THE MASS FIRE POTENTIAL OF
URBAN STRUCTURE AND FORM

Thesis for the Degree of M. U. P.
MICHIGAN STATE UNIVERSITY
STEPHEN. WARREN SCHAR

1969

 

1H is“;

“.m‘ "4‘

yv) I!

Limb I.’\-

Michigan «11 .

I
I: Universi“

 

..*__v

 
 
 
 
     
   

 
 

BINDING BYTE:

IIIIAII & SIIIIS'
BIIIIII BINDFRY INC.

LIBRA E DEBS

  

 

.-
~
d

 

N.-
a
3.3. -
.
.. '
I. -
0

.AA‘
50'.

In
,;I

“\I

V.
a

h-‘

5H

Q.‘

‘1

"r

 

ABSTRACT

THE MASS FIRE POTENTIAL OF URBAN STRUCTURE AND FORM
by Stephen warren Schar

This thesis addresses the potential for the spread of mass fire which
exists in many urban areas. The specific physical pattern of each
community possesses all the variants in form and structure possible, yet
some of these forms and patterns create hazards to the effective provision
of public safety and community needs. When the physical features of any
community, by their congestion, deteriorating condition or mere distribu-
tion, create a block to the ability of fire departments to end the fire,
then conditions of spread exist. In situations where uncontrolled fires
join, mass fire is the result. It is the primary objective of this
thesis to examine and identify the parameters which determine when,

where, and how a fire can be expected to spread to mass fire proportions.

The parameters which are identified are all descriptors of the physical
condition of the community as well as the atmospheric environment. It

is shown that all parameters concerned with fire can be classified as
either atmospheric conditions, fuel conditions, or tOpographical features.
These three major classifications are examined for the elements which are
critical in the spread of fire. Concepts of heat transfer by conduction,
convection, and radiation are examined for pertinence to urban fire spread,
and related to the types of fuels found "typically". These general

mechanics by which fire spreads are related to the urban "fire environment".

The importance of the expected volume of the fire in relation to the

expected water supply is discussed, and is shown important in the ability

 

1w?

Stephen W. Schar

of a community to contain the fire. The distribution of fire facilities
is briefly discussed in relation to the distribution of residential and
commercial areas of the community. Standards for the areal distribution
of control forces as well as water supply are presented as derived by

the American Insurance Association.

An important element in the identification of basic parameters for mass
fire spread is the schedule for fire grading and rating applied by fire
engineers in the determination of fire insurance rates. The importance

of "conflagration breeding blocks" as derived from this table lead directly
to the identification of those elements of the block's condition which
make it a breeder. These elements are applied directly to the basic
parameters of fire spread to provide descriptive fire spread parameters

for an urban area.

The ten parameters identified for basic fire spread are refined to five
directly applicable to areal description. This translation allows the
author to present the parameters in the form of two models, as taken

from fire research sources. The parameters of urban fire spread, as
applied within the models, involve (a) the construction features of
buildings and structures, (b) the likely configuration and intensity of
initial fire, (c) the intensity of development of the area, (d) the
expected atmospheric conditions, and (e) the fuel type. The models which
result are directed at the Hazard within blocks, and the Spread between

blocks.

These models are sensitive to differences in the types of land use found

within blocks, the amount and kind of Open space found, and the distribution

Stephen W. Schar

of buildings within the blocks. The concern of planning in these three
areas is no less important. Urban planning seeks to plan the physical
development of the community along principles which respect the health,
well-being, and safety of the community. This thesis identifies those
parameters and Ehgi£_critical values which would enable a planner to

determine the areas of maximum fire danger within the community.

Historical evidence of the occurrence of mass fire is cited as a
justification of this tOpic's relevance to urban planning. The relevance
lies chiefly, however, in the final determination of the importance of
physical planning to the prevention or attenuation of conditions which
cause fire spread in massive proportions. If individual fires spread
through relatively high density areas because of unfavorable structural
and design characteristics, a mass fire may logically be the result of

inadequate physical planning and corrective action.

THE MASS FIRE POTENTIAL
OF

URBAN STRUCTURE AND FORM

BY

Stephen warren Schar

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER IN URBAN PLANNING
School of Urban Planning and Landscape Architecture

1969

6 52573 5’0

M/m/W

Acknowledgement

Throughout the process of research and writing for this study, I received
valuable assistance from several institutions and individuals. I owe

a special debt of gratitude to Judson D. Blakslee, Director of Training
and Education, Region Four, Office of Civil Defense. His help and
inspiration typify the involvement and dedication of the top professionals
currently engaged in disaster preparedness. His influence was a welcome
element throughout my study at 0CD Staff College and 0CD Research Library.
I would further like to acknowledge the patience and support, both moral
and financial, of the Department of Urban Planning, Michigan State University.
In particular, the support of Professor Charles W. Barr and Dr. Richard
Anderson was extremely welcome at the latter stages of work. Lastly,

I wish to thank my wife, Ann, for her devoted patience.

ii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS . . . . . . . . . . . . . . .
LIST OF TABIES . O O O O O O O O O O O O O O 0
LI ST OF FIGUMS O O O O 0 O O O 0 O O O O O O I

INTmDUCT ION O O O O O C O O O O O O O 0 O . O 0

CHAPTER
I. THE CONCEPT OF MASS FIRE. . . . . . . .
Definitions. . . . . . . . . . . .
Control of Mass Fire . . . . . . .
II. THE URBAN MASS FIRE . . . . . . . . . .
Fire Loss. . . . . . . . . . . . .

Fire Spread Contributors . . . . .

 

III. THE URBAN FIRE ENVIRONMENT. . . . . . .
Fire Volume and Area . . . . . . .
water Supply . . . . . . . . . . .

Mutm 1 Aid 0 O O O O O O O O O O O

 

Fire Departments . . . . . . . . .

Department Distribution Standards.

Fire Environment Grading_and Rating.

Correlation to Community Population

Structural Condition of the Community

1. Area of the District

 

2. Street Widths

 

3. Accessibility of Blocks

 

4. Area in Qpen Space

iii

11

l4

14

15

17

19

21

24

24

29

5. Per Cent Built Upon
6. Height of Buildings

7. Frame Areas

 

8. Conflagration Breeding Blocks
9. Ex sure
Environmental Conditions. . . . . . . . . . . . . . . .
1. High Winds Frequency
2. Excessive Snowfall
3. Hot Dry Weather

4. Unusual or Exceptional Conditions

 

Divergency in water Supply and Fire Control
Capability. . . . . . . . . . . . . . . . . . . . . . .
IV. MODELING URBAN MASS FIRE SPREAD . . . . . . . . . . . . . . . . .
Urban Vulnerability Model. . . . . . . . . . . . . . . . . .
Basic Parameters. . . . . . . . . . . . . . . . . . . .
Classifying the Urban Area. . . . . . . . . . . . . . .

Forest Service Approach. . . . . . . . . . . . . .

 

Civil Defense Approach . . . . . . . . . . . . . .

StudygApproach . . . . . . . . . . . . . . . . . .

 

Conflagration Potential Model. . . . . . . . . . . . . . . .
Basic Parameters. . . . . . . . . . . . . . . . . . . .
Hazard Value Per Block. . . . . . . . . . . . . . . . .
Fire Loading by Use . . . . . . . . . . . . . . . . . .
Relative Potential Rating . . . . . . . . . . . . . . .
Application of the Model. . . . . . . . . . . . . . . .

Firebreak Potential Model. . . . . . . . . . . . . . . . . .

iv

35

36

38

38

4O

45

46

48

49

52

54

55

S7

58

62

64

V.

BIBLIOGRAPHY.

Area and Firebreak Typology . . . . . . . . . . .
Radiating Surface Area. . . . . . . . . . . . . .
StruCtural surface MOdel O O O O O O O O O O O O 0

Application of the Model. . . . . . . . . . . . .

CONCLU S IONS O O O O O O O O O O O O O O O O O O O O O O O 0

APPENDIX. .

A.

B.

THE MECHANICS OF FIRE AND FIRE SPREAD . . . . . . . . . . .

DATA

Fire Environment . . . . . . . . . . . . . . . . . . .
Elements of the Fire Environment . . . . . . . . . . .

Fuel. 0 I O O O I O O O O O O O O O O O I O O O 0

Air Mass. 0 O O O O O O O O O O O O O O O O O O 0

Tom raphy O O O O O I O O O O O O O O O O O O O O

The Mechanics of Spread . . . . . . . . . . . . . . .

AND SUPPLEMENTARY INFORMATION

1. SELECTED FIRES AND CONFLAGRATIONS SINCE 1910 . . .

2. RANKING OF PRINCIPAL FACTORS CONTRIBUTING TO SPREAD.

3. CHANGE IN RANKING BETWEEN TIME PERIODS . . . . . .

4. COMPARATIVE FIRE STATISTICS FOR THE UNITED STATES.
FOR 1966 AND 1967

5. COMPARATIVE FIRE STATISTICS FOR THE UNITED STATES.
BY CITY SIZE

6. POTENTIAL FIE MAS - 1967. O O O O O O O O O O O

7 0 WATER ”PL ICAT ION MTE O O O C O O O O O O O O O O

65

66

69

71

74

82

A-12

B-l

10.

11.

12.

13.

WATER SUPPLY DEFICIENCIES - 1967 . . . . . . .
MUTUAL AID RESPONSE - 1967 . . . . . . . . . .
REQUIRED FIRE FLOWS UNDER AMERICAN INSURANCE .
ASSOCIATION STANDARDS - 1967

RECOMMENDED AREAS SERVED FOR HYDRANTS-AIA-l967
MULTIPLIERS FOR BUILDING CHARACTERISTICS . . .

REQUIRED SEPARATION DISTANCES IN FEET. . . . .

vi

B-lO

B-ll

B-12

B-13

so

Table

10.

LIST OF TABLES

Principal Factors Contributing to Conflagrations
in the United States and Canada Since 1900. . . . .

Relationship of Fire Department Coverage to
Population Density - 1963 . . . . . . . . . . . . .

Interrelationship and Composition of Municipal
Classification and Specific Fire Insurance Rates. .

Distribution of Cities over 10,000 Population
According to Total Fire Insurance Classification. .

Penalty Schedule for Assessing Points for Area
of District in Open Space . . . . . . . . . . . . .

Penalty Schedule for Assessing Points for Per Cent
Of BlOCk area Built Upon O O O O O O O O O O O O O 0

Classification Divergence Penalty Scale For
water Supply and Fire Control Capability. . . . . .

Ranked Importance of Parameters Governing the
Destruction of Resources by Various Fire Types. . .

Variables for Calculating Mass Fire Potential
Numerical Values for Hazard by Block. . . . . . . .

Block Hazard Rating Model Application . . . . . . .

vii

Page

11

20

25

28

31

32

37

44

55

61

Figure 1

Figure 2

LIST OF FIGURES

Title Page

Relationship Between Insurance Classification and
POPUl ation o o o e e o o o 0 e o o o o o o o o o o o o 2 7

Radiating Area vs. Exposure Distance For a Rectangular
Mdiater O O O O O O O O O ' O O O O O O O O O O O O O O 68

viii

INTRODUCTION

This thesis is directly concerned with the passive, but vital, role
of urban physical form in the public safety of the community. The specific
physical pattern of each community possesses all the variants in form
possible, yet some of these forms and patterns create blocks and hazards
to the effective provision of public safety and community needs. The
extent of these restrictions, and their nature, are of the most urgent
concern in the planning and provision of those services which deal with
emergencies and the protection of life and prOperty. Such emergency or
public safety services generally include fire protection, law enforcement,
pollution control, and civil defense.

The specific emphasis in this study deals with the violent, disas-
trous emergency of fire. In each municipality, the organization most
directly involved with fire protection control and prevention is the fire
department. Yet in either peacetime or wartime situations where loss of life
or property may assume large and disastrous prOportions, civil defense is
also concerned. The factors which affect the operations or effectiveness
of these two agencies will be our major interest.

The objectives of the fire protection function are generally con-
sidered to be (1) the prevention of fire starts: (2) the prevention of
loss of life and property in the event a fire does start; (3) the con-
finement of a fire to the place of origin; and (4) the extinction of the
fire. These involve the services of those trained in fire prevention and

fire fighting. Both are strongly influenced in their effectiveness by the

physical form of a community and its pattern of land uses.
The objective of this thesis is to examine, in depth, the critical

parameters governing the spread of a fire of conventional magnitude to that

 

of a mass or group fire. These parameters must,of course, include large
human variables, such as the performance of fire department operations.

As elements in fire spread, human errors merit much consideration. The focus
of this study is upon physical, environmental variables however; and we shall,
therefore, limit ourselves to a discussion of community structure as an
element. Ambient fire conditions must be included, as we shall show.

The importance of this topic lies in its relevance to the intense
urbanization of our society today. Fire cannot be considered a small danger
to life and prOperty in developed areas, especially in areas of high intensity
of land use and structural development. People cause fires by commission
or omission and, therefore, where people congregate so may fires be expected
to develop. When a high probability of fire ignition is coupled with a
dense distribution of "fuel," fire may be expected to spread. The extent
of spread is our concern here. Our objective, as stated above, is a close
examination of the urban structure as a fuel array, a spread determinant,
and a critical parameter in the early attenuation of mass fire.

The treatment of mass as a fuel and space as a channel for heat
flow are somewhat alien to the urban planner and designer. He is unaccus-
tomed to having his well-structured setting of buildings and space referred
to as fuel loads and fire breaks. This concept is the target of this study.
Unless the urban planner is inclined to an understanding of the problems
faced by disaster experts and pre-planners, he may well be working at cross

purposes with them.

CHAPTER 1
THE CONCEPT OF MASS FIRE

The danger of uncontrollable fire exists in every area of the country,
be it urban or non-urban. Destructive fires may vary in nature and size
from those involving small amounts of fuel to some involving vast areas of
combustible material. While small fires may be extinguished with little
effort, the large-scale fire may continue to burn unchecked until the
available fuel is consumed. Somewhere between the two extremes of scale
an uncontrolled fire passes beyond the capabilities and investments of
human control. Such instances are termed major fires. If, however, such
fires involve several structures, areas or large quantities of combustibles

at one time, the term "mass fire" is rightfully used.

Definitions
Three terms are commonly involved in the discussion of major fires.

These terms are mass-fire, conflagration, and fire-storm. They differ

 

significantly in physical behavior, conditions necessary for their propa-
gation and continuance, and impact upon fire fighting efforts.

Mass Fire. Mass-fire is commonly taken to mean "a fire which
occurs when a single fire extends to cause simultaneous burning of many
individual structures or when several separate fires merge into a single
fire involving a large number of buildings."1 Mass fires need not involve
only structures in urban areas, however. Forest fires frequently evolve into
mass fires. Forest fires are, in fact, the most prevalent form of mass

fire in this country today.

 

1B.M. Cohn, L.E. Almgren, and M. Curless, A System for Local
Assessment of the Conflagration Potential of Urban Areas. (Chicago: Gage
Babcock Associates, Inc., 1965), p. 7.

Conflagration. The conflagration might best be defined as a

 

'mass-fire' involving many simultaneously burning structures and having

a moving front. The direction of spread is generally in the direction of

 

the prevailing wind and is influenced by topography as well as by the
availability of fuel and combustibles.”2 The point of importance here is
the "moving front," generally in the direction of the "prevailing wind."
Conflagrations, as a form of mass fire, can occur in both cities and
forests. In both, the moving front is long compared to its width and
ignites new fuel ahead of it and leaves smoldering ashes in its wake.

The designation of many large-scale fires as conflagrations is
frequently improper. In general practice the term is applied only to
fires extending over a "considerably large" area and destroying large
numbers of buildings. It is best to use the term conflagration conserva-
tively. For certain fires of a moving nature the term "group fire" may
be more descriptive. These include fires within the limit of an industrial
plant pr0perty even if several buildings are involved, and fires in a
group of mercantile buildings, particularly within a single city block. In
both such cases, buildings may be so close together that a fire may spread
from some of the buildings to adjoining ones, but it is unlikely to spread
outside the plant area because of fire wall barriers, streets, or other
Open spaces.

Conflagrations, as described by the above definition, generally
take one of four different forms in urban areas. The first of these include
fires which start in hazardous structures in congested and high building—
density areas. These spread in one or more directions before being brought
under control. These fires usually spread first to nearby structures of

similar quality and chiefly spread in the direction the wind is blowing.

 

2 Ibide’ Po 6.

V
n

1.”

 

 

   

 

 

2“.
I

 

 

e"

 

I
n
‘\
‘\I

A second type of conflagration includes fires which occur in primarily
residential sections, and spread beyond control due to closely built com-
bustible construction and wooden shingle roofs. Conflagrations which result
from extensive forest and brush fires entering a municipality over a wide
frontage comprise a third type. Finally, eXplosions and intense flame-out

3

may result in fire over a wide area.

Fire Storm. The fire-storm is the third of the large-scale fires.

 

It is a mass fire which involves many simultaneously burning structures

and has a stationary front. It is characterized by strong inward-rushing

 

winds created by the incredible demand for oxygen to support combustion
in a large area of fire. Complete destruction within the burning perimeter
is the result. "Essentially all the fuel over the fire area is simultane-
ously ignited and simultaneously burns, producing a thermal convection
column so strong that it completely dominates all normally important
atmospheric factors. The very strong inflow of air at the periphery
prevents any significant outward fire spread."4

Certain fundamental characteristics of a fire storm occur in any
fire, but are seldom of the right combination to produce a fire storm even
in large areas where a high density of combustibles exists. On a similar
scale is the familiar column of smoke and superheated gases that rise over
the fire, while air is drawn in at the sides. The difference is chiefly in
the volumetric scale of the inward rushing wind, and its speed. In either
case, total destruction is the result in the innermost area of the fire,

in the area of most complete combustion. The "front," as it were, is

 

3George H. Tryon, Editor, Fire Protection Handbook, Twelfth Edition,
(Boston: National Fire Protection Association, 1962), pp. 1-56, 57.

4National Academy of Sciences - National Research Council, A Study
of Fire Problems, (Washington: The Academy, 1961), p. 24.

 

concentrated on the rising column of flame and hot air. Although certain
buildings or areas may escape destruction in a conflagration (due to con-
vection currents, fire barriers, or fire fighting efforts), near complete
destruction is the result of the fire storm.

One of the major factors necessary for the creation of a fire
storm is the simultaneous ignition of many fires, and their joining. This
requirement was used during World War II on several German and Japanese
cities. In Dresden, Hamburg, Leipzig and a few other cities, fire storms
occurred with great loss of life. These fire storms were the result of
the merging of thousands of individual building fires. They were started
by saturation incendiary bombing with a mix of high explosives to "Open
up" the structures to fire bombs.

In one of these fire storms, the heavily built-up areas of the city
were blanketed with a high density of incendiary bombs and high explosives.
Within minutes of the first attack with these weapons, roughly two-thirds
of the structural units within a four and one-half square mile area were
burning. This entire area develOped as one mass fire. A vertical heat
column develOped over the central area in the absence of a strong ground
wind. This thermal column was estimated to have attained a height of more
than two and one-half miles with a diameter of one and one-half miles.

The rapid upward movement of "superheated smoke and burning vapors in the
thermal column induced strong indrafts of air from around the entire perim-
eter of the burning area. Streets entering the burning area became air
intake channels. The inrush of air through these channels assumed gale-
like prOportions."5 Within forty-eight hours, complete destruction was

accomplished.

 

5Lloyd Layman, First Fire Reseagch Coppelagion Conference. (wash-
ington: National Academy of Sciences - National Research Council, 1957),
p. 8.

--,"l

.U

---'.-
o

e
v--‘
r

...-r

a“.
..
.__'-
-_
- T:
“'~.~
I

I'.'
" T
“:V .
. ‘-v
.3d.:

‘.
F'.‘

h.‘

I‘-
4:

9%
v.1

J

I
a“

 

.l-
A
'h
N.»

Control of Mass Fire

 

There are several essential differences between peacetime and
wartime mass fires. These shall be discussed briefly later. Their simi-
larity rests upon one point, however. That is their effect upon fire
fighting capability which is geared to fighting normal fire problems.
Prominent fire-research groups have concurred in the philosophy that fire—
control action which can minimize the effects of these disasters involves
concepts inherent in most fire control practices today, but with added
emphasis on pre-planning and hazard reduction.

