[http://depts.washington.edu/vehfire/begin.

html]
Fire Science
Fire is familiar and frequently observed around us in settings as harmless and simple as
a candle on a table. Ignition can be as straightforward as striking a match, yet the
underlying phenomena are complex and involve numerous factors.

In this section the fundamentals of fire science will be covered.

The following pages will cover:

The Four Elements of Fire

Important Factors and Probability of Fire

Scientific Method
   Defining the question
   Collecting Data
   Forming Hypotheses
   Testing
   Conclusions

Practical Applications

Elements of Fire
There are four fundamental elements required for fire in any setting:

 Fuel
   Material to be burned

 Oxidizing Agent
   Generally oxygen in air

 Heat
   Heat energy to initiate the reaction.

 Propagation
   Unless the initial fire spreads beyond the initial fuel, it will self-extinguish

The interaction of these four elements of a fire can be demonstrated schematically in the
following manner:

(For an explanation of the diagram, use the mouse to point at the numbers in order.)

Important Factors and Probability of Fire

Fires in Automobiles Involve Random Factors

While the mechanisms of fire are clear, and in extreme conditions, easily predictable, the occurrence of
fire is probabilistic in nature. Particularly in vehicle fires, there are numerous factors, many of which (e.g.
wind, speed of travel, operating conditions) that occur in unpredictable ways. The probability of the
occurrence of a sustained fire is dependent on four major and commonly considered factors (or elements):

Probability of a sustained fire is determined by:

Fuel Mixture
Ignition Propagation
The actual number of underlying factors is much higher. Each of the additional factors results in a
probability distribution of its own related to conditions that will and will not support fire. The following is a
partial list:

Probability of ignition is determined by:

Fuel type Ambient temperature

Fuel volume Hot surface temperature
Fuel temperature Hot surface texture
Droplet size Residence time of vapors
Air flow around hot surface
Space geometry Location of spark or flame
Oxygen presence (volume) Energy of spark or flame
 

While tables may list the autoignition temperature of motor oil to be 500-700 degrees F, there are some
circumstances in which it will not ignite at 800 degrees F. However, given similar conditions, the probability
of autoignition of motor oil is greater at 800 degrees F than it would be at 500 degrees F.

Similarly, many factors influence the probability of fire propagation.

Probability of fire propagation is determined by factors including:

Initial energy of fire Droplet size

Fuel type Air flow
Fuel volume Geometry of space
Fuel temperature Ambient temperature
Adjacent fuel type Thermal responses of
 Adjacent fuel distance container surfaces
 

Consequently, there is a high probability that fire will not occur or propagate the first time some of the
conditions are present. Therefore, it is likely that some remaining general evidence may help define the
cause of the fire. For example, if there is an allegation of leaking fluid in the engine compartment that
fueled a non-collision fire, it would be likely that such fluid had leaked for some time before conditions
aligned to cause fire. Examination of the garage floor where the vehicle was parked may support or refute
this possibility. Or if a fire was caused by an electrical fault, it would be possible that prior to a fire, the
electrical circuit in question may have malfunctioned in some observable way. It is also true that fires can
occur even in the absence of predictive evidence.

Scientific Method
The process of logical and methodical investigation is often called “The Scientific
Method.” Although the name sounds as if there is a single agreed-upon approach, a short
internet search will quickly demonstrate that educational institutions from college level on
down use “The Scientific Method” with generally agreed-upon principles but some
variations as to the details. Because of the variations in the use of the terminology, we
will refer instead to “scientific methods.” Whatever vocabulary is used, the application of
scientific methods in vehicle fire investigation is of paramount importance.

Scientific methods include a combination of steps incorporating

both deductive and inductive reasoning:

1. Defining the question

2. Collecting data

3. Forming hypotheses

4. Testing hypotheses

5. Conclusions

Analysis is necessary between each of the five steps listed. In this section, we will review
the five steps and analysis used in scientific methods as applied to vehicle fire
investigation.

Defining the Question

The purpose of the investigation generally leads directly to the definition of the question
or problem. The “question” in vehicle fire investigation can commonly be broken down
into several individual questions: location of fire origin, ignition source and initiating fuel
source. Depending on the purpose(s) of the investigation, there may be other questions
as well. For example, if the investigation is being conducted to research the effect of a
particular component on fire propagation into the passenger compartment, then
propagation path will be an additional question added to the list.

Each question will require a methodical process of consideration using scientific

methods. Of course, in the field and in practice, the approach to all the questions will
overlap at least some of the time. Gathering data about the location of fire origin and the
initial fuel may well be conducted at the same time.

Collecting Data
After the questions are understood, and before the actual collection of data, analysis is
performed to identify the data needed and to design methods of collection. This may be
as straightforward as organizing inspection equipment and data collection checklists or it
may be complicated by the need to design disassembly protocols that maximize
documentation of conditions.

In fire investigation, even moving materials without the use of tools can disturb important
evidence. During one investigation of the total burn non-collision fire of a van with interior
engine access, careful removal of debris revealed that the remaining metal edge of the
polymeric engine cover was displaced at the floor level. Interviews later revealed that the
driver had engine trouble. While driving he had the engine cover set back so he could
adjust the carburetion while on the road. The documentation of the precise engine cover
position led directly to the identification of the cause of fire. Without a step-wise process
of inspection and data collection, key evidence would have been lost.

