International Experience in the Development and

Use of Performance-Based Fire Safety Design

Methods: Evolution, Current Situation and Thoughts
for the Future

BRIAN J. MEACHAM
Society of Fire Protection Engineers
73 15 Wisconsin Avenue, Suite 1225W
Bethesda, MD 20814 USA

ABSTRACT

Over the past fifty years, fire safety engineering* has grown and developed into an accepted.
if not fully mature, engineering discipline. This has been possible for a variety of reasons,
including an ever-increasing understanding of fire safety science, continuing development
of analytical methods for engineering analysis and design, technological advances in
computational tools, and the global movement towards performance-based building and fire
regulations. This paper reviews the current state of fire safety engineering by looking at
international experiences in the development and use of performance-based fire safety
design methods. It discusses the impact of increasing scientific knowledge and the
evolution of performance-based fire safety design methods and regulations, and it speculates
on what will be required for fire safety engineering to reach maturity as an engineering
discipline.

KEY WORDS: Performance-based fire safety design, fire safety engineering, performance-
based building regulations

INTRODUCTION
"Performance-based" has been the epithet of the 1990s throughout the fire safety
community. It seems that one cannot open a fire-related journal, set of proceedings, or other
publication without finding reference to performance-based codes, performance-based fire

As used in this paper, the term "fire safety engineering" is synonymous with the terms "fire engineering" and
"fire protection engineering," and is defined as the application of science and engineering principles to protect
people and their environment from fire.

FIRE SAFETY SCiENCE-PROCEEDINGS OF THE SIXTH INTERNATIONAL SYMPOSIUM, pp 59-76

Copyright © International Association for Fire Safety Science

safety design, or some derivative thereof. Why is this so? What has driven the fire safety
community to search for means and methods to measure, analyze. describe, and predict
performance? In many respects, the movement to performance-based analysis and design is
part of the natural evolution of an engineering discipline. By one definition. engineering is
the art or science of making practical application of the knowledge of pure science [I]. As
the knowledge base grows, the horizon for practical application of the knowledge increases
accordingly. When this knowledge base includes understanding how materials, ,'systems.
and buildings perform under a variety of conditions. it opens the door to performa<ce-based
analysis and design. This can be seen by reviewing the development of an allied profession.

THE IMPACT OF INCREASING SCIENTIFIC KNOWLEDGE ON THE

EVOLUTION OF PERFORMANCE-BASED STRUCTURAL ENGINEERING

Consider structural engineering. For centuries, most structures were built of wood,
masonry, concrete and other natural materials because these materials were readily available
and people had some knowledge about their properties and how to build with them.
However, a comprehensive technical understanding of material properties and forces acting
on structures did not exist. As a resuit, although these materials were used to build some
rather large structures, such as the Parthenon and the pyramids of Egypt, the size of
structural elements and the height of the structures were restricted. Lack of technical
knowledge and modern materials prevented builders in earlier times from erecting structures
common today, such as high-rise office buildings [2]. In short, the analytical tools were
lacking to understand and predict the performance of buildings, as is routine today.
However, beginning with the works of Newton and Hooke in the late 1600s, and continuing
with the work of Cauchy, Euler, and Young in the early 1800s, a scientific knowledge base
on properties of materials and their reaction to forces was forming [2]. By the mid-to-late
1800s, and continuing into the early 1900s, such advances in scientific understanding, and
their conversion to practical applications, led to the realization that other nlaterials could be
used, in some cases more effectively, in the building of structures. Examples from the time
period include the introduction of steel frame and reinforced concrete construction;
materials still widely used today.
The knowledge base that led to the widespread use of these materials accumulated over
several years, and was fed by the efforts of researchers in a variety of disciplines.
Furthermore, the research did not stop when new applications were realized at the end of the
19th century, but rather, became more focused at addressing unknowns and uncertainties in
the technology and the knowledge base. In the late 1800s and early 1900s, when new
structural steel applications were being tried. using the new understanding of forces and
strengths of materials, there was the realization that uncertainty needed to be addressed. For
the construction of important structures such as bridges, British and American engineers
began performing strength calculations. These included calculating the "highest possible
tensile" stresses in the structure, using the methods of the time, and ensuring these stresses
were less than the "official tensile strength" of the material. Knowing that they did not have
a complete understanding of highest possible stress or strength, the engineers made the
highest calculated working stress significantly less, as much as seven or eight times less,
than the strength of the material as determined by breaking a simple, smooth. parallel-
stemmed test piece [2]. This was known as "applying a factor of safety." and at the time.
any attempt to save weight and cost by reducing the factor of safety was considered a recipe
for disaster.
This approach of applying factors of safety was carried over into building codes, and was
widely applied to structural components, such as beams and columns. In the early 1900s.
such factors of safety were often in the range of three to four. In essence, these values were
used because they were known to work: no analytical methods were applied to derive them.
As more knowledge about materials was gained, and better analytical tools were developed.
the structural engineering community began decreasing the factors of safety specified in
design standards and codes of practice. This is evident in the Load and Resistance Factor
Design (LRFD) approach widely used today [4.5,6].

THE IMPACT OF INCREASING SCIENTIFIC KNOWLEDGE ON THE

EVOLUTION OF PERFORMANCE-BASED FIRE SAFETY ENGINEERING

Fire safety has been a concern for as long as people have been building and inhabiting
structures. This is especially true in urban areas where the threat of conflagration is highest,
as evidenced by regulations to minimize the likelihood of conflagration enacted following
the fire of London in 1666. However, the methods used to mitigate unwanted fires changed
little from that time until the early part of this century. Much like the structural engineering
situation, a comprehensive technical understanding of fire was lacking, and people generally
relied on experience as a guide.
By the middle-to-late 1800s, however, advances in scientific knowledge and the industrial
revolution created new challenges for fire safe design. New materials were being developed
whose burning characteristics were not well known; new mechanical technologies, such as
steam and internal combustion engines, presented explosion and fire hazards, and new
structural technology enabled larger and taller buildings to be constructed. In addition.
more and more people began moving from the countryside into growing cities. As a result
of these factors, large life-loss and large monetary-loss fires became a characteristic of the
early 20"' century. The magnitude of the problem for the United States is well represented
by data that show the ten deadliest fires occurred between 1865 and 1947, with six of the ten
largest fire losses (in millions of dollars) occurring during the same period [6].
The reaction to the fire problem in the early part of this century was not, as in structural
engineering, primarily motivated by increasing technical knowledge,' but was motivated by
societal concern, resulting in controls imposed through regulation; and by economic losses.
prompting research by industry and the insurance community. Large life-loss fires in the
early part of the century prompted the development of building and fire codes and standards.
including the first edition of NFPA 13, Installation of Sprinkler Systems [ 7 ] , in 1896. the
establishment of the NFPA Committee on Safety to Life in 1913 [8]. and the first edition of
the Uniform Building Code in 1927 [9]. Unlike the changes ongoing in the structural
community, however, the provisions being enacted in the fire community were primarily
prescriptive in nature, identifying what, where, and how some system or feature was to be
installed, with little emphasis on how the building should perform overall

