This article was originally published in Encyclopedia of Forensic Sciences, Second

Edition published by Elsevier, and the attached copy is provided by Elsevier for the
author's benefit and for the benefit of the author's institution, for non-commercial
research and educational use including without limitation use in instruction at your
institution, sending it to specific colleagues who you know, and providing a copy to your
institution’s administrator.

All other uses, reproduction and distribution, including without limitation commercial
reprints, selling or licensing copies or access, or posting on open internet sites, your
personal or institution’s website or repository, are prohibited. For exceptions,
permission may be sought for such use through Elsevier's permissions site at:

http://www.elsevier.com/locate/permissionusematerial

Lentini J.J. (2013) Fire Patterns and Their Interpretation. In: Siegel JA and Saukko PJ
(eds.) Encyclopedia of Forensic Sciences, Second Edition, vol. 3, pp. 396-405. Waltham:
Academic Press.

© 2013 Elsevier Ltd. All rights reserved.

 Author's personal copy

Fire Patterns and Their Interpretation

JJ Lentini, Scientific Fire Analysis, LLC, Big Pine Key, FL, USA
ã 2013 Elsevier Ltd. All rights reserved.

Glossary such as quantity and geometry, and in which adequate air

Flashover A transition phase in the development of a fire in for combustion is available.
a structure, in which all combustible surfaces exposed to Full room involvement A compartment fire in which the
thermal radiation ignite more or less simultaneously and fire entire volume is involved in the fire, but flaming will only
spreads rapidly throughout the space, resulting in full room occur when there is sufficient oxygen available.
involvement. The transition is from ‘a fire in a room’ to ‘a Ventilation-controlled fire A fire in a postflashover
room on fire.’ compartment, where all fuels are involved, and the heat
Fuel-controlled fire A fire in which the heat release rate and release rate of the fire is controlled by the availability of
growth rate are controlled by the characteristics of the fuel, oxygen.

Introduction and Overview help fire investigators determine the origin of small fires. As a
fire progresses to full room involvement, however, the ‘rules’ of
A fire pattern, also called a burn pattern, is the visible or fire pattern interpretation change.
measurable physical changes or identifiable shapes formed by Unfortunately, many fire investigators fail to understand
a fire effect or group of fire effects. Fire effects are the changes in underventilated fires and may misinterpret the most heavily
or on a material as a result of exposure to the fire. damaged area of the room as necessarily being the origin of the
The term ‘fire effects’ describes the artifacts left behind by fire. This is explained further when ventilation-generated pat-
many different processes, including dehydration, melting, terns are discussed.
color changes, oxidation, charring, loss of material, alloying,
spalling, expansion and deformation, annealing, soot and
smoke deposition, and clean burn. Even damage to an electri-
cal system can constitute a fire pattern. Fire investigators are Plume-Generated Patterns
thus presented with a vast array of different patterns, all of
which require some level of understanding of how they are Fire patterns can result from several different kinds of int-
created and what they mean. eractions between the fire and its surroundings. The best-
Accurate interpretation of fire patterns can lead the fire understood interactions are those that occur between the fire
investigator to the correct origin of the fire, and hopefully, to plume and a nearby vertical surface. Such patterns are often
the correct identification of the fire’s cause. However, accurate called ‘truncated cone patterns.’ (Fire investigators refer to
interpretation is often an elusive goal. ‘cones’ and ‘inverted cones,’ but their definition is the exact
For many years, fire investigators followed the simplistic opposite of the classic geometric definition of a cone. In fire
maxim that heat rises, and fire burns up and out. While it is investigation, the apex of the cone is at the bottom.) These
true that heat rises, it does so only until it reaches the ceiling. At patterns are typically recorded on vertical surfaces, and fire
this point, the behavior of the fire changes. investigators look for what they call ‘V’ patterns.
When the rising smoke and heat reach the ceiling, the Fire patterns evolve over the life of a fire. When the fire is
upward movement is stopped and a ceiling jet is created, new, the pattern on an adjacent vertical surface will assume the
sending the products of combustion in a horizontal direction, shape of the flame, and will exhibit what fire investigators call
which then become confined by the walls of the compartment. an inverted cone pattern. This triangular pattern is characteris-
A hot gas layer develops at the ceiling, and gradually, this tic of young fires. Figure 1 is a schematic representation of how
layer becomes thicker and more energetic. It radiates heat in all such a pattern is created, and Figure 2 shows a triangle-shaped
directions, including downward. When the temperature of the pattern produced by a test fire. As the fire continues to grow,
hot gas layer reaches 500–600  C, all exposed combustible the pattern becomes columnar, and the sides of the pattern are
surfaces in the room ignite more or less simultaneously. This roughly perpendicular to the floor. Columnar pattern produc-
transition from ‘a fire in a room’ to ‘a room on fire’ is known as tion is shown schematically in Figure 3 and an actual colum-
flashover. It is the fire investigator’s goal to determine the first nar pattern is shown in Figure 4. Columnar patterns have a
fuel package to ignite. That ignition necessarily occurred some very short life span, and change as soon as the fire begins to
minutes before flashover. Postflashover burning, however, is interact with the ceiling.
usually far more energetic than preflashover burning, and it Once the fire plume encounters the ceiling, or an interven-
has the capability to obscure or even obliterate patterns that ing horizontal surface, a ‘V’ pattern is the result. A semicircular
existed prior to flashover. pattern is created on the ceiling. Figure 5 shows how the
By visualizing the interaction of a conical plume with a pattern is produced when the burning fuel package is close to
vertical surface, one can understand the generation of some the wall, and Figure 6 shows an actual V pattern from a
fire patterns. Such patterns are common fire artifacts, and often fire scene. As can be seen, the schematics are quite a bit