Data collection may also include ancillary sources, such as service manuals, witness
interviews, collection of reports and logs from emergency service agencies, and vehicle
service history. These sources, when available, often provide valuable information.

Forming Hypotheses
When determining the cause of a fire, scientific methods require all hypotheses that can
be supported based on principles of fire science to be considered for a list of conclusions.
Collected data are analyzed in the context of each of the investigation questions. For
example, when investigating a crash fire, one might formulate a hypothesis that each of
the liquid systems disturbed in the crash was a potential source for the initial fuel of the
fire. Data are used to confirm or refute each hypothesis, when possible. When data are
lacking to either confirm or refute a hypothesis, the investigator evaluates the possibility
of gathering more data or conducting testing.

Testing
According to the New Oxford American Dictionary, “test” is defined as “a procedure
intended to establish the quality, performance, or reliability of something.” In the context
of vehicle fire investigation, testing is often conceptual as opposed to physical (such as
burn-testing vehicles or components). Hypotheses can be tested conceptually by
evaluating the consistency of all conditions and information with known principles – that
is, by using deductive reasoning.

Vehicle burn tests are expensive and logistically difficult within city limits of most
municipalities. There are also limitations regarding the ability of the investigator to define
or control all the variables of importance in an actual fire event. When possible, however,
burn tests can be very useful in evaluating specific conditions or hypotheses. Tests of
individual components can also be performed to assist in understanding the potential for
fire in various hypotheses. For example, if one hypothesizes that a fluid was ignited on a
hot exhaust manifold, temperature measurements can be made of the manifold in a
similar vehicle under similar operating conditions. If the manifold is not hot enough for
autoignition of the fluid in question, the hypothesis can be eliminated. If the manifold is
hot enough, the hypothesis is possible. Unless all other hypotheses are eliminated, the
exhaust manifold is one, but not necessarily the only, possible ignition source.

Analyzing results of conceptual or physical tests may lead to collection of additional data
and formulation of new hypotheses to test. In such cases, the methodical process is
iterated until either all questions are answered or all possible sources of information are
exhausted.

Conclusions
During vehicle fire investigation, evidence is often consumed, the number of variables
that can affect a given problem are high (vehicle conditions, environmental conditions,
human action) and the ability to define conditions is limited. Because of the complexity of
fire events, the occurrence of fire is a probability function dependent on random
variables. The probabilistic nature of ignition was evident in the test-to-test variation of
fluid ignition temperature measured even as the attempt was made to reasonably control
for all factors [1]. In some instances, evidence may be sufficient to suggest more than
one hypothesis while insufficient to eliminate all but one. In such cases evidence may or
may not also suggest a greater probability of one or more of the hypotheses available.

When determining the cause of a fire, scientific methods require all hypotheses that can
be supported based on principles of fire science to be considered for a list of conclusions.
For a final list of conclusions, these hypotheses need to be 1) proven true to be included,
2) disproven to be excluded, or 3) considered possible if they cannot be either proven or
disproven. These three categories may be described in terms of their probabilities. Such
qualified conclusions may still be useful. For example, if a vehicle fire is being examined
as part of a recall investigation, knowing what did not cause the fire may allow the
investigator to eliminate it from consideration as part of a trend. The use of the scientific
method and the evaluation of all possible hypotheses are stressed in court settings and
in National Fire Protection Association NFPA 921 Guide for Fire and Explosion
Investigations.

References

1. Colwell, J. D., et al., “Hot Surface Ignition of Automotive and

Aviation Fluids,” Fire Technology, 41, pages 105-123, 2005.

Practical Application
In practical application, the steps of scientific methods may, and often must be,
performed simultaneously. To collect data effectively during an inspection, the
investigator must consider hypotheses as to origin, ignition, fuel and propagation path so
that the appropriate systems can be documented. Depending on the nature of the
damage (observation), the investigator may formulate a hypothesis as to an ignition
source, for example, then consider any physical evidence consistent and inconsistent
with the hypothesis (deductive testing), and either form a conclusion or revert back to
data collection. This process is repeated numerous times in an investigation as different
questions and hypotheses are evaluated. In time-limited investigations, as most are, this
is frequently possible only to a small degree. Some necessary information may not be
available during the first inspection, or protocols may need to be developed and agreed
upon for disassembly or disturbance to evidence. Additional inspections may then be
required to continue data collection.

In court environments, additional requirements are imposed upon the systematic process
of investigation. Courts require:

 Evidence to be relevant to the questions and issues

 Techniques to be reliable as demonstrated by their being

  Tested

 Peer reviewed

 With known error rates

 Generally accepted

The goals of this vehicle fire investigation course are to provide:

 Fundamental understanding of fire investigation for hypothesis generation

 Consideration of all alternatives

 Peer-reviewed and accepted approaches to vehicle fire investigation

 Bases for hypothesis testing

 Exposure to data and literature

 End of Fire Science Section

 Congratulations!
 You have completed review of the Fire Science section. Clicking the "next" button on
this page will take you to the Fuels section.

 Fuel Sources
  
 Motor vehicle environments have rich repositories of combustible materials, each of
which may be considered as an initial fuel related to fire origin and as secondary fuels
in fire propagation. Fuel sources exist in the three material states of liquids, solids and
gases.
 This section covers flammability characteristics of each of these sources. Choose the
fuel source from the list below or continue by clicking “next.”
  
 Vehicle Fluids     Polymeric Materials     Gaseous Fuels

Vehicle Fluids
 

Almost every fluid onboard a motor vehicle is flammable

under certain conditions.