' Of course, the desire to build larger buildings was economically motivated as well
Nonetheless, the motivation to develop and enact building fire safety codcs and standards
for safety to life. and to develop strategies and technologies to minimize contents and
property damage due to fire. began a period of scientific investigation of fire and fire
properties of materials. An indication of this can be seen in a submittal letter for a U.S.
Department of Commerce Building Code Committee report on fire resistance dated August
21, 1930 (sic) [lo]:
In searching for test data to serve as a basis for conclusions the committee has been
impressed by the fact that the science of fire protection is in a stage of development
as compared with the knowledge of other properties of materials and co~lstructions.
The program of tests under way and contemplated by the Bureau of Standards has
advanced this science very greatly.
In the years that have followed. fire safety science research around the world has provided
the fire safety engineering profession with an ever-increasing understanding of fire
phenomena, of the fire performance of materials, and of the impact of fire on people,
property, business continuity. and the environment. Building on the seminal works of such
pioneers in fire research as Kawagoe, Thomas, Emmons. and Rasbash, great strides were
made in gaining fundamental scientific knowledge of fire safety science from the 1950s
through the 1980s. One needs only review the past proceedings of the symposia of the
International Association for Fire Safety Science, the SFPE Handbook of Fire Protection
Engineering, or such technical journals as Fire Safety Joztrnul to see the evidence of this.
As research and loss experience increased the technical understanding of fire and its
impacts, it became apparent that fire and fire impacts could be estimated or predicted. This
realization provided an impetus to the fire safety engineering community to look at methods
to predict fire and its impacts instead of simply designing fire mitigation measures as they
had in the past. To make this happen, the fire safety engineering community relied on the
research community, which had collected data; developed correlations, analytical methods,
and models to describe fire initiation. growth, and spread; and developed analytical methods
and models to predict possible fire impact. Although the fire safety engineering community
welcomed the input from the research community, many found the available correlations,
analytical methods, and models time-consuming to work with and not generally accepted by
the building and fire authorities.
More recently, however, advances in computer technology led to the development of
computer-based analysis and design tools for use by fire safety engineers. This resulted in a
change to how fire safety problems were approached, as the once time-consuming analytical
methods became faster and easier to use. As more of the computer-based analytical
methods were used, some in the fire safety engineering community accepted them as tools
for daily use, and began to advocate fire safety analysis and design using the computer-
based analytical methods instead of the traditional codes- and standards-based design
approach. In some respects, the increasing use and general acceptance of computer-based
analytical methods within the fire safety engineering community has contributed to the
movement to performance-based building codes and the concept of performance-based fire
safety design [ l l ] . As such, some people view "computer fire models" as synonymous with
"performance-based design" (although the author would argue against such an interchange
of terms), and some consider fire safety engineering to now be at a state of development
parallel to structural engineering (the author would disagree with this as well).
However, there remain some critical differences in the states of development between fire
safety engineering and structural engineering. Unlike structural engineering, fire safety
engineering still lacks a common framework for evaluation and widely accepted analytical
methods with known limits of uncertainty. Although significant effort has been aimed at
developing a common framework over the past ten years, it does not yet exist in a
comprehensive and integrated form. Such a framework is close at hand, however. and the
next section will overview historical developments to date in this area. Of greater concern
is the lack of understanding regarding the limits of uncertainty of the analytical methods,
and more importantly, of the broader system within which one applies these methods.

