceedings of The Fourth International

Symposium, International Association of Fire

Safety Science, 1994.

■ Human Behavior in Fire

10 Fire Protection Engineering NUMBER 16

50 Kennett, E. W., (Ed.) Proceeding of the 1980

Conference on Life Safety and the

Handicapped, Gaithersburg, MD: National

Bureau of Standards, 1982.

51 Levin, B. M., (Ed.) Fire and Life Safety for The

Handicapped, Conference and Preparatory

Workshop Reports, National Bureau of

Standards, Gaithersbug, MD: 1980.

52 Levin, Bernard M., & Nelson, Harold E., “Fire

Safety and Disabled Persons” Fire Journal, 75,

5, 1981, 35-40.

53 Proulx, Guyléne; Latour, John; & MacLaurin,

John, Housing Evacuation of Mixed Abilities

Occupants, Ottawa: National Research Council,

1994.

54 Proulx, Guyléne, & Pineau, Joelle, Review of

Evacuation Strategies for Occupants with

Disabilities, Ottawa: National Research Council,

1996.

55 Shields, T. J., & Dunlop, K. E., “Emergency

Egress Models and the Disabled” Conference

Proceedings of The Sixth International

Interflam Conference, 1993, 143-150.

56 Shields, T, J., Fire and Disabled People in

Buildings, Garston: Building Research

Establishment, 1993.

57 Yoshida, Yoshiyuki, “A Study of Evacuation

Behavior in The World Trade Center Explosion,”

Fire Technology 32, 2, 1996, 174-189.

58 Fahy, Rita F., & Proulx, Guyléne, “Collective

Common Sense: A Study of Human Behavior

during The World Trade Center Evacuation,”

NFPA Journal, 89, 2, 1995, 59-67.

59 Wenger, Dennis; Aguiere, Benigno; & Vigo,

Gabriela, Evacuation Behavior Among Tenants

of The World Trade Center following the

Bombing of February 26, 1993, College

Station, TX: Texas A & M University, 1994.

60 Juillet, Edwina, “Evacuating People with

Disabilities,” Fire Engineering, 146, 12, 1993,

100-103.

61 Fahy, Rita F., Enhancement of EXIT89 and

Analysis of World Trade Center Data, Final

Report August 1994 – August 1995,

Gaithersburg, MD; NIST, Building and Fire

Research Laboratory, NIST-GCR-95-684, 1996.

62 Jin, Tadahisa & Yamada,Tokiyoshi,

“Experimental Study of Human Behavior in

Smoke-Filled Corridors,” Fire Safety Science –

Proceedings of The Second International

Symposium, New York: Hemisphere Publishing

Corp., 1989, 511-519.

63 Jin, Tadahisa, “Studies on Human Behavior

and Tenability in Fire Smoke,” Fire Safety

Science – Proceedings of The Fifth International

Symposium, Melbourne: International

Association of Fire Safety Science, 1997, 3-21.

64 Paulson, Truls, “The Effect of Escape Route

Information on Mobility and Wayfinding Under

Smoke-Logged Conditions,” Fire Safety Science –

Proceedings of The Fourth International

Symposium, Ottawa: International Association

of Fire Safety Science, 1994, 693-704.

65 Owen, Matthew; Galea, E. R.; & Lawrence, P.,

“Advanced Occupant Behavioural Features of

The Building-EXODUS Evacuation Model,” Fire

Safety Science – Proceedings of The Fifth

International Symposium, Melbourne:

International Association of Fire Safety Science,

1997, 795-806.

66 Thompson, Peter; Wu, Jianhua; & Marchant,

Eric, Fire Safety Science – Proceedings of The

Fifth International Symposium, “SIMULEX 3.0:

Modeling Evacuation in Multistory Buildings,”

Melbourne: International Association of Fire

Safety Science, 1997, 725-736.

67 Proulx, Guyléne, “Movement of People: The

Evacuation Timing,” DiNenno, Philip J. (Ed.)

SEPE Handbook of Fire Protection Engineering,

2nd. ed., Quincy, MA: National Fire Protection

Association, 2002, 3-342 - 3-366.

68 Nelson, Harold E. (“Bud”) & Mowrer, Frederick

W., “Emergency Movement” DiNenno, Philip J.

(Ed.) SFPE Handbook of Fire Protection

Engineering, 3rd. ed., Quincy, MA: National

Fire Protection Association, 2002, 3-367- 3-380.

69 Richardson, J. Kenneth, “Performance-Based

Fire Codes, Why We Need Them” NFPA

Journal, 91, 1,1997, 72-75.

70 Fire SERT Centre, University of Ulster,

Engineering Fire Safety in the Process of

Design, Demonstrating Equivalency,

Jordanstown, N. I.: CIBW14 International

Symposium and Workshops, 1993.

71 Meacham, Brian J., The Evolution of

Performance-Based Codes & Fire Safety Design

Methods, Boston: Society of Fire Protection

Engineers, 1996, 76.

72 Richardson, J. Kenneth, in Selected Readings in

Performance-Based Fire Safety Engineering,

(Ed.) by Kenneth J. Richardson, pp. iii-vii,

Boston: Society of Fire Protection Engineers,

1993, iii-vii.

73 Proulx, G., & Sime, J. D., “To Prevent Panic in

an Underground Emergency: Why Not Tell

People the Truth?” Fire Safety Science –

Proceedings of The Third International

Symposium, New York: Elsevier, 1991, 843-852.

74 Proulx, Guyléne, “A Stress Model for People

Facing a Fire,” Environmental Psychology, 13,

1993, 137-147.

75 Proulx, Guyléne; Pineau, Joelle; Latour, John

C.; * Stewart, Lisa, Study of The Occupants’

Behavior During the 2 Forest Laneway Fire in

North York, Ontario, January 6, 1995, Ottawa:

National Research Council, 1995.

76 Sime, Jonathan D., “Perceived Time Available:

The Margin of Safety in Fires,” Fire Safety

Science – Proceedings of The First

International Symposium, New York:

Hemisphere Publishing Corp., 1986, 561-569.

77 Brennan, Partricia, “Timing Human Response

in Real Fires,” Fire Safety Science – Proceedings

of The Fifth International Symposium,

Melbourne: International Association of Fire

Safety Science, 1997, 807-818.

78 Saunders, Wendy I., Decision-Making Model of

Behaviour in Office Building Fire Evacuations,

Ph.D. Thesis, Department of Psychology,

Victoria: University of Technology, Melbourne:

2001.

79 Fahy, Rita F., & Proulx, Guyléne, “Toward

Creating a Database on Delay Times to Start

Evacuation and Walking Speeds for Use in

Evacuation Modeling,” Proceedings of The

Second International Symposium on Human

Behavior in Fire, MIT, 2001, 175-183.

80 Society of Fire Protection Engineers, SFPE

Engineering Guide to Performance-Based Fire

Protection Analysis and Design of Buildings,

Quincy, MA: National Fire Protection

Association, 2000.

81 Boyce, K. E.; Shields, T. J.; & Silcock, G. W.

H., “Toward the Characterization of Building

Occupancies for Fire Safety Engineering:

Prevalence, Type, and Mobility of Disabled

People,” Fire Technology, 35, 1, 1999, 35-50.

82 Boyce, K. E.; Shields, T. J.; & Silcock, G. W.

H., “Toward the Characterization of Building

Occupancies for Fire Safety Engineering:

Capabilities of Disabled People Moving

Horizontally and on an Incline,” Fire

Technology, 35, 1, 1999, 51-67.

83 Boyce, K. E.; Shields, T. J.; & Silcock, G. W.

H., “Toward the Characterization of Building

Occupancies for Fire Safety Engineering:

Capability of Disabled People to Negotiate

Doors,” Fire Technology, 35, 1, 1999, 68-78.

84 Boyce, K. E.; Shields, T. J.; & Silcock, G. W.

H., “Toward the Characterization of Building

Occupancies for Fire Safety Engineering:

Capability of People with Disabilities to Read

and Locate Exit Signs,” Fire Technology, 35, 1,

1999, 79-86.

85 Shields, T. J., (Ed.) Proceedings of The First

International Symposium on Human Behaviour

in Fire, Jordantown, NI: Fire SERT Centre

University of Ulster, 1998.

86 Proceedings of The Second International

Symposium on Human Behaviour in Fire, held

at Massachusetts Institute of Technology,

London: lnterscience Communications Limited,

2001.

87 Pauls, Jake, “A Personal Perspective on

Research, Consulting, and Codes/Standards

Development in Fire-Related Human

Behaviour 1969-1997 with an Emphasis on

Space and Time Factors, Proceedings of The

First International Symposium on Human

Behaviour in Fire, University of Ulster, 1998,

71-82.

■ Human Behavior in Fire

11 Fire Protection Engineering NUMBER 16

By Mark Chubb

F or much of the past quarter-cen- tury, study and attention to the influence of human behavior in

fire have focused almost exclusively on the actions people take in emer- gency situations. Despite very little ongoing commitment to such study, a great deal is now known about what people do in fires.1 We know a little less about why people do these things and still less about how to influence their actions, but we certainly know enough to influence the way we design buildings.

What some consider the unpre- dictable nature of observed human be- havior in fire situations belies its consis- tency, but such misunderstandings probably explain why many fire protec- tion engineers seek to control or mini- mize the influence of people on fire out- comes through their designs. This bias stands in stark contrast to other engi- neering disciplines, like aerospace, bio- medical, chemical process, and nuclear engineering, where human factors play a very significant role in defining and

controlling the technology and, conse- quently, play key roles in achieving and maintaining safety and reliability.

Keeping in mind the old adage that men, women, and children represent the three leading causes of fires, can we af- ford to ignore or even marginalize the roles people play in fire protection de- signs? Putting people first will help us prevent fires in the first place and, failing that, help encourage them to take con- structive and appropriate action when fires do occur.

PEOPLE AS CAUSE AND CURE

In the new era of performance-based design, consulting fire protection engi- neers spend most of their time con- cerned with two things: 1) buildings sys- tems and services, and 2) the codes and standards that regulate their design, con- struction, and use. People and fires often receive attention from engineers only when code compliance cannot be achieved by dealing with the building or its component systems alone. And the solutions proposed under such circum- stances often introduce unrealistic as-

sumptions about what people will do or what fires will occur (or not occur) in an effort to demonstrate that some building feature or another is not required. Con- struction-cost considerations seem to play a significant role in this sad state of affairs.

Most building fires involve faulty in- teractions between people and the things that go on inside buildings. If these fires heavily damage or destroy the buildings where they occur or cause injuries or death to the people inside, they do so not only because the build- ing did not respond well to the fire threat but because the people didn’t re- spond well.

Recently reported research by Thomas and Brennan2 offers a possible explanation for this and suggests why we should pay more attention to fire prevention in our designs. Their analysis of Australian fire incident reporting sys- tem returns indicates that the largest proportion of residential fire fatalities in Australia involved people responsible for the outbreak of fire in the first place. Although results of a similar assessment for nonresidential fires have not been

Managing Fire Safety by Putting People Back in the Picture

Building Expectations:

FALL 2002 www.sfpe.org 12

reported, anecdotal evidence would seem to suggest that similar circum- stances apply to commercial and indus- trial fire deaths as well.

Concentrating on the building’s reac- tion to fire, although logically and legally necessary, will ensure the barn door gets closed only after the cow gets out. Putting people first will help us achieve both aims: 1) preventing the outbreak of fire, and 2) managing its consequences when prevention fails.

RULE-BASED APPROACHES

In the past, fire protection engineer- ing practitioners spent a good deal of time looking at what made their clients tick. Insurance loss control specialists, in particular, paid a great deal of attention to process hazards in plants, manage- ment attitudes toward fire safety, and employee training and work practices. When financial incentives and disincen- tives influenced by underwriting prac- tices and premiums didn’t work, the in- dustry relied on rules.