396 Encyclopedia of Forensic Sciences, Second Edition http://dx.doi.org/10.1016/B978-0-12-382165-2.00213-0

Encyclopedia of Forensic Sciences, (2013), vol. 3, pp. 396-405

 Author's personal copy
Investigations | Fire Patterns and Their Interpretation 397

Triangle shaped
pattern on wall

Figure 1 Intersection of an incipient fire with an adjacent wall.

Copyright 2012 from Scientific Protocols for Fire Investigation, 2nd
Edition by John J. Lentini. Reproduced by permission of Taylor and
Francis Group, LLC, a division of Informa plc.

Figure 3 Intersection of a columnar fire with an adjacent wall.

Copyright 2012 from Scientific Protocols for Fire Investigation, 2nd
Edition by John J. Lentini. Reproduced by permission of Taylor and
Francis Group, LLC, a division of Informa plc.

Figure 2 Triangle-shaped pattern, also called an inverted cone pattern,

produced in a test fire. Copyright 2012 from Scientific Protocols for
Fire Investigation, 2nd Edition by John J. Lentini. Reproduced by
permission of Taylor and Francis Group, LLC, a division of Informa plc.

cleaner than the actual patterns, and sometimes fire investiga-

tors will disagree about whether a pattern is actually a V
(or half of a V).
When the fuel package burns farther away from the wall,
the interaction of the fire plume with the wall occurs at a higher
level, and the bottom of the intersection of the cone and the
wall is not as sharp. This results in a U-shaped pattern on the
wall, and a circular pattern on the ceiling. Figure 7 shows
schematically how a U-shaped and circular patterns are cre-
ated, and Figure 8 shows a classic U-shaped pattern from a fire
scene. Patterns on ceilings are often more difficult to recognize,
because ceilings are likely to collapse sooner than walls. Figure 4 A columnar fire pattern produced by a fire that was
Figure 9 shows a semicircular pattern above the origin of a extinguished prior to the formation of a V-shaped pattern. Copyright
fire that originated inside a wall. 2012 from Scientific Protocols for Fire Investigation, 2nd Edition by John
As a fire progresses, more fuel packages become involved, J. Lentini. Reproduced by permission of Taylor and Francis Group, LLC, a
and may result in new ‘V’ patterns or ‘U’ patterns being division of Informa plc.