THE EVOLUTION OF A FRAMEWORK FOR PERFORMANCE-BASED FIRE

SAFETY ANALYSIS AND DESIGN AND OF PERFORMANCE-BASED CODES

Much has been written about the evolution of performance-based fire safety design methods
and codes [11,12], so an extensive treatment will not be provided here. Rather, only a brief
overview of international activities in the development of performance-based fire safety
design methods and codes will be provided, with detailed discussion offered on some key
events of the past twenty-five years.
Developments toward a Framework for Fire Safety Design
Svstems and Risk-Based A~vroaches
The 1970s saw the beginnings of a dramatic shift in thinking from the traditional "complies
with the codeldoes not comply with the code" approach to a "systen~s" approach for
evaluating and designing building fire safety measures [I I]. In 1972, the U.S. General
Services Administration and the U.S. National Bureau of Standards jointly developed an
event logic diagram that showed alternative approaches to achieving building fire safety.
.4fter several revisions, this tree eventually became the basic reference guide of the GSA's
goal-oriented systems approach to building fire safety. This document, comrnonly referred
to simply as Appendix D, became a fundamental document for describing a risk-informed
systems approach to building fire safety design [13]. Major features include:
0 A concept of relative risk (the absence of risk is not feasible).
* Management goals as described in the context of acceptable levels of risk.
* Workable components of a fire safety system that can be adapted to any building.
An event logic tree expressing relationships among the different system components.
0 A method of calculation enabling the performance of alternative fire safety systems to
be compared.
e The use of probability to describe fire safety performance.
Following the publication of Appendix D, activities relating to risk-informed systems
approaches to building fire safety expanded considerably. One direct result was the
formation of the National Fire Protection Association Technical Committee on Systems
Concepts for Fire Protection. This committee's first action was to publish an event logic
tree related to fire safety in 1973, which although modified over time, can still be found in
NFPA 550, Guide to the Fire Safety Concepts Tree 1141.
Another result from the GSA Appendix D work is the Fire Safety Evaluation System (FSES)
used in NFPA 101A, Guide on Alternative Approaches to Lifi Safety (referred to hereafter
as 101A) [15]. The FSES is a schedule approach to determining eqtlivalencies to the NFI'A
101 Liji Safety Code for certain institutional occupancies [16]." It consists of a variety of
fire safety parameters (such as construction, hazardous areas, manual fire alarms, automatic
detection systems, and automatic sprinkler systems) for which designated point values have
been developed. In some cases. such as for health care facilities, there are also risk
parameters for patient mobility, patient density. fire zone location. ratio of patients to
attendants. and average patient age. It should be noted that the point values for the fire
safety parameters and for the risk parameters were not based 011scientific or statistical data.
but were developed from the "experienced judgment of a group of fire safety professionals
and represent the opinions of that panel of experts" [16]. For this reason. one should not
transfer the point scheme to determine equivalency with other codes, or assume that the
point scheme is valid outside of North America.
The development of the Bziilding Firesafety Evaluation Method (BFSEM) also followed the
publication of Appendix D. With its beginnings as a course taught by Rexford Wilson and
Robert Fitzgerald [17], the BFSEM has grown into a systems approach to building fire
safety analysis [I 8,191. It uses a structured framework to guide the process of performance
evaluation. Using network diagrams, the user evaluates such factors as ignition potential,
fire growth potential within the compartment of origin, barrier performance, fire spread
beyond the compartment of origin, automatic and manual suppression, smoke spread, and
occupant safety. A major attribute of the method is that one can compare different designs
on a common basis. Fire related factors, like fuel load and arrangement, and fire protection
features, like automatic and manual fire detection and suppression (including fire
department response), integrity of barriers, and operation of emergency systems, are
evaluated as to how the fire will develop and spread. This is accomplished through a
combination of the framework, engineering judgment. and deterministic calculation
methods. The way that the framework has been developed, the analysis can be done using
one's knowledge and experience alone (e.g., an audit), with the additional use of analytical
methods, and with the use of subjective or objective probabilities (from engineering
judgment or statistical data, respectively).
Another key effort that began in the 1970s was the development of a risk assessment model
by Beck [20,21,22]. The intent was to develop a system model that could identify cost-
effective building fire safety design solutions that achieved an acceptable level of occupant
fire safety. This required integration of a risk assessment model with various sub-models
that addressed fire, human response, and the like. The risk assessment system model (the
top level) is founded on an event-based modeling approach wherein events are characterized
in terms of discreet times and probabilities of occurrence. The risk assessment model is
used to characterize the outcome of a fire growth and spread scenario in terms of times to
reach untenable conditions. The model assesses the fire safety performance of a specific fire
safety design in terms of two decision-making parameters: The Expected Risk to Lifi (ERL)
and the Fire Cost Expectalion (FCE). These terms are defined in various publications on the
model and its computerized form (CESARE-Risk) [22,23,24]. To calculate the expected
risk to life and the fire cost expectation values, the computerized model, CESARE-Risk,
considers interaction between fire growth. fire spread, smoke movement, human behavior.
the response of building systems and the response of a fire brigade. The model uses
specified design fires to characterize the broad spectrum of fires that could be expected in
reality. As Beck has noted, due to the complexity and lack of sufficient understanding of
fire phenomena and human behavior, certain conservative assumptions in approximations
have been made in the mathematical modeling [23]. As a result, the predictions made by
the model should only be considered as approximate. and should not be used for absolute
assessments, life risks or protection costs. For comparative or relative assessments,
however, such as comparing a proposed design to a code-conforming design, the model can
be considered much more reliable and effective.
A similar computerized fire risk assessment model, FIRECAM, is currently being
developed and evaluated in Canada. FIRECAM is based on the concepts developed by Beck
(as outlined above), and although some differences exist in the sub-models, they are
essentially the same will not be discussed here [24].
A still different type of approach to risk-informed fire safety analysis and design developed
within the structural fire safety arena [25,26]. Using probability- and reliability-based design
concepts developed in the structural engineering community, these approaches began by
considering primarily the structural material, the likely fuel loading, and the likely time-
temperature output of a fire, with explicit safety and reliability factors to account for
uncertainties. Over time, the concepts that emerged from structural and other engineering
disciplines have been applied to the entire building fire safety problem [27,28,29]. As
discussed in a key publication on fire risk and uncertainty [27], fire safety engineering can
benefit greatly not only from advances in structural reliability engineering, but from
quantitative risk assessment (QRA) techniques used in chemical engineering and
probabilistic risk assessment (PRA) techniques used in nuclear engineering as well. A
central point is that risk-based design is useful in accounting for uncertainty and unknowns
associated with data and with models: areas that fire safety engineering need to address.
Guidance (Framework) Documents
Toward the end of the 1980s and into the 1990s, work began on the development of
guidelines for fire safely engineering that described the overall process and indicated how
risk-based and deterministic-based methods could be used in addressing building fire safety
concerns. Whereas many of the approaches described above focused on evaluation of
building fire safety, or evaluation of risk, these documents intended to provide design
guidance. The framework for many of these guidelines has its origins in outcomes of the
Australian Warren Centre Fire Safety and Engineering project 1301 and of the Draft
~\~ationulBuilding Fire Safety Systems Code [31]. The Warren Centre project brought
together some 70 project fellows from Australia's building, fire and research comlnunities
to discuss the need to base fire safety design on engineering technology. Two key outcomes
were that risk assessment models should be used as a basis for identifying cost-effective
combinations for fire-safety sub-systems for building design, and designers should adopt
appropriate fire safety engineering techniques for the design of fire safety systems in
buildings. Using these outcomes as a base, the Building Regulation Review Task Force
(BRRTF), which was established in 1989, developed the first draft of a performance-
oriented building code entitled the National Building Fire Safety System Code (NBFSSC)
[31]. A key feature of the Code was the use of sub-systems describing different aspects of
the fire safety problem. (It is interesting to note that these sub-systems correspond nicely
with those used by Fitzgerald.) Efforts to adopt and use the NBFSSC did not progress
immediately, and it was not until the Fire Code Reform Centre (FCRC) was established in
1994 that work in the area of fire safety design picked up again in Australia. However.
given the past work, it took little time to develop and publish one of the first fire
engineering design guides currently in use today [32] (the New Zealand Fire Engineering
Design Guide [33] preceded the Australian guidelines by about two years).
Although efforts to expand upon the NBFSSC slowed in Australia, groups developing fire
safety engineering guides in the C'K and within the International Organization for
Standardization (ISO) adopted the concepts of the NBFSSC. In the early 1990s. IJK fire
safety engineering professionals set out to develop a code of practice that would promote
the use of fire safety engineering principles in building design [34]. Over the next several