If
 Temperature is sufficient to vaporize a fluid

And

There is a region in which the mixture with oxygen will support combustion

And

There is a sufficient ignition source present in the same region

Then
Each of these fluids may ignite and propagate fire to adjacent materials.

When it is necessary to identify the originating fuel source, the investigator must examine
each vehicle system to determine whether any fluid was involved in the cause of the fire.

Flammable and combustible fluids

 Gasoline and diesel fuel

 Coolant

 Engine oil

 Power steering fluid

 Automatic transmission fluid

 Brake fluid

 Windshield washer fluid

 Refrigerants and lubricants

During a fire investigation, it is often necessary to assess the possibility that one or more
of these fluids was the originating fuel. To do so, it is important to understand the
characteristics of each. (Battery acid is the one fluid that is not flammable.)

Flammability Characteristics of Vehicle Fluids

 

The flammability characteristics of fluids are measured by standard laboratory tests. The
procedures for these tests are rigorously defined to maximize repeatability of the
measurements. Researchers have also performed tests of flammability properties of
vehicle fluids in conditions more similar to those found in real-world vehicle fires. The
results of both laboratory and real-world testing together provide the investigator with an
understanding of the range of possible fire characteristics for the fluids of interest.
Standardized tests of flammability characteristics

Before presenting flammability data, it will be helpful to review the vocabulary used to
characterize the fluids.

Flash point: The minimum temperature at which the liquid gives off sufficient vapor to
form an ignitable mixture with air. (At flash point, there may not be sufficient vapor
produced to sustain a fire after the vapors are ignited.)

Autoignition: The lowest temperature at which a substance ignites without the aid of an

external energy source such as spark or flame.

Fire point: The minimum liquid temperature for which sustained burning of the liquid
occurs after ignition of vapors. (Fire point is slightly higher than flash point.)

Flammability limits: Flammability requires a fuel and oxygen together. Flammability

limits are the boundaries of high and low fuel concentration, within which flammability is
possible.

Combustible fluid: A fluid with a flash point above 100 degrees F.

Flammable fluid: A fluid with a flash point below 100 degrees F.

For more detailed descriptions of the definitions and information about the standardized
tests used to measure the values, click here.:

……………….>>>>>

Flammability Characteristics of Vehicle Fluids

Detailed Definitions
 

Definitions from the previous page are repeated

here with additional technical details as to how
they are measured.
  

Flash point

The flash point of a liquid is the lowest temperature of a liquid, as determined by specific
laboratory tests, at which the liquid gives off vapor at a sufficient rate to support a
momentary flame across its surface [1]. ASTM (formerly American Society for Testing
and Materials) publishes open and closed cup test procedures [2-4, 6] for repeatable
flashpoint measurements; the value will vary considerably with conditions. Most Material
Safety Data Sheets (MSDS) for vehicle liquids list flash point.

Autoignition

The term “autoignition” is used differently in different sources and contexts. NFPA 921 [1]
includes the following definitions:

Autoignition: Initiation of combustion by heat but without spark or flame.

Autoignition temperature: The lowest temperature at which combustion

material ignites in air without a spark of flame.

Autoignition temperature, a material property, varies substantially with conditions; the

term “minimum autoignition temperature” is sometimes presumed to refer to the
temperature measured for a given material using an ASTM specified procedure [5].

For more information about the variation in the minimum autoignition temperature
measured using small variations in conditions, click here.

Fire point

Fire point is the lowest temperature at which a material can produce vapors fast enough
to support continuous combustion [4, 6-7]. Flashpoint does not require combustion to be
sustained; fire point is generally somewhat higher than flash point.

Flammability limits

Flammability requires a fuel and oxygen together. Flammability limits are the boundaries
of high and low fuel concentration, within which flammability is possible.

According to NFPA 921 [1], it is the upper and lower concentration limit at a specified
temperature and pressure of a flammable gas or vapor of an ignitable liquid and air,
expressed as a percentage of fuel by volume that can be ignited.

Investigators rarely have enough information to know the flammability limits of fires being
investigated, but the concept is still inherently important. Because vapors have both
upper and lower flammability limits, the conditions in which fire is possible are
constrained. For example, the concentration of gasoline vapor in a closed gasoline tank
is above the flammability limit (too rich) and therefore will not ignite.

Combustible and flammable liquids

There are two additional terms that are sometimes encountered in fire investigation:
 Combustible liquid: A liquid having a flash point at or above 37.8 degrees
C (100 degrees F). [1]

Flammable liquid: A liquid that has a closed-cup flash point that is below
37.8 degrees C (100 degrees F) and a maximum vapor pressure of 2068
mm Hg (40 psia) at 37.8 degrees C. [1].

To view references for this page before continuing click here,

Flash Point and Autoignition Temperatures of Common Vehicle

Fluids

Laboratory Measurements
 

Laboratory flashpoint measurements provide useful information regarding the

temperature at which a fluid may release enough vapor to sustain a flame in ideal
conditions. Autoignition temperature measurements require more interpretation. In the
laboratory, autoignition is measured by placing samples inside nearly closed chambers,
without airflow, and with the benefit of instrumentation to identify even fragile and fleeting
ignition events. The values shown in the following table tell us the lowest possible ignition
temperatures for the fluids listed when under ideal circumstances.