Beginning in 1927, the National Board of Fire Underwriters (NBFU) in the United States produced a model fire pre- vention code for use by its field engi- neers in their evaluations of local fire departments’ fire prevention programs.3

This document served as the basis for model code organizations’ efforts when NBFU’s successor organization, the American Insurance Association, sus- pended its publication in the mid-1970s. Fire prevention codes based on NBFU’s national model, unlike today’s fire codes which have become increasingly im- mersed in the building process and fire protection features, dedicated their pages to transcribing the lessons learned from past failures into prescriptions for preventing future catastrophes.

No one knows for sure how success- ful NBFU’s approach was in controlling or motivating human behavior. We can say with some confidence, however, that fire prevention code requirements influenced designers and users of tech- nologies to take account of active hu- man error (slips) in the operation and maintenance of complex systems whose hazards had already become apparent. Arguing that this approach neglects the problem of latent human error associ- ated with new or emerging technologies seems a fair criticism, but one easily ad-

dressed without relying on costly exter- nal controls that displace investment and marginalize the role of human operators.

Groner 4, 5 has suggested ways fire pro- tection engineers can take account of human factors in fire protection design by focusing on what people are trying to achieve through their actions. This ap- proach puts people at the center of sys- tems and forces designers to consider how the environment they create influ- ences people, and vice versa. Chubb6 of- fered a complementary approach, sug- gesting that incorporating a values-centered perspective into the de- velopment and application of regula- tions could foster an environment more conducive to controlling human error.

How can fire protection engineers put such principles into practice in the meantime? They can start by keeping in mind that most people who build don’t do so to save money. They build to make money or, at the very least, to pre- serve their wealth. This recognition brings with it an implicit acknowledge- ment that people accept the act of build- ing brings with it risks associated with unwise investments or poor construction practices. Making these risks clear to clients is among the fire protection engi- neer’s chief obligations.

Since clients rarely place fire safety – as distinguished from fire code compliance – among their chief concerns in the de- sign of new buildings, encouraging ap- propriate investments in fire safety re- quires fire protection engineers to obtain a clear understanding of the purpose a new building serves for the client. This requires looking beyond the occupancy classification to learn how the activities, intentions, and expectations of people will influence their actions with respect to the use and maintenance of the building.

SIZING UP THE SITUATION

A building that provides shelter and security against external threats like weather, earthquake, and theft is next to useless if it creates new and potentially greater risks of loss or damage from in- ternal threats. The enclosing nature of buildings creates new risks in the event of fire that owners and some fire protec- tion engineers may easily take for granted.

Many buildings designed even a few years ago now find their construction

out of step with owners’ and occupants’ expectations. Storage heights reaching towards the rafters, overstuffed closets filled with disused files and furniture, “open plan” office floors with cubicles and corridors that resemble rabbit war- rens, and rats’ nests of cables beneath floors and above suspended ceilings il- lustrate a few of the situations that occur when space does not magically expand to suit the users’ ever-changing notions of what they need.

If the building can’t cope with the users’ expectations under normal cir- cumstances, what chance does it have when a fire occurs? What chance do the people and property inside have? How will failure to account for these expecta- tions affect performance-based designs that take little account of the future?

Oddly enough, older buildings con- structed under prescriptive regulations or no regulations at all offer some inter- esting answers to these questions.

Maybe such structures are little more than the product of a simpler time, but many buildings designed and built at the end of the 19th and beginning of the 20th centuries have stood the test of time well, often finding themselves oc- cupied by many new, even novel, uses during their lifespans. Their stout con- struction and attention to detail reflected the romantic notion that people can conquer any obstacle through little more than the application of scientific princi- ples and perseverance. Without the so- phisticated tools considered common- place by contemporary engineers, designers built with a sense of place and permanence that embraced a scale sometimes larger than life itself.

It seems impractical if not outright im- possible to approach building design this way today. The reasons for this have nothing to do with technology or the science that produces it.

As faith in science grew, so did impa- tience with the status quo. Political up- heaval, a worldwide economic depres- sion, two world wars followed by political stalemate, and the unsettling re- ality of internecine conflict that has come about since the collapse of Com- munism have replaced optimism with pragmatism. Getting things right is less important than getting them done quickly and economically, before it’s too late.

What does all this have to do with fire safety? Nothing and everything.

13 Fire Protection Engineering NUMBER 16

Fire itself adheres to physical not hu- man laws. People ignore or neglect these laws at their own peril.

The infatuation with fast, the lack of practical impediments to finding scien- tific support for virtually any proposi- tion, the seemingly endless opportuni- ties to shift accountability so as to avoid responsibility, and the increasingly re- mote experience of fire as a cause of hu- man suffering have left those who own

and use buildings with little reason to pay it or fire protection engineers much heed. People aren’t intentionally indif- ferent about fire; they just have more im- portant things to worry about. Or so they think until a fire occurs.

BUILDING EXPECTATIONS

Sadly, people living in our postmod- ern world expect bad things to happen.

Most folks do not wish others harm, but they do hope that whatever ill next be- falls society won’t affect them directly. Far more often than not, they live on in a quiet yet mildly discomfiting confi- dence that this was in fact the case.

In the backs of our minds though, most of us live with some small fear that we could be next. Our limited attention to fire reflects the overwhelming need in terms of efficiency, effectiveness, and, indeed, sanity to attend only to those threats of which we are aware and that we can control. Anything else, particu- larly to the extent that it becomes a pre- occupation or distraction, represents wasteful inattention to the more imme- diate and pressing concerns of daily life.

Raising awareness of fire and engen- dering a commitment to fire safety among building owners and occupants need not require us to unduly alarm people. Indeed, distorting the likelihood or consequences of fire can either fuel unfounded fears or more likely instill in- difference in the form of learned irrele- vance.

Fire protection engineers can engage people in meaningful activities that pro- mote improved fire prevention and fire safety outcomes in a number of ways. Most of these approaches require atten- tion not only to assumptions about the roles people play and the biases that af- fect them, but the behaviors and actions we want people to display in different situations.

For example, discouraging the wildly variable premovement delays observed in evacuation trials and actual emergen- cies requires us to overcome the conse- quences of learned irrelevance in the in- stance of fire alarm signals. Believing that most alarms they experience are false because they saw no sign of fire, most people “learn” to disregard these signals and treat them as little more than another annoyance of modern life.

Something as simple as providing regular, perhaps even automated, feed- back to all building occupants, not just those who investigated the alarm, fol- lowing each activation would indicate what caused it and provide reassurance that the system does work, even if too well in the instance of things that only look like smoke but aren’t. What’s more, when such alarms are caused by avoidable circumstances such as im- proper siting of detectors, poor mainte- nance practices or defective apparatus, providing such information to building

■ Putting People Back in the Picture

FALL 2002 www.sfpe.org 14

occupants will usually stimulate expec- tations that someone will deal with the situation to prevent future inconve- nience.

Designing fire alarm systems to tattle on their owners has become common- place since the introduction of ana- logue/addressable systems with sophis- ticated microprocessor controls. But are we telling the right people when things aren’t working right?

The failure to maintain test regimes, perform routine maintenance, or re- spond promptly to dirty detectors un- dermines the reliability of fire alarm sys- tems required for the safety of the building’s occupants. Why not tell some- one who cares when these faults remain unaddressed for too long? A “Big Brother” approach need not become the norm either. Instead of notifying the lo- cal authority responsible for fire code enforcement, the system could just as easily and perhaps more effectively no- tify all the building occupants who rely on the system for their safety.

The idea that systems should engage

the active attention of their users by cre- ating intentional system states that com- mand attention is not new. Only its ap- plication to technology in the everyday world of homes and buildings is. How can we harness this tool for fire preven- tion?

Alerting people to the potentially harmful consequences of their actions can produce many positive benefits. Lately, some home appliance manufac- turers have incorporated audible warn- ing devices into their refrigerators to alert users when the door remains open for too long. Shutting the door saves en- ergy and prevents spoilage. Why can’t comparable devices be incorporated into electric and gas ranges for example to warn users who leave cooking unat- tended for too long?

A simple device of this sort could in- corporate a timer with limit states set in proportion to the heat output of the cooktop elements. A warning signal that intensifies as the time since activation in- creases would continue for a reasonable time limited by the burner heat output

before interrupting the power or fuel supply to the element or burner. This arrangement might not prevent all fires caused by unattended cooking, but it certainly wouldn’t make matters any worse than they already are.

Designing fire doors and smoke con- trol door assemblies so they sound a mild but annoying alarm when left open too long unless held by an approved device represents another useful appli- cation of this idea. By the same token, designing such doors with power-as- sisted openers and time-delay closers that permit the door to remain open for a time before returning to a closed posi- tion could aid the movement and safety of the elderly, those with mobility impairments, and staff in hospitals and similar care facilities without compro- mising fire safety.

More and more building owners em- ploy security systems to control access to their buildings and restrict move- ments within them to authorized areas. Clearly such security measures can be used to prevent incendiary fires or the

15 Fire Protection Engineering NUMBER 16

unauthorized disabling of fire protection systems. Integrating such measures with fire safety systems could also provide a ready and accurate indication of the number and location of people requir- ing evacuation. With some enhance- ments, these systems can provide tai- lored feedback to users and occupants reminding them of important fire safety messages or other information useful in their day-to-day use of the building.

BACK TO THE FUTURE

To some fire protection engineers, these ideas may sound a bit far-fetched, needlessly complex, or too costly to merit serious consideration. But these same ideas have already changed the way computers, home appliances, and telecommunications devices are de- signed. And now they are gaining in- creasing currency within the building in-

dustry as cybernetic building software and systems are developed to help building owners deploy cutting-edge telecommunications infrastructure and manage their energy costs to achieve ef- ficiency or gain new clients.

This leaves fire protection engineers with a choice between employing tech- nology to overcome what they consider human frailties or to improve human performance by putting people back in the picture.

In most situations, how people per- form will depend to a very great extent on what they expect to occur. This is clearly influenced by their relationship to the situation. Putting people back in the building fire safety picture will help us foster positive expectations of self-ef- ficacy and engagement with situations that not only achieve better performance in the event of fire but may actually help us prevent fires in the first place. ▲

Mark Chubb is with the New Zealand Fire Service.

REFERENCES

1 Bryan, J., “Human Behavior in Fire: The Development and Maturity of a Scholarly Area of Study,” in Human Behaviour in Fire, Proceedings of the First International Symposium, Shields, T. Ed., Textflow Ltd., Belfast: 1998.

2 Thomas, I., & Brennan, P., “Victims of Fire? Predicting Outcomes in Residential Fires,” in Human Behaviour in Fire: Understanding Human Behaviour for Better Fire Safety Design, Proceedings of the Second International Symposium, Shields, T., Ed. Interscience Communications Ltd. London: 2001.

3 Todd, A., A Spark Lighted in Portland: A History of the National Board of Fire Under- writers. McGraw-Hill, New York: 1966.

4 Groner, N., “Putting People Into the Performance-based Design Option,” Fire Technology 32 (3), pp 281-284, 1998.

5 Groner, N., “Intentional Systems Representations are Useful Alternatives to Physical Representations of Fire-related Human Behavior,” in Human Behaviour in Fire, Proceedings of the First International Symposium, Shields, T., Ed. Textflow Ltd., Belfast: 2001.

6 Chubb, M., & Williamson, R., “Value- based Fire Safety: A New Regulatory Model for Mitigating Human Error,” in Human Behaviour in Fire, Proceedings of the First International Symposium, Shields, T., Ed. Textflow Ltd., Belfast: 1998.

■ Putting People Back in the Picture

FALL 2002 www.sfpe.org 16

INTRODUCTION

I n situations where egress is complex or difficult, such as in high-rise buildings or large factories, human

voice is often used to provide informa- tion. Failure to understand message content can result in several ways. A message that is not intelligent may not be understood. A message spoken in Spanish to an audience that only under- stands Cantonese will not be under- stood. A person talking rapidly or with a speech impediment can cause a mes- sage to not be understood. Even a well- spoken, intelligent message in the language native to the listener can be misunderstood if it is not audible or if its delivery to the listener is distorted. These last failure mechanisms are the basis for the specification, modeling, and measurement of speech intelligibil- ity performance.