years, a framework was developed that described the interactions of various fire safety sub-
systems: fire growth and suppression. smoke spread, fire spread, detection, intervention, and
evacuation. The result is the British Standards Institute (BSI) document DD240. Fire Safety
Engineering in Buildings [35]. Although the intent was to develop a standard on fire safety
engineering, the document was published as a draft for development, as concerns were
raised over the current state of knowledge in fire safety engineering. The same intent and
same result occurred within I S 0 TC92 SC4, a group whose aim was to develop an
international standard for fire safety engineering. As with BSI DD240, however. it
appeared that the state of knowledge was such that a standard could not yet be produced.
Instead, I S 0 TR 13387, The Applicaiior? of' Fire Performunee Concepts to Design
Objectives [36], will be published as a technical report. and not a standard.
After reviewing the situation world-wide [11,37], the Society of Fire Protection Engineers
(SFPE) has also developed a guide to performance-based fire safety analysis and design
(one that resembles the Australian, British, ISO, and New Zealand guides) [38]. The fact
that the SFPE guide resembles the other approaches indicates that the fire safety engineering
community is close to having a common framework for fire safety analysis and design. The
fact that there is not yet a single guide, however, indicates that shortcomings in still exist.
This will be explored in some detail the final major section of this paper.
Development of Performance-Based Building Regulations
Although systems approaches to fire safety design were being developed in a number of
countries by the mid-to-late 1980s, they were not being widely applied. This was due in
part to the developmental stage of many of the approaches, as well as to the restrictions
imposed by the building regulatory systems in use at the time. However, changes began
occurring in the building regulatory arena in the late 1970s and early 1980s that helped
promote the acceptance of, and give order to, the use of systems approaches to fire safety
engineering. Changes in the building regulatory arena occurred for many reasons, including
broad government deregulation and a desire to promote better exchange of technology and
trade across national boundaries. To address these issues, many countries began
transitioning to a performance-based building regulatory system and away from the
traditional prescriptive-based system. The basis for many of the new systems was the five-
level structure developed by the Nordic Committee on Building Regulations (NKB) [39,40].
The NKB structure is as follows:
Level 1: Goals - essential interests of the community at large (society) with regard to
the built environment.
Level 2: Functional Requirements - qualitative requirements of buildings or specific
building elements.
Level 3: Operative Requirements - actual (quantitative) r e q ~ ~ i r c n ~ c nin~ s(CI-111s
. 01
performance criteria or expanded functional descriptions.
Level 4: Verification - instructions or guidelines for verification of compliance.
Level 5: Examples of Acceptable Solutions - supplements to the regulations with
examples of solutions deemed to satisfy the requirements.
Perhaps most important to the fire safety engineering community are Levels 3 and 4:
Operative (Performance) Requirements and Verification. If a building or a building element
can be described in terms of how it performs, normally and under fire-induced load, and
methods are available to verify this performance, a design solution that complies with the
regulation can be engineered. This regulatory structure, and these realizations, provided
additional motivation to explore how materials, systems, and people responded (performed)
in various fire situations (scenarios). and to look at the concept of overall fire safety
performance of buildings promoted by Nelson, Fitzgerald, Beck and others. The regulatory
structure also provided a framework within which fire safety engineering could be applied.
Although the NKB structure served as a common model for building regulatory
development, there was no common model for defining the performance of materials.
systems, buildings, or people when subjected to fire effects. Fire tests continued to be more
focused on the prescriptive-based regulatory system, and efforts with the International
Organization for Standardization (ISO) to harmonize fire test methods were proceeding at a
very slow pace. In addition, there were still few verification methods that had broad
acceptance within the fire safety community. The lack of common definitions, test
methodologies, and verification tools. among other non-technical issues, prompted different
countries to adopt different regulatory approaches.
The Javanese Avvroach
In Japan, the Building Standards Law, a highly prescriptive building code system. had been
in force since 1950. Although these regulations were deemed adequate in providing an
acceptable level of fire safety, it was also felt that the prescriptive nature "incurred the
undue increase of construction costs and restraint to building designs" [41]. Recognizing
this situation. the Building Research Institute (BRI) of the Ministry of Construction (MOC)
embarked on a five-year research project to develop a system where alternatives to the
prescribed fire safety requirements could be demonstrated to be equivalent to the objectives
of the Building Standards Law [41,42]. Although the project was successful, it was also
recognized that the outcome was a system that focused on equivalency to the Building
Standards Law, and that more could be accomplished if a performance-based system were
developed. To address this concern, and to expand the application of performance-based
fire safety design, the Ministry of Construction began a new project in 1993, "Development
of an Assessment Method for Fire Perfonnance of Building Elements" [41]. The aim of the
new effort is not to describe how to design fire safety, but to define the requirements and
standards with which buildings must comply. The format proposed for the new system is
based on the Nordic 5-Level Model described above, and includes Objectives, Functional
Requirements, Performance Requirements, Recognized Verification Tools, and Recognized
Specification of Structure and Component (performance) [43]. Key aspects of the proposed
system include multiple levels of performance standards (criteria) for verification, and built-
in safety factors that consider the knowledge base and experience with various protection
methods [41]. The performance standards (criteria) include performance-based (designated
P), complementary (C), deemed-to-satisfy (D). specification-based (S), and expert opinion
(E) [44]. The P criteria are unique in that they may be some limit state. such as the heat to
which escaping occupants are being subjected. expressed in terms of a computational
formula to be solved from an analysis of the fire conditions. This approach couples the
threshold performance criteria with a computational method. and does not permit the
designer to select either independent of the regulation (i.e.. all five levels of the NKB
structure are included in the base regulation).
The Amroach of England and Wales. New Zealand, and Australia
The situation in England and Wales began similarly, yet the results were quite different.
Until 1985, the building regulations for England and Wales were largely prescriptive and
rather restrictive. In an attempt to increase flexibility in design, and produce a more
intelligent system, a reform of the building regulations was undertaken in the late 1970s and
early 1980s. The result was dramatic. With its publication in 1985, the Building Regulutions
had been reduced from 307 pages to only 23 pages [45]. This was made possible, in part, by
applying the basic concepts of the NKB structure, using functional, or performance
language instead of prescriptive requirements. However. unlike the approach taken in Japan.
acceptance criteria and methods were not included in the regulation. This afforded engineers
the opportunity to demonstrate compliance using performancelacceptance criteria. safety
factors, engineering methods, and verification methods of their choice. However, due to the
complexities in gaining acceptance for methods that may not be understood or agreed to by
all, many designers and engineers chose to rely on the prescriptive guidance provided in the
"Approved Documents" and a series of British Standards (BS 5588 series) [46]. Thus, to
gain acceptance for engineered solutions, efforts were undertaken to develop a fire safety
engineering standard (the result of which is BSI DD240, as discussed earlier in this paper).
The transition to a performance-based regulatory system in Australia and New Zealand was
similar to situation in England and Wales [I! ,47,48]. Shortcomings were seen with existing
regulations, and more flexibility and better cost efficiency were seen as strong motivators to
change the system. In both countries, the regulatory systems follow the NKB structure, with
Levels 1, 2, and 3 in the regulations, and Levels 4 and 5 published separately from the
building code as engineering guidance documents and approved documents.
The Situation in the United States
The situation in the United States is somewhat more complex, as slightly different
approaches are being taken by the International Code Council (ICC) in the development of a
performance-based building regulation, and by the National Fire Protection Association
@IFPA) in the development of a performance-based Life Safety Code [! !.49,50.5 11. In
brief, the ICC is trying to define performance in such a way as to avoid requiring the use of
specific threshold values (e.g., upper layer temperatures, incapacitating carbon monoxide
levels) for broad classes of buildings, compartments. and occupant characteristics. The
concern is that with such a wide range of variability. the use of a specific value for a
generalized class of buildings, compartments, or people, may result in an overly
conservative design in some cases, and a potentially dangerous design in others. (A closer
look to the Japanese approach may help in this regard.) The intent is to have the design
engineers undertake appropriate analyses and select appropriate criteria, following the
requirements of the code in consultation with the approving authority, Somewhat
differently, the NFPA, is considering the inclusion of specific design fire scenarios and the
referencing of guidance on Fractional Effective Dose (FED) as recommended tenability
criteria, both of which may limit the design engineer's options if the specific project differs
significantly from the cases the code developers considered [52]. Regardless of this
dichotomy in regulatory approaches, the SFPE Engineering Guide to Performunce Bused
Analysis and Design [38] is intended to be compatible with both.