Such pre-planning might be directed along lines involving "(1)
the reduction of the number of potential ignitions, (2) the provision for
isolation or rapid extinguishment of fire starts to prevent formation of
serious fires, and (3) the minimization of fire-spread potential should
large-scale fires be produced."6 It should readily be apparent that in
situations in which equipment and/or manpower is incapacitated or insuf-
ficient to meet all fire starts, the form of the urban area itself may
become the most effective deterrent to further fire spread. Not only
should it be possible to reduce the probability of large-scale fire Spread
through develOpment principles, it is conceivable that once fires threaten
to Spread, their effects might be minimized through the passive nature of
urban design.

The mass-fire problem can certainly not exist in all urban areas.
We shall later examine the reasons why. What is important to realize
from the very beginning, however, is that the physical urban form, as shaped
by natural land forms and man-made development, is a prime determinant of

the relative danger of mass fire.

 

6National Academy of Sciences, A Study of Fire Problems, p. 24.

wO-t-

-
. .-
. y"
.‘Old

 

.0
.I‘

I
v

‘."_
...:

 

I
I...

 

 

 

CHAPTER II
THE URBAN MASS FIRE

The uncontrolled Spread of fire within human settlements is by no
means a recent phenomenon. The first recorded conflagrations and mass
fires occurred as far back as 2000 B.C., and reportedly destroyed Sodom
and Gomorrah. Troy was razed by fire, and prior to the Christian era,
Rome suffered six great fires. Rome's worst conflagration occurred in
64 A.D., and reportedly lasted seven days. London has likewise had its
share. As early as 1086, the city has been reported to have burned from
one end to the other. The Great Fire of London in 1666 destroyed 13,200
houses and many important buildings. That fire, which continued for
five days, was fought with little else than buckets, swabs, and little

two-quart hand squirts.7

Fire Loss

There can be little doubt that the factors which led to such large-
loss fires included the building methods of the day, and inadequate fire
control facilities. One would expect that the relatively SOphisticated
fire control equipment and building controls of today would make the
mass fire much less of a danger. Such a conclusion is contrary to fact.
Fire losses in the United States annually run in excess of $1.8 billion.
In 1966, losses included an estimated $1.5 billion damage to buildings
and contents, and another $.3 billion to vehicles, forests, and other

8

prOperty. Deaths from fire numbered 12,100. Much of this prOperty loss,

it must be noted, occurred in fires which can be classified as mass fires.

 

7Layman, p. 9.

8Percy Bugbee, "Fire Administration," The Municipal Yearbook 1967,
Orin F. Nolting, Ed., (Chicago: International City Managers Association,
1967), p. 374.

.
D

.3.

ecu—-

'UA‘

.p.
a
'5-

u.‘

 

.-

s‘.

 

{I

II.

During the period from 1926 to 1961 conflagrations and mass fires

in the United States and Canada caused destruction to "more than 6000

£2£l222223"9 Table B-1 in Appendix B, lists some selected mass fires and
conflagrations in the United States and Canada since 1910. Conflagrations
are still possible, especially so when the elements discussed in the previous
chapter occur in the correct proportions. The relative infrequent occur-
rence of the conflagration in prOportion to the total number of building
fires may tend to provide a false sense of security. It can be seen from
close examination of Table B-1 that many of the major fires occurred in
relatively small communities. There can be little doubt of their economic
and psychologic impact upon those areas.

This chapter is devoted to a study of the 25222 fire environment.
The reasons the fires in Table B-1 became conflagrations are therefore
important, for they shed valuable light on the means of controlling the
spread of mass fire. A study of the causes of spread of all the major
mass fires in the United States and Canada from 1900 to 1961 reveals the
major elements in the spread of these fires. Table 1 below lists these
for the periods 1901-1925, 1926-1961, and 1901-1961.

Note should be made, in examining fire loss tables, of elements
of inflation in monetary value, of increased fire reporting and recording
skill and accuracy, of changing definitions of loss typologies and cate-
gories, etc. The point to be taken, however, is that despite our techno-
logical improvements, fires are still capable of running uncontrolled and
destroying dozens of lives, hundreds of buildings and structures, thousands
of acres of forest land, and millions of dollars of real and personal

property.

 

9Tryon, pp. 1-58.

10

 

 

 

 

 

 

 

 

 

.00IH .0 ..a.m.0.z .«00H .coHuHom gumHoze .Hoonccmm coHuomuoua muHm

 

 

 

 

 

xoouu mo>flmonxo mo :ononxm
meadow suede swam

ucoeuummoo swam mo uncommon 3on
ousuuws EoHoans mo sowmonxm

moo Henson: omfiuaovwa mo soflmonxm

 

mmofloafion use: coflusummm> mun

oumx Renaud Eoum mooswn mowcwom
mmcfioHHSQ oofimuoo venoum mmmm
mmoaamdoo smog osmossumncoz
msmofluwdn .mooon .mxsovnuwsm
mcoflufiosoo Houses muo>om

common when some“:

money Hmnuo us Doosuusmoo swam
oofluoououm ouwm oum>awm ousovmomoH
mcHuanu muHm m>Huoommqu

wwmmdm nouns mo mwdawsm

osou essence swam among no phenom
wwm>oomflo swam :0 hmHmQ

mmmoom mowugmam swam voodoou ooflummmcou
EHmHs wow>0m ow moama

 

 

 

Hmnumo3 amp no no: wadedmoso
cofluomuoum oaanom mumovmomsH
soauoouowm wwdmomxm mo xosq

Emumam cowuonwuumflo umus3 musovmomcH
.o.m.8 on can» Housswm >ufioon> coax
moon mamcfinm poo:

 

 

 

000H 000H 000H R00 «0« asH
U “Null."
«.0 0.0 0 H H 0
«.0 0.0 0.0 H 0 H
«.0 0.0 0 H H 0
«.0 0.0 0 H H 0
«.0 0.0 0 H - H 0
0.0 0.0 0 « « 0
0.0 0.0 0 « « 0
5.0 0.0 «.H 0 H «
5.0 0.0 «.H 0 H «
0.0 H.H 0.0 v 0 H
H.H H.H «mm 0 0 «
0.H 0.H «.H 0 0 «
0.H H.H 0.« 0 0 0
0.« 0.0 0.0 0H 0 H
0.« 0.« 0.« 0H 0 0
0.« 0.« 0mM1. HH 0 0
e.« 0.0 0.H «H 0H «
H.0 0.0 0.« 0H 0H 0
H.0 0.0 0.« RH «H 0
H.0 0.0 0.« 0H 0H 0
0.0 0.0 0.« 0« «m. 0
0.0H v.0 0.0H 0v «« 0«
0.0H H.HH «.0H 50 0« 0H
0.«H «.«H 0.0H 00 «0 0«
«.0H 0.0H 0.«H «0 00 ««
0.0H «.0 0.0« 00 0« 00
H00H H00H 0«0H H00H HO0H 0«0H
on on on 0» on on
HO0H 0«0H HO0H H00H 0«0H H00H

 

Hence no cofiuwomoum

 

Houomm

 

w>Huonwwuooo needs no 009652

WEB ZH

coma mUZHm ¢Q¢Z¢U Qz¢ mmadfim QmBHZD

mZOHfidmwddmZOU OB OZHBDmHMBZOU mm090¢h AdeUZHmm

H mqmda

"mowoom

HNMfl'l-OKOBQOWO

 

":v:
v...'

0...
u

'l-‘U

song. I

In"

’I:
I”

[II

'F-\.
Jung

 

1.,

~V
l.“ ‘

 

.
.0
‘ ‘3.
s-
.g'
'0
n“ 7
IF 0.
' ~
‘0
.. .
“1‘“.
H
-
-.;- .
\ ~.
\.
.
\.‘
‘

11

Fire Spread Contributors

WOod shingle roofs have been the largest single contributor to
conflagratory fire Spread since 1900. However, the enforcement of building
codes, and general changes in construction methods in recent decades have
undoubtedly lessened the importance of this factor. When viewed as a
characteristic of a special "fuel," the flammability of the exposure surfaces
of a building or structure is of major concern--not only to the protection of
that structure-~but to the prevention of fire-brand ignition and creation.

It is important to note that the NFPA statistics show the contribu-
tion of wood shingle roofs to conflagration to be diminishing, while
the "lack of exposure protection" is shown to be increasing in importance
as a parameter. Exposure protection relates to the variables of distance
(from an exposed structure to a fire source) and construction material
and type on the exterior building surface. Both parameters pertain to fuel
characteristics, and both are subject to building restrictions and legal
standards.

Wind velocity is (Table B-2) a prime determinant in mass fire
spread. It would not seem reasonable to assume that wind speed has increased
over the years, but more likely that investigatory methods after mass fires
have become more thorough and scientific. The importance of wind in spreading
forest fires is well documented. The chief value of Tables 1, 8-2, and B-3
lie in the presentation of relatively stable elements in the spread of mass
fires. Those variables which have decreased within the last four decades
have been subject to public regulation and physical planning. As the public
services of a community are improved, the public safety likewise improves.
The improvement of the fire insurance based "rating and grading schedule"

for communities has allowed municipalities to compare their general "fire

12

environment" with that of other areas, and so improve along certain areas.
Failures in control practices have, therefore, decreased in importance.

We will later document the importance of "mutual aid" in achieving greater
fire fighting ability.

Factors not immediately subject to improvement (Table B-3), such
as weather conditions, have subsequently risen in importance as spread
parameters. In communities relatively isolated from other urban areas, the
danger of uncontrolled rural fire entering the town is of vital concern.
An example of such would include the destruction of isolated southern
California subdivisions by forest fire.

The congestion of large urban areas has frequently contributed to

the spread of fire by delaying the arrival of sufficient fire control

forces.10 In addition, the inaccessibility of buildings in highly develOped
areas has not only hampered control efforts, but increased the possibility
of transfer of fire by mass. Despite these conditions, however, a high
percentage of fires do not reach mass-fire proportions. Table B-4 presents
the average building fire loss as being approximately $1342 in 1967. The
annual loss per capita by fire amounts to approximately $5.90. This seems
a relatively small amount, unless you multiply it by 180 million persons.
The size of a community can be an important variable in the amount of fire
loss which might result. Quite often the generally higher fire control
capability of larger urban areas, and the greater reliability of water
supply are important fire loss factors. The per capita fire loss in 1967

for communities under 20,000 persons was $8.24, as compared to $5.90

 

10Such delay has not always been the result of congested auto traffic
and such as found under normal operating conditions. Under conditions of riot
and civil disorder, as existed in the Watts uprising in August, 1965, the presence
of rubble in the streets, burned automobiles, and the kindling and re-kindling
of numerous fires present monumental problems to fire services. - See H.F. Jarrett
Fire Data from the Watts Riot: Results of Preliminary Analysis and Evaluation,
(Santa Monica, California: System Development Corporation, May 10, 1966).

13

for those over 20,000 in size. Larger cities had fewer building fires

per 1000 population as well as a lower average building fire loss.

14

CHAPTER III

THE URBAN FIRE ENVIRONMENT

One of the prime determinants in the potential magnitude of a fire
is the size of the area vulnerable to the rapid spread of fire. In repeated
instances in past years, large-loss fires have involved extensive areas,
thus placing the distance fire will spread among the first parameters of

concern. The horizontal distance of possible fire spread is not the only

 

dimension of concern. The entire fire envelope, a gppig_space, must be
considered. This cubic space is most frequently bounded by structural members,
as in a building fire, but in the case of fires which spread beyond a closed
environment, the stratification of atmospheric conditions might conceivably

create an effective wall or ceiling for the fire envelope.

Fire Volume and Area

In the "usual" instance of an urban fire, active control measures
will be brought to bear. The effectiveness of fire fighting methods and
proper use of extinguishing agents are important factors in controlling the
spread and volume of fire. The exposure planes of a fire when considered
as a volume or cubic mass, rather than as a strictly linear function, carry
geometrical growth implications relating to both distance and intensity of
heat radiation. An examination of large-loss fire statistics will indicate
that the majority of large-loss (implying an individual loss of $250,000
or more) structural fires may be expected to occupy more than 100,000
cubic feet in volume.11 Such a volume, when abstractedly considered as
a fire envelope, can be seen to present large exposure planes for both

radiative spread and control measures. In this vein, the National Fire

 

11Warren Y. Kimball, "Control of Large-Loss Fires," Fire Journal,
Vol. 62, No. 6, November, 1968, p. 73.

15

Protection Service has indicated that 25,000 cubic feet is approximately
the maximum volume in which fire can be readily controlled and extinguished
with manual application of fire hose streams. 12 The availability of
additional water for control has been of little effect, since the require-
ments of volumes greater than this size outweigh the ability of control
groups to provide effective application.

Table B-6, abstracted from the Kimball NFPA study, presents data
gathered on 396 large-loss fires in the United States in 1967. It can be
determined from the table that in excess of 78 per cent (243 of the 311
fires where area could be determined) of the large-loss fires resulted in
loss of greater than 80 per cent of the property involved. Of these, 23
per cent caused extensive damage to other areas as well. The 80 per cent
destruction mark was taken in this study to indicate whether or not the
fire fighting forces were able to control the fire in the immediate areas
of involvement. Significantly, Kimball states that the majority of fire
fighting effort in these fires was aimed at protecting the exposure faces

of adjacent areas. 13

Water Supply
A prime purpose of the application of water to fires concerns the
lessening and absorption of the heat of the fire area. A usually acceptable
standard of pgg_gallon per minute per 100 cubic feet of fire area is seen
necessary for efficient containment and cooling of fire by hose streams. 14

Kimball states, however, that it is not unusual for fire fighters in urban

areas, where supply and flow are ample, to apply water at the density rate

 

12Ibid.

13Ibid., p. 74.

14Ibid., p. 76.

16

of £225 or more gallons per minute per hundred cubic feet of involved fire

envelope. Under fire conditions in the presence Of large-volume, large-area

fires, as shown in Table B-7, the water application rate does not as a rule

even approximate the one gallon per minute per hundred cubic feet ratio.

In fact, in only 20.4 per cent of the cases reported by Kimball did the

water density equal or exceed the 1.00 g p m mark necessary for cooling

and extinguishing the fires. The efforts of fire fighters in such instances

must be limited to the protection of exposure faces, and the blocking of

further expansion. Reflection as to the large number of fires involving

in excess of 80 per cent loss (Table 7) would lead one to suspect that the

inability to deliver enough water to the right place at the right time is

crucial to the spread and size of major fires. Had more water been available,

fire fighters would in all likelihood have applied it at greater densities.
Several reasons for the failure or inadequacy of water supply under

heavy draught conditions are presented by Kimball.15 In more than 20 per cent

of the instances involving fires in prOperties protected by public 25’

private water systems, inadequate supply at the hydrants was listed as a

problem. These hydrant deficiencies were found to include four chief factors:

1

Excessive distance to hydrants from fire site
800 to 1500 feet or more

2 - Poor hydrant distribution

 

3 - Nearby hydrant fenced or locked

 

4 Inadequate hydrant connections

 

 

Fire hydrant inadequacy is by no means a new problem. Yet it

can be generally stated that lay Opinion holds that fire hydrants are

 

15
Ibid.

l7

"everywhere" and well distributed. There is ample evidence to indicate
that many recent shopping center and industrial development sites are not
as well protected as they might be. Kimball cites instances of long hose
lays necessitated by fenced or "protected" hydrants--lays of as much as
1500 feet.

"That delay (caused by the need for long hose lays)
was considered one of the principal reasons for the
loss of the plant. At another fire, in a large
university dairy barn, only one hydrant was provided
on a large water main. Although the hydrant was
used to capacity, the amount of water applied on the
fire was needlessly limited."

The importance of water supply, and adequate distribution of

 

hydrants cannot be overstressed. Table B-8 indicates, however, that
although inadequate water supply at hydrants is chief among reported supply
deficiencies, laxity of upkeep, inattention to safety precautions, and

other human factors are of recurring importance.

Mutual Aid

 

The majority of large-loss fires in 1967 were fought by more than

one fire department. The fire services' concept of mutual aid, with appli-

 

cation also tO civil defense is applied daily by large as well as small
communities to augment and bolster their physical and human fire resources.
This mutual aid from neighboring communities must be supported by a water
supply system at the fire scene capable of meeting the demands of multiple
units. Kimball states that it is clear that those responsible for the fore-
sight Of mutual aid plans frequently neglect to plan and provide an adequate
fire flow to implement their plan. The plan, which on paper seems adequate,
is sometimes inadequate in practice. In such cases, the result is a waste

of manpower and equipment.

 

161bid., p. 75.

18

Where water supplies are adequate, a good mutual aid system can
generally give a small community the emergency resources needed. Table B-9
presents the mutual aid response to the 1967 large-loss fires reported by
Kimball. In 54.5 per cent of the reported fire cases, mutual aid response
was used. Conversely, however, this ability of nearly every community to
gather resources much in excess of its normal capacity may be responsible
for a major problem at many large-loss fires. The ability to rapidly
mobilize pumping and tanking capacities may be a causal factor in the in-
adequate provision Of water storage and consequent flow.

The implications of an inadequate water supply and a greater
dependence on mutual aid are broad for mass-fire situations. Such mass-
fire situations may require wholesale protection of mile-square areas. The
protection of eXposure flanks under such conditions, even assuming ideal
atmOSpheric conditions, will necessarily involve a massive water supply 35
‘ggll_2§_a high level of mutual aid. The lessening of a community's ability
to provide either of these control elements will require a reliance upon some
other element of protection-~an element which will not vary with deployment
decisions but remain an integral part of the community's form.

We earlier spoke in terms of fuel needs, fuel density, and the like.
These basic fire needs may well present the Opportunity for the least
expensive and most reliable form of mass-fire protections I refer, of course,
to the living pattern, the develOpment pattern of the community. Adequate
building codes, together with effective inspection and strict enforcement
provide the basis for a good fire prevention program for the community.

The Spacing requirements of all types of structures have been researched and
studied quite carefully in terms of individual structures. What has not

been so carefully researched is the overall distribution of the urban "fuel"

19

in terms of its mass-fire capability. The density of development, and

the "strategic" separation requirements of buildings and structures, are
of importance both in terms of fuel spread and protection distribution.
One need not think in terms of nuclear holocaust to realize the importance

of fire protection unit deployment.

Fire Departments

The degree of overt fire protection afforded a city is not so much
a function of its pOpulation as of its density of pOpulation and the inten-
sity and value of its developed property. As a corollary statement, people
cause fires, reflecting the direct prOportionality of an area's fire
"frequency" to the density of its pOpulation. Within cities and among
cities, however, the pOpulation density varies widely. Among the cities
in the United States with population in excess of 500,000, the population
per square mile varies from 2,000 to 25,000 persons.

Unquestionably, population density is related to land use, as is
the 33123 of develOped property in monetary terms. Areas with intense
pOpulation densities, implying much built-up land, multi-story dwellings,
and buildings of all types present a potential fire hazard. In light of these
concerns, Kimball points out that

"the average resident population per fire department

pumper company in New York City is 37,200 persons and

per ladder truck company, 61,500. A first-alarm response

of four pumper and two-truck companies involves fire

companies protecting an average of approximately 15,000

persons. These fire companies cover an average of about
six square miles."17

Table 2, below, presents the relationship between fire department

coverage and pOpulation density for twenty of our most populous cities.

 

l7Warren Y. Kimball, "Population Density and Fire Company Distribu-
tion," Fire Journal, Vol. 59, No. 2, March, 1965, p. 39.

 

20

.H0I00 .0

 

.momH .noumz .HmCHDOU 000m :.coHuonfluumHQ hcsmaou swam com wuflmcoo coflumHsmom= .HHOQEflM .M nouns: "mowoom

 

 

 

 

 

 

 

 

0.00 00.0H 0 00«.« 00« 000 000H0 000
0.0« H0.0 0 000.« 0.00« 000 00HH00
0.0H 00.0 «H 000.« 000 000 0000000
0.0H 0H.0 0 00H.0 0HH 0«0 0000H00 302
0.«H 0«.0 0H 000.0 000 000.« 00H0000 00H
0.0 00.« 0H 000.0 0.00 000 0H00000
0.0 00.« 0« 000.0 00 000 H0000H00HO
0.0 0«.« H« 00H.0 H.H0 H00 000003HH2
0.H 00.H 0« 000.0H H.00 000 H0000000H0
0.0 00mm0---:;2-:;;a;mm-::. 00«.HH 00 000 000H0>0H0
0.0 00.« . __ 00. _ 000.HH 0.00H 000.H 0H00000
0.« 00.H «0 000.«H 0.00 000 muoerHmm
0.« «0.H 00 000.«H H0 000 0H00H .00
0.0 00.H H0 000.0H 0.H0 000 .O.0.00000H0003
0.« H0.H 00 000.0H 0.00 000 0H00000
0.H 00.H 00 000.0H 0.00 000 000000
0.0 00.H 00 000.0H 0.0«H 000.« 0H00H000HH00
0.« H0.H H0 000.0H 0.00 000 000H00000 000
0.0 00.H 00 000.0H 0«« H00.0 0000H00
0.« 00.H 00 000.0« 0H0 000.0 000» 302
hcsmsoo xooua mcsmEoo Henson 0002 .wm Hem 0H0: «mm Mom Ammawz smmv Aooov muwu
use moaaz .om Hem moawz .ow woufim mcaoaflom cowumaomom 0004 pong coflumanmom

mo .02 ommwobm

 

mme

NBHMZMQ ZOHBdADmOm OE MO¢MM>OU BZHZHmdme MMHm ho mHmmZOHBQAmm

N mqmda

 

21

Cities with population densities in excess of 12,000 persons per square mile
average well under 2 square miles per pumper company. In cities with
population densities under the 12,000 figure, the average coverage per
pumper company is well gyg£_2 square miles. A notable exception, Pittsburgh,
has topography requiring much closer coverage, resulting in a 1.04 square
mile per pumper company.