THE PROBLEM

Fires such as the King’s Cross fire in London in 1987 and an apartment fire in York, Ontario, have been cited as situa- tions where the lack of intelligible voice communication to occupants was a con- tributing factor in the losses.1, 2 We often see paging systems in places such as air- ports and meeting spaces with speakers every eight to twelve feet (three to four meters). How will the speech intelligibil- ity of the adjacent fire alarm system com- pare when it has speakers spaced 40 to 70 feet (10 to 20 meters) apart?

No one argues that a tone signal must be audible to the listener and that a voice transmission must be intelligible. Dis- agreements regarding audibility led the fire alarm industry to adopt audio indus- try definitions and measurement methods. This moved the industry from using sub-

jective evaluations of audibility to objec- tive methods.

In 1997, the Notification Appliances Committee of NFPA 72 began working with the audio industry to learn more about speech intelligibility and how to establish objective performance require- ments for emergency voice alarm com- munication (EVAC) systems. The goal was to define speech intelligibility performance in a way that could be objectively measured, eliminating sub- jective evaluations.

WHAT IS SPEECH INTELLIGIBILITY?

Figure 13 is useful in understanding the path of a voice signal from a talker to a listener.

The figure shows the types of error that can be introduced into the message at each stage. Problems or faults have a cumulative effect on message under- standing. For example, a person might speak with an accent but still be under- stood by a listener who is face-to-face with the talker. The communications system might add distortion that results in

the message not being understood. Or perhaps it’s understood when there is lit- tle or no background noise, but not un- derstood when there is background noise.

Researchers are addressing the two ends of the communications chain shown in Figure 1.4, 5 For the purposes of this article, speed of talking, language, and talker articulation are not directly addressed. They are indirectly addressed because a system that reliably delivers a message, with a limited amount of distor- tion, reverberation, and echo, is more likely to be understood even when a talker introduces problems or when a listener has impaired hearing. A system with a higher degree of intelligibility can offset some, but not all, deficiencies introduced by the talker or the listener.

Speech intelligibility is the measure of the effectiveness of speech. The mea- surement is usually expressed as a per- centage of a message that is understood correctly.6 Speech intelligibility does not imply speech quality. A synthesized voice message may be completely under- stood by the listener, but maybe judged to be harsh, unnatural, and of low qual- ity. A message that lacks quality may still be intelligible.

FACTORS AFFECTING SPEECH INTELLIGIBILITY

For speech to be intelligible, it must have adequate audibility (sound pres- sure level) and adequate clarity.

For audibility, we are concerned with the signal-to-noise ratio. Voice is highly modulated, and so while intelligibility measurements do incorporate audibility, it is not to the same standards used for audibility of tone generating systems.

Speech Intelligibility

Talker Microphone Mixer Amplifier Room

Amplifier

Listener

Assumed normal

Assumed normal

Intelligibility measures

Language speed

articulation

Noise reverberation echoes Language hearing

Bandwidth distortion

Figure 1. Voice Signal Path (Courtesy of K. Jacob, Bose® Professional Systems3)

17 Fire Protection Engineering NUMBER 16

Thus, a tone and a voice message that are both perceived as equally loud may have considerably different readings on a dB or dBA meter using fast or even slow time constants. That is one reason that audibil- ity measurements are not required by the National Fire Alarm Code for voice signals.

Phonemes are the smallest phonetic unit capable of conveying a distinction in meaning in a particular language and are instrumental in accurate word recogni- tion.7 Examples are the m of mat and the b of bat in English. Clarity is the property of sound that allows phonemes to be dis- tinguished by a listener.8 Clarity is the freedom of these sound units from distor- tion introduced by any part of the sound system or environment. Recently, a major U.S. cellular telephone company has im- plemented a television ad campaign play- fully pointing out the very real problem of phoneme clarity.

Clarity can be reduced by: 1) amplitude distortion caused by the electronics/hard- ware; 2) frequency distortion caused by either the electronics/hardware or the acoustic environment; and 3) time domain distortion due to reflection and reverbera-

tion in the acoustic environment. Designers and engineers have the

greatest effect on speech intelligibility by their choice of equipment, the number and distribution of loudspeakers, and the power at which they are driven.

MEASURING SPEECH INTELLIGIBILITY

The system hardware and the acoustic environment cannot be separated when evaluating speech intelligibility. Installa- tion choices, such as wire size and rout- ing, affect power levels and induced noise. Mounting locations and surfaces affect sound fields, and construction materials and furnishings affect acoustic parameters. Thus, the performance metric for speech intelligibility must assess all of the requisite parameters.

International Electrotechnical Commis- sion (IEC) and International Standards Or- ganization (ISO) standards already incor- porate objective methods for evaluating speech intelligibility. The standard, IEC 60849, Sound systems for emergency pur- poses,8 is similar to NFPA 72. Some of the methods recognized in the standard are

subject-based, and others use instrumen- tation. ISO 9921 also references estab- lished methods.6 For each of the recog- nized methods, there already exists an internationally accepted standard for the test method/protocol.

The IEC standard includes a chart that equates the scales for each of the differ- ent test methods to a common scale called the Common Intelligibility Scale (CIS). Evaluation of speech intelligibility may use any one of several methods cited in the standard. Four of these methods use test instruments. Three subject-based methods are also permitted. One method has both a subject-based solution and an instrument-based solution. These are summarized below in Table 1.

For the four instrument-based solutions, at present there are at least six different instruments available from four different manufacturers. Consult the references for more detail on each of the test methods.

The recommended minimum perfor- mance level for EVAC systems is that the average CIS score, less one standard devia- tion, be 0.70 or greater. This permits devia- tions, does not require an exact score, and

Method

STI – Speech Transmission Index

RASTI – Rapid Acoustics Speech Transmission Index

PB – Phonetically Balanced Word Scores

MRT – Modified Rhyme Test

AI – Articulation Index

%ALcons – Articulation Loss of Consonants

Standard Ref. in IEC 60849

IEC 60268-16 The objective rating of speech intelligibility by speech transmission index, 1998

IEC 60268-16 The objective rating of speech intelligibility by speech transmission index, 1998

ISO/TR 4870 Acoustics – The construction and calibration of speech intelligibility tests, 1991

No reference given

ANSI S 3.5, Methods for the calculation of the articulation index, 1969 ANSI S 3.5, Methods for the calculation of the speech intelligibility index (SII), 1997

Peutz, V.M.A., “Articulation loss of consonants as a criteria for speech transmission in a room,” J. Aud. Eng. Soc. 19, 12, December 1971

Comments

This is an objective, instrument-based method. Requires hardware and software for measurement and solution. Available in a computer-based solution, as a feature of some multi- function audio analysis equipment, and as a handheld meter.

This is an objective, instrument-based method. Reduced STI method. Available in a handheld format.

This is an objective, subject-based method. ANSI S3.2 Method for measuring the intelligibility of speech over communication systems, 1989, is a better reference for evalua- tions using the English language. Notification Appliances Chapter permits ANSI S3.2 use, although ISO/TR 4870 is also permitted.

This is an objective, subject-based method. No standard listed. ANSI S3.5 notes that the method has the same limits as given in ISO/TR 4870 (PB). Good reference is ANSI S3.2 Method for measuring the intelligibility of speech over communication systems, 1989.

This is an objective, instrument-based method. The 1969 version is referenced. This has been updated to the 1997 edition. Requires hardware and software for measurement and solution.

This is an objective, instrument-based method or an objective, subject-based method. Available in a computer-based solution.

Table 1. Speech Intelligibility Test Methods

FALL 2002 www.sfpe.org 18

ensures that approximately 84% of the space has a score of 0.70 or better – as- suming a normal distribution of the results. The 2002 edition of the NFPA 72 Hand- book contains a discussion of why a CIS of 0.70 was used as a baseline.9

PLANNING, DESIGN, INSTALLING, TESTING, AND USING

A reliable communication system must be properly planned, designed, and in- stalled. Testing uncovers faults and allows corrections to be made, but also shows successful techniques for future reference. One issue that designers and authorities must face when planning a system is the question of where intelligible voice com- munication is needed.

In a large space used for public meet- ings, conventions, and trade shows, an EVAC system needs to be reliably intelligi- ble because it is intended to give informa- tion to the general public that is not famil- iar with the space. In large public spaces, a person should not have to move any great distance to find a place where they can understand the message.

However, in a high-rise apartment build- ing, is voice intelligibility required in all spaces? It may not be necessary for the EVAC system to be intelligible in all parts of the apartment even though it must be audi- ble in all parts. It may be sufficient to pro- vide a speaker in a common space to pro- duce an adequate audible tone to awaken and alert. When the voice message follows, it may not be intelligible behind closed bedroom and bathroom doors. The occu- pants, in a familiar space, can move to a lo- cation where a repeating message can be intelligibly heard. The same signaling plan may work for office complexes – a person may have to open their office door to reli- ably understand the message.

Once the design team plans to have some type of a system and decides that the system must be intelligible in certain spaces or areas, the fire alarm code’s re- quirements and recommendations for intelligibility may become part of the per- formance design objectives or goals. It is important that all of the stakeholders, in- cluding the code officials, agree on the design goals and objectives.

By agreeing upon specific design goals and objectives, multiple approaches can be used to achieve the desired perfor- mance. The National Fire Alarm Code permits designers to use any and all rea- sonable means to achieve the objectives.

Designers and installers who are new to the subject or who want to learn more about proper voice system design and in- stallation should consult more in-depth resources.10

Just as fire protection engineers can model fires, acoustic and audio engi- neers can model speech intelligibility before a building is built and before a system is installed. Acoustic properties of materials are well documented and result in reliable evaluations of proposed de- signs in the same way that a fire protec- tion engineer might evaluate flame spread and smoke contribution of mate- rials. The electronic performance of the communications system can be adjusted in the models based on data from the system manufacturer.

At this time, there is no requirement in the body of the National Fire Alarm Code that speech intelligibility actually be mea- sured. The measurement methods are discussed only in the Annex of the Code. However, if it is decided to measure intel- ligibility, how many tests should made in a particular space? Currently, there is no guidance for audibility measurements regarding the number and locations of test points nor for intelligibility measurements. With audibility, we have an intuitive sense of where a system might fail, and we tend to concentrate our testing plan in those ar- eas. How many designers, technicians, and authorities have such intuition regard- ing intelligibility? This is not an argument to not test intelligibility. That would just be a head-in-the-sand reaction. Rather, it means that we need to start testing and that we are likely to test a larger number of points initially as we gather experience.

As with audibility, there are methods to test when a space is not occupied and then “add in” the expected or measured noise level at a later time during analysis. This permits less-invasive testing. It is common practice to test the audibility of systems before a space is occupied. Expe- rience and available data permit us to estimate the expected noise level and compare it to the nonoccupied system performance. Similar procedures are done for intelligibility measurements. However, the required data may not be readily available or apparent to the fire protection engineer. Also, as with audibility mea- surements, intelligibility measurements in a space cannot be reliably made unless those parts of the interior finish that affect sound transmission and attenuation have been installed. ▲

REFERENCES

1 Proulx, G., “Is It Wise to Evacuate During a High-Rise Fire?,” National Fire Protection Association World Safety Congress, Denver, CO, May, 2000.

2 Proulx, G., “Highrise Evacuation: A Questionable Concept,” 2nd International Symposium on Human Behavior in Fire – Conference Proceedings, Interscience Communications, London, 2001.

3 Jacob, K., “Understanding Speech Intelligibility and the Fire Alarm Code,” Bose® Professional Systems, Framingham, MA. Presented at the National Fire Protection Association World Safety Congress, Anaheim, May 14, 2001. Available at www.rpsa-fire.com.