The values in the table are presented for general reference and as such do not need to
be memorized. The tabulated values provide a way to compare the relative flammability
of various fluids and can be used as a resource for practical fire investigations.

Autoignition
Fluids Flashpoint [12] oF
Temperature [13] oF
Automatic Trans. Fluid [ 2, 4] 302-383 410-417
Brake Fluid [ 2, 4, 10, 11] 210-375 540-675
Compressor Oil ( PAG and ester) [ 4, 8] 392-500 410-714
Coolant    
    Ethylene Glycol (100%) [ 1, 2, 4] 232-260 725-775
    Ethylene Glycol (90%) [ 2] 270 N/A
    Propylene Glycol (100%) [ 1, 4] 210-230 700
Diesel Fuel [ 1, 2, 3, 4] 100-204 350-625
Ethanol (in gasohol) [ 1, 3, 5] 55 685
Gasoline (50-100 octane) [ 1, 2] -36 to -45 536-853
Gasoline (unleaded) [ 4] -45 495-833
Engine Oil (conventional and synthetic)
300-495 500-700
[ 1, 2, 4]
Methanol (in windshield fluid) [ 1, 2, 3,
52-108 725-878
4, 5,14]
Power Steering Fluid [ 2, 4] 300-500 500-700
Refrigerants    
    R134a 140 KPa (5.5 Psig) [ 7] 350  
    R134a[ 7,15,16] 1370-1418
Not flammable at ambient temp. and
    Freon 12 [17] >1382
atmospheric pressure
    HCFC-22 [ 9] Flammable at 60 psig
    Hydrocarbon Refrigerants Flammable Flammable
Starter Fluid (ethyl ether) [ 5,18] -49 320

Note about the table: When various sources had different flashpoint or autoignition
temperature values for the same material, the range in the table was increased to include
all values found.

To utilize flammability characteristics in investigations, it is necessary to also have

measurements made in vehicle environments.

To view references for this page before continuing click here,

Autoignition Temperatures of Common Vehicle Fluids

Vehicle Environments
 

The autoignition temperature measurements important to fire investigators are those that
best reflect the circumstances found in the vehicle operating environments. Because
there are enormous variations in the environments, there are also wide variations in the
potential ignition temperatures. The following table gives values for the various fluids
under different conditions.

Fluids Autoignition Temperature oF

Automatic Trans. Fluid 580 to 1120
Brake Fluid (DOT 3) 520 to 1065
Coolant  
    Ethylene Glycol (100%) 950 to 1245
    Ethylene Glycol (50%) 775 to >1200
     Propylene Glycol (100%) 960 to 1020
    Propylene Glycol (50%) 975 to 1065
Diesel Fuel 950 to >1200
    Biodiesel (B100, B20) 705 to 1300
Ethanol 1260 to 1330
Gasoline (87-92 octane) 1135 to1550
    Gasoline blend (E85) 1300 to 1325
Lubricating Oil 580 to 1130
Power Steering Fluid 590 to >1200

To construct this table, all sources were combined to establish the ranges. While concise
in what it presents, this method also hides some inconsistencies between published test
results. For example, some researchers reported that gasoline would not ignite up to
1200 degrees F., while others had ignition at 1100 degrees F. (The notation, “>1200,”
indicates there was no ignition when the tests were stopped at 1200 degrees F.)

If you would like more information about fluid ignition measurements, sources, the test
conditions and results, click here.

Combinations of Fuels
 

The presence of gasoline mixed with other onboard fluids significantly reduces the fire
point of the other fluids [1]. The graph below shows that even a small amount of gasoline
reduces the fire point of power steering fluid. Similar results were found for engine oil and
automatic transmission, power steering and brake fluids, and are likely for other onboard
fluids as well.

Fire Point of Power Steering Fluid/Gasoline Mixtures

  

Diesel Fuel Contamination with Gasoline

In another source [2], it was reported that contamination of diesel fuel with 5% gasoline
lowered the flashpoint to 90 degrees F below that of pure diesel fuel (from 167 to 77
degrees F). Even a small degree of contamination can therefore reduce the flashpoint of
diesel fuel to fuel temperatures expected during normal operation in some parts of the
country. However, a contact at the American Petroleum Institute [3] reported that on
average, diesel fuel is contaminated with approximately 1 cup of gasoline per 4,000
gallons of diesel fuel; this level of contamination would not be significant.

References

1. Ohlemiller, T.J., et al., “Aspects of the Motor Vehicle Fire Threat from
Flammable Liquid Spills on a Road Surface,” Building and Fire Research
Laboratory, NIST, August 1998.
 2. Mowrer, F.W., et al., “Heavy Truck Fuel System Safety Study: Final Report ,”
August, DOT HS 807 484, 1989.

3. Personal correspondence with API, August 2001

Process of Investigation
 

1. Identification of initial fuel: In a collision fire investigation, each fluid system must be
examined for collision damage and resultant leakage. In a non-collision fire investigation,
odors, service history or pavement stains where the vehicle is typically parked may
provide supporting information about the leakage of specific fluids, and therefore their
availability for fire (to see an Interview check list, click here). Witness descriptions of the
fire and burn pattern may also help differentiate between fluids in selecting those
possible for fire origination.

2. There are additional material properties not covered in the tables provided in this
course material, such as the energy available in a given volume of the fluids (heat of
combustion). In the absence of meaningful test data, the authors will present anecdotal
reports of testing and experience. This information can be used to assist in differentiating
between fuels in an investigation.