The distribution of truck companies is, of course, related to the
number of multi-story buildings and structures in a given area. The median
distribution of ladder truck companies for these twenty cities is one truck
company per 50,000 peOple and for each 4.3 square miles. Dallas has one
truck company for each 23.4 square miles and each 56,700 persons. Boston,
on the other hand, maintains one truck company for every 1.7 square miles,
and each 24,800 persons. Policy in Boston requires, however, that two
trucks respond to every first alarm building fire. This would imply then
that Boston's truck coverage is based on 3.4 square miles and 49,600
persons. 18 The determinant of the "doubling" here is undoubtedly the
degree of congestion and structural quality. Both these factors have been
previously mentioned as crucial in the distribution of fuels as well as

protection units.

Department Distribution Standards

The American Insurance Association (AIA), which succeeded the
National Board of Fire Underwriters (NBFU) in 1965, has revised the distri-
bution standards for fire companies. Prior to 1963, the response distances
were predicated upon the type of district (high-value of residential) to

be protected, but revised standards are now also dependent on the volume

 

18 _
Ibid., p. 40.

22

of water (fire flow) required. Because of general improvements in streets,
equipment and other factors, the standard response distances have also been

adjusted.

The standard response distances depend to a large degree upon the
ability of public water systems to supply sufficient volumes of water at
hydrants. The AIA's Safety and Engineering Division, in reappraisal of
standards, formulated a schedule for water system adequacy based on
"average" conditions in communities of varying sizes. These standard
required fire flows are presented in Table B—lO. It is deemed essential
by the AIA that in setting required fire flows consideration must be given
to the congestion of buildings and structural conditions in the district
under consideration. If conditions of these two variables warrant, the
required flows may be adjusted accordingly. The AIA does not ignore the

differences between residential and high-value (commercial and industrial)

districts, either.

"For districts other than residential, outside the high-
value district considered, the required fire flow shall
be considered on the basis of structural conditions and
the congestion of buildings and . . . the required duration
shall be as indicated for four to ten hours . . . For
residential districts, the required fire flow shall be
determined on the basis of structural conditions and
congestion of buildings. In districts with about one-
third the lots in a block built upon having buildings
of small area and low height, at least 500 gpm is
required; if the buildings are of larger area or higher,
up to 1000 gpm is required; where districts are more
closely built or the buildings consist of high-value
residence, apartments, tenements, dormitories, or
similar structures, 1500 to 3000 gpm is required, and
in densely built districts with three-story and higher
buildings, up to 6000 gpm is required."19

These capacities, as well as those required by Table 8-10, do
not ignore the concept of area served, however. Adequacy cannot be con-
sidered only a function of the total supply of water deliverable to a

district, but must also consider the number of hydrants available as outlets.

 

19James F. Casey, Editor, The Fire Chief's Handbook, Third Edition,
(New York: Reuben H. Donnelly Corporation, 1967), p. 66.

23

The use of excessively long hose lines reduces the water pressure as well
as limiting flexibility of control efforts. Prior to 1963, the NBFU conducted
a study on the placement of hydrants, in which they expressed a preferance
for the area-served distribution rather than the lineal spacing method.
The AIA adOpted these proposed figures for spacing as shown in Table 8-11.
Implicit in these figures is the variation based not only on community
size (reflecting the overall demands on the water system and the probability
of multiple fires) but on the degree of high-value development (which may
result in lowered water pressures and an inadequate fire flow) and the
physical congestion more likely to be found in higher level communities.
In addition, the use of automatic sprinklers in high-value commercial and
industrial structures is now commonplace. The AIA has recognized the supply
to these sprinklers can be taken from the municipal water mains without
providing for a secondary supply. This use of primary supplies may present
problems of system failure in older structures, necessitating changes
in supply procedures. Each such change affects the total effectiveness
of the water system.20 The importance of the fire flow available in any
response district has become a prime factor in pre-planning for fire actions
as well as the placement of pumper and truck companies.
The AIA has set the following standards for the distribution of
fire companies.21
"High-Value Districts. The standard response distance. . .
is now 15 miles for districts requiring fire flows less
than 4500 gpm. The standard distance is 1 mile for districts
requiring fire flows of 4500 gpm or greater but less than
9000 gpm. The standard distance is 3/4 mile for districts

requiring fire flows of 9000 gpm of more.

"The standard response distance for the first-due ladder
company. . . is 2 miles for districts requiring . . .

 

201010., p. 68.

211010., p. 71.

24

less than 4500 gpm. The standard . . . is 15 miles for
districts requiring . . . 4500 gpm. . . but less than
9000 gpm. The standard distance is 1 mile for districts
requiring fire flows of 9000 gpm or more.

"Residential Districts. The standard response distance
is 2 miles for the first-due engine company and 3 miles
for the first-due ladder company. For sparsely built
residential districts with buildings having an average
separation of 192_feet or more, the standard response
distance for both engine and ladder service is 4 miles.
For closely built residential districts requiring more
than 2000 gpm fire flow or having buildings three or
more stories in height, including tenement houses,
apartments or hotels, the standard response distances
are 1% miles for engine companies and 2 miles for
ladder companies, but are to be reduced to l and 15
miles, respectively, when the life hazard is above
normal."

 

Fire Environment Grading and Rating

The emphasis which we have placed on the provision of adequate
water supply and fire company distribution is not only of concern to pre-
fire planning. It is reflected strongly in the municipal fire grading and
rating system used by insurance companies in determining fire insurance
rates. The Standard Schedule for Grading Cities and Towns of the United
States with Reference to Their Fire Defenses and Physical Conditions is
administered by the AIA. It classifies communities on a relative basis.
Under its classification system, a Class 2 city is a better risk than a
Class 3 city, but poorer than a Class 1 city. Revision is periodic, and
generally thorough. Table 3, below, presents an abbreviated listing of the
items of concern to fire engineers.
Correlation to Community POpulation

For various reasons, the determination of classification for a
city has had strong correlation to the ranked size of that community.
Figure l purports to show a significant relationship between insurance clas-

sification and population growth. As a relatively small and unprotected

a-Z/N Znu- .~.<.Jh .._ — “Han/00.5.0 JAN.— — . 0 - Z. .2 5:.-.» ZFOW.~..~ v.10 0:50.170 0.0/«ah a h :U.¢r..h:.fl pt...- ..-.....:II.

II u...-I.<.-.

25

 

 

 

 

 

 

 

     

 

 

 

0H000000 00H0000 HmoHoHasz

.00 .m .0OmH ..ocH .MHHoccoo .m conoom 0x00» 302 .xoonocsm m.m00£0 muwh 0:9 .hmmmo .m nosed "mousom

0000 0000 0000 0000 000m 000m ooom coma OOOH oom .I.oe aucOHowumo

Home Hoov Homm Hoom HomN doom HomH HOOH 00m 0 _.I.zomm mo mucaom

OH 0 m 0 m m 0 m m H mmsqu mHm
msmm Hssz u . ooom wooa 0NH mamuoa
III III H mocomuo>wo
cofluoououm mus>0um + III III m mcofluwpcou 000m5000
monomomxm + on 00H 00 mcowuflocou Handboouum
museums + 0mm 00 m coflucm>mum swam
00000000O + 00« 00 0 030a 00H0HHsm
coHuoowumooo + omm wad mm ahead 000m
00H0000 0309 000H 000 00 0000u00000 mnHm
oona 00m mm mammsm nouns

.mucfiom OoHs> cwoocoo owmaoomm mo cumocou
oaooosom mcflusm uocoaoaumo o>wusaom maeuH mo Honasz mo some assocmw

 

 

mmadm m024MDmZH MMHh UHmHUmmm

02¢ ZOHB¢UHmHmm4AU AflmHUHZDZ m0 ZOHBHmOAZOO Q24 mHmmZOHfidflmMImmBZH

m mqmdfi

£15:

.
,H'IHI

. I . I. a .1 .I . 1 . 0

up". ”u" ”I a 0h 0 MI— .«m .0 h Saw I. .nlh
.ac a 0: .h‘ an- uhu A..." J.- p...- p\~ 0‘

1." u.“ up $0.9 .7:- V- c v 0 . - 0n- .0 0 as. 3

Q.“
_ .
\‘-.
b..':

26

town, a community is likely to be in Class 9 or 10. As a village with some
water supply and a fire department, it may find itself in Class 7 or 8.

As general improvements are made and municipal services are added, the
classification tends to advance toward Class 3. Table 4, below, presents
data for 1967, and would tend to support the general conclusions one might
draw from Figure 1. In 1967 only 18 cities (out of the 1542 reporting)

are in Class 2. No cities qualify for Class 1, chiefly because of
"deficiency points resulting from a failure of cities to provide specific

restrictions on building construction."22 A further element was "climatic"

conditions, which create greater fire spread risks and mobility problems.

It is important to remember that the variations in achievement.
by different communities as measured by the fire defense grading is as much
dependent on the city's financial and provision of public service status
as its physical improvements. Fire protection is still a community service,
and as such, is subject to variances in management, internal pricing and
allocation of resources, and changes in municipal policy. The point is
that the sudden demands of a large scale mass-fire emergency are much more
likely to be successfully met by a well-run fire protection system in a well-
graded city. In that the adequacy of the municipal water system and the
responsiveness of the fire agency represent nearly 65 percent of the grading
scheme, (Table 14), a high rating may be taken as a good approximation
of ability to meet the disaster.

As a counter to this point, however, it must be noted that the
"preparedness" parameter is only one of a dozen which determine the extent

of mass fire spread. The contingencies which inevitably result in any

 

22Orin F. Nolting and David S. Arnold, Editors, Municipal Year_
Book, 1967, (Chicago: International City Managers Association, 1967),
p. 392.

 

HIMWH<0~HJ .‘62 h .H.<V.-

RATING CLASS

27

Figure 1

Average Insurance Rating Class

BY

Community Size

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10 '
9 (H.
_— +—
8
c—qp—q Li». p—n
l———q.
7
6
c—II—q I—r Inn—1 h—q I'- ”-1
5
—I—_ Inn—n Hr h—I I'-I 4 .
4
--- -q> -u re PI h- -H
3 If,
—— Inn-1 F" I—q — I—q» L—II -— I—T
2
-I— r —- —-I III-- —II :11» L1 In.“
1
Under 500 31001 2001 4001 10,000 25,001 Over
500 to to to to to to 50,000
1000 2000 4000 10,000 25,000 50,000

POPULATION SIZE OF COMMUNITY

NOTE: CLASS 1 insurance rating reflects the highest possible
fire preparedness, CLASS 10 the lowest.

Source: The International City Managers Association, Municipal
Fire Administration, (Chicago: The Association, 1956),
p. 32.

5000

DEFICIENCY POINTS

28

 

.Nmm .m .0000 .0000050 .c00um0OOmm¢ 00000002 >000 00:00ussw0uc0 .00m0 .xoom 000» 0smwO0csz 0:9 ”0OHsom

 

 

 

mc0uuom0m

000 000 00H 00 0« H« 00.00H «00.H H0009

0 0 0 0 0 0 00.0 0 0H

H 0 0 0 0 0 0H.0 H 0

0«H H 0 0 0 0 00.H 0« 0

0HH 0H 0 H 0 0 00.0 00H 0

000 00 0H. H 0 0 00.0« 000 0

H0« 0«H 00 0 0 0 0«.0« 0«0 0

0HH 00H 00 00 0 H 0«.0« HH0 0

0« 00 00 00 H« 0H 00.HH 00H 0

0 0 0 H 0 0H 0«.H 0H «

0 0 0 0 0 0 00.0 0 H
000.0« 000.00 000.00H 000.00« 000.000. 000.000 000H0 0000 000H0 0000 000Ho

00 on on on o» w0>o 000000 :0 0mmus0ow0m :0 000000 00cmwomc0
000.0H 000.0« 000.00 000.00H 000.00« 000000000 0000

 

 

00000 an 000000 ms0uuom0m mo c000005mom

 

ZOHBdUHmemflqu MUZ¢MDmZH NMHM 0&808 OB GZHQMOOU¢ 000.00 mm>0 wWHBHU ho ZOHBDmHMBmHQ

v mamdfi

 

 

.,,._..,
:0-vm.

H...
0"
in. II- b

0,- csr
w u»-
—

 

‘V.
-v 0.“,

D v
.5! o 9'
ti. 5.

 

v.3.
0...
u;

. ,
- 31

 

I

~ I
:--. ‘ u
k...

 

29

emergency can easily offset the ability of a water supply to adapt to
large volume demands. The elements of community form and structure are
fully considered in the AIA fire defense grading schedule. When combined
with building law considerations, the structural conditions of a community
account for 18 per cent of the final municipal grading.
Structural Condition of the Community
The 14 per cent of the fire defense grading schedule allotted
to structural conditions is designed to be applied in a "district" wide
manner of study. In large cities, a separate grading may be applied to
each high-value district; in smaller cities, it is designed for application
to the principal commercial, high-value area. The important factors used
for the grading of this section include the size of the area or district
bounded by fire breaks or elements of the city's form which will act as
fire breaks, the widths of streets, and the accessibility of blocks.
Narrow streets, the inaccessibility of buildings, congestion of the district
and of individual blocks, poor general structural conditions and eXposures
from surrounding sections all increase the probability of sweeping fires.
Buildings of fire resistant construction, sprinklered brick buildings,
fire breaks, and fire barriers all tend to constrict a spreading fire.
Briefly, then, the important elements of the structural conditions
division of the AIA fire defense grading schedule are as follows:

1. The Area of the District. Since an undivided area increases

 

the probability of fire spread, breaks and barriers in the area are

important. To be recognized as substantial breaks the specific

 

breaks should have a total width of at least 150 feet. Such
breaks would include rivers, parks, streets, railroad tracks,

vacant land, railroad embankments, depressed or raised freeways,

30

and possibly groupings of mutually supporting fireproof

or sprinklered structures which effectively subdivide a
district. Since overall size of the subject district is
important, deficiency points are allocated for areas greater
than 10 acres in size. For example, a SO-acre contiguous dis-
trict is assessed 4 points, while a 400-acre contiguous district
is charged with 50 deficiency points.23

2. Street Widths in the District. The critical street width

 

dimension for grading purposes is 50 feet. Where buildings

are uniformly set back of the street line, the width of the
street may be "assumed" (in the schedule's terminology) as the
distance from building front to building front. Deficiency
points are charged on the basis of the percentage of total
street length within the district of width less than 50 feet.
For instance, "for each 10 per cent of total length 50 feet

wide or less, assess 5 points."24 Such factors as street

widths in high value districts may be seen as having mixed value,
especially when viewed in light of the wide range of construc-
tion materials possible, window exposure and the like. The
principle of the distance of separation of "fuel piles" is an
important one, however.

3. Accessibility of Blocks. A block "shall be considered
inaccessible if more than 50 per cent of the numbér of buildings
have only one side accessible from a street, alley, driveway,

or courtyard and other Open spaces readily accessible from the

 

tration,

23International City Managers Association, Municipal Fire Adminis-
(Chicago: The Association, 1956), p. 398.

24Irid.

31

street."25 The inability of fire control equipment to reach

an area of fire, or to set up a fire line to control a spreading
fire is crucial. The presence of closed "fuel piles" is a
definite invitation to spreading.

4. Area of District in Open Space. The mitigating value of

 

undeveloped land for fire breaks is again recognized here.

The deficiency point system is based upon the per cent of the
area of the district in streets and Open spaces which cannot

be built upon, including one half the width of the district-
bounding streets. This measure, of course, recognizes those
areas prohibited from development or structural use by codes

and ordinances. It assesses points on the basis of the schedule

listed below:

TABLE 5 PENALTY SCHEDULE

 

50 per cent Open or over 0 penalty points

40 per cent 20 penalty points

50 penalty points

30 per cent

l 20 per cent 90 penalty points

10 per cent 130 penalty points

(Source: Municipal Fire Administration, 1956, p. 399.)

 

i
___1

It can be seen that an extremely high value is placed on the degree

of Open space in a high-value district. This is, of course,

 

25110541., p. 399.

32

consistent with basic combustion principles. It would also

seem that the high degree of development of large central core
cities would, through this grading factor, prevent that core city
from achieving a Class 1 grading, even if its other elements rank
high in relation to other cities.

5. Per Cent of Block Area Built Upon. As a corollary to item

4, this factor attempts to assess deficiency points and award
credit points for the degree of development by blocks. An

example of the assessment scale is below:

TABLE 6 ASSESSMENT SCALE

 

Per Cent Built Upon

0 per cent = 140 credit points i

 

30 per cent 40 credit points

50 per cent = 0 points

.__ ~——--—“..-a———. -—-

80 per cent 35 deficiency points

100 per cent 70 deficiency points

(Source: Municipal Fire Administration, 1956, p. 399.)

 

 

 

 

Parking spaces, though a develOped use, are not considered as
built upon here. It can be seen that 50 per cent of an area

in developed use, and likewise 50 per cent in open space is

seen as the desirable norm by the standard grading schedule.
Obviously, the configuration of the space around the structure is

an important variable. The schedule implies, however, that

33

where less than 50 per cent of the block is built upon, the lack

of congestion is considered below "normal, and "credit" or
minus-deficiency points are allowed.

6. Height of Building_. The height of buildings influences not

 

only the fire environment of a district, but also that district's
classification as to density of population, intensity of use,

and level of value. In determining the number of points to be
charged for the character of the district, however, the schedule
bases its scale upon the previously measured quality of the Water
Supply and Fire Department. Undoubtedly, it has been determined
by fire engineers that the ability of a well-trained fire control
group to apply an adequate stream of water and other control
measures to upper stories of frame buildings and non-fireproof
structures and buildings will offset a certain amount of the
problems caused by these large buildings. Where water supply
and fire control fall below a certain level, points are assessed
on the basis of frame buildings 2 and 3 stories high, frame buil-
dings 4 stories and over, frame buildings 6 stories and over,
non-fireproof non-frame buildings over 2 and over 5 stories, and
for non-fireproof non-frame and semi-fireproof buildings over 6

26

stories. These assessments of deficiency points all attempt

to recognize the exposure problem of upper-story fires as well

as the difficulties encountered in fighting multi-story fires.

 

26
Ibid.

34

7. Frame Areas. A frame area "shall be construed as including

 

continuous frame buildings which do not have a separation equal
to 2 feet of clear space or a brick-filled wall with no Openings
in the wall.27 This portion of the schedule attempts to recognize
the extreme hazard of such frame areas as breeder areas as well
as easily ignited spread areas. The total lack of exposure
protection (either in terms of geometric distance or physical
non-flammable barrier) is emphasized not only in the grading
schedule section on building laws and fire prevention, but here
under structural conditions as well. It assumes much importance
for matters of mass-fire prOpagation and spread as well as for
individual control efforts. Points of deficiency are assigned
under this category in terms of the percentage of the district
in contiguous frame construction as well as the size of the area
in such construction.

8. Conflagration Breeding Blocks. Closely allied to item 7 ,
this evaluation section attempts to single out blocks within

the district which "have a hazard distinctly greater than normal
for the district, and are grouped; that is, the separating space
is less than 100 feet . . ."28 The system assesses 5 deficiency
points for each block in groups of two or more adjoining blocks.
Thus, blocks which are penalized under the frame construction,
section are quite likely to be additionally penalized here.

9. Expggure to the District. The fire engineer here attempts

to take into account the prevailing wind direction and the

 

27Ibid., p. 400.

ZBIbid.