4 Proulx, G., & Sime, J.D., “To Prevent ‘Panic’ in an Underground Emergency: Why Not Tell People the Truth?,” Proceedings of the 3rd International Symposium on Fire Safety Science, Elsevier, London, 1991.

5 Pauls, J.L., & Jones, B.K., “Building Evacuation: Research Methods and Case Studies,” Fires and Human Behaviour, John Wiley and Sons, New York, 1980.

6 ISO 9921-1, Ergonomic assessment of speech communication – Part 1: Speech interference level and communication dis- tances for persons with normal hearing capacity in direct communication (SIL method), International Standards Organization, 1996.

7 Mannell, R.H., “Natural and synthetic speech intelligibility and quality testing,” 1984, Speech, Hearing and Language Research Centre (SHLRC), Macquarie University, Sydney, Australia, http://www.ling.mq.edu.au/~rmannell/slp8 07/synth_assessment/intelqual.shtml.

8 IEC 60849, Sound systems for emergency purposes, 1998.

9 Moore, W., & Richardson, L., editors, NFPA 72, National Fire Alarm Code Handbook, National Fire Protection Association, Quincy, MA, 2002.

10 Schifiliti, R.P.,“Intelligibility – Extended Bibliography,” R.P. Schifiliti Associates, Reading, MA, June, 2002. Available at www.rpsa-fire.com.

Editor’s Note – About This Article This is the third in a continuing series of articles that is supported by the National Electrical Manufacturer’s Association (NEMA), Signaling Protection and Communications Section, and is intended to provide fire alarm industry-related information to members of the fire protection engineering profession.

19 Fire Protection Engineering NUMBER 16

By Rita F. Fahy, Ph.D.

INTRODUCTION

E vacuation models are important tools for the evaluation of engi- neered designs, because such

evaluations require the estimate of safe evacuation time for the occupants. The time needed to escape must be less than the time available to escape, which would be estimated using fire and smoke transport models. The

designer or engineer has a range of options for predicting evacuation time.

The available tools range from fairly simple equations to the evacuation counterparts to CFD fire models. Each option has its own advantages and dis- advantages. The user needs to weigh the assumptions embedded in each method in order to determine which is most ap- propriate for the application at hand.

Just as fire growth models can predict the spread of smoke and other toxic products throughout a structure, evacua-

tion models can predict the location of people as they move through a struc- ture. Used together in the evaluation of a design, these models can provide some indication of the risk that occu- pants might face under a modeled sce- nario.

The engineer or designer also needs access to, and an understanding of, the currently available data that should be used either in combination with the methods or as input to methods like computer models. Finally, the engineer

Tools for the Simulation of

Human Behavior

FALL 2002 www.sfpe.org 20

or designer needs to understand the role of safety factors and other tools for properly and conservatively interpreting results.

This article will discuss some different types of tools that are available to model human behavior in fires and the issues that must be considered in making the choice. A more complete discussion of the handling of human behavior in engi- neered designs will be found in the forthcoming SFPE guide.1

WHY AND HOW HUMAN BEHAVIOR IS MODELED

Methods for calculating evacuation time and modeling human behavior vary widely in complexity and sophisti- cation. Before choosing a method, it is essential for the user to understand the components of total evacuation time.

The entire span of time it takes to leave a building actually starts at the on- set of the fire event. At a later point in time, the first cues reach the occupants. Some additional time may elapse before the occupants react and first respond to the fire cues. The initial reaction may re- sult in activities such as investigation (to confirm that there really is a fire), fire- fighting, alerting others, etc. Once the occupants decide to leave, additional time will probably pass while they make preparations, collect valuables, discuss options, etc. Then, once they begin to move to the exit or a place of refuge, the time it takes to travel to those points will pass. To account for total evacua- tion time, the engineer or designer must estimate:

• time to notification, • reaction time (time to perceive the

cues and realize that some action is needed),

• pre-evacuation activity time, and • travel time. Time to notification could be mod-

eled using fire and smoke models, and models for detector activation, although human detection will always be a possi- bility and should be considered. The other three components require a com- bination of modeling human behavior and predicting travel time.

True predictive models of human be- havior in fire do not yet exist. There has not been sufficient research on human behavior in fire conducted so that mod-

els can predict reaction and response, and the associated delays, with any de- gree of accuracy. However, a great deal of data has been collected that provides observed delay times in evacuations, in both fire and drill situations.2 Unfortu- nately, not all types of properties, for example, industrial properties and healthcare facilities, are well-represented in the data sets.

TYPES OF CALCULATION METHODS

There are different types of evacuation models. There are simple, straightfor- ward calculation methods for estimates of evacuation times. These equations or simple computer models may be based on observed movement from drills and experiments. The SFPE Handbook of Fire Protection Engineering provides excel- lent discussion of calculations to be used to estimate evacuation time, and the reader is referred to that discussion. 3, 4

The next level of complexity is net- work flow models that handle large numbers of people. These models are useful for benchmarking designs and can indicate potential bottlenecks that would negatively impact an evacuation, but they cannot be used to predict what any one person might experience, since they treat the occupants like water in a pipe rather than as individuals. As a re- sult, they cannot be used to determine the effects of a fire on occupants as they move through contaminated spaces. These models can be used to calculate optimal evacuation times because they move occupants in the most efficient manner. If such a model predicts that occupants will not evacuate a building before unsafe conditions exist, it is un- likely that an evacuation in real life could have a shorter evacuation time, and design changes will be necessary.

Behavioral simulation models are the most complex, treating more of the vari- ables related to both movement and be- havior. Their added complexity requires tremendous amounts of data for their development, if the assumptions they contain regarding behavior are to be based on reality rather than expediency. Their users also need a fuller under- standing of the components of human behavior in fire in order to choose ap- propriately among available options. Al-

though these models may produce visu- ally attractive results, the user must be very careful to understand the basis of assumptions embedded in the model, since so little data on occupant behavior actually exists in the literature. These models often have quite extensive input requirements, sometimes requiring in- formation on occupant characteristics, such as mobility or patience, that are not measurable and whose distribution in the general population is not known.

WHAT A POTENTIAL USER SHOULD LOOK FOR IN A CALCULATION METHOD

An evacuation model that attempts to include a huge range of occupant char- acteristics into movement or behavior calculations raises the question of where the data come from. What are the ap- propriate values or distribution of values to use, and where will the user find them? For example, it can be persua- sively argued that a person’s patience, motivation, and agility play a role in their actions and behavior during an evacuation, but if a model specifically requires a value for each occupant’s “pa- tience,” “motivation,” and “behavior,” what data should be used? What are the appropriate values, and where have they been reported? What ranges of val- ues are there for the general population that should be used when modeling a building whose design is under consid- eration? What are the uncertainty bounds around the values, and how do they correlate with each other? In the end, is precision gained or lost by using them explicitly?

On the other hand, a simpler model is implicitly making the assumption that certain details are not essential to the calculation of estimation of total evacua- tion time. For example, a model that does not have the ability to predict or estimate the delays that will occur for some people in an evacuation while they are moving out of the building is making the implicit assumption that people neither lose time during their evacuation nor change direction. Unless some other adjustment is made to com- pensate, either by the model or in the model input, such a model will consis- tently underpredict or underestimate the total evacuation time. The justification

21 Fire Protection Engineering NUMBER 16

for this, however, is that the data on who will delay, for how long, and with what effect are not available, at least not in sufficient detail and quantity to credi- bly model or predict that behavior. Simi- larly, the data necessary to predict who will change direction, at what point, and to what effect are not available.

A list of questions that should be asked by an engineer or designer in choosing an appropriate calculation method was compiled for the SFPE Handbook.4 Some of the major issues addressed on that list will be mentioned here but all are important considera- tions.

The first question the user has to con- sider is whether an optimization model or a more detailed behavioral model, or a risk assessment method, is most ap- propriate. In the course of working on a building project, an engineer might use a combination of models – a hand calcu- lation or a network flow model to obtain a quick assessment of the evacuation time necessary, and then a more compli-

cated model to test the impact of changes in assumptions about the fire scenario or occupant characteristics.

Behavioral models attempt to realisti- cally predict the actions and decisions made by occupants during an evacuation. These models are attractive because they seem to more accurately simulate evacua- tions. However, due to the scarcity of be- havioral data, they tend to rely heavily on assumptions, and it is not possible to gauge with confidence their predictive accuracy. Users of such models need to establish confidence in the assumptions before relying on the results.

To set up the travel options for occu- pants, evacuation models use either a network of nodes and arcs or a mesh structure. When spaces and travel paths are defined using nodes and arcs, the movement of occupants is restricted to those paths, and the predicted move- ment is seldom smooth. (Occupants tend to jump from node to node, for ex- ample.) A mesh structure lays a grid of “tiles” over entire enclosures on a floor

plan, and occupants are able to occupy or move from tile to tile. This allows the more precise location of occupants throughout spaces. Setting up the de- scription of a floor plan using nodes and arcs can be very time-consuming, but once done, it does not have to be rede- fined unless the structure is redesigned. CAD packages can often be used to in- put the floor plan description for a mesh structure, but models that use a mesh are generally more time-consuming to run, due to the complexity of the calcu- lations used to move people through such open spaces.

The degree to which behavior is sim- ulated varies extensively among avail- able models. Some require the user to estimate and input the premovement de- lays that are appropriate for a particular scenario and structure. Others include behavior “rules” and will predict behav- ior according to those rules. The engi- neer must be cautious in choosing a model. The complexity of some models implies a greater predictive capability,

■ Simulation of Human Behavior

FALL 2002 www.sfpe.org 22

but again the scarcity of data available on behavior means that a great number of assumptions are imbedded in the models, and the appropriateness of those assumptions is critical in evaluat- ing the validity of a model’s results.

Of particular importance are the questions concerning the appropriate- ness of a model to the task at hand. The user must be clear on the assumptions that are explicitly stated by the model developer. Even more importantly, the user must be aware of the assumptions embedded in a model. For example, if a model uses a constant travel speed for occupant movement, the user must un- derstand the source of that value and its applicability to the scenario being mod- eled. If all occupants will move at the same speed, the user must be able to justify the absence of differently abled occupants. The user must determine whether the model has been validated; if so, how and to what extent was the validation performed? The validation should address these questions and

assist the user in choosing appropriate input values and understanding the impact of those choices.

The user must take into considera- tion the issue of safety factors: do the model results incorporate safety fac- tors? If so, how is that done? If safety factors must be applied to the results by the user, that must be specifically stated, and appropriate methods for doing so must be described. However, remember that a true safety factor is intended to compensate for the uncer- tainty in the calculation, not for a bias leading to systematic underprediction, as in most simple models. The same multiplier cannot be validly used for both, and the uncertainty and bias associated with evacuation prediction may each be longer than those associ- ated with common models of fire effects. That means traditional safety factors may be too small and may need to be derived from direct examination of the uncertainty and bias of the evac- uation models used. ▲

REFERENCES

1. SFPE Engineering Guide to Human Behavior in Fire, review draft, Society of Fire Protection Engineers, Bethesda, MD, 2002.

2. Fahy, R.F., & Proulx, G., “Toward Creating a Database on Delay Times to Start Evacuation and Walking Speeds for Use in Evacuation Modeling,” Proceedings of the 2nd International Symposium on Human Behavior in Fire, Interscience Communications Ltd., London, 2001.

3. Proulx, G., “Movement of People: The Evacuation Timing,” SFPE Handbook of Fire Protection Engineering, 3rd edition, P.J. DiNenno, editor, NFPA, Quincy, MA, 2002.

4. Nelson, H.E., & Mowrer, F.W., “Emergency Movement,” SFPE Handbook of Fire Protection Engineering, 3rd edi- tion, P.J. DiNenno, editor, NFPA, Quincy, MA, 2002.

FALL 2002 www.sfpe.org 23

By Guyléne Proulx, Ph.D.

DECISION-MAKING DURING A FIRE

D ay-to-day decision-making such as choosing a meal from a menu or picking the best

road to get to a meeting on time is quite different from decision-making during an emergency. Even major life change decision-making, such as the choice of a career, buying a house, or getting married is still different from decision-making during a fire.