Gasoline: Most researchers found gasoline difficult to ignite on a hot

surface such as an exhaust manifold, particularly one without shielding
that may harbor vapors. However, once ignited, a gasoline fire burns with
an intensity greater than many of the other fuels and tends to propagate
quickly. [1]

In the Ignition section there is a case study of a

gasoline pool fire that spread quickly and blocked
occupant egress. To see it now, click here.

Engine oil, power steering and transmission fluids: These lubricants

tend to ignite readily on hot surfaces and by spark ignition. The authors
have also observed instances in which these fuels were able to sustain a
small fire for some time prior to propagation. When present adjacent to a
hot surface, these fluids have also been observed to re-start readily after
being extinguished.

Coolant: In tests of shielded manifolds, coolant was observed to readily

ignite on hot surfaces. However, the resultant ignition sometimes burned
the available vapor cloud quickly without propagating to other materials.
Coolant apparently was not likely to sustain a small, stable flame in the
manner of lubricants, and is thus considered less likely to propagate a
fire.

3. Trends: Autoignition temperatures vary by many factors. While no single autoignition

temperature value can be cited for a fluid, there are trends that can be relied upon. a)
The minimum autoignition temperature measured using the ASTM procedure can be
considered the lowest value probable. [2, 3] b) Autoignition temperatures in vehicle
environments are likely to be much higher than the ASTM values. [2] c) To estimate the
autoignition temperature in a given circumstance, it is helpful to review the range of
values measured in tests of various conditions. d) Air flow increases the autoignition
temperature. [4] e) Geometry that increases residence time decreases autoignition
temperature. [3]

 4. Fires are relatively infrequent events requiring the alignment of specific conditions. If a
defect is suspected to be the cause of fire for a vehicle in a fleet, one would expect to find
many such defects in similar vehicles that have not had a fire. As part of the
investigation, one should study similar vehicles for their history related to the suspected
problem.

 5. Assessment of a liquid as the initiating fuel for a fire must be made while
simultaneously considering the ignition sources available. Ignition sources are covered
separately in the course (see the Ignition section in the index on the left of the screen, or
read ahead with the "next" button to reach the section later in the course).

References

1. Santrock, J., “ Evaluation of Motor Vehicle Fire Initiation and Propagation Part
4: Propagation of an Underbody Gasoline Pool Fire in a 1996 Passenger
Van”, NHTSA 98-3588-143, 1998.

2. Shields, L, Scheibe, R., “Computer-Based Training in Vehicle Fire

Investigation-Part 2: Fuel Sources and Burn Patterns,” SAE 2006-01-0548,
2006.

3. Colwell, J. D., et al., “Hot Surface Ignition of Automotive and Aviation Fluids,”
Fire Technology, 41, pages 105-123, 2005.

4. Santrock, J. “Flammability Properties of Engine Compartment Fluids Other

than Gasoline,” NHTSA 98-3588-193, 1998.

1991 Mitsubishi Eclipse: Case Study

Overview

 
 Collision damage was limited. Some roof crush from the rollover was evident.
 

 1991 Mitsubishi Eclipse

 At 70-75 mph, drove off right shoulder to
allow ambulance right of way
 Lost control, override of culvert, rolled on
roof
 Undercarriage damage, low to moderate
Delta V
 One occupant, back injuries

A 1991 Mitsubishi Eclipse was traveling eastbound on a two-lane road in mid afternoon.
There were no known adverse road conditions at the time of the incident. The driver said
he was traveling 70-75 mph. He pulled off to the right on a gravel shoulder to allow an
ambulance the right of way, and lost control of the vehicle at the edge of the roadway.
The driver steered left to re-enter the roadway and overcorrected with a right steer; the
vehicle left the right side of the roadway and the undercarriage struck culvert. The vehicle
rolled, coming to rest on its roof off the roadway. A passing motorist stopped and
informed the driver that the vehicle was on fire. The driver exited the vehicle through the
door then walked away from vehicle and lay down due to back pain. The driver was the
only occupant in the vehicle.

The driver first turned and observed the fire after he walked approximately 15 to 20 feet
away from the vehicle. He said he saw the fire in the center of the (upside down) vehicle
toward the front of the car. The fire was accompanied by smoke. By the time the driver
walked approximately 50 feet and laid down, a passerby with a fire extinguisher
extinguished the fire.

 
There was no fire or collision damage evident in the upper engine compartment.
 

Vehicle Damage
Impact with the culvert and subsequent rollover severely damaged the front suspension,
engine oil pan and exhaust system. The right front suspension strut was pulled from its
upper mount. The oil pan was breached and spilled its contents onto the exhaust pipe
located adjacent to the oil pan. Black oil was observed on the grass in the culvert two
days after the incident. Portions of the exhaust pipe including the catalytic converter were
missing and appear to have been separated from the vehicle as a result of the impact
with the culvert. Other collision damage to the vehicle was confined to the front fascia
(bumper area), roof, rear fenders and tail lamps.

Injuries
This vehicle is equipped with a 2 point, motorized shoulder belt and manual lap belt
restraint system. The occupant was wearing the shoulder belt portion of the system only
at the time of the incident. The driver was taken by ambulance for emergency room
treatment and released. Injuries included pulled back muscles, a bruised left knee and
scratched left foot.