35

prevalence of frame construction and wooden shingle roofs in

the eXposing sections. He considers each of the four exposure

sides or planes separately, and then assigns from 5 to 20 points

per side.
Environmental Conditions

It can be seen that the consideration of structural conditions
attempts to measure not only the type of materials and conditions existing,
but also the placement Of those materials throughout a district, and the
quantity of each condition in each block of the district. If up to date
and well done, it is an exceptionally comprehensive analysis of the fire
environment of the community as well as the municipalities' ability to
control and extinguish fire. Since two subject cities of possibly similar
construction in two different areas of the country may be exposed to
different climatic situations, however, it is necessary that a consideration
of various restricting climatic conditions be taken. The AIA fire engineers
determine climatic conditions or data for each municipality in four distinct
categories. These are:

l. The Frequengy of High Winds. The frequency of high winds as
well as their duration are important not only to the spread Of fire, but to
the humidity of an area as well. Certain areas are exposed to continual
or high winds for the major portion of the year, and so are penalized a
certain number of points.

2. Snowfall in Excess of 10 Inches per Month. The amount of
snowfall received can hamper fire control efforts by restricting control
apparatus access, damaging water supply, or causing structural failure and
consequent exposure of building contents to fire hazard. An allied hazard

is the severity of cold weather. The severity of a winter, as well as its

36

length can immobilize storage of water supplies as well as fire fighters.
It presents a very real problem in the freezing of water during fire fighting.

3. Hot Dry Weather. Hot dry weather is an exceptionally dangerous

 

atmospheric condition, especially in forested areas. The lack of humidity
in the air increases the danger of spontaneous combustion as well as reduces
the amount of humidity in fuels and makes them more susceptible to ignition.

4. Unusual or Exceptional Conditions. These conditions include

 

those disasters or natural phenomenon which are not measurable by the above
and which offset protection and increase the probability of fire starts.
These include the frequency by which forest fires may enter a city, the
probability of tornadoes, or hurricanes in the region which may result in
numerous fires and interruption of fire control mobility. Blizzards and
severe snow storms impede the fire Operations. Earthquakes may injure
buildings, rupture water supply mains and cause fires.
Divergence in Water Supply and Fire Control Capability

Finally, the NBFU (and now the AIA) have determined that the rela-
tive strength of the community water supply and the relative ability of the
fire department should n2£_differ by a significant amount, in order to insure
that a lack of supply or an inefficient control group will not be tolerated
by the community. Where the Class assigned to water supply (that is, Class 1
through 10) differs by more than 2 Classes from that assigned to the fire
department (that is, Class 1 through 10), there shall be added to the total
points of deficiency (of the community) a certain number of points varying
"29

with the amount of divergence between the classes of the two factors.

The table is represented below in part.

 

291bid., p. 401.

37

TABLE 7

CLASS DIVERGENCE PENALTY SCALE

 

-.-
-.-

 

 

-..— -H_.—-.._-~ ...

S Divergence in Class Additional Deficiency Points
2 o

i

g 3 45

l 4 90

E 5 150

i 8 420

’ 10 680

 

It can be seen that a community which develOped a sound water
supply and attempted unwisely to rely heavily on mutual aid agreements would
find itself penalized additional deficiency points. The purpose of the
standard fire defense grading system then would appear to be the insurance
of a balanced ability by the community to meet a fire disaster. As we
shall see, the public service orientation of much of the grading system must
be accompanied by a recognition that the physical form of the community is
equally as important in the control of massive Spreading fires. The grading
schedule considers the community structure, but places little emphasis upon
the overall form. This is perhaps understandable, for it is undoubtedly
difficult to establish parameters of physical form which are meaningful to
mass-fire control not to mention the determination of critical levels for
those parameters. The establishment of critical parameters requires an
examination of the basic prOperties of fire prOpagation and spread as well
as an understanding of the urban fuel environment. With such a basis, a

discussion of the parameters of mass fire can begin.

38

CHAPTER IV
MODELING URBAN MASS FIRE SPREAD

The most desirable form for a section on parameters would be a
simple listing, with modeled results attached indicating their respective
degrees of variability from the ideal. Such a form would require exten—
sive experimentation and observation. Much of this has taken place, yet
on a scaled version not yet sufficient for our purposes. We are forced,
therefore, to either undertake our own research, or to search the
literature for an examination of the critical parameters concerning mass-
fire prOpagation and spread. There is a small amount of literature avail-

able, and some of it is excellent.

Urban Vulnerability Model

Perhaps the most direct and applicable is a study carried out
at the United States Naval Radiological Defense Laboratory in San Francisco,
California.30 It not only presents a vast amount of scientifically ob-
tained research material, but presents conclusions and hypotheses in a
convincing and logical manner. The authors realize that entirely different
sets of parameters may dominate the fire environment under different
atmospheric, natural, and cultural conditions. It would be possible
to imagine large group fires which could jump all bounds of reality and
consume vast areas of an urban area. Such conditions might result under
certain types of nuclear thermal pulse and damage. Yet this is an extreme
case, in which the conditions would all favor unlimited expansion of the
blaze. Such an extreme must be considered less probable than other combina-

tions of parameters.

 

30R.H. Renner, S.B. Martin, and R.E. Jones, Parameters Governing
Urban Vulnerability to Fire From Nuclear Bursts, Phase I, (San Francisco, United
States Naval Radiological Defense Laboratory, June 30, 1966), pp.70-86.

39

A second type of extreme would involve the damage done by blast
from a nuclear weapon, or a phenomenally large area of damage done by an
earthquake. Under this extreme, though large amounts of fuel are exposed,
the conditional parameters are not found in the right combinations and
sequences, and massive fires which might be expected do not materialize.
The extreme might also occur in urban areas where the fuels are generally
"hardened," and although all other conditions seem favorable for fire
spread, the fuel is not suitable.

In both of these instances the extreme conditions of fuels and
ambient conditions affect the outcome of the fire not only as sensitive
parameters, but also by bringing about a "gross change in character of the
fire from that of the more usual range of values" of the parameters.31
Therefore, the result is likely to be an extensive reordering of the sensi—
tivity of the parameters to their critical combinations and interactions.

If we remove from consideration both these extremes of end result, the
outcome is liable to be, as Renner, Martin, and Jones determined, the

"cases where the final fire outcome is in large measure determined by the
magnitude of the primary fire."32 In fires resulting from nuclear blast,
this will bear direct relation to the area "flashed over" by initial thermal
radiation. In the non-nuclear case, with which we are concerned this will
not hold true since we are primarily involved with the spread of fire
beyond the primary ignition. This assumes that the end result fire was

born of a single fire, lost beyond control. In the event of non-nuclear

 

311bid., p. 73.

321bid., p. 74.

40

multiple ignitions, which might result from natural disasters which precede
and cause fire starts, this postulate of self limitation may be applicable.
Multiple initial fires may mark the perimeters of the fire, or merge and
consume the urban kindling fuels within their immediate range. The end
expression of the mass fire in such a case may have been well outlined by
the initial fires.
Basic Parameters

In the terms of the previous discussion on fuel geometry and
ambient atmospheric conditions, the parameters in the propagation of mass
fire might be stated as those below. They are listed in order of decreasing
sensitivity, that is, in the order of the decreasing magnitude of their
effect upon the furtherance or attenuation of mass fire.
1

1. Fuel Concentration

2. Size of Initial Fuel Area
Determine Occurrence

 

3. Initial Fire Density
4. Fuel Type i
5. Surface Wind

6. Distribution of Initial Fire

Determine Severity
7. Atmospheric Structure

1

8. Topography J

The first four of these parameters determine whether a mass fire
will occur, and "influence its magnitude of severity.“ The last four
"determine whether it will behave as a conflagration or a firestorm."33
These basic parameters must be "translated" to the urban environment,

however, to be of value to a substantive study. When the parameters are

thus translated, and applied to a specific type of fire, they are quite

 

33Ibid., p. 72.

41

likely to shift somewhat in order of importance and sensitivity. Renner,
et al., postulate that these parameters do indeed shift in emphasis varying
with the type of mass fire and its agent--namely conflagration, firestorm,
or spreading fires of conventional magnitude.34

In these three types of mass-burn there is little relationship
between the number of structures initially fired and the number ultimately
destroyed. Indeed, the extent of fire vulnerability of the area is a
function primarily of the spread and ultimate magnitude of the fire's
physics. Spreading fires of conventional magnitude may well be an early
stage of conflagrations or firestorms, dependent on the satisfaction of
several factors. The parameters for conventional fire Spread must be
satisfied, yet the conditions for mass fire spread must not be achieved.
The fuel configuration must evince a high concentration and contiguity of
buildings and fuels, combined with buildings and structures which have
combustible exteriors. Further, atmospheric conditions must be considered
as hazardous fire weather.

Conflagrations, as differentiated from spreading fires of conven-
tional magnitude, require a "high density of fuel loading, a brisk surface
wind, a large number of structures in (the) fire area simultaneously on fire,

"35

(and) a large fire area. The Naval researchers estimated that surface

wind velocities would have to be greater than 8 miles per hour, with an
initial fire density of greater than 50 per cent of the structures in the
initial fire area on fire. They further postulate a required minimum initial

fire area of .5 square mile.36

 

34Ibid., p. 33.

351bid., p. 84.

36Ibid.

42

Firestorm start criteria include requirements for a high fuel
density, a low initial surface wind (as contrasted to the high wind required
for conflagrations), a large number of buildings and structures in the fire
area simultaneously on fire, and a large, roughly circular, fire area.

The surface wind would probably be below 8 miles per hour, with initial
fire density and fire area similar to that for conflagrations.

In all three cases we are concerned (in the urban area) with the
spread of fire either from structure to structure or from an exterior
fuel to a structure. The parameters governing these two types of spread
in urban areas vary to a slight degree in their order of importance and
sensitivity. In the instance of spreading fires of conventional magnitude
the parameters have been fairly clearly delineated by Renner, Martin, and
Jones. For example:

Structure to Structure Spread 37

l. The specific construction features of the structures are

the major parameter in this spread type. Of concern are the

number, size, and location of outer wall Openings. The combus-

tible nature of the outer coverings, both roof and side wall

as well as the roof type all affect the exposure criteria.

Further, the overall building dimensions and the shielding it

provides for interior fuels are elements of concern.

2. The degree of intensity of the immediate urban "sub-area"

is the second parameter of concern. This parameter must con-

sider building densities, height of structures, and the separa-

tion distances between structures.

 

37Ibid., pp. 79-80.

43

3. Fuel type is a vital parameter, determining the general
cOmbustibility of an area. Under this parameter are grouped
concerns with the density, size, age, composition and other
factors governing the ignition thresholds, burning times, and
heat concentration and radiative properties of the structural
fuels.

4. The building fuel load reflects not only the quality of

the structure, but the general quantity and type of interior
fuels.

5. Of a non-structural nature, the configuration and intensity
of the initial fired-structure is a vital factor. The place-
ment of the ignitors determines the path the fires will follow,
and hence the specific extent of the ultimate burn.

6. The moisture content of urban fuels is chiefly determined
by atmospheric quality. The atmospheric measures of relative
humidity, recency of precipitation, and general air temperature
are all elements determining the general susceptibility of fuels
to flame.

7. The wind speed and direction not only determine the area
of final burn but the rapidity of spread and the duration of
the fires as well. The ability of winds to cause "jumping"
fires is crucial in this type of fire.

8. The number, geometry, weight and life times of firebrands
are affected both by fuel quality and wind geometry. If mass—
transfer of fire occurs, the likelihood of jumping fires is
very great. Further, the susceptibility of structures to firebrands

is an important element.

 

 

JNwImh .mm .uamzmb ..0 000:0 .000050 0000052
5000 0000 O0 h00000000:05> :0000 m:0cuu>00 0000060000

.mGCOh. fin.“ UCM sGHHHMZ .mom sHQCGmm cm.“ «mugom

00:0000020 0000000002 no 0:02 “Iv

00:0000050 000050000 00 0:0xcmm 000:00: A00 "0902

 

44

 

I]:

 

 

I I «H I 0000Hom an 00000 00 0000Hm000“
I I «0 I 0005005000 :0 0:0:050 000000xm 00 :0000000m
I I 00 I 0:0000x005omw
I I m I 00:0:000 005005000 00 00000 000000xm mo >00s0xoumm
I I b I :000000000000 0:0 0000065: 0>00000m
I I o I 00050 000000xm mo 0000000000000£0 0:0000000
I I N I 0mhe 0:0 0000 0050 000000xm
I m 0 I 0005005000 00 00050 000000xm mo >00E0xonm
HH 0H 0H 0H . 0000000000
I 00 O0 00 00050 00.:0000ucou 0:00080.
O0 0 I 00 :0005000000O 0050
I 0H 0H 0H 00>oo 3000
0 00 I «0 00000 0>on¢ 00000 00 300m
0 0 0 HH 0000000000 0o 00000000000oo 000 00003
I m I O0 00:0000000 00 00000000000050 0005005000
0 m0 I m 00:0:000 scum 0w500m mo 0&030 0:0 000000
I N I 0 00:0000000 no 00005000 0:0 000252“
0 0 0 0 000000000 000 00000 0003 M
m0 O0 I 0 00050 00 0:00:00 00500002
m 0 I 0 00000 00000c0 mo h000:00:0 0:0 :000005000:00
0 0 I 0 0000 H000 0000H000.
0 00 I 0 090.0 0050
H HH 0 m 0000I050 00000 00 0000050H000”
m 0 m 0 0005005000 00 00050000 :00005000:00~
mc0xcmm Imc0xcmm mc0xcmm m:0x:0m M
000000 000umw 005005000 005005000 0000000000 “
000000000 :000000000:00 O0 00 m
000000xm 005005000 M

 

0000mm 00:000:0>:00.

mmmHm Hm memDOmmm m0 ZOHBUDMBme mma 02Hzmm>00 mmmafl2¢M€m
mom WUZ<BMOAZH m0 UZHMde

m Namdfi

45

9. The width and configuration of firebreaks is not as important

in this fire type as in the others. Since the fires do not

join and are of conventional magnitude fire breaks are not put

to the true test.

The spread of fire from exterior fuels to structures, and in the
instances of conflagrations and firestorms involve essentially the same
parameters as above, but in different orders. Table 8 attempts to list
the various parameters in order. From examination of this table, one
can deduce that the parameters most likely to be of major importance in
the fire's spread, regardless of type of fire, are the construction features
of the buildings and structures of the area, the configuration and.intensity
of the initial fires, the intensity of "builtupness" of the urban sub-area,
the building fuel load, the wind speed and direction, and the fuel type.

We shall examine these parameters more closely.
Classifying the Urban Area

The application of specific parameters (which we have identified)
to an urban area requires that we are able to typify an urban sub—area
or district in some type of homogeneous classification. This is necessary
so that we may attempt to examine within each homogeneous area the degree
to which a parameter of physical form is adaptable. In other words, the
variance between different districts or geographical areas of a large city
or urban area may be large in all matters of socio-economic, cultural, and
natural assets. This is likewise true in terms of physical form, general
structural types, and general fire environment. A means of generalization
is then desirable and necessary to enable us to simplify descriptions of

differing portions of the urban area. Were we not to apply a general

46

descriptor to homogeneous areas of the physical community, we would be
.forced to consider the structural condition of each building and structure.
At this point in our discussion of parameters, and in view of the charac-
teristics of mass fire, such an individual treatment might well be avoided
for a generalization of conditions. Thus, each generalized sub-area can be
described as possessing certain general properties.

The Naval research team points out the applicability of techniques
of remote sensing to apply spot analyses to each area, and therefore
determining a distribution level or value for each of the parameters within
an analysis area.38

Urban planning has relied heavily upon broad classifications
of land usage in the past. This division has been done with specific goals
in mind, and with the understanding that constraints and conceptual limita-
tions to the system existed. Land use as a criteria for generalization
might be very useful, since data is not only readily available in most
instances, but the general conceptual divisions have considerable merit
in terms of the degree of fire risk possessed by each type of use. Fire
engineers and experts have used the basic land use classifications in the
past to examine the conventional fire danger in different areas of study
cities.

Forest ServiceAApproach. F.M. Sauer, Craig C. Chandler, and

 

Keith Arnold, in a 1953 USDA study of post attack fire ignitions, divided

risks into two broad classes as shown:39

 

3811301., p. 90.

39F. M. Sauer, Craig C. Chandler, and Keith Arnold, Primaringni-
tions Following Atomic Attack on Urban Targets, (washington: United States
Department of Agriculture, 1953).

 

47

Class 1 - Residential

Within this class, quality of housing as well as construc-
tion features were used to rate the fire risk. Elements of
congestion, fire breeding, and general inhabitant housekeeping
evidently led to the three sub-categories of Good, Poor, and

Slum.

Class 2 - Commercial

This broad category included a range of commercial-
industrial enterprises, defining risk areas not so much by

their general housekeeping features but by the known or sus-

pected degree of fire risk or explosiveness of the interior

fuels and the structural distribution of the establishments.

The sub-classification included specific risk levels for

Large Manufacturing, Small Manufacturing, Wholesale Distri-

bution, Downtown Retail Distribution, Neighborhood Retail

Distribution, and Waterfront land uses.

This system used a mixed and perhaps not-too-thorough categoriza-
tion, using modes of qualification as well as quantification.

A second group of fire experts devised a system which appears to
have gone too far towards generalization, in that it attempts to categorize
land uses in terms which may lead to problems of application. Chandler,
Story, and Tangren apply the five categories of Light Residential, Heavy
Residential, Comercial, City Center, and Massive Manufacturing to the

entire urban area.40 Clearly there is a duplication of categories, such

 

40
Craig C. Chandler, Theodore G. Storey, and Charles D. Tangren,
Prediction of Fire Spread Following Nuclear Explosions, (Berkeley: Pacific
SW Forest and Range Experiment Station, United States Forest Service, 1963),
p. 29.

48

as between Commercial and City Center. Such a classification seems to

have been influenced by such theories of urban development as the concentric
and sector notions, and vastly overgeneralizes. The three authors apply
burning speed data to the five categories and derive a set of possible
burning times.

Civil Defense ApproaCh. The Office of Civil Defense, in Technical

 

Manual TM-B-l, presents a listing of the nine categories which they use in
determining fuel loading, estimating damage and loss, and estimating other

weapon effects essential to emergency planning.41 These categories are

based not only on differences as land uses, but also on potential for
emergency develOpment or as areas of high life or property loss. These
nine areas are:

Residential

Commercial

Industrial

Transportation

Storage

Institutional

Special

Recreational

Unused Land

These areas can each be thought of as broadly "homogeneous" in

terms of their fire vulnerability, relative degree of fuel loading, value
of recovery (that is, post-attack or post-disaster production recovery)

potential, and degree of open space available for emergency shelter and

 

41Federal Civil Defense Administration, Civil Defense Urban Analysis,
Technical Manual TM-B-l, (washington: FCDA, 1953).

49

fire-safe environment. It would be necessary, however, for mass fire study
purposes to evaluate sub-areas within each analysis land use district, since
the critical fire parameters will exist in varying mixes.

Study Approach. Conceptually, the importance of land use as a

 

criteria for assigning homogeneity cannot be denied. Yet other elements
must be considered; elements which might be used as sub-categories within
a land use system, or as a separate system for assigning homogeneity. Our
concern here is with the fire vulnerability of general area§_of the urban
scene, and as such we must consider the impact of fire upon adjacent, but
not necessarily homogeneous, land use areas. Residential and commercial
areas of nearly equal density and fire loading are much less likely to
present a fire load boundary. It would be relatively easy to derive a set
of subjective criteria for homogeneity, yet the derivation of a comprehensive
set is much more difficult. Possible areas for consideration might include
distinctions between variables of:

Land Use (human activity).

Occupancy of Buildings

Density of Structures

Fireload per Unit Area

Economic Divisions

Population Divisions

Political Divisions

Ground Cover and Vegetation Cover

Insurance Rate Zones 42
The decision to use one criteria or another will be dependent primarily on

the sensitivity of each of the basic fire parameters to variable conditions

 

42Renner, p. 91.

50

within homogeneous areas of the urban scene. Any of the areas, or sub-
areas within the areas, will be termed homogeneous when they can be
described by identical or near identical values of the respective defined
fire-spread parameters. The Naval researchers postulate that the occurrence
of each parameter within a district may be described along a distribution
curve determined by spot analyses and surveys.43 The degree of homogeneity
is then relative to the variance of the parameters from a constant value

for that "type" of district.