There are three main reasons which differentiate decision-making during a fire from other decisions. First, there is much more at stake in a fire. The con- sequences of a decision could deter- mine the survival of the decision-maker and of the people he or she values the most. Second, the amount of time avail-

able to make a decision is limited. Of- ten the decision-maker will feel that a decision should be made quickly be- fore crucial options are lost. Third, the information on which to base a deci- sion is ambiguous, incomplete, and un- usual. It is also usually impossible for the person to find more appropriate in- formation due to the lack of time and the means to get information.1

Differences of decision-making during a fire from everyday life decision-making

• More at stake, possible life and death outcome for oneself or loved ones

• Limited time to make decisions which might be irrevocable

• Ambiguous limited information on which to base decisions

STRESS IS NOT PANIC

During a fire, the nature of the infor- mation obtained, the limited time to re- act, and the assessment of danger will create a feeling of stress. This stress will be felt from the moment ambiguous in- formation is perceived until well after the event when the person has reached safety.2 During the course of the event, the intensity of stress experienced will vary as a function of newly perceived information and assessment of the ef- fectiveness of the decisions taken.

The media and general public often mention the potential of mass panic during fires, imagining a crowd that suddenly wants to flee danger at all cost, possibly getting trampled or crushed in the process. Although these types of behavior are extremely rare in fires, the expectation that people will panic is very strong. This belief is very much nourished by the media and movie industry, which plays on strong emotional images. In fact, panic in the form of irrational behavior is rare dur- ing fires, and researchers have long ago rejected this concept.3, 4

Actual human behavior in fires is somewhat different from the “panic” scenario. What is regularly observed is a lethargic response to the fire alarm, voice communication instruction or even the initial cues of a fire.5 Except for low-rise residential buildings, where

Cool Under Fire

24 Fire Protection Engineering NUMBER 16

occupants feel that it is their responsi- bility to investigate an unusual smell, noise or movement, occupants are usu- ally not very responsive in the initial moments of a fire. People are often cool during fires, ignoring or delaying their response to the initial cues of an actual emergency. Once occupants de- cide that the situation requires moving to an area of safety, the time left could be minimal.

If the person eventually assesses the situation as an emergency, instead of panic what is most commonly observed is an increased level of stress. Stress is not panic. It is considered that every person involved in an emergency will eventually feel some stress regardless of their age, gender, past experience, training, or cultural background. This stress is not an abnormal reaction or a negative response; on the contrary, stress is regarded as a necessary state to motivate reaction and action.6 The per- formance of a person dealing with an increased level of stress will depend on the task demands, the environmental conditions, and the person himself or herself.7 Decision-making under stress is often characterized by a narrowing of attention and focusing on a reduced number of options. This explains why training is so important, since people are unlikely to develop new solutions under heightened stress; a well-run de- cision plan learned and practiced be- forehand is easier to apply under stress.

IMPACT OF SEPTEMBER 11

Unless very well trained, occupants in high rise buildings are usually reluc- tant to leave their floor and are pre- pared to stay in place when the fire alarm goes off. Phased evacuation or a protect-in-place approach are seen as less disruptive and more efficient by high rise building occupants and man- agement. Staying in place during actual fires is sometimes the official fire safety plan in high rise buildings8 or the cho- sen response by occupants.9

To make a decision during an emer- gency, people will process information, both perceived in the environment and drawn from past experience.10 Part of this past experience now includes the events of September 11, which received unprecedented and sustained media coverage. Everyone has repeatedly

seen the airplanes hit the towers, the fires, and the following collapses of the buildings; these terrible images may have changed the public perception of fire risk.11 People may be fearful that a similar attack will be made on their building or that a fire could cause their building to collapse. If people have de- veloped a new attitude toward safety, their response, in case of an emer- gency, might be different from our past expectations. On the one hand, it is possible to imagine that instead of a lethargic response to initial fire cues, people will have a different response in light of the events of September 11. They might refuse to protect-in-place and proceed en masse toward the stair- wells to evacuate. On the other hand, the public is known to be fairly resilient at changing their behavior and attitude; they might have already forgotten and resume their past behavior. These are contradictory speculations. To better understand the impact of the events of September 11, it is essential to conduct thorough studies. It is important to know if a new attitude toward fire safety exists and the potential conse- quences of different evacuation scenar- ios. The design of buildings, fire safety features, and fire safety plans might need to be modified to accommodate this new attitude toward fire safety, or maybe nothing has changed. We need to know.

LOOKING AT THE BIG PICTURE

To study and understand occupant behavior in fire, it is essential to take into account the characteristics of three interacting dimensions of the event. These dimensions are the occupant, the building, and the fire. The first di- mension that should be considered is the occupant characteristics. The char- acteristics of the occupant are an es- sential component that impact on their likely response and egress from a fire. Research has demonstrated, for exam- ple, that gender, age, physical ability, and group formation will have a sub- stantial impact on response time and speed of movement.12 Intuitively, it is known that a number of other occu- pant characteristics such as past fire ex- perience or familiarity with the build- ing should play a part in the response to a fire, but there is limited data to

back up these insights. A second dimension of the fire event

is the building where the situation is happening. In building codes, occu- pancies are usually classified according to their size and use. These classifica- tions, although extremely useful for professionals in the construction indus- try, are ill-defined to look into human behavior in fire. The occupant response to a fire alarm signal in a theater, a mu- seum, or an airport is likely to be differ- ent although these are all assembly buildings. Instead of looking at build- ings from their occupancy classifica- tion, the focus should be on some spe- cific building characteristics that could explain occupant response. For exam- ple, the information provided through a voice communication system is a better indicator, than the occupancy type, to predict occupant response. The type of activity that occurs in a building is also an important factor that could explain occupant response. For example, if the fire alarm goes off at an airport, while a person is walking toward the exit with her bags on a trolley, she is likely to have a different response compared to a man that is waiting for an interna- tional flight inside the security gates.

The management of a building is an- other very important building charac- teristic that can contribute to the suc- cess or failure of occupant evacuation. It is documented that the activation of the fire alarm signal is rarely sufficient to trigger evacuation movement in pub- lic buildings, unless this signal has been supplemented with well-trained staff or live information provided through a voice communication system.13 However, the information content that is most likely to prompt a specific response from occupants is not very well known at the moment.

Fire characteristics are the third ele- ment that will impact the occupant de- cision-making. Some preliminary stud- ies show that the smell of smoke is probably not sufficient to waken sleep- ing adults.14 How people tend to react to smoke, based on its color, smell, acridity, or thickness, is still an impor- tant area of research. What is known at the moment is that people are prepared to move through a fairly significant quantity of smoke, however, the char- acteristics of the smoke, the distance traveled and the conditions surround-

■ Cool Under Fire

FALL 2002 www.sfpe.org 25

ing such a decision are not well under- stood.12 Fire scenarios have to be thought through and taken into ac- count when assessing the likely behav- ior of building occupants.

To predict human response, it is nec- essary to take into account how the oc- cupant, building, and fire characteristics will interrelate in a specific case. This is not an easy task.

MISSING DATA

Some data on human behavior in fire is already available, but if misused, it can lead to fuzzy deductions. For ex- ample, it is acknowledged that a good marking system will support way find- ing in a building; however, a well-lit fire exit sign above a door may not be enough to cause occupants to leave through that door during an emer- gency. Understanding decision-making in fire is necessary to envision occu- pants’ likely response during an emer- gency. Practitioners have to acknowl- edge that there is a lot more than simple common sense to human be- havior in fire. Human behavior in fire is a scientific field that identifies facts, concepts, and relationships established through systematic observation and ex- perimentation. Drawing conclusions about occupant behavior on the ac- count of the fire chief or a few by- standers is not sufficient.

A number of areas of human behav- ior in fire require more systematic data to be collected using sound scientific methodologies. In the short term, re- search should be concentrated on a few main topics. One of these topics is to determine the responses that differ- ent cues will trigger from occupants. Traditionally, the expectation was that everybody starts evacuating when the fire alarm signal activates. Numerous cases have demonstrated that it is not always likely to occur.9 The cues or in- formation that will instigate the best re- sponse according to the occupant and the building characteristics is still not fully known.

Response time is another area where more research is needed. A few case studies have been conducted to mea- sure occupant delay time between noti- fication of a fire and the beginning of evacuation. These studies represent only a small sample that can hardly be

generalized at the moment. It is sus- pected that training can have a major impact on occupant response, but no data on this dimension seems readily available yet. The presence of staff on the premises is another important factor that can influence the behavior of occu- pants, but there is no way to take this factor into account at the moment. The social interaction among people in- volved in the event can certainly influ- ence the occupant response but this di- mension lacks data. The impact of occupants with a disability is another area where knowledge is slowly build- ing up and where more work is re- quired. Finally, we can question if all the data already accumulated is trans- ferable from different continents and if cultural differences actually exist. As in many young sciences, the field of hu- man behavior in fire requires more data to be gathered through sound method- ologies in order to eventually come up with solid models and theories that can be verified and validated.

GETTING DATA

There are a variety of means to con- duct research in human behavior. They all complement one another and add to the knowledge base of this scientific field. Invaluable data can be accumu- lated following actual fires. There should be a systematic method used to obtain victim and fire service accounts to create a large statistical data bank. This would help in drawing compar- isons among countries and identifying trends and recurring events. Specific in- terviews and a walkthrough with fire victims, contrasted with expert accounts are invaluable sources of data. Record- ings from closed-circuit televisions should be used if present in a building at the time of a false or genuine alarm. Conducting field studies such as orga- nized fire drills can also provide good insight into the possible initial behavior and movement of occupants. Con- trolled experiments could obtain pre- cise data for well-defined conditions. It is the accumulation of this data from different sources that will help develop a better understanding of human be- havior in fire. ▲

Guylène Proulx is with the National Research Council of Canada.

REFERENCES

1 Proulx, G., “A Stress Model About People Facing a Fire,” Journal of Environmental Psychology, Vol. 13, 1993, pp. 137-147.

2 Lazarus, R., & Folkman, S., Stress, Appraisal, and Coping, Springer, New York, 1984.

3 Sime, J., “The Myth of Panic,” in D. Canter (Ed.), Fires and Human Behaviour, John Wiley & Sons, Chichester UK, 1980, pp. 63-81.

4 Keating, J., “The Myth of Panic,” Fire Journal, May 1982, pp. 57-61.

5 Proulx, G., “Occupant response to fire alarm signals,” National Fire Alarm Code Handbook. National Fire Protection Association, Quincy, MA, 1999.

6 Selye, H., “The Stress Concept and Some of Its Implications,” in V. Hamilton & D.M. Warburton Eds., Human Stress and Cognition, John Wiley & Sons, London, 1979.

7 Wesnes, K., & Warburton, M., “Stress and Drugs,” in G. R. J. Hockey (Ed.), Stress and Fatigue in Human Performance, John Wiley & Sons, New York, 1983.

8 Proulx, G., “The Impact of Voice Communication Messages During A Residential Highrise Fire,” Human Behaviour in Fire – Proceedings of the First International Symposium, University of Ulster, Belfast, UK, 1998.

9 Proulx, G., “Critical Factors in High-Rise Evacuations,” Fire Prevention, No. 291, July/August 1996, pp.24-27.

10 Janis, L., & Mann, L., Decision-making, The Free Press, New York, 1977.

11 Finucane, L., “Public Perceptions of Risk,” The Skeptic, Vol.19, No. 2, pp. 8-9, 16, 1999.

12 Bryan, J., “Behavioral Response to Fire and Smoke,” SFPE Handbook of Fire Protection Engineering, 3rd Edition, National Fire Protection Association, Quincy, MA, 2002.

13 Proulx, G., & Sime, J., “To Prevent ‘Panic’ in an Underground Emergency: Why Not Tell People the Truth?,” Fire Safety Science – Proceedings of the Third International Symposium, G. Cox & B. Langford (Eds), Elsevier Applied Science, London, 1991.