 
 Fire Damage, Location and Propagation
Post fire damage on the vehicle was minimal. From witness accounts and inspection of
the vehicle, the fire was confined to the area of the exhaust system adjacent to the oil
pan. No other heat damage was evident on the vehicle.

1991 Mitsubishi Eclipse: Case Study

Fuels Available

The front of the vehicle is left in this photograph of the underside. The oil pan is
torn away exposing the crank shaft. The exhaust pipe (bottom of photograph) is
scratched, heat shields deformed, and steel braiding torn (lower right) exposing
the corrugated/convoluted portion of the exhaust pipe.
 

 Undercarriage damage breached oil pan and spilled onto

ground and exhaust system

 Oil was the only source of fuel

The oil pan was breached from impact with a culvert. While most of the engine oil fell to
the ground, some splattered onto the exhaust system. Coolant and clutch fluid spilled
after rollover. Since the vehicle came to rest on its roof, these spills would have fallen to
the ground, remote from any evidence of fire damage.

1991 Mitsubishi Eclipse: Case Study

 Ignition and Propagation
 

Ignition
 Exhaust system shielded by a braided wire
covering
 Fire sustained by oil vapor provided by the
hot surface of exhaust pipes
 Entrapment of oil increases likelihood of
ignition
 Fire damage confined to exhaust system

Exhaust pipe with torn braided wire cover.

 

The exhaust pipe in the area adjacent to the oil pan was partially enclosed by a metal
shield and braided wire covering. The pipe geometry under the braided wire cover was of
convoluted/corrugated pipe design. Entrapment of the oil in exhaust pipe covers may
have facilitated autoignition of oil vapor on the hot pipe surface. The shielded exhaust
pipe during 70-mph operation was likely to have been hot enough to ignite engine oil.

The only other potential ignition source for the oil vapor was mechanical spark resulting
from interaction between the culvert and undercarriage components. However, a hot
surface provides a more continuous ignition source for combustible mixtures. The
mechanical sparks due to impact were instantaneous; if a combustible mixture was not
present at that precise moment, no ignition would have occurred. Therefore, mechanical
spark ignition was considered less likely.

Propagation
 Passing motorist informed driver of fire
 Fire witnessed in center of overturned
vehicle toward front as driver exited vehicle
 Fire extinguished as driver walked 50 ft.
from vehicle
 No consumables above exhaust system
reduced propagation

Close up of area on convoluted/corrugated exhaust pipe blackened by fire.

Post fire damage on the vehicle was minimal. From witness accounts and inspection of
the vehicle, the fire was confined to the area of the exhaust system adjacent to the oil
pan. No other heat damage was evident on the vehicle. After the vehicle came to rest, a
passing motorist stopped and informed the driver (while he was still in the vehicle) that it
was on fire. The fire was promptly extinguished by a passer-by.
 Because the vehicle came to rest on its roof, the exhaust system was a high point on the
vehicle, with no consumable materials above the flame. Therefore, the fire was sustained
by the oil vapor generated by the exhaust pipes, and did not propagate prior to being
extinguished. It is unknown if the fire would have propagated or simply used the available
fuel supply (oil) and self extinguished. Engine oil appears to have been the only fuel
available in this incident. Evidence indicates that the majority of oil was spilled while the
vehicle was on its wheels and very little oil was found splattered in the engine
compartment that would have facilitated propagation of the fire to other areas of the
engine compartment. Even the undercarriage of the vehicle aft of the oil pan exhibited
little evidence of oil spray.

1991 Mitsubishi Eclipse: Case Study

Summary of Fire Investigation

 

Propagation
 Fuel: Engine oil
 Ignition: Hot surface likely, mechanical
spark possible
 Initiation: Immediate
 Propagation time: Fire extinguished before
propagation to passenger compartment

The oil pan was breached by collision with the culvert after the vehicle left the road.
Engine oil was splattered directly on the hot exhaust system adjacent to the oil pan. Oil
collection occurred in the mesh covering around the exhaust pipe and was likely ignited
by the hot surface. Propagation of fire was minimal because of its location (high on the
upside down vehicle) and because it was extinguished rapidly.

Polymeric Materials
 

The use of polymers (plastics) in motor vehicles has been increasing steadily in recent
years. Although there are benefits for vehicle weight, fuel economy, styling and other
features, the increased use of polymers also means more fuels in the vehicle
environment. One trade journal provided the following summary of polymer use for 1996
model vehicles.

Polymer type Abbreviation Usage Average Wt

(lb/’96 car)
Polyurethane PU Body panels, fenders, bumpers, headliners, 44 lb
upholstery
Polypropylene PP HVAC, fan & shroud, battery tray, cowl, ducts, IP, 40
package shelf
Polyvinyl Chloride PVC Bumper trim, electrical wiring, boots, bellows, seat 21
cover, steering wheels, floor
Polyethylene PE Fuel tank, bumper, electrical wiring, reservoir, fuel 20
filler
Nylon PA Fuel system, fuel lines, gas cap, grill 18
Acrylonitrile butadiene ABS Bumper, fascia, headliner, duct 16
styrene

References

1. Plastics Newsletter, Market Search Inc., April 1996

2. Polymer Flammability Standards

3.  
4. There are standards and ratings for polymers that are used by designers in selecting
materials for the motor vehicle environment.
5. Manufacturers are required to comply with Federal Motor Vehicle Safety Standards
(FMVSS) for all vehicles sold in the United States. FMVSS 302 specifies that materials
used in vehicle interiors have a limited burn rate.
6. Underwriters Laboratories rates material burn rates in tests with samples mounted
horizontally and vertically (UL 94). When choosing materials, such as wiring insulation,
designers can choose the rating they believe is suitable.
7. To review FMVSS or UL standards now, click here.