Very little has been said up to this point about the suspected
size of these homogeneous areas. The size of each area would be determined
not only by the amount of continuous identical land use and fire load,
for example, but also by the existence of natural barriers and firebreaks.
Firebreaks might be considered to be streets, bodies of water, large vacant
areas, or very low-fuel-load areas. In other words, inconsistencies in
fuel loading over the urban area would in large measure create boundaries
to fire spread, and render impractical the designation of a fire area which
crossed such a boundary. The designation of fire zones, containing areas
of near equal or homogeneous mass fire risk will most frequently be
accomplished by essentially linear systems of open space, of varying width
and make up. This concept of fire zones, and basically homogeneous fire
risk districts, is neither new nor novel, yet it is not frequently applied
explicitly. Rather, its implementation is a side effect of much of urban
development. Existing examples of the past use of fire zones in legal

control of building development undoubtedly reflect concepts of fuel

 

431bid., p. 93.

51

concentration and fuel loading held at the time of construction. Whether
these concepts of the value of fuel loading variance are understood or
applied in legal tools today is questionable. Principles and methods for
the attenuation of conventional means of fire spread are successfully
applied in many areas, yet the difference in parameters of spread for
conflagrations and mass fires indicates a need to examine the measures held

effective for conventional fire spread.

52

Conflagration Potential Model

 

The potential for the spread of mass fire may be determined by an
application of the parameters for fire spread which we previously presented.
These parameters, when applied within a homogeneous area of study should
give an estimate of the potential mass-fire hazard for that area. Consi-
derable work has been done on this aspect of the mass—fire problem,
specifically by the aforementioned Renner, Martin, and Jones, and also by
Messers. Cohn, Almgren, and Curless of Gage Babcock and Associates. The
Gage Babcock report series, entitled A System for Local Assessment of the

Conflagration Potential of Urban Areas, seeks to develOp an empirical

 

system which will not only assess the conflagration potential but designate
fire breaks as well.44 Such a proposal is consistent with the objectives
of this paper. It will be remembered that the parameters of concern in the
spread of mass fire were as follows:
1 - The concentration of fuel, its density, its configuration,
its fuel load, and its nearness to other fuels.
2 — The size of the initial fire area, and its location in
reference to the SUpply of fuels.
3 - The initial fire density, and the configuration it possesses
in terms of continuity of fires.
4 - The fuel type of the initial fire as well as the fuel type
of the suspected end fire area.
5 - The surface wind, measured in terms of speed, direction in

relation to fuel, and relation to upper atmospheric movement.

 

443.0. Cohn, p. 30.

53

6 - The distribution of the initial fire, eSpecially in regards

to the area of the suspected end fire.

7 - The atmospheric conditions, in terms of humidity, recent

precipitation, prevailing upper strata winds, and climate.

8 — The topography, particularly as regards land forms, which

may act as channels for promotion of fire or barriers to it.

Basic Parameters

These parameters were evaluated in light of conventional, conflagra-

tory, and fire-storm types of fire, and from the list were chosen five

elements most sensitive to conditions of the conflagratory and fire-

storm fire environments. These five, in terms more equatable to the

urban environment are:

l-

The construction features of the buildings and structures.

This relates directly to the elements of fuel concentration
and type.

The configuration and intensity of the initial fires, which
is related not only to the reason for fire start, but also

to possible delays in alarm, or in control unit arrival at

the fire scene.

The intensity of the use of the area, or the density of the
buildings. This relates to the "land fuel load" as well

as to the element of "individual-fire" barriers and breaks.
Congestion and the existence of fire-breeders enter at this

level of consideration.

The building fuel load is important as distinguished from

the land fuel load expressed above. It determines not

only the original combustibility of the structure, but its

54

burning time, probability for creating fire brands and general
fuel level.

4 - The wind speed and direction not only determine in large
part whether a mass fire will develop as a conflagration or
firestorm, but also helps determine the general extent of
the end fire.

5 - The fuel type determines the general combustibility of an
area and its resistance to ignition. The fuel type may well
be important in the designation of certain areas as fire
barriers or fire breeders.

The majority of these factors (items 1, 3, and 5) are concerned
with the day-to-day condition of the study area, and can as such be
termed "static" factors. Since changes in material and structure quite
likely occur, these elements are not truly static, yet will be so termed
for our purposes. As variables which may be more easily measured they
may be used to determine a conflagration or mass-fire hazard. The study
by Cohn, Almgren and Curless examines these three factors and arriVes at
a schedule as shown in Table 9 which incorporates values for the
Occupancy Fire Load, the Density of Building Coverage, the Height of
Structures, and Construction of Floor, Roof, and Exterior walls, and a
Slope Multiplier.

Hazard Value Per Block

The result of this series of value assignments for hazard is a
. model which measures the relative hazard value per block. The model (and
its value scale) is basically representative of the AIA fire defense

grading schedule, yet it does present the concept in a different manner.

(A)

(B)

(C)

(D)

55

TABLE 9

VARIABLES FOR CALCULATING MASS FIRE POTENTIAL

NUMERICAL VALUES FOR HAZARD BY BLOCK

 

Scale
Occupancy Fire Load Values
1 - Vacant or Non-Combustible- O
2 - Light Fire Load 10
3 - Moderate Fire Load 20
4 - Heavy Fire Load 30
Scale
Multiplier
Density of Building Coverage_ Values

 

0 to 5 per cent
6 to 20 per cent
21 to 30 per cent
31 to 40 per cent
41 to 50 per cent

Ground area
covered by
buildings or

HOOOOOOO
OkDCDOIb-Nl-‘O

 

 

51 to 60 per cent . combustible

61 to 70 per cent . storage

Over 70 per cent .
Scale
Multiplier

Height of Buildings Values

Vacant Land 0 Approximately

1 to 25 stories 1 each 12 ft. in

3 to 5 stories 2 height counts as

Over 5 stories 3 one story = l

' additional point

Scale

Floor Construction Values

Fire Resistive or Non—Combustible 0

One or more Floors all or Partially
Combustible lO

56

"Table 9 continued"

 

 

Scale
(E) Exterior wall Surface Values
Standard Masonry O
Substandard Masonry 10
Non-Combustible 10
Non-Combustible on Combustible
Supports 15
Combustible 30
Scale
(F) Roof Construction Values
Protected Non-Combustible O
Unprotected Non-Combustible 10

Non-Combustible on Combustible
Supports 20

Combustible 30
Scale
Multiplier
(G) Terrain Multiplier Slope Values

 

10 per cent or less 1
11 to 20 per cent 1
21 to 40 per cent 1.
41 to 60 per cent 1
61 to 80 per cent 1
Over 80 per cent 2

Source: B.M. Cohn, L.E. Almgren, and M. Curless, A System for
Local Assessment of the Conflagration Potential of Urban Areas,
Chicago: Gage Babcock and Assoc., Inc., 1965, pp. A24 - A28.

 

 

57

The model requires the estimation of a good many qualitative factors

(such as light, moderate or heavy fire occupancy load). The value for
density of buildings or construction is intended as a multiplier, giving
density, in effect, the highest weighing value of the factors. Such an
assignment is consistent with all the concepts we have examined. Symboli-
cally, the model as presented by Cohn, Almgren and Curless takes the

following form:45

((O+F+W)H +R_P__. . D .T =
M )100 M M R

is the occupancy fire load value

is the Floor Construction value

is the Exterior wall Construction value
is the Height Multiplier value

where

23:712'110

is the Roof Construction value
is the Construction Density Percentage
is the Density Multiplier

CD

is the Terrain Multiplier
and R is the Final Block Rating

*3
3

Fire Loading by Use

The authors of this model go to considerable detail in cXplaining
their classifications of such qualitative terms as negligible, light,
moderate, and heavy (as found in part A, Table 9). Indeed, an understanding
of various fire loads typically found with certain land uses may be
extremely helpful to us here. A negligible fire occupancy load rating
usually implies a vacant area, or a structure with essentially non-combustible

contents. Examples would include:

 

451bid., p. 31.

58

“Machine Shops and Metal Working

Storage of Metal Implements of Machinery
Boiler Houses, Power Houses

Brick Storage, Stone Crushing

water Treatment and Sewage Disposal Plants

"46
Light fire loading would commonly include these land uses:
Apartments, Houses, Hotels, Hospitals,
Libraries (metal shelving),'Telephone Exchanges,
Funeral Parlors, Halls, Gymnasiums,
Schools, Laboratories, Offices, Police and Fire Stations
Moderate fire loading is characteristically found in these land
use types:
Bowling Alleys, Theaters, Churches, Restaurants
Department Stores, Retail Stores and Shops
Drug Stores, Laundry and Dry Cleaning Shops
Most Manufacturing Plants
High fire occupancy loading is typified by the authors as including:
"Aircraft Hangars, Petroleum Refineries,
Whiskey Warehouses, Paint Factories, Stock Yards
Rubber Tire Storage, Warehouses

Crowded Department Stores, Feed Mills
Textiles, Clothing and Mattress Manufacturing and Storage."

47
These categories, as assumed by the authors, are consistent with
the AIA fire defense schedule, as well as with the standards published by
the National Fire Protection Association (NFPA).
Relative Potential Rating
The model for measuring the conflagration potential of each block
is valuable primarily because it allows a fire hazard evaluator to rate

areas of the city relative to their location and degree of hazard. The

application of such a model should provide a municipality with reasonable

 

46Ibid., p. A24.

47Ibid.

59

estimates of the relativegpotential for mass fire spread as well as areas

 

of confinement. The model is intended to be used as a block by block
measuring device, yet could be applied in a sampling design to a large,

relatively "homogeneous" land use area as discussed above. In any case it

 

is important to remember that such subjective evaluatory methods should be
viewed in a relative manner, comparing block ratings to other blocks.
Cohn, et al., caution that the ratings should not be viewed "as individual
absolute measures. On the basis of the ratings alone it cannot be predicted,
for instance, which or how many buildings in a specific block might be des-
troyed in a conflagration, nor can the ratings be used to predict the actual
extent of a conflagration which might occur some day in the future."48
The model ideally identifies potential, and rates severity of mass fire
hazard. Since the comparisons within the model are primarily on the basis
of the fire loading evaluation, the model does not imply a measure of the
speed with which a fire might spread through the block. This shows instead
the "propensity of the block for simultaneous burning of several buildings,
i.e. a conflagration" or fire storm.49
The result of an application of this model to each block in a
community would be a rating for each block. Presented in map form, with
gradients applied for differing hazard ratings, the data would pictorially
represent different hazard potential areas. These relative assignments
of ratings, when coupled with assumptions of fixed climatic conditions
and standard atmospheric flows, could be grouped into ranges which would

broadly define the potential for mass fire. Cohn, Almgren, and Curless

applied the following breakdown in a general case:

 

481bid., p. A9.

49Ibid.

6O

”Blocks with ratings up to 20: §2_group fire or conflagra-
tion potential, but possibility of (conventional) fire spreading
to adjacent buildings.

Blocks with ratings of 21 to 40: Low potential for group fires
and conflagrations, but moderate to high probability of (con-
ventional) fire spreading to adjacent buildings.

Blocks with ratings of 41 to 70: Moderate potential for group
fires and conflagrations.

Blocks with ratings over 70: High potential for conflagration.

Blocks with ratings of 15 or less are classified as firebreaks'.‘50

In the determination of block ratings, it will be common to find
a wide variance in all the elements within each block. Averaging must,
of course, occur. The authors provide one example which may be valuable.
"Of the 1-2 story buildings in a certain block, the ground
floor area of 10 per cent is warehouses . . . (negligible fire
load); 40 per cent is apartment housing (light fire load);

30 per cent is in retail stores (moderate fire load): and
20 per cent is in lumber yard (heavy fire load).

10 per cent x 0 (rating) = 0
40 per cent x 10 = 400
30 per cent x 20 = 600
20 per cent x 30 = 600
TOTAL 1600
1600/100 = 16 = block rating

 

The average occupangy rating for that (evaggatory) portion of
the block (rating procedure) is 16 . . . "

This averaging routine is carried out for each of the elements of
the model, and might result in the example continued below. The buildings
and structures are divided into groups determined by height, with ratings
determined separately. The height-separate ratings are then accumulated

and grouped to derive the block rating.

 

5°Ibid., p. A10.

511bid., p. A23.

61

TABLE 10

BLOCK.RATING MODEL APPLICATION

 

 

 

 

 

Category 1-2 3-5 Over 5
Stories Stories Stories
(0) Occupancy Fire Loading l6 13 10
(W) Exterior Wall Construction 7 0 O
(F) Floor Construction 4 8 0
(HM) Height Multiplier (x) l 2 3
(R) Roof Construction 26 24 0
(D) Construction Density Percentage(x) 43 43 14
TOTALS 2279 2838 420
Total/100 23 28 4
Totals of Columns = 55
(DM) Density Multiplier (x) 0.8
(TM) Terrain Multiplier (x) 1.0
Result 44
(R) Final Block Rating = 44
(() + w + F) HM + R) D/lOO ' DM . TM - = R
(55) (0.8) (1.0) = 44

Source: B.M. Cohn, L.E. Almgren, and N. Curless, A System for Local
Assessment of the Conflagration Potential of Urban Areas,
Chicago: Gage Babcock and Assoc., Inc., 1965, p. A. 33

 

 

62

The system is relatively straightforward. Since it describes the
block in terms of mass (fire loading, structural size, etc.) in relation
to the total space available, it could become an important tool for the
fire evaluator.

Application of the Model

The authors, Cohn, Almgren, and Curless, made no attempt to describe
a typical rating for varying land uses, yet they do imply that since the
various parameters do vary wi£h_variance in land use, the model will
impute generally basic differences in rating between land use types. A
few examples of blocks studied by the authors may be relevant.

Case 1: This block consists of single-family, one- and two-

story wood frame dwellings. The buildings cover approximately

25 per cent of the area of the block, and the terrain slope

is less than 10 per cent. The buildings have a light fire

load, and have combustible exterior walls, floors, and roofs.

The fire hazard rating for this block is $9! placing

it well below the group fire potential category, but not low

enough to quality as a fire break.

Case 2: This analysis area contains a high-rise residential

development, such as that found in an urban renewal area. The

high-rise units are integrated with scattered 2-story townhouse
developments. Construction is fire resistive in all cases.

The 2-story townhouses comprise a light fire load. The over-

S-story high-rise buildings have a light fire occupancy load,

a large number of Openings in the masonry walls and a fire

resistive roof and floor rating.

63

The final "ratings of blocks of this nature seldom exceed
10, reflecting their exceptionally low potential for a mass
fire."52 In addition, these blocks generally qualify as fire
breaks.
Case 3: This case involved an intensely built up block com-
posed of commercial and industrial uses. It has a density
of nearly 95 per cent. One- to two-story buildings occupy
about 10 per cent of the block area, with a generally moderate
fire load, masonry walls, and combustible floors and roofs.
The 3 to 5 story buildings are generally a moderate to heavy
fire load, with roofs non-combustible but unprotected. Two
6-story warehouses of heavy fire loading but non-combustible
protected roofs, floors and exterior walls cover 40 per cent
of the block area. The application of the model would yield
results of §_for the l to 2 story buildings, §2_for the 3 to
5 story, and 2§_for the over-S-story buildings, for a total
of 23_for the block. This places this block in the category
of high_potential for mass fire. 53

These examples would seem to indicate that the density of develop-

ment of individual blocks or analysis.areas plays an important part in

the determination of the rating. This is indeed a valid conclusion.

The determination of density figures varies widely in actual practice,

primarily around the pros and cons of including streets and alleys in

the calculations. Since streets vary widely in width, as do building

 

521bid., p. A38.

53Ibid., pp. A37-A39.

L1.

64

set back lines, and the calculation of barriers to fire spread is included
in the second portion of the fire assessment model, the determination of
total block area is taken to include that area within lot lines and to
include the alleys as an integral part of the block. Within this defined
block area buildings, combustible storage and parking lots are taken as
the basic elements for deriving a density figure.

Numerous special conditions can, of course, exist which do not
easily fall within the category of buildings for the determination of
fire load, construction and height. The storage of highly flammable materials
above or below ground presents a fire hazard rating much in excess of that
which is derived from the model. These blocks, whether used for above—
ground storage of flammable liquids or gases, subject to flooding by these
fuels, or for the storage of explosives are assigned a standard fire hazard

54 The concern for these areas is reflected in the intense

rating of 150.
research carried on by the NFPA in its preparation of standards.

Lumber yards are generally treated as combustible buildings, with
each 12 feet in height counted as a story. Automobile parking is considered
as a one-story building of light fire loading, unprotected non-combustible
walls, floors, and roof and of 40 per cent density. In addition, properties

with superior protection are treated as l-story buildings, fire resistive,

with a light fire loading.

Potential for Firebreaks
Up to this point we have discussed the potential for mass fire

within blocks or analysis areas. To relate these Specific units to the

 

54mm. , p. A30.

65

entire urban fire area, we must determine where these units might combine
to form contiguous rating groups, and where limits to these contiguous
groups might be found. The effectiveness of the limits for attenuating
differing degrees of fire size is a vital concern also. Our problem, then,
involves the determination of potential fire breaks and a model for "testing"
their effectiveness.
Area and Firebreak Typology
In the determination of potential fire breaks, we must turn to

our previous allocation of ratings of fire hazard for each block. These
ratings have been grouped in categories of:

Firebreaks

No Mass-fire potential

Low mass-fire potential

Moderate mass-fire potential

High mass-fire potential
In addition to these ratings assigned on a block—to-block basis, the
community is composed of numerous open spaces. Cohn, Almgren, and Curless
determined that open areas of a minimum width of 120 feet would be referred
to as potential primary fire breaks. These Open spaces could include
"freeways, railroad rights-of-way, rivers or bodies of water, parks,"
or private open Spaces.55 Thus, potential primary fire breaks might be
defined as permanently Open spaces not less than 120 feet wide, or as pre-

viously mentioned, blocks having a fire hazard rating of 15 or less.

These blocks must be a "minimum of 300 feet" in width. Cohn, et al.,

 

55Ibid., p. A41.

66

further define secondary fire breaks as being composed of Open spaces
not less than 80 feet in width or of blocks with a fire hazard rating
under 20 and no less than 300 feet wide.56

The basic strategy of the model begins by the mapping of all
seven of the above area types. This involves the creation of designated
multi-block areas for high potential hazard, and for moderate conflagration
potential. The addition of potential fire breaks to the mapping process
should result in a division of the community into separated, aggregate
areas of high and moderate mass-fire potential, divided by lines of open
space. The end goal is the creation of a series of fire-potential—free
paths running across the community, connecting with open spaces of 22:
pgtential blocks to isolate the high and moderate potential areas. The
result at this point will be a map delineating potential fire breaks, not
only of Open-space nature, but of no-fire potential blocks. These "strips"
will undoubtedly vary in width. The test for effectiveness of breaks need
only be applied to areas less than 500 feet in width, since in the Opinion
of Cohn, Almgren, and Curless this is the maximum width required.
Radiating Surface Area

The elemental parameter in the determination of the effectiveness
of a fire break is the radiating surface area. The shape of the radiating
surface, its orientation to the exposed subject fuel, as well as its
distance to that fuel are all crucial determinants of the critical para-
meter level. Much fire engineering is devoted to this subject, especially

in Fire Technology, a publication oriented specifically to the fire

 

56Ibid., p. 32

67

protection engineer.57 In the instance of building fires, with the conditions
of fire frontage facing a potential fire break, three surface shapes are of
concern. The square radiating area, the rectangular area, and the excep—
tionally long horizontal rectangular radiating area are each created by a
specific combination of the fuel array, the organization of window Openings
in the fire structure, the length of the building and the wind pattern.
The wind factor demands an element of adjustment, since average wind condi-
tions vary widely in communities. Cohn, Almgren, and Curless state that:

"The calculation of total radiating area, which is necessary

to find the required distance separation for an acceptable

fire break, is the most complicated step... in the method..."

Figure 2 presents a rough means of determining the required
exposure distance for an acceptable fire break. It is based upon the radi-‘
ating surface area of a fire under three levels of normal wind condition,
assuming the wind is blowing toward the exposed building. This is the
basic element in testing the effectiveness of a fire break under the
assumptions of conflagratory conditions along one edge of the break.

A set of physical variables are found in the analysis of the
exposed surface area as well as the fire surface area. Cohn, Almgren,

and Curless determined a series of critical levels for each of these

parameters and found those in the following list subject to simple

 

57One example of particular interest is an article by H.E. Anderson,

"Fire Spread and Flame Shape", Fire Technology, Vol. 4, No. 1, February, 1968,
pp. 51-57. The author states that "in laboratory tests, the size of a flame
front can significantly effect the rate of fire spread")p.51). He adopts an
analog approach in testing, and concludes that "we can expect a fire moving
up a 30° lepe to be about five times faster than one backing d8wn the slope.
Wind tunnel tests have shown that fires with flames inclined 30 from the
horizontal will have rates of spread nearly 18 times faster than those for
similar fires with no wind present." (p. 56).