14 Lynch, J., “Nocturnal Olfactory Response to Smoke Odor”, Human Behaviour in Fire – Proceedings of the First International Symposium, University of Ulster, Belfast, UK, 1998.

FALL 2002 www.sfpe.org 26

By Brian D. Black

INTRODUCTION

I mmersed in the thousands of articles, papers and reports on the tragedy of the attack on the World Trade

Center were items addressing the issue of life safety for, and evacuation of persons with, disabilities. Stories were told of people in wheelchairs who died in the resulting collapse of Towers 1 and 2, and anecdotal report- ing told of at least a few saved before the buildings were lost.

Most professionals agree that as hor- rendous as September 11 was and will always be, the field of fire protection engineering cannot and will not be dri- ven by that single event. Indeed, for those of us involved in the field of life safety for persons with disabilities, the World Trade Center is becoming the anomaly that masks the true loss of life that consumes disabled persons day af- ter day across the United States. This ar- ticle steps away from our nation’s losses in downtown Manhattan to take a broader view of the challenge of life safety and the disabled.

“Disability” as a topic poses addi- tional problems and raises issues that far exceed the limits of this article. Un- der federal statutes, disability includes the loss of sight experienced by an 85-year-old retiree, the deafness of a child exposed to rubella in utero, even the drug or alcohol abuse of an unem-

ployed man in a single-room occu- pancy in the “bad” section of town. This article is limited to the fire protec- tion and life safety needs of persons with mobility impairments – specifi- cally, those unable to use stairs as an independent means of egress from buildings.

U.S. HISTORY OF ACCESSIBILITY AND LIFE SAFETY FOR PERSONS WITH DISABILITIES

The movement for accessibility in our nation’s public buildings and facili- ties began after World War II and the return of soldiers who had incurred spinal cord injuries during battle. As America entered the postwar construc- tion boom of the 1950s, veterans sought to change our nation’s construc- tion practices to include what was referred to as barrier-free design.

Life Safety, Fire Protection, and

Mobility- Impaired Persons

27 Fire Protection Engineering NUMBER 16

In 1959 the American Standards As- sociation, acting on the request of the President’s Committee on Employment of the Handicapped, called a general conference of groups interested in ac- cessible design. This conference initi- ated a project to develop prescriptive requirements for access, and Making Buildings Accessible to and Usable by the Physically Handicapped, A117.1- 1961 was born.

All of eleven pages long, even this early document made passing reference to the problem of egress for persons with mobility impairments. It stated:

“Because entrances also serve as ex- its, some being particularly important in case of an emergency, and because the proximity of such exits to all parts of buildings and facilities, in accordance with their design and function, is essen- tial (see 112 and 2000 through 2031 of American Standard Building Exit Code, A9.1-1953), it is preferable that all or most entrances (exits) should be acces- sible to, and usable by, individuals in wheelchairs and individuals with other forms of physical disability herein applicable.” 1

The concept had a number of flaws from the start:

• even in single-story buildings, it is not necessarily true that all en- trances serve as exits (or that all exits are normal building en- trances);

• the note is in nonmandatory lan- guage, and suggesting something is “preferable” does not make it a requirement;

• “entrances” are normally at or near grade, and the question of how to address exiting from stories above or below grade remained unan- swered.

And while many experts agreed that such provisions were vague and inef- fective,2 United States federal guidelines3 and standards4 would con- tinue to include similar perfunctory language regarding egress for persons with disabilities without adequately addressing the issue.

Why this lack of attention to a funda- mental problem of life safety occurred is in many ways a uniquely American phenomenon. When requirements for accessibility were first introduced in the country, they applied to a very small

number of buildings, typically owned by a government entity or funded by the federal government.5 Private offices, restaurants, hotels, and similar occu- pancies continued to be built with stairs and narrow doors that rendered access into the buildings impossible and the question of egress by mobility-impaired persons moot. However, the 1970s ush- ered in a disability rights movement in which nondiscrimination on the basis of disability (and a resulting access into more and more buildings) was per- ceived as a civil right, similar to those rights afforded others for whom dis- crimination on the basis of race or eth- nicity was prohibited. The battles for these newly claimed rights were waged in the legislatures and courts, and thus not exposed to the health, safety, and general welfare tests imposed by build- ing codes and standards. And because life safety concerns were often seen by disability advocates as red herring excuses for not installing ramps and elevators, “How will you get out once you’re in?” became a challenge to a civil right and not a legitimate concern for building design. Laws and regula- tions in virtually every state mandated access into more and more buildings and facilities, increasing the likely exposure of persons with mobility im- pairments to fires or other catastrophic events, while correlating requirements for egress were kept off the books.

By comparison, accessibility in other countries was paired with concomitant regulations for “egressibility” for per- sons with disabilities. British Standard BS5588 noted in its forward: “A basic tenet of building law is that access pro- vision has to be linked to egress provi- sion...,” 6 while the Public Works Canada Barrier-Free Design Standard included comprehensive provisions for accessible means of egress as early as 1985.7

By 1989, the model code groups in the United States began to tackle the problem of egress by persons with dis- abilities from multistory buildings. The Council of American Building Officials’ Board for the Coordination of Model Codes (BCMC) developed both a stan- dard set of accessibility criteria that was soon adopted by the model building codes8 and accessible means of egress requirements that were included in both the model building codes and the NFPA Life Safety Code. The provisions

addressed in Engineering Solutions (be- low) are now found in the “Means of Egress” chapter of the International Code Council International Building Code (IBC)9 and are scheduled to be adopted by reference in the 2002 Amer- icans with Disabilities Act Accessibility Guidelines.

PERCEPTIONS VS. REALITY

Throughout the development of the model code accessible means of egress package, a shared perception drove the BCMC deliberations. However, a cursory review of the history of fire in America and its effect on persons with disabilities suggests a far different reality.

The events of September 11 and the attack on the World Trade Center have served to aggravate this difference be- tween perception and reality, and while the risks mobility-impaired peo- ple may face in high-rise office build- ings cannot be ignored, neither should the real dangers of fire and disability in the home.

In the years 1989 through 1998, 70% to 75% of civilian fire deaths and injuries occurred in residences.10 Data on the number of these suffered by persons with disabilities are scarce as neither the National Center for Health Statistics (NCHS) nor the National Fire Incident Reporting System (NFIRS) indi- cate whether a victim was disabled.11

In those reports where disability was noted, the incidence of physical disabil- ity was 7%,12 though even in these

PERCEPTION

A 30-year-old paraplegic business executive dies in a tragic conflagra- tion in a high-rise office building, trapped in the fire and smoke because she cannot use the eleva- tors or stairs to exit the building.

REALITY

A 54-year-old double-amputee with alcohol-related diabetes falls asleep while smoking in bed in his apart- ment, causing a fire that kills not only himself but three others in the building.

Continued on page 41

■ Mobility-Impaired Persons

FALL 2002 www.sfpe.org 28

cases it is difficult to determine the degree to which mobility impairment was a factor in the resulting casualty.

Nonresidential civilian deaths or injuries remain in the minority. The National Fire Data Center reports that of all nonresidential structure fires in 1998, stores and offices were involved in approximately 18% of the incidences.13

Couple this with research that indicates that persons with disabilities remain significantly unemployed or underem- ployed, even after passage of the Americans with Disabilities Act in 1990, and the exposure of people with mo- bility impairments to potentially lethal office fires remains extremely low. Conversely, most disabled adults live in apartments or houses, places where risks remain relatively high.

The juxtaposition of perception and reality remains a problem and threat to persons with disabilities. Both U.S. fed- eral regulations and the model building codes require life safety provisions for persons with disabilities in nonresiden- tial building design, yet neither ad- dresses the problem in single-family construction.

ENGINEERING SOLUTIONS

Many of the advances in fire safety in the built environment that benefit all building occupants benefit persons with disabilities to an equal or greater degree. Referred to as the “macro-ap- proach – improving building design for all building users without positive or negative discrimination,”13 requirements for compartmentalization, fire-resistive construction, and sprinkler protection make buildings safer for all occupants and provide additional safeguards for persons who cannot use exit stairs to evacuate a fire floor or building.

Special accessible means of egress provisions now included in the ICC In- ternational Building Code and NFPA Life Safety Code represent a “micro-ap- proach,” and afford mobility-impaired persons with an additional and neces- sary layer of safety.

The IBC9 states that every accessible space in new construction must be served by at least one accessible means of egress. Where two or more means of egress are required from a space, at least two accessible means of egress must be provided. Where areas of ref- uge are used as part of accessible

means of egress, travel distances for the occupancy in question also apply, which may demand more than two ac- cessible means of egress from larger spaces.

Accessible means of egress include one or all of the following components:

1.Accessible routes; 2. Stairways within exit enclosures,

48 in. (1220 mm) wide as mea- sured between handrails to accom- modate a three-person carry of someone in a wheelchair;

3. Elevators equipped with firefighter service and standby power;

4. Platform lifts in very limited appli- cations;

5. Horizontal exits; and 6. Smoke barriers. Like all means of egress, these com-

ponents when linked together must lead ultimately to a public way.

In buildings not equipped through- out with an automatic sprinkler system, areas of refuge are part of an accessible means of egress from levels above and below the level of exit discharge. Each must be immediately adjacent an acces- sible means of egress stairway or eleva- tor and be sized to accommodate at least one wheelchair for each 200 occu- pants or portion thereof. Equipped with a two-way communication system linked to a central control point, an area of refuge becomes a staging area for a person to be evacuated from or a location to “protect in place,” separated from the remainder of the story by a smoke barrier. Finally, an exterior area of rescue assistance may be provided where an accessible exit discharge is not practicable, providing a level of protection for the disabled person simi- lar to that provided by areas of refuge.

Obviously, accessible means of egress from levels above or below a level of exit discharge are significantly different than the regular means of egress used by ambulatory persons. Simply, an elevator on an accessible means of egress will often not be avail- able for independent operation by a disabled person, having gone into fire- fighter service and been recalled to a designated floor. And stairs are never negotiable for the mobility-impaired persons for which accessible means of egress are intended. Accordingly, build- ing code requirements and facility design will fail to accomplish their in- tended result of an equivalent level of

safety for persons with mobility impair- ments if procedures to use these new egress systems are not put in place and used.

BEYOND BRICKS AND MORTAR

In 1995, the United States Fire Ad- ministration funded “Emergency Proce- dures for Employees with Disabilities in Office Occupancies,”14 a guide devel- oped by the National Institute of Stan- dards and Technology and the National Task Force on Life Safety and People with Disabilities. In the fall of 2001, the group that helped to develop this docu- ment met again to develop a guide for firefighters and other first responders who will face evacuating disabled per- sons from buildings involved in fire or other emergency conditions. All partici- pants agreed a standard set of proce- dures or protocol is needed to ensure that the assets provided in our built environment by the building codes provide an equivalent level of safety for persons with disabilities.

For a number of years, the Eastern Paralyzed Veterans Association (EPVA) has suggested the following protocol for evacuation of mobility-impaired persons from new and existing build- ings.

1. Protect in place. This is common in high-rise construction, where zoned evacuation is the norm. Because re- moving a person who uses a wheel- chair demands extraordinary (literally, beyond the ordinary or norm) activity and because leaving the building is not always necessary to the safety of the disabled occupant, protecting in place should be the first option.