8. Definitions
9.  
10. There are a few terms that are helpful in discussing the flammability properties of
polymers.
11. Amorphous: Non-crystalline and without long range order. Many polymers do not have
crystalline structures, and therefore cannot be said to “melt.”
12. Melting temperature: In common language, we say that a solid “melts” when it turns to
a liquid. In more formal language we generally understand “melt” to refer to heat-
induced motion of molecules that causes the breakdown of crystalline structure.
13. Decomposition temperature: A measurement of the thermal stability of a polymer.
Oxidation begins at the decomposition temperature. The decomposition temperature is
shown in one of the tables to follow.
14. Polymer: Large molecules of nonmetallic elements composed of many repetitive units
(mers). Commonly called plastics.
15. Vaporization temperature: Some polymers begin to vaporize before they appear to
oxidize. In such cases, the vaporization temperature is a measure of thermal stability.

Properties of Polymers
 

Polymers may have vastly different flammability characteristics depending on their

chemical composition, the structure (e.g., the amount of cross linking or chemical
bonding between adjacent polymer chains), and other factors. The investigator does not
generally have information as to the specific polymers used in the vehicles examined.
Therefore, investigators can simply examine the range of melting, decomposition and
ignition temperatures possible and assess the circumstances accordingly.

The melting temperature, decomposition temperature, and ignition temperature for

various polymeric materials commonly found in automotive applications are provided in
the following table.

 
Temperatures (oF) [1, 2]
Polymer
Melting Decomposition Ignition
PU (rigid) 248-320 N/A 590
PU (foam) 852-1074 N/A N/A
PP 331-361 793-806 829
PVC 167-415/ Amorph 491-516 675-945
ous
PE 251-286 824 829-910
Fiberglass 802-932 N/A 1040

According to one source [2], PVC has a melting temperature as low as 167 degrees F.
The authors consulted a materials scientist who works with polymers and two other
sources [1,3] and were unable to confirm such a low melting point.

 
 Polymeric gasoline tank from Explorer (pictured earlier). Note
that though vehicle appeared to be a complete burn, there
were some consumables remaining, including the majority of
the fuel tank (which show drips of molten material in this view).
The top of the tank was consumed, but a fluid level line inside
the tank was visible when viewed from above. When the inside
of the tank was inspected, the bottom portion was clean and
top portion exhibited soot deposits, indicating that gasoline
remained in the tank during the fire.

References

1. Tewarson, A., et al., “Characterization of the Ignition Behavior of Polymers

Commonly Used in the Automotive Industry,” NHTSA-98-3588-71, 1998.

2. NFPA 921, Guide for Fire and Explosion Investigations, 2004 Edition.

3. Osswald, T. A., Menges, G., Materials Science of Polymers for Engineers,

Hanser Publishers, 1996.

4. Process of Investigation
5.  
6. 1. Investigators can use the general pattern of burn damage to polymers to assess the
point of origin of fire.
7. An example can be found in the case of a 1993 Ford F150 to be covered later in the
course. If you want to see the relevant pages of the case study, click here.
8. 2. Ignition temperatures of polymers are low enough for autoignition on a hot exhaust
system. Fire can occur if wiring insulation or other polymeric parts contact hot exhaust
system components during vehicle operation or crash deformation.
9. 3. Differences in polymer properties may also complicate interpretation of the evidence.
Sometimes the variation in burn damage of adjacent parts may be caused by
 differences in flame resistance of the polymers involved rather than the characteristics
of the fire. Identification of such differences can be investigated through controlled tests
or by research of manufacturering specifications.
10. 4. Degradation of polymeric components can lead to fluid leakage, causing non-
collision fires as well. Evidence of the original leak is likely to be destroyed early in the
fire. Identification of degraded parts leading to fire may be aided by inspection of similar
non-fire vehicles.
11. 5. The temperature history of polymers (including paints) can sometimes be
determined through laboratory analysis. For example, paints provide characteristic
“signatures” when submitted to a process called Thermal Gravimetric Analysis (TGA)
that may give a clue as to prior temperature exposure. This is because there are
certain irreversible changes that the paint may undergo during heating, such as
solvents that are driven off at particular temperatures. Similarly, other polymers
(especially “thermoset” plastics) may suffer heat exposure degradation, which is not
visually apparent. Analytical techniques such as Fourier Transform Infrared (FTIR)
spectroscopy can assess the characteristic fingerprint of molecular bonding, often
affected by heat exposure. Other methods, such as Differential Scanning Calorimeter
(DSC), provide information on melt history of thermoplastic materials. In such cases, it
is usually helpful to have an undamaged sample of the polymer for comparison in order
to aid interpretation of the findings.

Gaseous Fuels
 

Current applications of gaseous fuels in vehicles

Propane: In current vehicle applications, there are many recreational vehicles with
propane powered appliances and a modest number of vehicles that have been converted
to run on propane to power the vehicle.

Hydrogen and Methane: Manufacturers are currently exploring the use of hydrogen and
methane (natural gas) as fuels to power general production vehicles. There are some
vehicles and fleets on the road that have been converted to methane power.