 

531bid., p. 33.

‘RADIATING ABBA IN SQUARE FEET

68

Figure 2

Radiating Area vs. Exposure Distance

100,000

50,000)
40,000

30,000:

20,000"

10,000

5,000 3'1'
4,000-1‘

3,000 * .

2,000 “”YI

1,000 ” '

 

NOTE:
1 - Light Wind Conditions — °
2 - Average Wind Conditions- 0
3 - High Wind Conditions -— 0

Source: B.M. Cohn, L.E. Almgren, M. Curless, A System for'Iocal Assessment
of the Confla ation Potential "for 'Urban Area, (Chicago: Gage

Baboockand Associates, Inc., 1 ), p.’ 32.

   

()9

measurement or derivation. The analysis of these parameters is, of
course, applied to both sides of the firebreak. In simple statement,
the parameters are:

l - The Height of the Buildings

2 - The Amount of Wall Openings or Windows

3 - The Roof Construction of the Buildings

4 - The Set Back of the Buildings from the Exposure Lot Line

5 - The EXposure Width of the Building

6 - The Width of Vacant Areas.
The application of these elements to the buildings on either side Of a
potential fire break will form the first portion of a modeling sequence
to determine the effective radiative surface of that segment of analysis..
Structural Surface Model

The model develOped by Cohn, Almgren, and Curless uses a set of
multipliers based on a combination of the elements of roof construction,
height of building, and amount of wall Openings. Table 8-12 presents
this multiplier set for fire classifications by roof construction. A
comparison between the multiplier values for Class 1 fire resistive roofs and
Class 5 peaked wood roofs indicates the importance placed upon roof struc-
ture and fuel type as a measure in the attenuation of fire, and the consequent
restriction of radiative area.59 The specific multiplier which applies
to the building under analysis is applied to the building width to yield
the radiating area of a potential fire in that building. If the building
is set back from the lot line, a Factor of Contribution is applied to the
radiating area to modify for the effect of increased distance. The result
is the adjusted radiative area. In simplified notation, the model takes the

following form:

 

59Ibid., p. A46.

70

O (M), where (M'W'C) (W ) = A
' ‘Wiv' F

= Height of Building in Stories

Degree of Wall Openings

Roof Classification

Derived Multiplier (Table 8-12)

Building Width in Feet

Contribution Factor Derived from Set Back Distance
Width of Vacant Area in Feet

= Adjusted Radiative Area in Square Feet

where

><OZZNOZE
II

and

In addition to determining the area of the radiative surface, Cohn,
et al., found it important to determine the shape of the area as well.

This was accomplished by

E=F
W
where AF = Final Adjusted Radiating Area in Square Feet
W = Building Width in Feet
F = Average Flame Height in Feet
V

= Width of Vacant Areas in Feet

Apply W+V = a ratio on the order of
F
1 to 1.5 = square shape
1.6 to 8 = rectangular
Over 8 = long rectangular

This derivation of the area of the radiative surface as well as the
shape of the surface allows the evaluator to determine the required
effective fire break or exposure distance from Figure 2. In addition,’

Table 8-13 presents the material in Figure 2 in a tabular form.60

Several of the parameters of the model require examination at this

point. The classification of an exposure surface by the amount of Openings

 

60
Ibid., p. A51.

71

it presents is a very important element in evaluating the potential for
ignition. The degree of exposure given to interior fuel loads by window
Openings is the chief element in determining their probability of ignition
by radiative heat. The primary source of radiative heat is visible flame;
hence, the amount of flame which breaks out of its closed "interior" fuel
environment through windows and other wall Openings is the basis for our
concern with radiative area. Since buildings and structures vary widely
in the amount of Openings in their walls, a subjective classification is
perhaps the most workable for our evaluatory efforts. The report by Cohn,
Almgren, and Curless classifies Openings in the terms of none, few, average,
many, all. The determination is basically the estimation of what portion
of that wall will 2 E shield exposed buildings from radiative heat for any
length of time.

A classification of pppg_for "wall openings" implies a protected
masonry wall with no doors, windows or wood frame combustible elements.
Frontage walls with the standard NFPA (and fire hazard rating) classification
of non-combustible or combustible are classified as Ell_openings. This
is due to their suspected inability to stand for any prolonged period of
fire life.

Application Of the Model

In the previous section on the estimation of conflagratory hazard
within blocks, three cases or examples were discussed. Case 1 involved
a residential neighborhood of single-family, one- and two-family homes, of
wood-frame construction. These dwellings comprised 70 per cent of the
frontage on the 650 foot sample block, or 455 feet of building width.

Roofs were predominantly peaked to 15 feet, and of wood construction.

72

(Table 18). Since these dwellings are of wood construction, the wall
opening category is stated as all. By the application of the model,
the adjusted radiative fire area is found to be approximately 22,000
square feet, and of a long rectangular shape. In an area of average
normal velocity winds, the required separation distance (see Table 8-13)
is 265 feet.61
The crucial test of this procedure occurs in the comparison of the

derived figure for required effective separation distance and the actual

separation. If the required distance is at least equal to the actual dis-

 

tance, the Cohn, Almgren, and Curless report would designate that area
as a primary fire break. In all cases, the maximum required distance of any
structural type in a block is the distance used as the required "block"

distance. Where the actual separation is at least two-thirds of the

 

required distance for separation "that increment should be considered a

secondary fire break."62

The end result of the procedure embodied in this model is intended
to be a comprehensive examination of the micro-environment of mass fire--
or at least the potential for the group fire. Fire is itself a very tran-
sitory element; in that it is an emergency event, it is totally dynamic in
both its ignition and spread. The preparation for controlling mass dynamism
such as a mass fire must examine static elements as inputs to the spread.
As inputs, these elements are defined as parameters, and hence our involve-
ment with physical structure leads us to a required analysis of the arrange-
ment, concentration, and composition of structure as a fuel--a basic element

in the triangle of fire.

 

611bid., p. 335.

621bid., p. A50.

73

The methods for computation of block mass-fire hazard and determina-
tion of fire spread limits have considered the building as the major fuel
element in urban areas. This does not imply an ignorance of vegetative
fuels nor of volatile storage fuels, but rather a concern with the more
"usual" urban environment. We have discussed building density, height,
and volume, and briefly alluded to building arrangement and separation.
Building separation normally refers to the average distance separating the
walls of adjacent buildings. This separation generally decreases as
building density increases. The arrangement of these individual separations
is of considerable importance for the period immediately following first
ignition. If the initial fire spreads past this first separation, its
perimeter increases exponentially, presenting larger control fronts. Thus,
a description Of the arrangementof buildings within each block may be in
order.

The fire hazard models ignore an explicit parameter of arrangement
for two reasons. First, the explicit presentation of arrangement is found
implicitly in the parameters of percentage develOped and density. Second,
the conflagration limit model includes a provision for building set back
(hence frontage arrangement) as a multiplier value. An unstated assumption
of the models' formation may relate to the scale of the expected mass

fire and the reliance on block and areal concepts of spread.

74

CONCLUSIONS

The stated objective of this study was an examination of the
critical parameters in the spread of mass fires within an urban area. Such
an examination must be made in terms of the urban environment as well as
the fire research environment. The differences in scale of fuels as well
as fuel reactions to ambient conditions are inherent in modeling ventures,
and must be applied to the "real" area. If this is not possible realis-
tically, it must be accomplished symbolically.

The models which were examined in the fourth chapter were not an
explicit outgrowth of the specific parameters examined in Chapter Three.
The models were easily adaptable, however, because it was based on a theory
of fuel state and ambient dynamics. Because the models recognize the
unique character of each sub-area of each urban area and rely on basic
concepts of fuel loading and disbursement, they arrive, I believe, at a
very simplified, workable predictor of potential hazard. Their simplicity
cannot be used as a criticism of their comprehensiveness. A model is, by
definition, a simplification of a real state. In that the model purports
only to rate potential for disaster, it is a success. Attempts to discredit
such a modeling approach to study as having failed to predict the area of
spread or pattern of spread, fail in themselves by a lack of understanding
of the basic purpose of the study. That purpose is merely the identifica-
tion of potential.

Such an argument is not a weakening of a position of advocacy for

the model, but rather a round-about means at describing the necessarily

75

stochastic nature of non-deterministic modeling. It must be understood
that when research deals with a dynamic process, such a fire spread,

it must consider a random element of variability as inherent in all its
parameters. The element of probability must be considered not only in the
expression of potential for fire breeding, spread, and delimitation, but
also as implicit in the model's structure and parameter definition. To
attempt removal Of the stochastic element and substitute precise historical

conditions ipse dixit, in effect creates a deterministic model of a non-

 

certain situation. Such determinism, by relying on non-stochastic descrip-
tors of past elements, is forced to ignore parameters of the immediate
situation. The result is a model which steps beyond its stated constraint,
yet fails to achieve its explicit objective.

The outcome of the discussion concerning the parameters of the
model in Chapter Four was essentially a relation of the importance of the
fire spread parameters to the existing fire environment. The model is
basically expressed as a description of mass and space in a Special environ-
ment. Such terms are by no means strange to the urban planner and designer.
The treatment of mass as a fuel and space as a channel of heat flow are
somewhat alien to the planner, however. The designer is unaccustomed to
having his well-structured setting of buildings and space referred to as
fuel loads and fire breaks. Such a discrepancy in language has been the
precise target of this study. Unless the urban planner is inclined to an
understanding Of the problems faced by disaster experts, he may well be
working at cross purposes with them.

Disaster is not just a pre-occupation of men waiting-for-the
worst. A re-examination of Table 8-1 will impress upon the reader the

magnitude of loss by mass fire. Many of those cities mentioned had

76

excellent fire control teams, and perhaps acceptable conditions for the
normal attenuation of conventional fire spread. Yet something wanvwrong.

Planning the form of a community is especially serious in the
light of threatened disaster. Interest in mass-fire protection was intense
immediately following World War II. An important conclusion drawn from
the records and analysis of bombing surveys stated that the major destroyer
was fire; fire storm and conflagration. The advent of the nuclear weapon
has not changed this. Small fire breaks did not stop the Spread of mass fire.
Areas of lowbdensity buildings actually fed fires, and allowed conflagrations
to sweep across cities. The reaction after the war was a loud cry for the
acquisition of large, thousand—yard wide areas of open land throughout the
urban areas, set off to act as fire breaks. This meant the linking of existing
parks and reservations with each other and with water courses, preserving
hill barriers and hastening the clearance of areas for a very broad freeway
rights-Of-way. These elements are a part of the overall pattern which some
schools of planning have advocated for decades.

The problem is still based upon economics, however. It is difficult,
if not impossible, to convince pe0ple to pay for something which may not happen
at all, or to sell their land when it is returning a profit. This is no
new problem. The answer may well lie in the multi-use concept of land
development. Under such a concept, land is explicitly acquired for some
purpose, while creating benefits for an additional number of other purposes
which are more difficult to value or to "sell" to the public. An examina-
tion of the critical parameters of the previously derived models may lend

several clues to such multi-uses.

77

The need for more improved enforcement of building regulations
may be recognized in governmentally Spurred programs of industrialized
building methods, federal construction-loan requirements, review of archi-
tectural and Site design, and Special materials research programs. These
program types are not new; in fact, they are all Operative or proposed.
The degree of local involvement may be a crucial element. If responsible
planning-oriented officials within local government understand the needs
of mass-fire attenuation, and the areas of potential hazard within their
community, they may be able to creatively enforce and regulate programs.
This is, of course, an idealistic concept and perhaps oriented toward the
use of administrative law. The thought of well-directed, responsible,
comprehensively motivated planning on the part of local officials is
nonetheless attractive.

The use of open Space as a means of attenuating spread is a
fortunate by-product of much recent development. The creation of freeway
rights-of-way across urban areas has become a reality, yet these large
'divided' areas of the community are frequently without sufficiently
wide (150 feet minimum) areas which might qualify as fire breaks. The
institution of multiple-use Open Space is especially important in such
areas. A primary use of these areas is most likely to involve recreation
or tran3portation, yet it could just as easily involve purification lagoons
for water and sewage treatment plants, industrial "park" areas, shopping
center parking, airport or heli-port landing sites, or university campuses.
Whatever the primary use, its accomplishment satisfies a major parameter
in the halting of mass-fire spread.

Planning is of necessity a futuristically oriented activity, and

presents no assurances that expenditures today will be relevant in the

78

30- to 50-year future. The use of multiple-use tactics are, therefore,
all the more attractive in that they require relatively little explicit
expenditure for projects which are viewed by the public as of too small
a probable occurrence. The planning of yet undeveloped areas of the
community is no less vital an area of concern for the disaster planner.
During WOrld War II, numerous urban areas were the targets of
incendiary and high eXplosive bombs. One of the hardest hit cities was
Hamburg, Germany. It was subjected, in 1943, to a two-day firestorm of
astounding prOportions. The records of the fire protection service of
that city were preserved intact by Dr. Hans Brunswig, Fire Chief. In a
recent translation he states that
"For three decades, city planners and architects have
been discussing such ideas as 'relaxation,‘ the 'city in rural
surroundings,' and 'satellite cities.‘ Their goals are based
on sociological and hygenic grounds, and they have been able
to realize these ideas many times in the last ten years of
reconstruction. These trends are consistent with effective

fire protection, regardless of whether a fire originates
in war or peace."63

 

Dr. Brunswig undoubtedly has a great regard for the effect of passive
urban form on the Spread Of mass fire. In 1966, he observed with regret
and concern that many of the structural and land development preventative
measures "for achieving structural fire protection have not yet become self
evident conclusions...64 The importance of examining urban fire environ-
ments on a block basis was further emphasized, in his statement that

"If individual fires...are not prOperly fought..., or

if these forces are unable to effectively combat these fires
because of unfavorable structural and design characteristics,

 

63Dr. Hans Brunswig, Practical Experiences of Fire Protection
Services, Part 1, Trans. by Curtis E. Harvey and Wilham C. Truppner,
Institute for Defense Analyses, June, 1966), p. 134. (My emphasis).

64_
Ibid., p. 163.

79

the danger of the fusion of these fires into a row fire emerges.

Row fires in turn can create block, area, and city fires.,...

pushing fire fighting services from an offensive to a defensive

position."
The relevance of the loss of an offensive position in regard to spy.
emergency or disaster must be evident. The crucial part played by a pas-
sive element of design (an element which ggp_be controlled) must be
recognized, dealt with, and improved. The importance of an elemental
parameter is no less great because that parameter requires mpgh_public
coordination and action, nor because that parameter is critical to an
event which may never occur.

There can be little doubt that an urban planner trained in several
analytical fields of survey and technical research; devoted to the concepts
of public interest, welfare and safety; and well aware of the needs (the
intimate needs) of each public service operation of the community could
perform a vital service to that community. There are few such men available,
however. The demands upon a physical planner's time may well force him to
treat dispassionately all but his own physically-oriented land use plans.
The socially oriented planner may well reject totally all concern with land
use planning and effective privision of non-welfare oriented public services.
In the end, the only persons concerned with public safety from fire may be
fire safety personnel, the budget officer, and the insurance companies.
In effect, this Situation frequently exists today.

Planning's concern in the fire realm has been with the distribution

of fire houses, the passing (not enforcement) of building codes and zoning

ordinances, and the promotion of subdivision design techniques which ignore,

 

65
Ibid., p. 391. (My emphasis).

80

if anything, the total fire potential of the community. Safety standards,

to a large extent, have been depressed, ignored, or changed to effect the
same private interests which inevitably develop the community in uneconomical
as well as unsafe forms.

Several questions which were unanswered in this paper might be
asked here. Is zoning, as now practiced, an effective tool in grouping
land uses in non-hazardous configurations? Is it being used to its full
force? Are communities really getting their moneys worth when they allow
extensive, hard-to-service development? Why hasn‘t a strong relationship
developed between working urban planners and the technically oriented,
expert groups concerned with public finance, fire safety, water quality, and
the like? Can the planner, trained as a generalist, hOpe to effectively
communicate with Specialists in fire control, etc.? Is non-contact with
service bureaus of the city government in effect non-support and thence
opposition to thoSe agencies? Can the blame for non-enforcement of building
and housing codes be laid to the planner in part?

One area of concern is of course the financial arena. Many plan-
ning ventures of the past have been predicated wholly on the provision of
federal and state monies. Since financial concerns underwrite much of the
efforts toward data collection, it would seem only elemental that agencies
seek out all available funds. In the instance of some federal programs,
local planning agencies have entered into stage programs, and produced as a
final result yet another paper plan. The continuation of the planning
process, and perhaps even the imaginative expansion of the process have
not frequently occurred.

Cynically speaking, what has not transpired is the dedication of

planning effort to the provision of public safety from fire, with notable

81

exceptions. The question is why. The answer may well be that local planning
agencies fail to recognize that pre-planning for fire occurrence is not
merely a matter of deriving ratios of equipment to pOpulation, but a whole
series of needed enforcements of ordinances, codes, and statutes already

on the books and all regarding as primary tools of the urban planner.

 

The answer may well have to come from local fire officials who are able to
swing sufficient public attention toward the possibility of fire spread.
Scare tactics may well be abhorred, yet effective. In any instance, they

are much desired over the real thing.

82

BIBLIOGRAPHY

 

Anderson, H.E., "Fire Spread and Flame Shape," Fire Technology, Volume 4,
No. 1, February, 1968. '

Arnold, R. Keigh, "Mass-Fire" The Disaster We Face," Michigan Conservation,
Vol. 37, No. 4, July-August, 1968.

 

Assembly Interim Committee on Municipal and County Government, Fire
Grading and Ratipg, Volume 6, No. 14, 1959-61, Sacramento: State
of California, 1961.

Blunt, Keith R., J.L. Lifter, and P. VanDaam, Implementation of An Ordered
Urban Sprawl Configuration, (Los Angeles: Planning Research
Corporation , January 16, 1967.

 

Bond, Horatio, Editor, Fire and the Air war, (Boston: National Fire
Protection Association, 1946).

 

Brown, William M., Alternative Civil Defense Postures, (Harmon—on-Hudson,
New York: The Hudson Institute, June 11, 1964).

Brunswig, Dr. Hans, Practical Experiences of Fire Protection Services,
Part 1, Translated by C.E. Harvey and W.C. Truppner, (Institute
for Defense Analyses, June, 1966).

Casey, James F., Editor, The Fire Chiefs Handbook, Third Edition, (New York:
The Reuben H. Donnelly Corporation, 1967).

 

Chandler, Craig C., Theodore G. Storey, and Charles C. Tangren, Prediction
of Fire Spread Following Nuclear EXplosions , (Berkeley: Pacific
SW Forest and Range Experiment Station, United States Forest
Service, 1964).

 

Cohn, Bert M., The Conflagration Potential in San Jose and Albuquerque,
(Westchester, Illinois: Gage-Babcock and Associates, Inc.,
October, 1966).

 

Cohn, B.M., L.E. Almgren, and M. Curless, A System for the Local Assessment
of the Conflagration Potential of Urban Areas, (Chicago: Gage-
Babcock Associates, Inc., 1965).

 

Countryman, Clive M., Mass Fires and Fire Behavior, (Berkeley: Pacific SW
Forest and Range EXperiment Station, United States Forest Service,
1964).

Craig, David W., "Take A Fire Chief To Lunch," ASPO Newsletter, Vol. 23, No. 10,
November, 1967, (Chicago: American Society Of Planning Officials).

 

83

Federal Civil Defense Administration, Civil Defense Urban Analysis, Technical
Manual TM-8-l, Federal Civil Defense Administration, 1953.

Folkman, William 5., Residents of Butte County, California:. Their Knowledge
and Attitudes Regarding Forest Fire Prevention, (Berkeley: Pacific
SW Forest and Range EXperiment Station, United States Forest Service,
1965).

 

 

Grinberg, I.M., and A.A. Putnam, A Model Study of the Interaction Effects
of Mass Fires, (Columbus, Ohio: Batelle Memorial Institute,
September 28, 1964).

 

 

Haswell, D.B., D.W. Williams, and T.J. Cummings, A Disaster Planning Manual
for Small Communities, (Washington: Office of Civil Defense,
April, 1965).