Encyclopedia of Forensic Sciences, (2013), vol. 3, pp. 396-405

 Author's personal copy
398 Investigations | Fire Patterns and Their Interpretation

Circular pattern on ceiling Circular pattern on ceiling

U shaped
pattern on wall

V shaped
pattern on wall

Figure 5 Intersection of an established fire plume with an adjacent wall.

Copyright 2012 from Scientific Protocols for Fire Investigation, 2nd Figure 7 Intersection of a fire plume located some distance from
Edition by John J. Lentini. Reproduced by permission of Taylor and the wall with the wall and ceiling. Copyright 2012 from Scientific
Francis Group, LLC, a division of Informa plc. Protocols for Fire Investigation, 2nd Edition by John J. Lentini.
Reproduced by permission of Taylor and Francis Group, LLC, a
division of Informa plc.

Figure 6 Typical V pattern in a residential fire. The fire originated in a

plastic trash can located against the wall in the background. Copyright
2012 from Scientific Protocols for Fire Investigation, 2nd Edition by John
J. Lentini. Reproduced by permission of Taylor and Francis Group, LLC, a
division of Informa plc.

Figure 8 Typical U pattern in a residential fire. The fire originated

produced. In the past, some fire investigators have interpreted behind the door at the left side of the photo. Copyright 2012 from
multiple ‘V’ patterns as evidence of multiple origins, but this is Scientific Protocols for Fire Investigation, 2nd Edition by John J. Lentini.
an interpretation that can rarely be justified once a fire be- Reproduced by permission of Taylor and Francis Group, LLC, a division
comes fully involved. Such interpretations are not justified in of Informa plc.

Encyclopedia of Forensic Sciences, (2013), vol. 3, pp. 396-405

 Author's personal copy
Investigations | Fire Patterns and Their Interpretation 399

Figure 9 Semicircular pattern on the underside of the floor above the

origin. The fire originated at an overheated switch inside the wall cavity.
Copyright 2012 from Scientific Protocols for Fire Investigation, 2nd
Edition by John J. Lentini. Reproduced by permission of Taylor and
Francis Group, LLC, a division of Informa plc.

cases of full room involvement or in any fire scene where all of

the fire damage is contiguous. It is only when there is no
communication between the fire patterns generated by sepa-
rate fire plumes that one can state with any validity that there
were two (or more) origins.
Plume pattern generation is simple to understand and
explain; the difficulty lies in locating the first plume pattern
generated. Further, once a compartment progresses beyond
flashover, interaction with the fire plume may not be
Figure 10 An example of a typical smoke horizon. Photo courtesy of
the dominant means of pattern generation on building
Mick Gardiner, Gardiner Associates, with permission.
surfaces.

Confinement Patterns

Like plume-generated patterns, confinement patterns are easily

explained. Confinement patterns are caused by the hot gas
layer being trapped beneath a ceiling, and interacting with
the walls. The result can be either a smoke horizon or a heat
horizon, depending on how well developed the fire becomes
in the compartment where the pattern exists. Figure 10 shows a
smoke horizon and Figure 11 shows a heat horizon.
Confinement patterns can provide a fire investigator with
sequential data because if there is a smoke horizon or a heat
horizon on a wall, then in all likelihood, that pattern was
generated before the failure of the ceiling above it. This can
be very useful for determining the origin in a multilevel struc-
ture, and in some cases, the confinement pattern may be the
only pattern remaining. Even the absence of a confinement
pattern can help to eliminate a particular level of the structure
as the location of the origin. Figure 11 An example of a typical heat horizon. Photo courtesy of Mick
Consider a structure where everything has collapsed into Gardiner, Gardiner Associates, with permission. Copyright 2012 from
the basement. If there is a smoke or heat horizon on the Scientific Protocols for Fire Investigation, 2nd Edition by John J. Lentini.
basement wall, it is unlikely that the fire entered the basement Reproduced by permission of Taylor and Francis Group, LLC, a division
late in its development. Conversely, the absence of a smoke or of Informa plc.
heat horizon in the basement tends to eliminate the basement
as the origin.
A different kind of confinement pattern is sometimes pro- the location of the studs. Figures 12 and 13 show both sides of
duced when the heat of the fire was confined in a wall space such a confinement pattern. Similarly, fires burning upward
between two vertical studs. While the pattern on the side of the through a floor will be influenced by the structural members
wall where the fire originated may be unclear, the pattern on under the floor and patterns with parallel edges may be the
the opposite side of the wall may show a clear delineation of result. A fire penetrating downward through the floor, on the