This is obviously a contentious pro- posal in the disability community, as well as in the profession of firefighters and other first responders. If everyone else is asked to leave a building, dis- abled people should be evacuated as well. But areas of refuge (and similar provisions in existing buildings) are provided specifically with the intent that people who cannot use stairs will remain in a building longer than their able-bodied counterparts. “Philosophi- cally, in addressing the problem of life safety for people with disabilities, the goal should be equal opportunity of life safety – not necessarily equal possibil- ity of egress.”2

■ Mobility-Impaired Persons

29 Fire Protection Engineering NUMBER 16

2. Elevator egress or evacuation. This option, while questionable to some, has been used in high-rise construction for decades as a means of evacuating people in wheelchairs from the upper stories of buildings. Understanding the risks of losing elevator operation due to loss of building power or a shutdown of the elevator due to water in the hoistway, first responders will still be in a position to determine whether an elevator car, lobby, hoistway, or machine room are either in- volved in a fire or in eminent risk from fire or smoke. They can determine when an elevator is a safe way to evacuate a building and assist mobility-impaired persons accordingly.

3. Stair evacuation. This should be the “option of last resort” and used only when a person with a disability is at im- mediate risk. Many people with disabilities have conditions such that transferring them to an evacuation device or bump- ing them down a stair in a three-person wheelchair carry could cause serious (and even fatal) injuries. For those untrained in evacuation procedures, attempting to carry some- one down a stairway could cause injuries to themselves as well. Last is a moral dilemma. Evacuating a person in a wheel- chair in an exit stairway will render that exit unusable to everyone else. Emergency personnel are thus faced with effectively removing an exit for use by all other building occu- pants while evacuating a mobility-impaired person down a stair or waiting until the building is empty before saving the

disabled person (and themselves). Clearly, a usable and safe elevator is the preferable option.

CLOSING REMARKS

Requirements for accessible means of egress in our nation’s codes are less than ten years old. As of this writing, similar requirements in the Americans with Disabilities Act require- ments have yet to be adopted. Clearly, these provisions will be revisited and modified as they are tested in real-world fire situations.

This being said, there are still a number of “frontier” issues in the field of life safety for persons with disabilities that beg further attention. Should evacuation chairs for controlled de- scent down exit stairways be required in accessible buildings? Do the events of September 11 indicate some type of accessi- ble means of egress should be required in existing buildings? Should accessible means of egress from accessible dwelling units be required?

Disability advocates, code developers, property owners, and fire protection engineers all have valuable expertise to lend to these issues. ▲

Brian Black is with the Eastern Paralyzed Veterans Association.

REFERENCES

1 ASA A117.1, 1961.

2 Pauls, J., & Juillet, E., “Influencing the Life Safety of People with Disabilities,” The Building Official and Code Administrator, July/August 1990.

3 Architectural and Transportation Barriers Compliance Board. Minimum Guidelines and Requirements for Accessible Design. Federal Register, Vol. 47. No. 150. 1982.

4 Uniform Federal Accessibility Standards. Federal Register, Vol. 49, No. 153. 1982.

5 U.S. Architectural Barriers Act of 1968, 42 U.S.C. 4151-4157.

6 British Standards Institution. British Standard BS 5588: Part 8, Fire Precautions in the Design and Construction of Buildings. Part 8, Code of Practice for Means of Escape for Disabled Persons, 1988.

7 Public Works Canada. Barrier-Free Design Standard, 1985.

8 BOCA National Building Code, Building Officials and Code Administrators. Standard Building Code, Southern Building Code Congress International. Uniform Building Code, International Conference of Building Officials.

9 International Code Council, International Building Code, 2000.

10 United States Fire Administration. A Profile of Fire in the United States, 1989-1998, 12th edition.

11 TriData Corporation. Fire Risks for the Mobility -Impaired, pro- duced for the United States Fire Administration, 1999.

12 National Fire Protection Association. NFPA Online.

13 Pauls, J., Life Safety for People with Disabilities, A State-of-the-Art Summary with an Emphasis on Codes and Standards, Public Works Canada, 1988.

14 United States Fire Administration. Emergency Procedures for Employees with Disabilities in Office Occupancies, 1995.

■ Mobility-Impaired Persons

30 Fire Protection Engineering NUMBER 16

Service Entrance Cable

Edwards Systems Technology has added speak- ers and speaker-strobes to its Genesis family of emergency signals. Extending only one inch from the wall, the UL-listed speaker-strobes are compact and have features including clear audio output and field-configurable light and speaker output settings. Both settings remain visible after installation, allowing at-a-glance verification and simple adjustment, which means contractors

have fewer parts to stock and adjustments can be made on the fly. www.est.net

—Edwards Systems Technology (EST)

The Servomex 4012 gas analyzer achieves fast response and precise oxygen, carbon dioxide, and carbon monoxide measure- ments necessary to achieve accurate calorimetry results. Several Servomex models are available. The Servomex 4100 may be fitted with up to four gas sensor transducers in a 19-in. rack-mounted instrument and can simultaneously mea- sure up to four gas streams.

www.servomex.com —Servomex Group Ltd.

Tyco Thermal Controls announces the System 1850-SE™ Service Entrance cable system, an alternative to concrete encase- ment of service entrance cables. The MI cable, manufactured from copper and mag- nesium oxide (both inorganic materials), will neither release toxic fumes nor propa-

gate flame during a fire. Installed in an enclosed steel tray, the system has gained approval for use in major cities including New York, Philadelphia, and Washington, DC.

www.tycothermal.com —Tyco Thermal Controls

UL-Listed CPVC Pipe BlazeMaster® CPVC pipe and fittings are now UL-listed for use with listed light- hazard, quick-response, extended-cover- age sprinklers and may be installed with- out protection (exposed) when subject to the following additional conditions: 1) exposed piping is to be installed below a smooth, flat, horizontal ceiling construction, and 2) when using fittings 11⁄2-in. or greater, only schedule 80 fittings may be used.

www.blazemaster.com —BlazeMaster® Fire Sprinkler Systems

New D8125INV Wireless Interface Module works with Radionics’ 9000

Series panels. It is an integrated interface module and keypad that allow users to connect Inovonics FA400 Receivers and up to 238 wireless transmitters. Each unit can support up to two FA400 Receivers. The key- pad is used to access the programming and diagnostics functions of the wireless portion of the system. The D8125INV has a self-prompting user interface and can be used to troubleshoot system RF transmitters.

www.radionicsinc.com —Radionics, Div. of Detection Systems, Inc.

Permanent, Watertight Seal The Metraflex MetraSeal is a fast, watertight, rubber seal that takes minutes to install around round pipe, including concrete, steel, plastic, cast iron, and copper, with a snug, permanent seal. No drying or curing time is needed. The seal is made of a blend of inter- connected rubber links and is available in a variety of sizes (rated to 20 psig). It is engi- neered to be a permanent, positive, mechan-

ical seal resisting aging, water, ozone, and sunlight for up to 25 years. www.metraflex.com

—The Metraflex Company

Speakers & Speaker-Strobes

The “Smart Module” for low-voltage uses not only allows for snap-in, non-hard-wiring installation, but also lets service technicians and end-users know at a glance that it is working. When the unit sacrifices itself to protect the devices it is hooked to, an indica- tor light will let the technician know that it needs to be replaced (with no tools or test equipment required). Easy to remove and replace, the module has min- imal arc circuitry.

www.ditekcorp.com —DITEK

“Smart” Surge ProtectorsWireless Interface Module

Gas Analyzers

The Adjustable Deadband Pressure Switch (ADPS) has independent set and reset points that are adjustable throughout the entire operating range. It utilizes a 15A SPDT mercury-free microswitch with a manual test actuator. Two models are available: the 300 psi model (12 psi mini- mum deadband) and the 600 psi model (25 psi minimum deadband). Complies with NFPA 20 fire pump controller codes.

www.pottersignal.com —Potter Electric Signal Co.

Adjustable Deadband Pressure Switch

Products/Literature

FALL 2002 www.sfpe.org 31

Preaction System

Anvil International has published a comprehensive catalog of pipe hangers and supports. Formerly known as Grinnell, Anvil offers a complete selec- tion of engineered hangers and supports, as well as custom design and manufacturing services. The catalog contains specs, illustrations, photos, and software information. Products highlighted include variable springs, constant supports, sway struts, hydraulic snubbers, rigid hangers, and special fabricated steel supports.

www.anvilintl.com —Anvil International

The new OD850 Outdoor TriTech PIR/Microwave Intrusion Detector features patented Motion Analyzer II (MAP II) sig- nal processing. It also employs a new tech- nology, Linear Travel Distance (LTD), which allows it to measure the linear dis- tance traveled by a target, and eliminates false alarms from objects such as tree branches and hanging signs. It includes

two user-selectable sensitivity levels, a timed relay output (adjustable from two seconds to 10 min.), and a user-selectable Day/Night mode.

www.dsworld.com —Detection Systems, Inc.

The SureFire™ noncycling preaction system has built-in features that ensure protection while preventing water from entering the system. This allows it to operate as a regular preaction system under normal conditions or as a dry system if primary and backup power fail. Available with both single and double interlock electric release in 11⁄2 - to 8-in. sizes.

www.vikingcorp.com —The Viking Corp.

Flanged Circuit-Balancing Valves New 10- and 12-in. flanged Circuit- Balancing Valves are available in both angled and straight configurations. They feature pressure-differential readout ports on both sides of the valve body to allow for easier installation, a tamper-proof memory for ease of return to the balance setpoint, and a multi-turn adjustment range for maximum controllability. The

new design also provides a maximum working pressure of 375 psi and has a body made of ductile iron.

www.armstrongpumps.com —Armstrong Pumps Ltd.

Pipe Hanger Catalog

Acclimate Plus™ is a new multicriteria tech- nology smoke detector with the enhanced capabilities of NOTIFIER’s Flash-Scan™ technology. Acclimate provides an opti- mum balance of stability and early warning to maximize fire detection while rejecting unwanted nuisance alarms, and responds to a broad range of fires at a fast speed. Flash-Scan protocol groups and scans 10 devices at a time, and constantly monitors so it can report an alarm within one second and turn on all outputs within two seconds.

www.notifier.com —NOTIFIER

High-Speed Detectors

Intrusion Detector

The TFD-75 heavy-duty fire damper is a flanged-framed, airfoil-bladed damper that has been tested to British Standard BS476, Part 20 (stays closed in extreme heat), and operability requirements of NFPA 130 (remains fully func- tional after being exposed to a temperature of 482°F for one hour). They may be installed in high fire potential areas, industrial applica- tions, and as isolation dampers to prevent fire from spreading. Standard version has an 8-in. deep channel frame; other versions are available.

www.american-warming.com —AWV

Heavy-Duty Fire Damper

FALL 2002 www.sfpe.org 32

Fire Protection Engineering in Building Design, Jane Lataille, P.E., FSFPE $40.00 members $49.99 nonmembers

Information on mitigating potential fire damage through proper

design and construction of buildings, industrial processes and utility

systems is included. Specific examples are given as to how fire pro-

tection engineering can be integrated with mechanical, electrical,

structural, and chemical engineering. The book also includes a sec-

tion on writing fire protection specifications as well as a comprehen-

sive reference list. Published by Elsevier Science.

Principles of Smoke Management, John H. Klote, P.E., FSFPE and James A. Milke, P.E., FSFPE $79.00 members $99.00 nonmembers

Principles of Smoke Management is the quintessential reference for

those involved in smoke control design. It contains an exhaustive

treatment of smoke management, including pressurized stairwells,

pressurized elevators, zoned smoke control management in atria and

other large spaces. It includes numerous example calculations and a

CD of computer software for analysis of smoke management systems.

Published by ASHRAE.

ISO Fire Suppression Rating Schedule Handbook, 2nd Edition, H.E. Hickey, Ph.D. $109.00 members $129.00 nonmembers

The first edition, issued in 1993, was well received as a necessary re-

source by thousands of Fire Chiefs, City Managers, and others inter-

ested in public fire protection. This new edition has six new chapters,

as well as extensive revisions to make the Handbook even more

helpful. Fire protection is the only municipal service where improved

capability can result in financial savings to taxpayers through reduced

property insurance premiums. This book will explain the process and

help the reader understand and prepare for the grading evaluation.

Published by Chicago Spectrum Press.