 

 

., Small Community Fire Fighting Resources for Major
Fire Disasters, (washington: Office of Civil Defense, April, 1965).

 

 

 

Heyman, Daniel P., Least Cost Location of Fire Suppression Equipment,
(Berkeley: Operations Research Center, University of California,
March 9, 1964).

 

The International City Manager's Association, Municipal Fire Administration,
Sixth Edition, 1956, (Chicago: The Association, 1956).

 

Jarrett, H.F., Fire Data from the Watts Riot: Results of Preliminary
Analysis and Evaluation, (Santa Monica, California: System
Development Corporation, May 10, 1966).

 

 

Keller, J.A., A Study of World War II German Fire Fatalities, (Albuquerque,
New Mexico: The Dibewood Corporation, April, 1966).

 

Kimball, Warren Y., "Comparative Levels of Municipal Fire Fighting
Service," Fire Journal, Vol. 59, No. 4, July, 1965.

 

., "Control of Large-Loss Fires," Fire Journal, Vol. 62,
No. 6, November, 1968.

 

 

., Fire Attack, (Boston: National Fire Protection
Association, 1966).

 

., "POpulation Density and Fire Company Distribution,"
Fire Journal, Vol. 59, No. 2, March, 1965.

 

Lommasson, T.E., Preliminary Investigation of Firestorm Start-Criteria,
(Albuquerque, New Mexico: The Dibewood Corporation, June 15, 1965).

 

84

Miller, Carl F. and James F. Kerr, Field Notes on WOrld War II German
Fire Experience, (Menlo Park, California: Stanford Research
Institute, October, 1965).

 

Moulton, Robert 8., "Blind Spots in Building Codes," Fire Journal, Vol. 62,
NO. 2, July, 1968.

National Academy of Sciences., A Study of Fire Problems, Publication #949,
(washington: National Academy of Sciences, National Research
Council, 1961). '

., Fire Research-First Correlation Conference,
Publication 475, (Washington: National Academy of Sciences,
National Research Council, 1957).

 

., Methods of Studying Mass Fires, Publication 569,
(Washington: National Academy of Sciences, National Research

 

., The Use of Models in Fire Research, Publication 786,
(Washington: National Academy of Sciences, National Research
Council, 1961).

 

Nelson, Ralph M., The National Fire Danger Rating System: Derivation
of Spread Index for Eastern and Southern States, Research Paper
SE-13, (Asheville, N.C." Southeastern Forest Experiment Station,
United States Department of Agriculture, Forest Service, September,
1964).

 

Nolting, Orin F., and David S. Arnold, Editors, The Municipal Yearbook,
1967, (Chicago: The International City Managers Association, 1967).

Planning Research Corporation, Urban Form as a Passive Defense Variable,
(Los Angeles: Planning Research Corporation, January 16, 1967).

Pryor, A.J., and C.H. Yuill, Mass Fire Life Hazard, (San Antonio, Texas:
Southwest Research Institute, September, 1966).

 

Renner, R.E., S.B. Martin, and R.E. Jones, Parameters Governing Urban
Vulnerability to Fire from Nuclear Bursts, Phase I, (San Francisco:
United States Naval Radiological Defense Laboratory, June 30, 1966).

Salzberg, F., G.L. Maatman, and F.J. Vodvaska, An Approach to Trans-Attack
Fire Suppression in Urban Areas, (Washington: Office of Civil
Defense, March, 1964).

 

85

Sauer, F.M., Craig C. Chandler, and Arnold Keith, Primary Ignitions Following
Atomic Attack on Urban Targets, United Stated Department of
Agriculture, 1953).

 

 

Shelberg, W.E., and E.T. Tracy, Countermeasure Concepts for Use Against
Urban Mass Fires from Nuclear Wegpon Attack, (San Francisco: United
States Naval Radiological Defense Laboratory, August 14, 1967).

 

 

Shorter, G.W., "The Fire Protection Engineer and Modern Building Design,"
Fire Technology, Vol. 4, No. 3, August, 1968.

 

Stecher, Gilbert E., and Harry N. Lendall, Fire Protection and Prevention
Fundamentals, (Philadelphia: The Spectator, 1953).

 

 

Tryon, George H., Editor, Fire Protection Handbook, Twelfth Edition, 1962,
(Boston: National Fire Protection Association, 1962).

United States Forest Service, Winds Over Wildlands - Ayguide for Forest
Management, Agriculture Handbook No. 272, (Washington: United
States Department of Agriculture, 1964).

 

Woltman, H.T., and E.C. Goodrow, Civil Defense Implications of Current
and Future Urban Configurations, (Los Angeles: Planning Research
Corporation, December 20, 1963).

 

 

Woltman, H.R., Barbara Woodfill, Keith Blunt and James Dei Rossi, The
Economic Feasibility of Decentralized Metropolitan Regions, (Los
Angeles: Planning Research Corporation, October 29, 1965).

APPENDICES

THE MECHANICS OF FIRE
AND FIRE SPREAD

DATA AND SUPPLEMENTARY
INFORMATION

APPENDIX A
THE MECHANICS OF FIRE AND FIRE SPREAD

Research by private and public organizations concerned with fire
behavior and control has resulted in a large body of knowledge on the
behavior of fire in urban and forest environments. Much of this has been
contributed and sparked by the Forest Service of the Department of Agricul-
ture and the National Fire Protection Association, (NFPA), but the
experiences of municipal fire departments have also been invaluable.
Foresters and wildlife managers have come to regard fire as an essential
tool in certain stages of silviculture, in land improvement, and in fire
hazard pre-planning. Much of their experience is directly applicable,
or at least interpretable, to the urban situation.

Much of the available knowledge concerns small-scale fires, and
those of low intensity. In such cases, the available fire control efforts
are sufficient, and the combined elements of fuel, weather, and tOpo-
graphy have not favored fire spread. In the urban environment, single-
structure fires are also the most common in occurrence. Because of this,
research on mass or large-scale fires has had to be conducted on a model-
ing basis, either in the Open or in laboratory environments. Modeling
situations are vital, for they permit close control and the measurement
of experimental options, and can thereby allow accurate analysis of
basic fire characteristics . AS in all modeling situations, the validity
of the extrapolation of small-scale data to a large-scale occurrences
is Open to question.

The propagation of certain fire characteristics has been Shown to

be at least partially dependent upon the Size or scale of the fire itself.

Because some of these relationships are too vague to be detected in small
fires, or because they may not be present at all, it would seem scien—
tifically valid to eXpect a series of thresholds in fire scale at which
relationships would occur.

It has generally been conceded by fire researchers that fire
behavior at any one point is highly dependent upon fire action at all
other areas in the fire. This has been referred to as a "pattern pheno-
menon."1 Since fire behavior is also largely dependent on the state of
the environment in which burning occurs, it is vital that a fire researcher
concern himself with both the fire's "pattern" behavior and the environ-
ment's effect on that behavior. In other words, a fire must be considered
as an element in an environmental system. The end result of such a
relationship is the conclusion that large-scale inferences cannot be
strictly drawn from small-scale fire occurrences without allowances

for the fire environment, and its effect on fire growth.

Fire Environment
The environment of a fire is a crucial element to its propagation,
further spread, and control. It is commonly described in terms of the

relationships between factors of air flow, fuel nature, and topographic

 

fire surface. As in all environments of a dynamic nature, it is never

 

fixed, but may vary in numerous ways--among them space and time.
The descriptive size of a fire's environment is of crucial impor—
tance, for it may both limit a fire or be a cause for its further violent

expansion. As with all real environments, the fire environment is

 

1Clive M. Countryman, Mass Fires and Fire Behavior, (Berkeley:
Pacific Southwest Forest and Range Experiment Station, United States
Forest Service), p. 4.

three dimensional, extending vertically as well as horizontally. Country-
man has described two broad types of fire environment, which he has termed
closed and gpep,2 Inasmuch as we are here primarily concerned with the
urban situation, this is a vital distinction. A closed environment would
normally isolate the fire from conditions outside its immediate envelope.
A fire burning inside a building might well Operate independent of condi-
tions outside the building. The environment within the building is
characterized by the arrangement of fuels, and their characteristics.
The movement of air and the degree of moisture within the building might
well be the function of the heating and ventilation system of the
building. The environment of the fire is confined to the closed space
of the building. In general terms, however, the fire environment of
the city or urban area is not confined. Rapid and full changes can occur
in fuel conditions, weather behavior and air movement. Topography may
play a big part. By and large, the city is an gpen environment. A fire
Situation will be influenced easily by conditions outside its immediate
environment.

It should be understood that a fire, originally started and
furthered by a closed environment, may break out of that environment.
In such a situation, the outside conditions can influence the fire's
spread to other fuel, and to enlarge it in both size and intensity. As
we Shall see later, the point at which a fire escapes its closed environ-
ment and enters the Open urban area may be the turning point in the

development of a mass-fire.

 

21pm. , p. 7.

Elements of Fire Environment

The three major elements of the fire environment, and the factors

which influence the degree of fire spread are fuels, air mass, and topo-

 

3
graphy. The study of mass fire must include these three variables,
and focus primarily on their interrelationships in the fire environment.
The manner in which they affect fire characteristics and behavior is the

substance of the majority of fire research, and of this paper.

Fuel

 

For our purposes, fuels may be described in terms of:
l. The type of fuel
2. The distribution of fuel
3. The eXposure of the fuel
4. The environmental conditions surrounding the fuel.
These elements are not complete descriptors of and by themselves.
The interrelationships between these characteristics play a vital role also,
and must therefore be considered. Such interplay among the descriptors
would include:
1. The variance of fuel characteristics with the time of day,
week,or year.
2. The variance of fuels in regard to geographic location, both
locally and nationwide.
For our purposes, it may seem clumsy to speak of the urban area

in terms of "fuel." Yet in the fire environment, that is exactly what

 

3Ibid., p. 9.

structures, buildings, and natural cover comprise.4 The type of fuel
which is found in the fire environment is of crucial importance. The
emphasis of this paper, however, is directed more toward the arrangement
of fuels, and so must devote only limited space to fuel types. Much
scientific research is being conducted with the fuels of fire, yet it is
on a level of chemical and physical research which places it beyond our
SCOpe. The varying degrees of combustibility of different fuels are of
importance to the concepts of this paper, but will not be discussed further
here.

In any one instance, fuel will be found in varying degrees of

density. The degree of continuity with which a fuel is Spread over an

 

urban area will influence strongly the likelihood of fire Spread. In
any one urban environment flammable fuels may be spread continuously over
an area, only sparingly with non-flammable areas between, or may completely
surround non-flammable or bare areas. The factor of "continuity" of
Spread as such is very important in the urban fire environment.

Also important to the distribution of fuels is the concept of

arrangement of fuel particles. This refers not only to the horizontal

 

arrangement of fuel, but to the vertical as well. Fuels may be closely
packed or in a very loose configuration. Finally, the gmpunt of fuel
which is found in an area must be considered. Allied with this factor is
the consideration of proximity to different non-fuels.

One major difference between urban fuels and wildland fuels is

that in the urban fire environment fuels are distributed among non-fuels.

 

4R.H. Renner, S.B. MArtin, and R.E. Jones, Parameters Governipg_
Urban Vulnerability to Fire from Nuclear Bursts, Phase I, (San Francisco,
United States Naval Radiological Defense Laboratory, June 30, 1966), p.151.

Therefore, a discussion of fuel must consider non-flammable areas. Urban
fuels are primarily located in either structures or vegetation. Structural
fireloads are usually the heaviest, containing the larger volume of fuel.
In the majority of urban fire environments, fuels are primarily found
between ground level and the second story.

The exposure of a fuel is of prime importance in the study of
fire spread. As will be discussed later, radiative energy_is a major factor
in the Spread of flame. As far as "interior" fuels are concerned, their
location within a structure, the amount of opening in a structural barrier,
the height of the structure above the ground, and the shape of the
structure are all critical factors in the amount of radiative heat energy
reaching the fuel. In the urban fire environment, elements which will
most commonly limit exposure of exterior fuels would include structures
and buildings, other fuels (such as vegetation, trees), and topography.6

The environmental conditions of the fuel and of its surroundings
are usually described in terms of several generalizations. These treat
the age of the fuel, the temperature of the fuel and of the air, the
moisture content of the fuel and of the air, and the wind conditions in
the immediate area of the fuel. Most of these ambient conditions are

weather dependent, and can therefore be described in terms of weather

 

variations, such as seasonal cycles, annual cycles, and day-night changes.

These include changes of a local nature as well as climatic cycles.

 

51bid., p. 157.

61bid., p. 161.

7

Ibid., pp. 162-165.

Air Mass

The air mass which surrounds a fire environment, and is part of
it, is perhaps the most variable of the elements. The air mass may interact
with both the topography and the fuel near the surface of the earth.
There are, of course, several variables which are critical in defining
the air as it affects a fire. Among these factors are:

1. The temperature of the air mass, specifically at the varying

 

levels above the fire.
2. The stability of the air mass, and the presence or absence
of winds at the varying altitudes.

3. The degree of precipitation in the air.

 

4. The humidity of the air in the fire environment.
5. The amount of wind in the fire area.

6. The atmospheric pressure in the fire environment.

 

7. The presence of local conditions such as ground fog.

 

In reality, when we speak of the air mass, we are referring to
the meteorological conditions in the fire area. Each of the elements given
above, which are atmospheric descriptors, can be shown to have various
degrees of effect on fuel conditions, fire spread, and fire termination.

The temperature of an urban area (which conceivably forms the
total urban fire environment) is instrumental in determining the amount
of moisture the air is capable of holding before precipitation occurs. The
latitudinal location of a city, its basic topography, and its distribution
between land and water are all contributing factors to this temperature.
Besides daily temperature variances, of course, annual variations are

important to the fire environment.

1mg

 

The relative humidity of the air is a vital parameter in the fire
environment. Very moist air has been verified to contain up to 4% water
by volume.8 The effect of humidity on fire, however, is largely of an
indirect nature. The moisture content of many fuels is very closely
associated to the relative humidity. In forest fire research, it has
been found that relative humidity may be taken as a fairly reliable indicator
of fire hazard.9

The major meteorological factor in fire Spread is generally
considered to be wind. Reports of major urban conflagrations have frequently
cited wind velocity, presistence, and instability as the major problem
in confining the fire. Quite apart from its role of supplying oxygen to
a fire area, wind can drive burning firebrands far ahead of the control

10

lines to start new Spot fires. Apart from surface winds, upper strata

currents may affect the development of large fire convection columns,
thereby changing the energy release pattern of a high intensity fire.11
The indirect effects of wind must, of course, include a drying effect on
the moisture content of fuels through the evaporation process.

The effect of wind in urban fire environments may be partially
due to the shape of the urban area. One source states that the winds in
urban fire areas are strongly influenced by:

l. The arrangement of buildings and structures on the ground,

the resultant artificial topography. Such structures must
also be thought of in terms of "fuel piles."

2. The height of structures, and the variations in these heights.

3. The density of urban structures, creating a "uniform"

 

9Countryman, p. 14.
loIbid.

llIbid.

 

'I~ -_

plain of heat-producing units, or perhaps a spotty pattern which
would influence ground convection currents.

4. The distance between structures.

5. The size of structures, in terms of their physical dimensions,
and of their Shape.

6. The arrangement of the ground-surface, and the changes in
elevations caused by man-made structures or natural land forms.
Such "shapes" would create channels of wind flow, both in
normal and fire environments.

7. The texture or perhaps grain of the urban area, in the reference
of a frictional surface which would slow air movement, creating
disruptions of smooth patterns, and micro-air environments.

8. The topography of the surrounding region, which would determine
the manner in which air flows were directed at the urban area,
and from which direction and altitude. This would also contri-
bute to the meteorological factors, describing the humidity,
moisture content, and temperature of the air.12

It Should be vital to an understanding of a specific fire environment

to determine wind patterns. Winds in an urban area are generally set by
the overall regional atmospheric conditons, yet are strongly influenced by
the urban form. Of prime concern to the fire environment is the rotational
pattern of wind. Such rotation may be started where ordinary surface and
low-level winds are retarded by structures, vegetation, or land forms, or

where there are local thermal sources.

Topography

The third of the major fire environment descriptors is topography.

 

12Renner, p. 146.

A-lO

It can be said to have both a direct and an indirect effect upon the

spread of fire. In the general case, fires spread more rapidly pp_a

SlOpe than on level ground. This is primarily due to the heating effects of
fire radiation on up-Slope fuels, and the effect of wind which is generated
'py the fire in uphill heating.l3

Conversely, the Spread of fire downslope is less than that on level

ground, due to the same factors of lessened radiative heating and wind.

Rarely in an urban area, however, is a simple SlOpe Situation found.

The usual tOpography consists of Short, broken Slopes. In these areas,

fire is generally considered to move slowly, and with a more erratic behavior.
It should be emphasized that the effect of windflow may greatly increase

or decrease the spread of fire over all forms of topography.

Importantly for the fire environment, tOpography affects local
weather and micro-climate. The greatest effect here may be in the thermal
differences found on Slopes of differing sun orientation. Such thermal
variations might well affect the fuel found on that slope.14
Likewise, the channeling of windflows may be quite important, resulting
in increases in wind velocity in restricted valleys.

In a description of the tOpography as an element in the fire environ-
ment, we are concerned with four factors. These include:

1. The Eype_of surface, that is, land or water.

2. The degree of angle of steepness of the Slope.

3. The direction of Slope of the land.

4. The elevation of the land in reference to the general area.15

 

13Countryman, p. 13.
14Ibid., p. 14.

15Renner, p. 100.

Clearly, there are certain tOpographic surfaces which cannot be
thought Of as fuels (i.e., concrete in the normal sense), but their existence
as non-flammables is equally important when considering fire Spread. Such
surfaces may provide fire-breaks, or gaps in a continuous area of fuels.

The dimensions Of such gaps will be of vital importance in the following
chapters.

The SlOpe of the topography is an all important variable in fire
control as well as fire spread, for it can be an effective block to water
supply as well as a barrier to fire spread. Certainly the orientation
of slopes is vital in foreseeing the path of possible fire spread, espec-

ially if natural boundaries to fire spread are formed by Slope ridge lines.

The Mechanics of Fire Spread

The spread of fire through a complex Of fuels is a process involving
stages Of both ignition and combustion. The crucial point for all discussions
concerning fire spread is the point separating non-ignition and ignition
of a fuel. The line separating these two stages can be drawn easily only
when the fuel being consumed is arranged in a discontinuous pattern.
When fire is burning through a continuous fuel array, the point of ignition
may well be obscured by the rapid and changing advance of fire. In
nearly all cases, however, the unignited fuel is heated to its point of

ignition by conduction, convection, and/or radiation from the heat of

 

 

combustion of ignited fuel.16 The importance of each of the three modes
of heat transfer is dependent upon the fire environment; that is, upon the
elements of fuel, tOpography, and the air mass.

The process of conduction, which involves "body-to—body" contact

 

between fuels plays little part in the spread of fire through discontin-
uous fuel arrays. The urban fire environment presents, for the most part,
a discontinuous fuel array. The element of conduction would remain a
minor factor in the mass Spread of urban fire, therefore, since discon-
tinuous fuel arrays must rely upon means of heat transmission which are
not dependent on phySical continuity.

In the presence of large non-fuel separations between fuel
elements, the dominant means of heat transfer is radiation. This is
likewise true in cases of "downward prOpagation" of a fire. In such

instances, the exposure of a fuel to the heat source is a critical factor

 

16Ib1d., p. 247.

17
Ibid., p. 248.

in ignition. Importantly for our purposes, distance from the heat source

is also vital.17 The method of heat convection has in the past been the

 

cause of fire spread in many major disasters. The transmission of heat
through bodies of highly heated air accounts for much upward fire spread.
Convective: currents have frequently carried burning solid fuel, or fire
brands, long distances to create new fires. In the case of convective
transfer, exposure and position of the unignited fuel in relation to the
fire source is the prime factor in determining ignition.

In the discontinuous fuel array, then, there may be a clear
distinction between ignited and unignited fuels. Likewise, the time and
point of ignition can be easily marked. Both these factors are vital to
the control of fire spread. We shall later examine this facet of fire
control in greater detail.