Encyclopedia of Forensic Sciences, (2013), vol. 3, pp. 396-405

 Author's personal copy
400 Investigations | Fire Patterns and Their Interpretation

Figure 12 An example of a horizontal confinement pattern. Hot embers Figure 13 Opposite side of the wall shown in Figure 12. The outline of
from above fell onto the horizontal member, where they caused damage the stud space can be clearly seen. Copyright 2012 from Scientific
to the drywall. Copyright 2012 from Scientific Protocols for Fire Protocols for Fire Investigation, 2nd Edition by John J. Lentini.
Investigation, 2nd Edition by John J. Lentini. Reproduced by permission Reproduced by permission of Taylor and Francis Group, LLC, a division
of Taylor and Francis Group, LLC, a division of Informa plc. of Informa plc.

other hand, is unlikely to be significantly influenced by struc-

tural members below the floor.

Movement Patterns

When the fire moves from one room to another, it frequently

records movement patterns at or near the doorways. An exam-
ple of such a pattern is shown in Figure 14. Movement patterns
are useful for tracing a fire back to its origin and for helping to
eliminate the room into which the fire moved as the compart-
ment of origin. Movement patterns typically appear as diago-
nal patterns moving upward from a doorway.

Irregular Patterns

Irregular patterns are typically found on floors in compart- Figure 14 Typical movement pattern through a doorway. Photo
ments where fires have gone to flashover and burned for some- courtesy of Mick Gardiner, Gardiner Associates, with permission.
Copyright 2012 from Scientific Protocols for Fire Investigation, 2nd
time thereafter. For decades, fire investigators thought that they
Edition by John J. Lentini. Reproduced by permission of Taylor and
could identify irregular patterns on floors as having been
Francis Group, LLC, a division of Informa plc.
caused by the presence of ignitable liquids burning there.
This misconception resulted from training exercises designed
to teach participants to ‘recognize arson.’ These training exer- Figure 15 shows such a pattern, which can be characterized as
cises almost invariably involved the extinguishment of the fire ‘an obvious pour pattern.’
before it became fully involved. In such a case, it is indeed Figure 16 also shows an irregular fire pattern, but there was
possible to recognize a pattern caused by a flammable liquid. no accelerant involved in the test fire that generated this

Encyclopedia of Forensic Sciences, (2013), vol. 3, pp. 396-405

 Author's personal copy
Investigations | Fire Patterns and Their Interpretation 401

Figure 15 Irregular pattern caused by the burning of a medium Figure 17 This fire pattern was erroneously attributed to the burning of
petroleum distillate on a carpet. This is a rare instance where the cause of ignitable liquids. The unburned areas were protected by objects on the
the pattern can be interpreted as having been caused by ignitable liquids floor. Photograph courtesy of David M. Smith, Associated Fire
on the basis of appearance alone. Copyright 2012 from Scientific Consultants, with permission.
Protocols for Fire Investigation, 2nd Edition by John J. Lentini.
Reproduced by permission of Taylor and Francis Group, LLC, a division
pattern. This pattern was produced when the carpeted floor
of Informa plc.
was exposed to radiant heat from above. The carpet failed by
randomly tearing open and shrinking back in response to the
heat. Some parts of the floor were protected by the carpet and
other parts were exposed. The lines of demarcation between
burned and unburned areas were sharp, continuous, and
irregular.
It is these alternating exposure and protection patterns that
occur in fully involved compartments that are capable of con-
fusing fire investigators, particularly those who have not
watched test fires burn beyond flashover. The patterns shown
in Figures 17 and 18 were both interpreted by fire investigators
as indicating that flammable liquids burned on the surface of
the floor. Even though the laboratory samples tested negative
for the presence of ignitable liquid residues, arson and homi-
cide charges were brought against survivors of both fires. Con-
victions resulted in both cases.
A peer-reviewed study of test fires set with gasoline and
kerosene revealed that patterns produced by ignitable liquids
on floors often appear different than what fire investigators
might expect. The pattern shown in Figure 19 is similar to one
that an investigator characterized as ‘puddle-shaped,’ and
therefore caused by flammable liquid burning on the floor.
An actual gasoline fire on a similar parquet floor that resulted
Figure 16 A more typical irregular pattern caused by the random failure
of the carpet, resulting in alternate areas of exposed and protected in a pattern with an entirely different appearance is shown in
subflooring. Copyright 2012 from Scientific Protocols for Fire Figure 20.
Investigation, 2nd Edition by John J. Lentini. Reproduced by permission The attribution of holes in the floor to the burning of an
of Taylor and Francis Group, LLC, a division of Informa plc. ignitable liquid on top of the floor is almost always in error.