THOUGHTS ON THE CURRENT SITUATION AND THE WAY FORWARD

At the present time, fire safety engineering, often under the label "performance-based fire
safety design," is being practiced around the world. Applications range from historic
structures [53] to a pyramid-shaped casino with a volume of 595.000 cubic meters [54],
from shopping centers and malls [55] to high-rise atria buildings [56], from apartment
buildings to sports arenas, and from cleanrooms to power generating facilities [57].
Analyses performed range from egress times to time until structural failure, from sprinkler
activation to smoke exhaust capacity, and from material control to fire department response.
As discussed in the previous sections, this growing use of perfoimance-based fire safety
design concepts has been made possible due to advances in the scientific principles
underlying fire and fire safety systems, the development of risk-informed systems
approaches to fire safety analysis and design, and the implementation of performance-based
regulations.
In many respects, the fire safety science and engineering community can be pleased with the
current situation and progress to date. However. there is pet a long way to go. Consider the
following summary of observations made by Cornell on the maturation of a newly emerging
engineering area (581, and reflect on the current state of performance-based fire safety
design described by this paper.
In its earliest stages, a new area is characterized by uncoordinated relationships between
practice and research, and between needs and solutions. Problems that have been
addressed by researchers tend to reflect personal tastes, ease of formulation or solution.
and simple chance. Those applications in practice that do exist tend to be small parts of
larger problems, isolated and resolved without reference to a broader framework because
the framework does not yet exist. In contrast. maturity is characterized by a smooth
interaction between research and practice: a vocabulary has evolved, a general framework
exists, and the capabilities and limitations of the area are widely appreciated. Virtually all
practitioners have received exposure to the subject and are accustomed to recognizing the
kinds of situations in which the method is applicable and even to articulating their
problems in the language of the area, and most research is being conducted in response to
obvious needs of practice.
In between these two stages there is room for the uneven levels of development that
characterize adolescence. New practice-generated problems for research are being
identified, not on!y by growing numbers of experienced researchers, but also bq the
engineers in practice who have begun to appreciate which of their old problems the neu
area can help. These initial reactions from practice are often poorly articulated and often,
unfortunately, too optimistic. Nonetheless one can see the establishment of certain
consensus positions that determine both a framework and a viable set of solutions for at
least some rather broadly defined problem areas. However, the development and internal
 coordination of the area are still largely incomplete at this stage: some topics are virtually
untouched, limits of effectiveness of the parts or the whole are not well understood, some
applications are rather naively formulated, and some practical applications have begun to
address a larger framework, but not yet with the confidence or the wisdom of experience.
Given the current status of the field, it can be argued that fire safety engineering is a healthy
adolescent. There is not yet a single, generally accepted framework for undertaking a
performance-based approach to building fire safety analysis and design, despite the
extensive list of approaches and guides from around the world. In addition, there is
disagreement as to how performance should be defined, the addition of "people" into the
equation increases the uncertainty of the system considerably, and fire safety is effectively
non-testable (the "system" is only tested when there is a fire). Nonetheless, a review of a
various approaches currently available indicates that a comprehensive framework is close at
hand that, at a minimum, should address the following fundamental concerns [I 1,371:
1. There is a need to consider the level(s) of tolerable risk (personal and societal).
2. There is a need for clear speciiication of, and agreement to, fire safety goals and
objectives, and performance and design criteria.
3. There is a need to understand how fire initiates, develops and spreads.
4. There is a need to understand how various fire safety measures (active and passive),
including fire department operations, can mitigate potential fire losses.
5. There is a need to understand how people react in a fire situation.
6. There is a need to have, and to apply credible data, tools and methodologies in the
determination of the above factors.
7. There is a need to consider the financial impact of fire safety decisions.
8. There is a need to address uncertainties in the analysis and design process.
1. Level@) of Tolerable Risk
Consensus does not yet exist on what levels of risk are tolerable, on how they should be
quantified and expressed, or on how they should be addressed in performance-based fire
safety design [59]. This is due in part to the fact that tolerable levels of risk are value issues
that scientists and engineers should not be deciding alone [28,59,60]. Society, through
participation in the development of reguiations. legal actions, and client demands. sets the
boundaries of acceptabi!ity. To date, most performance-based regulations couch risk
acceptability in terms of broad objectives and functional statements (NKB levels 1 and 2),
and leave decisions regarding who is at risk, from what. and at what levels, to design
engineers. Unfortunately, this can lead to considerable variability in risk analyses and final
design solutions. Questions arise as to what population group is the design based upon.
what criteria are appropriate to evaluate safe conditions for this group, and how sure is one
that the resulting design is suitable for the range of population groups expected in the
building. To move forward, consensus is needed on what risks are of concern, who is at risk
and how, what levels of risk are tolerable, how risk should be quantified and expressed, and
how risk should be addressed in performance-based fire safety design [61]. Scientists and
engineers should provide facrs to support decisions in these areas, and obtain broader input
from society to help set the ultimate limits. If this is not done, the fire safety engineering
community could see itself faced with designing to levels of risk and performance
unacceptable to society.
2. Fire Safety Goals, Objectives and Performance (Design) Criteria
Scientific consensus on the definition of performance and for design criteria also does not
yet exist. As such, how can performance-based fire safety design be widely accepted? Is
performance a function of the building, its systems, its occupants, or all of these factors'?
How does one select appropriate performance criteria. and what makes them appropriate?
For smoke detector activation, a given optical density may seem suitable, and for sprinkler
activation, an upper layer temperature may seem appropriate, but should there be a time
component involved as well? What are appropriate criteria for occupant safety? Regulations
and guides tend to shy away from the use of "failure" criteria, such as death, and instead
tend toward avoidance of untenable conditions. This, however, leads to such questions as
how is tenability defined and for what target population? The fire science and engineering
community can play a significant role by developing a technical consensus on performance