From the above discussion, it can be discerned that the three
methods of fire transfer are based on two processes. These are 2333:.

transfer and mass-transfer. The two are quite dissimilar. AS a

 

fuel element burns, the heat that is generated is given Off in several
ways. AS we have seen, radiation of heat from the flame is a dominant
means, and is governed by the temperature and scale of the flame. Radia-
tive heat is omni-directional from the flame zone. When radiant heat
strikes an unignited fuel, it is absorbed by the fuel and raises the
temperature of the fuel. The remainder of the radiant heat is absorbed by
non-combustible surroundings, or by the ignited fuel itself. In the

study of fire spread, we are primarily concerned with the amount of heat
absorbed by the unignited fuel. This amount can be generally Said to

be determined by geometric factors, that is distance from the heat source

and the amount of eXposure of the fuel to the flame source. Generally
Speaking, the nearer a fuel to the heat source, the greater the contributed
amount of radiant heat to that fuel. Likewise, the nearer a fuel is to
a heat source, the greater the amount of high intensity surface exposure.1
The radiant energy which is emitted into the immediate air mass
contributes to the direct transfer of energy in molecular motion. The
resultant motion is released in turbulent eddies and currents of air
directly above the flame of the active fire. The result of the tur-
bulence of air above a flame is a "plume" of heated air and burning gases
commonly termed a "convection column." Frequently this very buoyant
column transports solid, incompletely burned fuel far above a fire,
out of the convection current. In the event these solid brands continue
active burning, and land elsewhere, they create a new fire. The super-
heated gases and air, however, also carry great potential for starging fire.
In the event the convection column comes into contact with an un-
ignited fuel, it envelops the exposed surface of the fuel with a heated
layer of air, raising the temperature of the fuel. The fuel itself is
heated by molecular transfer through conduction. The degree to which
the fuel absorbs the heat of the column is dependent upon the temperature
difference between the two elements, the surface characteristics of the
fuel, the transfer nature of the fuel, and the nature of the flow of the
heated gases past the fuel.19 Since the flow of the buoyant convection
column will be greatest in a vertical direction, unignited fuel above the
fire source will be greatly affected by this means of heat transfer. The

importance of tOpography should be self evident at this point.

 

 

18mm.

lgIbide ) Po 2490

A-lS

The Spread of fire by the transfer of mass plays an important part.

Mass transfer is usually associated with (1) the flow of burning liquid

 

fuels downhill, (2) the flow or emission of burning gases, and (3) the
transfer of burning solid fuel. The importance of these means of spread
lies in the manner in which they are carried away from the fire. Burning
solid and liquid fuels travel by the force of gravity. All three forms
of matter may be forcibly ejected from a fire by explosion of a container,
or by the explosion of hot gases. Both burning solids and gases may be
carried by the buoyant action of heated air.

The slope of the ground will determine much of the direction and
extent of fire Spread by solid or liquid transfer. The stability of the
air mass and the disturbance caused by the convective column can carry
burning fuel many miles. The horizontal wind velocity within the fire
environment is a large factor in determining the distance over which
brands can travel.

The Spread of fire, then, is more than just a function of
Simple elements of distance between fuels. It is a complex interaction
of all the elements of the fire environment, and of the determinants

of fire intensity.

 

2oIbid., p. 251.

TABLE B-l

B-lO

B-ll

B-12

B-13

APPENDIX B

DATA AND SUPPLEMENTARY INFORMATION

Selected Fires and Conflagrations Since 1910
Ranking of Principal Factors Contributing to Spread
Change in Ranking During Time Periods

Comparative Fire Statistics For the United States
for 1966 and 1967

Comparative Fire Statistics for the United States
by City Size

Potential Area of Fires - 1967
Water Application Rate - 1967
Water Supply Deficiencies - 1967
Mutual Aid Response - 1967

Required Fire Flows Under American Insurance
Association Standards - 1967

Recommended Areas Served for Hydrants Under American
Insurance Association Standards - 1967

Modeled Multipliers For Building Characteristics

Required Separation Distances in Feet

.mp-n .mmua .mm ..«.A.m.z .mpms .coauapm nuuflmsa .xoonpcmm consumuoua muse

ooo.ooo.oa
ooo.ooo.om
ooo.ooo.oa

s s

ooo.ooo.om
ooo.ooo.am
000.com
ooo.ooo.m
ooo.ONo.m
ooo.ooo.~
ooo.eso.v
000.com.a
ooo.ooo.~
ooo.ooo.H
ooo.ooo.o
ooo.oow.a
ooo.ooo.m~
000.com.m
ooo.ooo.va
ooo.omm.m
000.com.m m

 

mmoq
mumuocoz

 

pmaaax NH .Hmnnumspca
macapnnsn mom

.mwpan 0HH

«use Ammuou ..mmpan ova
mmcflHH03p coma

pmnanx mos .ucouunoumz
menopansn mam
UGOHMHOHSS

00am happen ..mmpan NNN
mxooan suau om
moum>xooum

mmcapflasn omm

mmcnpansn omm

mmcnpansn was

mmcapsnsn owe

mxoodn ow ..mmpan oooa
pmaaax mmm ..mmpan ooov
mmcnpsflsn mmmn

pmasax p ..mmpan oops

mocnpaflsn mam
mmcaoaacn mo mouom mm

 

mxumamm

momcmu ocm moumum owuaco

.SM .maaa>mnsoq
.uaamu .mmaomc< moo
.mno .mndnmmom
cacnOmaamu .coaamz
monam amonom scam:
mcst .auao mcxoa
.m.z .mmm gonna
.6.z .aunu hmmnmn
.waamo .moaomcm moq
mannad .mEoz

.HHH .ommoanu

moan: .cuonsm

.m.z .msnmmz

.mq .uuomc>cunm
.oaamo .smflmxumm
.U.z .cuom 3oz
monam ummnom snowmccaz
.mo .mucoaum

.mmmz .Emamm

.xum .mmcaumm uom
scam: .uomcmm

Nuau

Oama MUZHm mZOHfidmwdquOU QZ¢ mmmHm Qmaumamm

aim mamdfi

u OUHDOW

moma
aoma
mmma
omoa
hvma
hvma
mvma
avma
mmma
vmma
vmma
mmma
omma
mNma
mmma
NNma
mama
hama
oama
mama
aama

 

use»

TABLE B-2

RANKING OF PRINCIPAL FACTORS CONTRIBUTING TO SPREAD

OF CONFLAGRATION IN THE UNITED STATES AND CANADA

 

 

Period 1901 to 1925 Percent of Total
1 - Wood Shingle Roofs 25.4%
2 - Inadequate Water Distribution System 13.0
3 - Inadequate Public Protection 13.0
4 - Wind Velocity Greater than 30 m.p.h. 12.5
5 - Lack of Exposure Protection 10.2
6 - Delay in Giving Alarm 2.8
7 - Congestion Reduced Fire Fighting Access 2.8
8 - Failure of water Supply 2.8

Period 1926 to 1961

l - Wind Velocity Greater than 30 m.p.h. 15.3%
2 - Inadequate Water Distribution System
3 - Lack of Exposure Protection

4 - WOod Shingle Roofs

5 - Inadequate Public Protection

6 - Unusually Hot or Dry Weather

7

8

9

0

...:
N

[.4
..a
O

(DONOUJ-h-bNI-‘N

- Delay in Fire Discovery

— Delay in Giving Alarm

- Congestion Reduced Fire Fighting Access
- Forest or Brush Fire Entered Town

wa‘lUlmCDKO

1

Source: Fire Protection Handbook, Twelfth Edition, 1962, N.F.P.A., p. 1-64.

TABLE B-3
CHANGE IN RANKING OF FACTORS BETWEEN TIME PERIODS

1901 - 1925 and 1925 - 1961

Those Increasipg in Importance in Spread of Conflagrations

 

1. Lack of Exposure Protection
10.2% to 11.1%

2. Wind Velocity Greater than 30 m.p.h.
12.5% to 15.3%

3. Delay in Giving Alarm
2.8% to 5.0%

4. Congestion Reduced Fire Control Access
2.8% to 4.6%

5. Unusually Hot or Dry Weather
8.4%

6. Forest or Brush Fire Entered Town
3.8%

Those Decreasing in Importance in the Spread of Conflagrations

 

1. Wood Shingle Roofs
25.4% to 9.2%

2. Inadequate water Distribution Systems
13.0% to 12.2%

3. Inadequate Public Protection
13.0% to 8.4%

4. Failure of Water Supply
2.8% to less than 1%

Source: Fire Protection Handbook, Twelfth Edition, 1962, N.F.P.A., p. 1—64.

 

TABLE B-4

COMPARATIVE FIRE STATISTICS

FOR THE UNITED STATES

1966 and 1967

125.5. 19.51
Alarms per 1000 pOpulation 20.1 20.0
Fires per 1000 pOpulation 10.8 10.4
Losses per capita $5.60 $5.90
Buildings fires per 1000 pOpulation 4.2 4.1
Average building fire loss $1241 $1342
Source: "Fire Record of Cities, 1967," Fire Journal, Vol. 62 #4,

 

July 1968, p. 35.

TABLE B-5

COMPARATIVE FIRE STATISTICS

FOR THE UNITED STATES

BY CITY SIZE

Under 20,000

20,000 and over
Alarms per 1000 pOpulation 16.6 20.0
Fires per 1000 population 11.6 10.4
Fire losses per capita $8.24 $5.90
Building fires per 1000 population 4.8 4.1
Average building fire loss $1605 $1342

Source: "Fire Record of Cities, 1967," Fire Journal, Vol. 62 #4,

 

July, 1968, p. 36.

TABLE B-6
POTENTIAL FIRE AREAS - 1967

(Cubic Feet)

 

 

# of Fires Involving # of Fires Involving # of Fires
Fire Area Less than 80 Per Cent More than 80 Per Cent Extending to
(Cubic Feet) of Potential Fire Area of Potential Fire Area Other Properties
Under 50,000 3 2 1
50,000 to 99,999 2 11 0
100,000 to 249,999 8 45 12
25C),000 to 499,999 16 55 13
500,000 to 999,999 16 44 11
l,()00,000 to 1,999,999 10 16 11
2,000,000 and over 13 14 8
TOTAL 68 187 56
Area not Reported 9 27 7
Extinguished by
Sprinklers 8 0 0
Area not a Factor or
No Data 34 0 0

 

ource: Warren Y. Kimball, "Control of Large-Loss Fires," Fire Journal, November ,
1968, Vol. 62 #6, p. 73.

 

TABLE B-7
WATER APPLICATION RATE

GALLONS PER MINUTE PER 100 CUBIC FEET OF FIRE AREA

 

Water Density Applied

 

 

Per 100 Cubic Feet of Number of Fires
Fire Area, GPM Reported - 1967

None 2 0.5% '

Under 0.10 33 8.3%

0.10 to 0.24 48 12.1%

0.25 to 0.49 65 16.4%

0.50 to 0.74 36 9.1%

0.75 to 0.99 27 6.8%
Total Under 1.00 211 53.3%

1.00 to 1.49 27 6.8%

1.50 to 1.99 13 3.3%

2.00 to 2.99 24 6.1%

3.00 to 3.99 11 2.8%

4.00 and over 6 1.5%
Total Over 1.00 81 20.4%

Fires where not a factor 10 2.5%

Fires where data not available 94 23.8%
Total Reported Fires 396 100.0%

Source: Warren Y. Kimball, "Control of Large-Loss Fires," Fire Journal,
November, 1968, Vol. 62, #6, p. 74.

TABLE B-8

WATER SUPPLY DEFICIENCIES - 1967

Cause No. of Reported Occurrences

 

Inadequate supply at hydrants 2

Small dead-end mains

Water storage depleted during fire

Private water supply inadequate

Public and private water storage quickly
exhausted

Valve closed and pumps not operating

Valve partly closed

Closed valve reduced pressure

Valve found closed when fire started

Low pressure

Water lost through large pipes in fire
building

.5me

PIPIP*F’F‘P'

Failure to start fire pumps

Failure of power to city pumps
Building standpipe inadequate

Small, low-pressure main

Volume adequate for only one hydrant

P'P'P‘P‘P‘P'

Source: Warren Y. Kimball, "Control of Large-Loss Fires," Fire Journal,
November, 1968, Vol. 62, #6, p. 77.

 

 

B-9
TABLE B-9
MUTUAL AID RESPONSE - 1967
|
Number of Outside Fire Number of F
Departments Responding, Fires (
1 35 “
2 36 j‘
3 30 ..
4 24
5 23
6 - 9 38
10 or more 10
Total with Mutual Aid Response 216
No Mutual Aid Used 164
No Data 16
Total Reported Large Loss Fires 396

Source:

 

 

Warren Y. Kimball, "Control of Large-Loss Fires," Fire

Journal, November, 1968, Vol. 62, #6, p. 76.

TABLE B-lO

REQUIRED FIRE FLOWS UNDER AMERICAN INSURANCE

 

ASSOCIATION STANDARDS - 1967

Population Required Fire Flow Duration

of Community (gpm) (mgd) (hours)
1,000 1,000 1.44 4
1,500 1,250 1.80 5
2,000 1,500 2.k6 6
3,000 1,750 2.52 7
4,000 2,000 2.88 8
5,000 2,250 3.24 9
6,000 2,500 3.60 10
10,000 3,000 4.32 10
13,000 3,500 5.04 10
17,000 4,000 5.76 10
22,000 4,500 6.48 10
27,000 5,000 7.20 10
33,000 5,500 7.92 10
40,000 6,000 8.64 10
55,000 7,000 10.08 10
75,000 8,000 11.52 10
95,000 9,000 12.96 10
120,000 10,000 14.40 10
150,000 11,000 15.84 10
200,000 12,000 17.28 10

Over 200,000

Source: James F. Casey, Ed.

 

12,000 plus 2,000 to 8,000 gpm additional
for a second fire for a 10-

hour duration

The Fire Chief's Handbook, New York:

Reuben H. Donnelly, Corp., 1967, p. 66.

gpm = gallons per minute

mgd = million gallons per day

TABLE B-ll

THE RECOMMENDED AREAS SERVED FOR HYDRANTS

 

 

(AIA) - 1967
Fire Flow Required Average Area per Hydrant
(gallons per minute) (square feet)
1,000 120,000
2,000 110,000
3,000 100,000
4,000 90,000
5,000 85,000
6,000 80,000
7,000 70,000
8,000 60,000
9,000 55,000
10,000 48,000
11,000 43,000
12,000 40,000

Note: 1 acre = 43,546 square feet

Source: James F. Casey, The Fire Chief's Handbook, New York:
Reuben H. Donnelly, Inc., 1967, p. 67.

TABLE B-lZ

MULTIPLIERS FOR BUILDING CHARACTERISTICS

 

 

Class 1 Roof Construction - Fire Resistive,

2-Hour or Better

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Story w A L L o p E N I N c 5

Height None Few Average Many All

1 .4 1.8 3.6 7.2 1 12

2 .7 3.6 7.2 14 24

3 1.1 5.4 11 22 36

4 1.4 ' 7.2 14 29 4s

5 1.8 9 18 36 6O

6 2.2 11 22 43 72

7 2.5 13 25 50 84

8 5 Over 2.9 14 29 58 96

Class 2 Roof Construction - Noncombustible or Fire Resistrve»

Story A w A L L 'o p s N I N c 3

-Height None . Few Average Many All

1 10 11 12 14 18

2 10: 12 , 14 17 27

3 10 13 15. 21 35

4 10 14 17 24 44

S 10 15 19 28 52

6 10 15 21 32 60

7 10 17 23 _ 35 69

8 5 Over 10 17 24 39 77

Class 3 Roof Construction - Wood, Flat or Peaked Up to 15 Feet

Story W A L L O P E N I N G S

Height None r Few Average Many All

1 30 31 32 34 38

2 30 32 34 37 47

3 30 33 35 41 55

4 30 34 37 44 64

5 30 34 39 48 72

6 30 35 41 52 80

7 30 36 43 55 89

8 & Over 30 37 44 59 97

 

 

 

 

B-l3

TABLE B-12 Continued

’—

Class 4 Roof Construction - WOod, Bow String Truss or Peaked 16 - 25 Feet

 

 

 

 

Story WALL OPENINGS
Height None Few Average Many All
1 45 46 47- 49 53
2 45 47 49 . 52 62
3 45 48 50 56 70
4 45 49 52 59 79
S 45 49 54 63 87
6 45 50 56 _ 67 95
7 45 - 51 58 70 104
8 6 Over 45 52 59 74 112

 

 

 

 

 

 

 

 

.r—vw A

[Class 5 Roof Construction — Wood, Peaked 26 Feet and Over

 

 

 

Story ‘ wALL OPENINGS
Height None Few Average Many A11
1 60 61 62 64 - 68
2 '60 - 62 64 67 77
3 60 63 65 71 85
4 60 64 67 74 - 94
5 60 64 69 78 102
6 60 65 71. 82 110
7 60 67 73 85 v 119
8 6 Over 60 ’ -67 ' 74 89 127

 

 

 

 

 

 

 

 

Source: B.M. Cohn, L.E. Almgren, M. Curless, A System for the Local
Assessment of the Conflagration Potential For Urban Areas,
(Chicago: Gage. Babcock and Associates, Inc., 1965), p. A48.

B-l4

TABLE B-13

REQUIRED SEPARATION DISTANCES IN FEET

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[Average Windn> Low Normal High
L Velocity (7 mph or less) (18 mph or less) , (31 mph or less)
fi.hape of . Rectan- Lon ' Rectan- Long Rectan- Long
' Radiating M5? Square gular Recz. Square gular . Rect. Square gular Rect.
1500 120
1600 . 120 120
2200 . 130 130 120
2300 120 ' 130 130 120
2600 125 120 140 135 125
3200 ~ 135 130 120 150 145 135
3800 120 145 140 130 160 155 145
4100 125 120 1§0 145 135 165 160 150
5000 135 130 120 160 155 145 180 170 160
6000 150 145 130 175 170 155 190 , 185 170
7000 160 155 140 185 180 165 200 195 180
8000 170 i__}65 150 "_1931_ 190 175 210 205 190
9000 180 175 160 205 200 185 220 215 200
10,000 190 185 165 215 210 195. 230 225 205
11,000 200 195 170 225 220 200 240 — 235 215
12,000 210 200 180 235 y_4225 205 250 245 220
137600 220 210 185 240 235 215 260 250 230
. 14,000 225 220 195 250 245 220 270 260 235
9 15,000 235 225 200 260 250 225 275 265 240
, m 16,000 240 31-239_____HZQ§ 265 . 260 235 280 11.3721.11_Z§9__
17,000 250 240 210 275 ‘ 265 240 290 280 255
. 18,000 255 245 220 280 270 245 295 285 260
<3 19,000 265 255 225 290 280 250 305 295 265
m 20,000 270 260 230 295 285 255 310 300 270
- 21,000 275 265 235 300 290 260 320 305 275
- 22,000 285 270 240 310 295 265 325 315 280
4 23,000 290 275 245 315 305 270 330 320 285
m 24,000 295 285 250 - 320 310 275 340 .325 290
m 25,000 305 290 255 325 315 280 345 330 295
2 26,000 310 295 260 330 320 285 350 335 300
27,000 315 300 265 340 325 290 355 340 305
0 28,000 320 205 270 345 330 295 360 345 310
2 29,000 325 310 275 350 335 300 365 350 315
F, 30,000 330 315 280 355 340 305 375 355 320
‘. 32,000 340 325 290 365 350 315 385 365 330
‘ 34,000 350 335 300 375 360 325 395 375 340
.4 36,000 360 345 305 385 370 335 405 385 350
l a 38,000 370 355 315 395 380 340 415 395 355
1 . 40,000 380_ 365 320 405 390 350 425 405 365
f " 42,000 390 375 330 415 395 355 435 415 375
3 m 44,000 400 380 340 425 405 365 440 425 380
1 46,000 410 390 345 430 415 370 450 ' 430, 390
I ‘3 48,000 420 395 355 440 420 380 460 440 395
, °1 50.000 425 405 360 450 430 385 470 445 400
z 6' 52,000 435 415 365 460 440 395 475 455 410
i to 54,000 440 420 375 465 445 400 485 460 415
a 56,000 450 430 380 475 450 405 490 470 420
6 58,000 460 435 385 485 460 415 500 475 430
o 60,000 465 445 390 490 465 420 485 435
2 62,000 470 450 400 500 475 425 , 490 440
64,000 480 455 405 480 430 495 445
66,000 485 465 410 y 490 440 500 450
68,000 495 470 420 495 445 460
70.000 500 480 425 500 450 465

 

 

 

Source: B.M. Cohn, L.E. Almgren, M. Curless, A System For The Local Assessment
of The Conflagration Potential of Urban Areas, (Chicago: Gage Babcock and
Associates, Inc., 1965): p. A51.

 

 

"'CITI'ITH‘Ijl’iIIIflfijflfilMillfili‘lflil’nflt(I'HTITH’I‘ITI'ES

69 3959