Encyclopedia of Forensic Sciences, (2013), vol. 3, pp. 396-405

 Author's personal copy
402 Investigations | Fire Patterns and Their Interpretation

out very quickly, and in the greater scheme of the overall fire,
provide relatively little fuel or energy.
It has been the misidentification of patterns similar to those
shown in Figure 16 that have caused numerous false convic-
tions for arson.
Laboratory analysis is required to accurately identify all
but the most obvious ‘pour patterns,’ and even in the case of
obvious patterns, a laboratory analysis certainly does no harm.

Spalling

Spalling is the chipping or pitting of concrete or masonry

surfaces in response to heat. The spalled area can range in
Figure 18 This fire pattern was erroneously attributed to the burning of
ignitable liquids. The unburned areas were protected by objects on the size from a few square centimeters to a few square meters.
floor. Because the result of spalling is a depression in a concrete
surface, spalling is often ‘puddle-shaped.’ For years, fire inves-
tigators attributed spalling of concrete to the intense heat
generated by the burning of an ignitable liquid. This is incor-
rect on a number of levels. First, spalling can be caused by the
differential expansion of the aggregate and the cement matrix,
by the presence of steel reinforcement bars, or by the vapori-
zation of the waters of hydration that are a part of the concrete.
More importantly, a flammable or combustible liquid
burns no more intensely than other carbonaceous fuels such
as wood. Burning ignitable liquids are no more likely to cause
spalling than they are to burn holes in wooden floors, instal-
ling is frequently seen on surfaces where no flammable liquid
could exist, such as a concrete ceiling.
Numerous experiments have been conducted to try to cause
Figure 19 A ‘puddle-shaped’ pattern, similar to one that was attributed spalling using ignitable liquids, and most of those experiments
to the burning of a puddle of ignitable liquid. Actually, this pattern was have failed. On the other hand, setting a well-ventilated wood fire
caused by the burning of a cardboard box. Copyright 2012 from Scientific above a concrete floor will almost invariably result in spalling.
Protocols for Fire Investigation, 2nd Edition by John J. Lentini.
Reproduced by permission of Taylor and Francis Group, LLC, a division
of Informa plc. Electrical Damage

When a fire damages the insulation of an energized electrical

conductor, the first thing that happens is the formation of
carbonized insulation, which can conduct electricity from
one conductor to another. The resulting current flow may or
may not be sufficient to activate the overcurrent protection,
that is, trip a circuit breaker or melt a fuse.
A fire investigator may observe multiple areas where this
‘arcing through char’ has occurred during the course of a fire.
Arcing through char will leave behind artifacts in the form of
globules melted copper. The first conductors that are likely to
be compromised by a fire are power cords to electrical appli-
ances. Branch circuit conductors behind walls and ceilings can
Figure 20 A fire pattern on a parquet floor that was caused by the also be damaged.
burning of 1 l of gasoline. By noting the location of arcing artifacts on various circuits,
the investigator may be able to infer the sequence of the arcing
events. Certainly, all the arcing events will have taken place
Ignitable liquids are incapable of existing at temperatures before the disconnection of power to the structure, and the
above their boiling point, so as long as liquid remains, the events on a particular circuit will have occurred before the
temperature on the floor will be less than the boiling point of activation of any overcurrent device protecting that circuit. If
the flammable liquid. Once the liquid has completely vapor- a particular conductor is severed, then any arcing event that is
ized, there will be no more heat from that source. Arsonists use electrically downstream of the point of severance can be logi-
ignitable liquids to cause a fire to spread more rapidly than it cally inferred to have occurred before the event that caused the
would otherwise. The ignitable liquids themselves tend to burn conductor to be severed.