criteria for ail aspects of fire safety engineering. To support shortcomings in the
profession's knowledge, advice and input should be solicited from appropriate
professionals, such as medical doctors or human behavior specialists. The resulting criteria
can then be submitted to regulatory developers who can decide which criteria should be
used in the regulations and how. The fire safety engineering community can then develop
tools and methodologies that use these criteria in an appropriate manner.
3. Fire Initiation, Development and Spread
Although more fundamental research will always be useful, the knowledge base on fire
initiation, growth, and development is currently sufficient to allow fire safety engineering to
be used for a variety of specific applications. However, the movement to performance-
based codes is encouraging the use of performance-based fire safety design on a broader
basis. For widespread use of performance-based design in structural engineering, well-
defined loads and resistances were required. In fire safety engineering, such well-defined
fire design loads and resistances do not yet exist. Presently, engineers tend to characterize
fire loads on an individual enclosure or building basis. This results in different fire loads
being used in different buildings, and does not provide a uniform or predictable level of
safety throughout individual classes of buildings. For performance-based fire safety design
to enjoy the same success as performance-based structural engineering, an effort is required
to develop design-basis fire loads that can be applied uniformly across regulated buildings.
This cannot be done by selecting a single characteristic design fire, or even a set of design
fires, but may require the fire impact to be disaggregated into component loads. For each
component load, a definable resistance should exist, as well as methods to calculate the
resistance. With definable loads and resistances, relationships can be developed such that
the overall fire performance of a structure can be evaluated in terms of resistance to
component loads: in much the same way as structural engineering is currently performed.
Allowable loads could then be defined in building regulations, based on building use,
occupant characteristics, or other risk factors. These loads would, by default, reflect various
fire impacts (as the use of various fire scenarios tries to accomplish today), and provide
uniform measures for consistent code applications. Although some work is being done in
this area [62],the concept is young, and if deemed valuable, will require considerable effort
to bring it to fruition.
4. Mitigation of Potential Fire 1,osses
Loss experience, testing and research have provided a wealth of understanding of how to
mitigate potential fire losses. As with the preceding discussion, mitigation technologies
exist in usable forms; however, to be effectively addressed in a performance-based system.
the mitigation technologies need to be understood in terms of the applied loads. In many
respects, the information required to craft the relationships between the loads and how the
technologies provide resistance already exist. and will fall into place as the other parts of the
overall framework are developed. Keys to successfully addressing this item relate back to
defining performance, clearly defining performance criteria, and understanding the level of
fire impact (risk) tolerable to society. Addressing system reliability as part of performance
will be critical, and should be a key component of performance-based fire safety design.
5. Human Behavior and Reaction to Fire
Better understanding of how people react in fire situations will provide valuable information
for determining who will be at risk, from what, and at what stage of a fire. Such
information would then lead to better selection of human factors criteria for use in
regulations and design, with appropriate uncertainty bounds, and lead to better analysis and
design tools. As with the discussion above, the fire science and engineering community
needs to reach out to other professions, and to society in general, for assistance in this area.
An encouraging step in this direction was taken with the First International Symposium on
Human Behaviour in Fire, and it is hoped the interdisciplinary dialog begun at that
symposium will continue [63].
6. & 8. Data, Analytical Tools, Evaluation and Design Methodologies, and Uncertainty
Fire scientists can continue to contribute significantly to the development of credible data
and engineering tools and methodologies for use in performance-based fire safety design.
As discussed earlier in this paper, the discipline has already made great strides in this area.
However, more than ever, there is a need to understand, quantify, and address the
uncertainties and unknowns in the data, tools, and methodologies. Fire test data can and
should be provided with uncertainty bands, and should come with information regarding the
test conditions, the test room, the instrumentation, and how the data was collected and
recorded. Guidance on this exists outside of this profession [64] and within [65]. Similarly,
engineering tools, be they simple analytical equations or complex computerized models,
should be made available with discussion of applications, limitations, and uncertainties
166,671. More important, perhaps, the uncertainties and unknowns in the overall
performance-based fire safety design process need to be clearly identified and addressed
where found to be significant. If it is found that the critical factors in a performance-based
fire safety analysis and design involve people more so than the combustion process, should
not more effort be focused on human behavior modeling instead of combustion modeling?
What are the dominant factors with which fire safety engineers should be most concerned,
and how concerned should they be? Although some work has been carried out in the area of
uncertainty [35,36,37,68], much more is required. and it is likely to be one of the dominant
need areas in the attempt to attain a robust and widely-accepted performance-based fire
safety design framework.
7. Economic Impact (Benefit-Cost) Analyses
Benefit-cost analyses need to be considered as part of the development of a performance-
based fire safety framework as well. Although perhaps not as critical to the fire science
community as focusing on performance criteria, there is a need for engineers to understand
what benefits are provided at what cost. and to be able to articulate this to regulators, clients,
and regulatory developers. Again, the biggest challenge may be to find a balance between
fire safety risks and costs that is at a level tolerable to society.
On a final note, given the current state of development of performance-based fire safety
engineering, one would hope that only truly qualified and ethical engineers were practicing
the profession, and that resulting designs always erred on the side of safety from everyone's
perspective. Unfortunately, this is not always the case. There are relatively few fire safety
engineers and code officials with the background needed to undertake or to review and
approve performance-based designs [69.70.71,72,73]. This is particularly distressing in light
of the shortcomings in design guidance and engineering tools. Even when qualified fire
safety engineers are engaged, the lack of guidance documents, widely-accepted criteria, and
evaluation methods makes each project different. This can make it difficult to provide
consistency between projects, and result in questions from the local authorities [74]. In
short, more and continuing education and supporting documentation are urgently needed.