PE Exam Video Series $495.00 members $695.00 nonmembers

Worcester Polytechnic Institute (WPI) and the Society of Fire Protec-

tion Engineers (SFPE) jointly announce the availability of a new series

of VHS videotapes produced specifically to assist professional engi-

neers in preparation for the PE exam in Fire Protection Engineering.

This is a series of 4 VHS format videotapes with a total of 9 hours of

instruction. The tape series covers all technical categories on the PE

exam in Fire Protection Engineering. Included in the tapes are the fol-

lowing topics: Water Supplies; Building Systems; Water-based Sup-

pression Systems; Non Water-based Suppression Systems; Detection

and Alarm Systems; Fire Prevention; and Hazard and Risk Analysis.

Resources SFPE Fire Protection Engineering Handbook, 3rd Edition $202.50 members $225.00 nonmembers

Work with the essential desk reference for all engineers and engineer-

ing students, architects, and system designers involved in fire protec-

tion. More than 10 all-new chapters, including: Flammable Liquid

Spill Fires; Water Mist Systems; Introduction to Fire Risk Analysis; Un-

certainty and Safety Factors... and more. That’s why an up-to-date

copy of this trusted text should be the cornerstone of your technical

library.

SFPE Engineering Guide to Piloted Ignition of Solid Materials Under Radiant Exposure $35.00 members $50.00 nonmembers

The focus of the guide is the techniques and data available to engi-

neers for predicting the time to piloted ignition of solids exposed to

flame radiation and the determination of the safe separation distances

required to prevent ignition. The guide reviews the concept of mini-

mum ignition level and reviews five methods to calculate the time to

ignition under constant radiative heat flux. The guide includes sample

results for each method and an Appendix of relevant material proper-

ties.

ICC Performance Code for Buildings and Facilities, International Code Council, Inc. $31.45 members $34.95 nonmembers

This long-awaited code and commentary is the first broad-based,

stand-alone performance code in the nation. It clearly defines the

objectives for achieving the intended outcomes regarding occupant

safety, property protection and community welfare. It provides a

framework to achieve the defined objectives in terms of tolerable

levels of damage and magnitudes of design events, such as fire and

natural hazards. Distinctly different from a prescriptive code, it allows

the user to systematically achieve various solutions. Published by

International Code Council, Inc.

Industrial Fire Protection Handbook, 2nd Edition, R. Craig Schroll $72.00 members $89.95 nonmembers

The second edition of this publication offers practical techniques and a

performance-oriented approach to effective fire safety. Promotes un-

derstanding of the issues involved with fire loss prevention and con-

trol and the applications of proven techniques to a work environment.

It establishes best practices that often go beyond basic regulatory and

code compliance. And it includes concrete examples from real work

environments. It also includes discussions of new extinguishing

agents, including wet chemical and clean agents designed to replace

halon. Published by CRC Press LLC.

SFPE New2002Publications

33 Fire Protection Engineering NUMBER 16

Resources U P C O M I N G E V E N T S January 27-28, 2003 8th International Fire and Materials Conference Info: www.intercomm.dial.pipex.com

February 25-27, 2003 Workshop on Fire Suppression Technologies Mobile, AL Info: [email protected]

February 26-18, 2003 Fire Asia 2003 Hong Kong Info: [email protected]

April 6-8, 2003 Taipei International Exhibition on Fire, Safety, &

Disaster Management Taipei, Taiwan Info: www.secutech.com

May 8-10, 2003 Strategies for Performance in the Aftermath of the World Trade Center Kuala Lampur, Malasia Info: www.cibklutm.com

May 18-22, 2003 NFPA World Safety Conference and Exposition Dallas, TX Info: www.nfpa.org

June 8-13, 2003 Third Mediterranean Combustion Symposium Marrakech, Morocco Info: www.combustioninstitute.it

June 22-25, 2003 13th World Conference on Disaster Management Toronto, Canada Info: www.wcdm.org

August 20-22, 2003 2nd International Conference in Pedestrian and Evacuation

Dynamics (PED) Greenwich, London Info: http://fseg.gre.ac.uk/ped2003/

September 8-12, 2003 4th International Seminar on Fire and Explosion Hazards Northern Ireland, UK Info: www.engj.ulst.ac.uk/4thisfeh/

2002 Price Changes for Publications: (Due to Publisher Price Changes)

Enclosure Fire Dynamics, Bjorn Karlsson & James G. Quintiere $76.50 members $89.95 nonmembers

Introduction to Mathematical Fire Modeling, Marc L. Janssens, Ph.D. $128.00 members $160.00 nonmembers

SFPE Engineering Guide to Performance-Based Fire Protection Analysis and Design of Buildings,

$53.00 members $59.00 nonmembers

Structural Design for Fire Safety, Andrew Buchanan $61.00 members $75.00 nonmembers

34 Fire Protection Engineering NUMBER 16

• Advanced Fire Technologies...................Page 19 • Ansul Incorporated..................................Page 15 • BlazeMaster® Fire Sprinkler Systems .....Page 21 • Chemetron Fire Systems............................Page 2 • Chemguard.................................................Page 5 • Decoshield Systems, Inc..........................Page 19 • Detection Systems, Inc. ...........................Page 20 • Edwards Systems Technology..........Pages 26-27 • Fire Control Instruments .........................Page 46 • Gast Manufacturing, Inc. .........................Page 48 • Grice Engineering....................................Page 18 • Kidde Fire Systems ..................................Page 22 • Koffel Associates, Inc. .............................Page 39 • NOTIFIER Fire Systems...........Inside Back Cover

• OCV Control Valves.................................Page 30 • Potter Electric Signal Company ..............Page 36 • The Protectowire Co., Inc. ......................Page 12 • The RJA Group .......................Inside Front Cover • Reliable Automatic Sprinkler...................Page 40 • Ruskin.......................................................Page 42 • Siemens Building Technologies, Inc.

Fire Safety Division................................Page 32 • SimplexGrinnell .......................................Page 51 • System Sensor ..........................................Page 49 • Tyco Fire Products .............................Back Cover • The Viking Corporation ..........................Page 11 • Wheelock, Inc..........................................Page 43 • Xerxes Corporation..................................Page 31

Index of Advertisers

Solution to last issue’s brainteaser

Find the values of A, B, and C in the following arithmetic sequence:

AB4, B03, B3C, BA1

A = 6, B = 7, and C = 2. The difference between subsequent terms is 29.

You have six weights. One pair is red, one pair is white, and one pair is blue. In each pair, one weight is slightly heavier than the other, but otherwise looks exactly like its mate. The three heavier weights all weigh the same, as do the three lighter weights.

How can you identify the heavier weights in only two separate weighings on a balance scale?

Developed by Paul Curry, provided by Jane Lataille, P.E.

B R A I N T E A S E R CORPORATE 100

The SFPE Corporate 100 Program was founded in 1976 to strengthen the relationship between industry and the fire protection engineering community. Membership in the program recognizes those who support the objec- tives of SFPE and have a genuine concern for the safety of life and property from fire.

BENEFACTORS Rolf Jensen & Associates, Inc.

PATRONS Code Consultants, Inc. Edwards Systems Technology Hughes Associates, Inc. The Reliable Automatic Sprinkler Company Schirmer Engineering Corporation SimplexGrinnell

DONORS Gage-Babcock & Associates, Inc. National Fire Protection Association Underwriters Laboratories, Inc.

MEMBERS Altronix Corporation Arup Fire Automatic Fire Alarm Association BFPE International Cybor Fire Protection Company FM Global Corporation Fire Suppression Systems Association GE Global Asset Protection Services Harrington Group, Inc. HSB Professional Loss Control Hubbell Industrial Controls James W. Nolan Company (Emeritus) Koffel Associates, Inc. Marsh Risk Consulting National Electrical Manufacturers Association National Fire Sprinkler Association Nuclear Energy Institute The Protectowire Co., Inc. Reliable Fire Equipment Company Risk Technologies LLC TVA Fire and Lifesafety, Inc. Tyco Services, Pty Wheelock, Inc. W.R. Grace Company

SMALL BUSINESS MEMBERS Bourgeois & Associates, Inc. Demers Associates, Inc. Fire Consulting Associates, Inc. Grainger Consulting, Inc. J.M. Cholin Consultants, Inc. MountainStar Enterprises Poole Fire Protection Engineering, Inc. Risk Logic, Inc. S.S. Dannaway & Associates, Inc. The Code Consortium, Inc. Van Rickley & Associates

35 Fire Protection Engineering NUMBER 16

T he purpose of licensure as a profes- sional engineer is to demonstrate that an engineer has at least the

minimum level of competence necessary to protect public health, safety, and wel- fare in their area of practice. Licensure as a professional engineer typically entails a combination of education, experience, successful completion of the Fundamentals of Engineering, or “EIT,” exam, and a Principles and Practices, or “PE,” exam.

Recognizing that engineers employed in different job settings may perform different types of tasks, PE exams are intended to cover all aspects of engineering that are crit- ical to protecting public health, safety, and welfare within the engineering discipline in which an engineer wishes to become li- censed. Periodically, the specification (which describes what types of problems should be asked on the exam) of each PE exam is reviewed to ensure that the exam accurately reflects the tasks performed by engineers and the knowledge needed to conduct those tasks.

The Society of Fire Protection Engineers recently went through a process where the

specification of the fire protection engineer- ing PE exam was reviewed. This process began in the spring of 2001, when a commit- tee was assembled to develop a survey instrument called a Professional Activities and Knowledge or “PAK,” questionnaire. The committee that developed the PAK question- naire consisted of licensed engineers with a diverse distribution of age, gender, geo- graphic location, practice size, years of education, and years of practice. The PAK questionnaire that the committee developed was intended to cover all the important tasks that fire protection engineers perform, sepa- rated into practice areas called “domains,” and the knowledge that fire protection engineers need to perform those tasks.

Once the PAK questionnaire was com- pleted, it was mailed to everyone who is listed in the SFPE database who has indi- cated that they hold registration in the U.S. as a professional engineer. Respondents were asked to rate each of the tasks and knowledge areas in terms of their impor- tance in the protection of health, safety, and public welfare. Survey respondents were also asked to provide a recommendation of the weighting of the test content from each of the domains and demographic back- ground information. A detailed statistical

analysis was performed on the responses to the PAK questionnaire.

Another carefully balanced committee was formed to review the results of the sta- tistical analysis, and to decide what should be tested on the FPE PE exam and what per- centage of the total number of problems on the exam should come from each domain. This new exam specification, which will take effect with the October 2004 FPE PE exam, is presented in Table 1.

In some ways, this new specification is very similar to the specification that is the basis for the current FPE PE exam (avail- able from http://www.ncees.org/profesional/ pp_fire). However, there are some notable differences. The percentage of the questions on the exam that will cover building systems has approximately doubled, while the percentage of questions on water-based suppression systems has been reduced by approximately 50%.

However, and more remarkably, the per- centage of the exam that is dedicated to haz- ard and risk analysis and the fundamental principles that support hazard and risk analysis (fire science and human behavior) has approximately tripled. This change speaks volumes about the maturation of fire protection engineering as a discipline.

New Specification for the Fire Protection Engineering PE Exam

Morgan J. Hurley, P.E. Technical Director Society of Fire Protection Engineers

from the technical director

Domain Subdomain Content Weight ContentWeight Knowledge Areas

20% I. FIRE PROTECTION ANALYSIS 12% A. Types of Analysis 8% B. Information Sources for Analysis

10% II. FIRE PROTECTION MANAGEMENT

15% III. FIRE SCIENCE & HUMAN BEHAVIOR 10% A. Fire Dynamics 5% B. Human Response

35% IV. FIRE PROTECTION SYSTEMS 12% A. Water-Based Fire Suppression Systems 5% B. Special Hazard Systems 9% C. Fire Detection and Alarm Systems 5% D. Smoke Management Systems 4% E. Explosion Protection and Prevention Systems

20% V. PASSIVE BUILDING SYSTEMS 12% A. Building Construction 8% B. Means of Egress

Table 1. Specification for the FPE PE Exam, which will take effect in 2004