Encyclopedia of Forensic Sciences, (2013), vol. 3, pp. 396-405

 Author's personal copy
Investigations | Fire Patterns and Their Interpretation 403

This mapping of arcs can result in a pattern that allows an Intensity patterns may be very confusing, because of the
investigator to determine which parts of the electrical system difficulty in determining whether the pattern was caused by a
were compromised first, and provide clues to the origin of fire of longer duration or of higher intensity. Attributing more
the fire. intense damage to longer duration frequently will result in
determining an incorrect area of origin. The test fire shown in
Figure 23 was set in the kitchen, which is at the left side of the
Clean Burn photo. Because of more fuel and better ventilation, however,
there was much more severe damage in the living room, shown
Clean burn is a fire pattern on surfaces where the soot has at the right side of the photo.
either burned away, or was never deposited because the sur- Often, fire patterns are a combination of movement, inten-
face was too hot. It stands to reason that if the surface had sity, ventilation, and plume-generated patterns.
soot deposited on it and then the soot was burned away,
chances are the fire originated elsewhere. The problem of
fire patterns is that they do not come with time stamps on
them. If the fire begins close to a wall, it may heat the wall to
Ventilation-Generated Fire Patterns
the point where soot is never deposited on it, yet there is a
In the early stages of a fire, the rate of heat release increases as
white area that is indistinguishable from a clean burn that
more fuels become involved. The plumes from these burning
occurred later in the fire.
fuel packages pump smoke and energy into a hot gas layer
Figure 21 shows a clean burn located directly above the
trapped beneath the ceiling. As the fire grows, the hot gas
origin of a test fire, on the left wall between the bed and the
layer becomes thicker and hotter. It radiates heat in all direc-
chair. It is unlikely that there was ever any significant amount
tions, including downward. When the temperature of the hot
of soot deposited in the white area. Video of the test showed
that the area indicated by the arrow was an area of early flame
contact with the wall. The white areas to the right of the
nightstand and to the right of the bed on the rear wall were
actually not burned. The wall above the nightstand caught fire
before the nightstand. This test fire was extinguished after only
10 s of postflashover burning. Figure 22 shows another venti-
lation-generated pattern from a similar test fire. It is interesting
to note that the pattern assumes a V-shape, but the pattern was
caused by a cloud of fuel-rich gases burning from the top
downward.

Intensity Patterns

Fires produce patterns as a result of the response of materials to Figure 22 Clean burn resulting from a fire that burned for
approximately 2 min beyond flashover. The V-shape developed from the
heat exposure. The various heat effects on materials can pro-
top downward. The fire originated between the bed in the chair. Photo
duce lines of demarcation. These lines of demarcation may be
courtesy of ATF Fire Research Laboratory and Steven W. Carman,
helpful in determining the characteristics and quantities of fuel Carman Fire Investigations, with permission.
materials, as well as the direction of fire spread.

Figure 21 Clean burns resulting from a short-lived test fire. The wall
between the bed and the chair never accumulated any soot. The wall and Figure 23 Although this fire started in the kitchen at the left, the living
ceiling above the nightstand were damaged by the burning of a fuel-rich room at the right exhibited far more damage. Copyright 2012 from
cloud of smoke produced by the chair. Photo courtesy of ATF Fire Scientific Protocols for Fire Investigation, 2nd Edition by John J. Lentini.
Research Laboratory and Steven W. Carman, Carman Fire Investigations, Reproduced by permission of Taylor and Francis Group, LLC, a division
with permission. of Informa plc.