SUMMARY
Performance-based fire safety engineering has recently emerged as "the" approach to
building fire safety design. This has been made possible primarily by advances in fire
science and engineering and the global movement to performance-based building codes.
However, the discipline of fire safety engineering and the concept of performance-based fire
safety engineering are still developing, and much is required before the discipline can be
considered mature. This paper has outlined the evolution of performance-based fire safety
design, the current situation. and thoughts for addressing current shortcomings. The
discussion of the shortcomings is not meant to imply that performance-based fire safety
design should be stopped, but rather, that fire safety engineers proceed cautiously, with an
understanding of what is acceptable to society, and with recognition of the limitations in the
current technologies. As more information is developed, and fire scientists and engineers
continue to work together and with allied professionals, performance-based fire safety
engineering will grow into a mature, stable, and widely accepted discipline.

ACKNOWLEDGMENTS

The author would like to extend his sincere appreciation to the Programme Committee of
the 6"' International Symposium on Fire Safety Science for the invitation to present this
paper. In addition, the author acknowledges the contributions and feedback from his
colleagues in the Society of Fire Protection Engineers and in the International Association
for Fire Safety Science. Specifically, the author extends his sincere appreciation to Robert
W. Fitzgerald, John Hall, Jr., Matti Kokkala, Jane Lataille. Sven Erik Magnusson, Caird
Ramsay, and Ai Sekizawa for their careful review and constructive comments on this paper.
Any errors and omissions are solely attributable to the author. Finally, the author would like
to thank the staff and Board of Directors of the Society of Fire Protection Engineers for
supporting the development and presentation of this paper.
REFERENCES
1. The Random House College Dictionary, Revised Edition, Random House, NY, 1980.
2. Gordon, J.E., Structures: Or Why Things Don't Fall Down, De Capo Press, NY, 1978.
3. Ellingwood, B., "Reliability Basis of Load and Resistance Factors for Reinforced Concrete Design." NBS
Building Science Series 110, NBS, February 1978.
4. AF&PAIASCE 16-95, Standard for Load and Resistance Factor Design (LRFD) for Engineered Wood
Construction, American Society of Civil Engineers, New York, 1995
5. Smith. J.C., Structural Steel Design: LRFD Approach, John Wiley % Sons, New York, 1991.
6. Hall, J.R.Jr. and Cote, A.E., "America's Fire Problem and Fire Protection," Fire Protection Handbook, 17"'
Edition, NFPA, Quincy, MA, 1991, p. 1-9.
7. NFPA 13, Standard for the Installation of Sprinkler Systems, NFPA, Quincy, MA, 1994, p. 13-1
8. Op cit., pp. 1-19
9. Uniform Building Code, International Conference of Building Officials, Whittier, CA, 1994, pp. i-iii.
10. Hatt, W.K, Chairman, "Recommended Minimum Requirements for Fire Resistance in Buildings," Report
of the De~artmentof Commerce Building Code Committee, US GPO, Washington, DC, 193 1.
11. Meacham, B.J., The Evolution of Performance-Based Codes and Fire Safety Design Methods, NLST-GCR-
98-761, November 1998.
12. Hadjisophocleous, G.V. et al., "Literature Review of Performance-Based Fire Codes and Design
Environment," JFPE,Vol. 9, No. 1, 1998, pp. 12-40.
13. General Services Administration, Building Fire Safety Criteria. Appendix D: Interim Guide for Goal-
Oriented Systems Approach to Building Firesafety, GSA, Washington, DC, 1972.
14. NFPA 550. Guide to Systems Concepts for Fire Protection, NFPA, Quincy. MA, USA, 1995.
15, NFPA 101A. Guide on Alternative Approaches to Life Safety, NFPA, Quincy, MA, USA, 1995.
16, Watts, Jr., J.M., "Fire Risk Ranking," SFPE Handbook of Fire Protection Engineering, 2nd Ed., Section 5,
Chapter 2, SFPE and NFPA, Quincy, MA, USA, 1995.
17. Personal communication, Robert W. Fitzgerald.
18. Fitzgerald, R.W., "An Engineering Method for Building Fire Safety Analysis," Fire Safety Journal, Vol. 9,
No. 2, pp. 233-243, July 1985.
19. Fitzgerald, R.W., "An Engineering Method for Translating Fire Science into Building Design,"
Proceedings of the CIB W14 International Symposium and Workshops. Engineering Fire Safetv in the Process
of Design: Demonstrating Eauivalency, University of Ulster. Fire SERT Centre, Jordanstown, Northern
Ireland, 13-16 September 1993.
20. Beck, V.R., "Performance-Based Fire Safety Design--Recent Developments in Australia," Proceedings of
the Technical Conference "Fire Safety bv Desien: A Framework for the Future," Fire Research Station,
Borehamwood, UK, I0 November 1993.
21. Beck, V.R. "Outline of a Stochastic Decision-Making Model for Building Fire Safety and Protection," Fire
Safety Journal, Vol. 6, No. 2, 1983, pp. 105-120.
22. Beck, V.R. and Yung, D., "Building Fire Safety Risk Analysis," SFPE Handbook of Fire Protection
Engineering, Section 5, Chapter l 1, SFPE and NFPA, Quincy, MA, USA, 1995.
23. Beck, V.R., "Performance-based Fire Engineering Design and its Application in Australia," Invited
Lecture, in Fire Safetv Science - Proceedings of the Fifth International Symposium. IAFSS, 1997, pp. 23-40.
24. Beck, V.R. and Yung, D., "The Development of a Cost-Effective Risk-Assessment Model for the
Evaluation of Fire Safety in Buildings," Fire Safety Science - Proceedings of the Fourth lnternational
Symposium, IAFSS, 1994. pp. 8 17-828.
25. CIB W14, "A Conceptual Approach towards a Probability-Based Design Guide on Structural Fire Safety,"
Fire Safetv Journal, 6, 1983.
26. Pettersson, O., "Rational Structural Fire Engineering Design Based on Simulated Real Fire Exposure,"
Fire Safety Science - Proceedings of the Fourth International Symposium, IAFSS, 1994, pp. 3-25.
27. Magnusson, S.E. et al., Fire Safety Design Based on Calculations: Uncertainty Analysis and Safetv
Verification, ISSN 1102-8246; Report 3078, Lund University, Sweden, 1995.
28. Magnusson, S.E., "Risk Assessment," Invited Lecture, in Fire Safetv Science - Proceedinas of the Fifth
lntemationalSvmoosium, IAFSS, 1997, pp. 41-58.
29. Frantzich, H., Uncertainty and Risk Analysis in Fire Safetv Engineering, Report LUTVDG/(TVBB-I016),
Lund University, Sweden, 1998,
30. Project Report and Technical Papers. Books 1 and 2. Fire Safetv Engineering Proiect, The Warren Centre
for Advanced Engineering, The University of Sydney, Sydney, Australia, December 1989.
31. Beck, V.R.. et. al.. Draft National Building Fire Safetv Svstems Code, Department of Industry.
Technology and Commerce, Canberra, Australia, May 1991.
32. Fire Code Reform Centre, Fire Engineering Guidelines, Sydney, Australia, March. 1996.
33. Buchanan, A , , Fire Engineering Design Guide, Centre for Advanced Engineering. University of
Canterbury. Christchurch, New Zealand, July 1994.
34. Cooke, G., "The Philosophy of Fire Engineering Being Developed by the British Standards Code Drafting
Panel," Proceedings of the CIB W14 lnternational Svmoosium and Workshoos. Engineering Fire Safety in the
Process of Design: Demonstrating Equivalency, University of Ulster. Fire SERT Centre, Jordanstown.
Northern Ireland, !3-1 6 September 1993.
35. Fire Safetv Engineering in Buildings, DD 240: Parts 1 and 2: 1997, British Standards Institute, 1997.
36. I S 0 TR 13387, Part I: The Aoplication of Fire Performance Conceots to Design Obiectives, 1999.
37. Meacham, B.J., Assessment of the Technological Requirements for Realization of Performance-Based Fire
Safety Desian in the United States: Final Reoort, NIST GCR 98-763, NIST, Gaithersburg, MD, 1998.
38. SFPE Engineering Guide on Performance-Based Fire Safetv Analysis and Design, draft for public
comment, SFPE and NFPA, January 1999.
39. The Nordic Committee on Building Regulations (NKB), Programme of Work for the NKB. Report No. 28,
Stockholm, 1976.
40. The Nordic Committee on Building Regulations (NKB), Structure for Building Regulations, Report