Encyclopedia of Forensic Sciences, (2013), vol. 3, pp. 396-405

 Author's personal copy
404 Investigations | Fire Patterns and Their Interpretation

gas layer reaches 500–600  C, it radiates sufficient heat down- identify the quadrant of origin but of the 64 who did attempt
ward (on the order of 20 kW m 2) to ignite every exposed to identify where the fire started, only 44 correctly identified
combustible surface in the compartment. This begins the tran- the quadrant of origin.
sition phase known as flashover. In the third burn cell, the fire burned for 3 min beyond
The burning of all of this exposed fuel dramatically in- flashover. Seventeen investigators declined to make a determi-
creases the heat release rate within the compartment, but nation. Of the 53 investigators who responded, only 13 cor-
even more importantly, it consumes most of the oxygen. At rectly identified the quadrant of origin. This amounts to a
this point, the fire can burn only in those places that have 75% error rate, which is no better than if 53 untrained in-
sufficient oxygen. This fire is known as a ‘ventilation- dividuals had picked the quadrant of origin at random.
controlled’ or ‘ventilation-limited’ fire. Clearly, these snapshots of origin determination accuracy in-
The compartment in this case is full of fuel in the form of dicate that there are serious problems with the training of fire
combustible vapors, aerosols, and gases, as well as particles of investigators.
unburned fuel. The entire atmosphere is flammable. In an In an effort to understand what was causing the confusion,
atmosphere of air, one can create a flame by introducing and the experimenters reconstructed three burn cells at the ATF
igniting a flammable gas such as methane or acetylene. If the laboratory Fire Research Laboratory in Ammendale, Maryland.
atmosphere is methane, however, as occurs on some extrater- The burn cells were equipped with thermocouples and radi-
restrial bodies, one could create a flame by introducing a ometers and the burning was carefully monitored. The results
stream of oxygen. This is exactly what happens in ventilation- indicated that what was happening was that the fire was burn-
limited fires. ing most intensely at a point directly across from the single
There is likely to be more intense burning around ventila- opening, a doorway, as a result of the vitiation (lack of oxygen)
tion openings such as doors and windows. What has been of most of the compartment and a jet of clean air coming in at
learned lately, however, goes far beyond that and has impor- the bottom of the door.
tant ramifications for the validity of fire origin determination. The test fire was modeled using FDS (fire dynamics simulator)
Patterns are likely to be created not only around the ventilation and the output was translated using smokeview, a program
openings, but, because of the way a ventilation-limited fire that allows visualization of the numerical output of the model.
behaves, they may also be created on walls and other surfaces Figure 24 shows that the radiant heat flux opposite the doorway
opposite to the ventilation openings as a result of a jet of air was 150 kW m 2 on the wall. Figure 25 shows a ‘slice’ of the
being drawn in by the fire. smokeview output, where red indicates oxygen at a concentration
Ventilation-generated patterns can create significant confu- of 21% (air) and blue indicates a concentration close to zero.
sion. Fire investigators very often will use the level and inten- These results help us to understand what is happening in
sity of the fire patterns to determine the area of origin. The the production of fire patterns after flashover, but far more
‘lowest and deepest char’ is often singled out as the point where research is needed. This was a simple test with a single opening
the fire must have burned the longest, but this is not true. and a single compartment. Simple tests similar to this are
In 2005, agents of the Bureau of Alcohol, Tobacco and necessary to understand what is happening, but the conditions
Firearms (ATF) conducted an experiment at a fire investigation of this test seldom obtain in a typical structure fire. Many times,
seminar in Las Vegas, Nevada, to replicate similar experiments the initial ventilation source (before the windows break) is
they had conducted at their training facility in Glynnco, Georgia. from inside the structure. Many times, there are multiple ven-
They built two identical rooms furnished as bedrooms, and then tilation sources, especially after the windows break.
set them on fire and allowed them to burn for 7 min. The consumption of all the oxygen in the room can cause
They then asked 53 participants in the seminar to observe the fire at the origin to burn much more slowly there than
the fire patterns and determine which quadrant of the room elsewhere. If the origin is located in a place where a ventilation-
contained the origin. Three of the participants identified the generated pattern is overlaid, then information about the ori-
correct quadrant in room number one and three different gin may be lost.
participants identified the correct quadrant in room number Fire investigators’ confusion about how to correctly locate
two. Although it was argued that some of the participants the origin of the fire is not a trivial matter. It is generally
were not qualified fire investigators, and that time and resources accepted that if one does not correctly identify the origin,
were limited and investigators were not allowed to touch the one is likely to incorrectly identify the cause. Even more
evidence, the low success rate of the participants was a cause for concerning is the fact that when a fire investigator identifies
concern. an area of origin and then finds no competent ignition source
The ATF agents conducting the experiment formed a hy- or even a competent fuel source in that area, he is likely to
pothesis that the amount of postflashover burning would in- conclude that the fire was set using an open flame that was
fluence the investigator’s ability to correctly identify the removed from the scene, and that the initial fuel was an
quadrant of origin. The experiment was repeated with some ignitable liquid or some other highly flammable fuel that
modifications. was introduced and completely consumed. A false determi-
In 2007, three burn cells were constructed in Oklahoma nation of arson is likely.
City, and 70 fire investigators were asked to identify the quad- This is new knowledge for many fire investigators, and
rant of origin. In the first burn cell, the fire was allowed to burn upon being confronted with it, they are likely to reject the
30 s beyond flashover, and 64 of 70 investigators correctly scientific basis of the experiments and insist on their ability
identified the quadrant of origin. to determine an origin even in a fully involved compartment.
In the second burn cell, the fire was allowed to burn for 70 s What should happen is that once a compartment is identi-
beyond flashover. In this case, six investigators declined to fied as having been fully involved, every potential ignition