No.34, Stockholm. 1978.
41. Tanaka, T., "The Outline of a Performance-Based Fire Safety Design System of Buildings," Proceedings
of the 7th lnternational Researcil and Training Seminar on Regional Development Planning for Disaster
Prevention. Improved Firesafety Systems in Developing Countries, United Nations Center for Regional
Development, Tokyo, Japan, 1995.
42. Wakamatsu, T., "Fire Research in Japan - Development of a Design System for Building Fire Safety,"
Proceedings of the 6th Joint Panel Meeting, UNJR Panel on Fire Research and Safetv, Tokyo, Japan, p.882.
10-14 May 1982.
43. Nakava. ..
. . I. and Hirano. Y.. "Jaoan's Aooroach Toward the Building Code and Standards of a new
Generation," in Proceedings. 1996 international Conference on ~erformaice-basedCodes and Fire Safety
Design Methods, SFPE, 1997. pp. 15 1- 158.
44. Babrauskas, V., "Ensuring the Public's Right to Adequate Fire Safety Under Performance-Based Building
Codes," Proceedings of the 1998 Pacific-Rim conference and Second lnternational Conference on
Performance-Based Codes and Fire Safetv Desien Methods, ISBN 1-58001-01 1-3, ICBO and SFPE, 1998,
pp. 239-247.
45. The Building Regulations. Department of the Environment, London. England, 1985.
46. Law, M., "Fire Safety Design Practices in the United Kingdom - New Regulations," Proceedings of the
Conference on Firesafety Design in the 21st Century, WPI and SFPE, May 1991.
47. Bowen, N., "The Performance Building Code of Australia: A Study of its Development," Proceedinzs of
the First lnternational Conference on Performance-Based Codes and Fire Safety Design Methods, SFPE,
1997, pp.39-50.
48. Hunt, J.H., "Performance-Based Codes: The New Zealand Experience." Proceedings, 1996 lnternational
Conference on Perfonnance-Based Codes and Fire Safety Design Methods. SFPE, 1997, pp. 17-28.
49. Tubbs, B., "Status of the ICC's Performance-Based Code Development, Proceedines of the 1998 Pacific
Rim Conference and Second lnternational Conference on Performance-Based Codes and Fire Safetv Design
Methods, ISBN 1-5800 1-0 1 1-3, ICBO and SFPE, pp. 195-204.
50. Meacham, B.J., "A Performance-Based Building Regulatory System: Providing a Means to Streamline the
Building Regulatory Process, Promote Innovation in the Construction Industry, and Improve lnternational
Economic Competitiveness." Proceedings of the NCSBCS,%ISTIHUD Joint Technical and Research
Conference and the NCSBCS 3 1" Annual Conference: Innovations in the Construction lndustrv and Building
Regulatory Process, NCSBCS, Dana Point. CA, 4-7 November 1998.
51. Watts, J., "Performance-Based Life Safety Code," Proceedings. 1996 International Conference on
Performance-Based Codes and Fire Safetv Design Methods, SFPE, 1997, pp. 159-169.
52. Life Safety Code, Draft for Public Comment, January 1998.
53. Watts, J.M. Jr. and Kaplan. M.E.. "Performance-Based Approaches to Protecting Our Heritage,"
Proceedings, 1996 lnternational Conference on Performance-Based Codes and Fire Safety Design Methods.
SFPE, 1997, pp. 339-347.
54. Evans, D.H., "Luxor Hotel and Casino: An Application of Performance-Based Fire Safety Design
Methods," Proceedings of the 1998 Pacific Kim Conference and Second lnternational Conference on
Performance-Based Codes and Fire Safetv Design Methods, ISBN 1-5800 1-01 1-3, pp. 477-494.
55. Timms, G.R.. Ramsay, G.C., and Horasan, M.B.N.. "Australia's Performance-based Regulations - A Fire
Safety Engineering Experience," Proceedinrs of the 1998 Pacific Rim Conference and Second lnternational
Conference on Performance-Based Codes and Fire Safetv Design Methods, ISBN 1-58001-01 1-3, lCBO and
SFPE, pp. 379-390.
56. Tanaka, T. et al., "Case Study: Performance-Based Fire Safety Design of a High-Rise Office Building,"
Proceedines of the 1998 Pacific Rim Conference and Second International Conference on Performance-Based
Codes and Fire Safetv Design Methods, ISBN 1-58001-01 1-3, ICBO and SFPE, pp. 563-574.
57. Capaul, T., Meacham, B.J. and Fontana, M., "The Application of a Performance-Based Evaluation and
Design Approach to a Structural Fire Resistance Problem," Proceedines of the lAFSS First European
Symposium on Fire Safety Science, ETH, Zurich, Switzerland, 21-23 August 1995.
58. Cornell, C.A., "Structural Safety: Some Historical Evidence that it is a Healthy Adolescent," Proceedings
of the 3'd lnternational Conference on Structural Safety and Reliability, Trondheim, Norway, 198 1, pp. 19-29.
59. Meacham, B.J., "Risk-Informed Decision-Making in Performance-Based Building and Fire Code
Development," Proceedings of the NFPRF Fire Risk and Hazard Assessment Research Application
Symposium, National Fire Protection Research Foundation, San Francisco, CA, 24-26 June 1998.
60. Hall, J.R. Jr., "Societal Issues in Performance-based fire Safety Design," Proceedings of the VUT
CESARE and SFPE Workshop on Performance-Based Fire Safetv Design, 28 February 1997.
61. To begin addressing these issues, a joint National Science Foundation (NSF)/Private Sector Initiative has
been launched to develop a process that building and fire codes and standards development committees can
use to help them identify, address, and incorporate risk concepts and risk-based solutions into building and
fire regulations. For more information, see: Meacham, B.J., "Incorporating Risk Concepts into Performance-
Based Building and Fire Regulation Development," Proceedings of the I999 Conference on Firesafetv Design
in the 2 1" Century, WPI and SFPE, June 1999.
62. Fitzgerald, R.W., "Chapter 3: Fire Design Loadings," The Evaluation of Building Fire Safety, Volume 3,
Draft 3, January 1999.
63. Proceedings of the 1st International Symposium on Human Behaviour in Fire, University of Ulster,
Antrim, Northern Ireland, 3 1 August - 2 September 1998.
64. American National Standard for Expressins Uncertaintv - U.S. Guide to the Expression of Uncertainty in
Measurement, ANSIR\ICSL 2540-2-1 997, National Conference of Standards Laboratories, 1997.
65. ASTM E 1591. Standard Guide for Data for Fire Models, ASTM, 1994.
66. Assessment and Verification of Mathematical Fire Modeis, I S 0 TR 13387, Part 3, 1999.
67. ASTM E1355, Standard Guide for Evaluating the Predictive Capability of Deterministic Fire Models,
draft to replace E 1355-92, American Society for Testing and Materials, November 1996.
68. Notarianni, K.A. and Fischbeck, P., "A Methodology for the Quantitative Treatment of Variability and
Uncertainty in Performance-Based Engineering Analysis and/or Decision Analysis with a Case Study in
Residential Fire Sprinklers," Proceedings of the Second International Conference on Performance-Based
Codes and Fire Safetv Design Methods, ISBN, 1-58001-01 1-3, ICBO and SFPE, pp. 297-3 11.
69. Fleischmann, C., "Education for Performance-Based Codes," Proceedines of the First lnternational
Conference on Performance-Based Codes and Fire Safety Design Methods. SFPE, 1997, pp.101-106.
70, Beever, P., "Performance Versus Prescriptive Fire Codes: Education for Cultural Change," Proceedines of
the Second International Conference on Performance-Based Codes and Fire Safetv Design Methods, SFPE
and ICBO, 1998, pp. 205-212.
71. Meacham, B.J., Concepts of a Performance-Based Building Regulatorv Svstem for the United States,
NlST GCR 98-762, NlST Gaithersburg, MD, November 1998.
72. Inter-jurisdictional Regulatory Collaboration Committee, Guidelines for the Introduction of Performance-
Based Building Regulations, ABCB, Canberra, Australia, April 1998.
73. Woiski, A,, Dembsey, N.A.. and Meacham, B.J., "Application of Acceptable Risk Principles to
Performance-Based Building and Fire Safety Code Development," Proceedings of the Second lnternational
Conference on Performance-Based Codes and Fire Safety Design Methods, ISBN. 1-58001-01 1-3, ICBO and
SFPE, pp.269-284.
74. Fleming, J.M, "A Code Official's View of Performance-Based Codes," Proceedings - 1996 lnternational
Conference on Performance-Based Codes and Fire Safety Design Methods, SFPE, pp. 257-269, 1997.