Encyclopedia of Forensic Sciences, (2013), vol. 3, pp. 396-405

 Author's personal copy
Investigations | Fire Patterns and Their Interpretation 405

Bndry
Smokeview 4.0.6 - Sep 15 2005 gauge
kW m-2

150

135

120

105

90.0

75.0

60.0

45.0

30.0

15.0

0.00

mesh: 2
Frame: 322
Time: 289.8

Figure 24 Smokeview output of an FDS model, showing the effects of ventilation. The fire originated to the right of the headboard, but air flowing in
through the bottom of the door resulted in a radiant heat flux of more than 150 kW m 2 on the wall opposite the door. Photo courtesy of ATF Fire
Research Laboratory and Steven W. Carman, Carman Fire Investigations, with permission.

Smokeview 4.0.6 - Sep 15 2005 Slice

X_02
mol mol-1

0.20

0.18

0.16

0.14

0.12

0.10

0.08

0.06

0.04

0.02

0.00

mesh: 1
Frame: 600
Time: 270.0

Figure 25 Smokeview slice showing the oxygen concentration in the plane of the door. Photo courtesy of ATF Fire Research Laboratory and Steven W.
Carman, Carman Fire Investigations, with permission.

source in that room should be examined, unless there is some Further Reading
compelling evidence that allows the origin to be narrowed
further than the entire confines of the compartment. An eye- Carman SW (2008) Improving the understanding of post-flashover fire behavior.
Proceedings of the International Symposium of Fire Investigation. Sarasota,
witness to the beginning of the fire or perhaps a videotape of
FL: NAFI.
the fire might meet these criteria. Certainly, the simple tradi- Lentini JJ (2006) Scientific Protocols for Fire Investigation. Boca Raton, FL: Taylor &
tional interpretation of fire patterns in such a situation is likely Francis.
to lead to error. NFPA (2011) NFPA 921: Guide for Fire and Explosion Investigations. Quincy, MA:
National Fire Protection Association.

See also: Chemistry/Trace/Fire Investigation: Analysis of Fire

Debris; Chemistry of Fire; Interpretation of Fire Debris Analysis; Relevant Websites
Physics/Thermodynamics; Thermal Degradation; Investigations:
Evidence Collection at Fire Scenes; Fire Scene Inspection CFITrainer.net – Sponsored by the International Association of Arson Investigators.
Fire.NIST.gov – The National Institute of Standards and Technology Bureau of Fire
Methodology; Types of Fires. Research website.

Encyclopedia of Forensic Sciences, (2013), vol. 3, pp. 396-405