Standards

Vehicle manufacturers must comply with federal standards for crash integrity of
compressed natural gas systems. FMVSS 303 regulates fuel system leakage in crashes
of compressed natural gas vehicles and FMVSS 304 regulates compressed natural gas
fuel container integrity when subject to over-pressurization, pressure cycling and bonfire
testing.

In the Resource Section there is a listing of relevant standards. To see it

now, click here.
 Characteristics of Gaseous Fuels
Hydrogen, methane and propane are all capable of being ignited by relatively
small sparks from shorted wires, static discharge or electric motor operation.
Characteristics of each are presented in the table below.

Hydrogen: Odorless and lighter than air, it will disperse unless contained.

It can be generated by battery charging (for more information about
batteries and fire, click here)

Methane: Also known as natural gas, it is odorless by itself, but when

manufactured it often has a rotten egg odor added to aid detection.
Methane is lighter than air.

Propane: Propane by itself has a sweet odor, but often has an added

odor of rotten egg to aid detection in the event of a leak. Propane is
heavier than air and will gather in low areas if released.

  Autoignition Flammability Minimum Ignition Vapor

(oF) Range (%) Energy (mJ) Density
(air=1)
    Lower Upper    
Hydrogen 932 4 75 0.019 0.1
Methane 999 5 15 0.28 0.6
(natural gas)
Propane 842 2.1 9.5 0.25 1.6

The course section on Alternative and  Hybrid Vehicles will further explore vehicles with
gaseous fuels.

References

1. NFPA 921, Guide for Fire and Explosion Investigations, 2004 Edition.

2. NFPA 497, Recommended Practice for the Classification of Flammable

Liquids, Gases, or Vapors and of Hazardous (Classified) Locations for
Electrical Installations in Chemical Process Areas, 2004.

3. NFPA 325, Guide to Fire Hazard Properties of Flammable Liquids, Gases and
Volatile Solids, 1994.
 End of Fuels Section
 

Congratulations!

You have completed review of the section covering ignition sources in motor vehicles.
Clicking the "next" buttons on this page will take you to the section called "Ignition."

 Ignition

Identification of ignition sources is often the key to understanding the cause of a fire or
how such fires may be prevented in the future. When burn damage is extensive,
evidence of initiating ignition sources can be consumed, making ignition source
identification difficult. Even in such circumstances, it may be possible to eliminate some
potential causes.

This section will cover ways of identifying ignition sources in the vehicle environment, and
differentiating which are applicable to a given case.

Each ignition source has characteristics by which it can be ruled in or ruled out of
consideration for a specific vehicle fire. While the number of ignition source types is
small, each is multifaceted, with a potential for fire origination at many locations. Each
may leave little or no evidence for the investigator. Using scientific methods the
investigator may examine each source and evaluate the probability of involvement, but
often cannot definitively identify a single source.

Types of ignition sources in motor vehicles to be covered:

Hot surface (autoignition)

Electrical

 Arcing

 Resistance Heating

 Static

Mechanical

Autoignition / Hot Surface Ignition

Hot surface ignition is governed by the temperature of surfaces in the vehicle,
the autoignition properties of the flammable materials, and factors that influence these
parameters. While autoignition temperature is measured in a laboratory, the actual
temperature of ignition in a vehicle is significantly affected by many factors. Autoignition
is dependent on the property of the material being ignited (tendency to ignite at a given
temperature) [1], the temperature of the surface in question, and other environmental
factors. Environmental factors include air flow, surface condition, and availability of
oxygen. Both air flow and surface condition may influence the potential for ignition on
surfaces of the same temperature. Air flow may affect air/fuel mixture and the time for
transfer of heat. A rough or scaly surface may provide pockets that can increase the time
fuel remains in the presence of heat (residence time).

This section will provide information necessary to assess the possibility of autoignition in
fire investigations. The following is a list of subtopics covered.

Frequency of autoignition fires

Underhood surface temperatures

Process of investigation

Case study

Examples of recalls for autoignition

References

1. Shields, L, Scheibe, R., “Computer-Based Training in Vehicle Fire

Investigation-Part 2: Fuel Sources and Burn Patterns,” SAE 2006-
01-0548, 2006

Autoignition: Frequency of Fires

Many studies have published estimates of the proportion of vehicle fires that resulted
from autoignition of gasoline. The vast differences in results are evidence of the difficulty
of identifying ignition sources.

Butler, 1987 90% of engine compartment fires due to leaking

[1] fuel, mostly hot-surface ignition after stopping.

20% of 1,376 fires due to gasoline igniting on

Fray, 1991 [2]
hot surfaces.

Hrynchuk, Test conclusion: Hot manifold not sufficient to

1983 [3] ignite gasoline.

 
 Nevertheless, there is little doubt that such fires do occur and are also misattributed. It is
therefore necessary to understand the mechanism and potential for autoignition of
vehicle fluids.

References

1. Butler, J.A., "Motor Vehicle Fires in Ireland and the United

Kingdom,” Fire Prevention Journal, UK, 1987.

2. Fray, M., Southall, D., Galer, I.A.R., "Reducing the Risk of Non-
Electrical Car Fires," SMMT, UMTRI - 82645, 1991.

3. Hrynchuk, R.J., "A Study of Vehicle Fires of Known Ignition

Source," International Association of Arson Investigators, Alberta
Chapter, Edmonton Alberta, January, 1983.

4. Surface Temperatures: Underhood

5. Underhood surface temperature is one of the many factors upon which autoignition is
dependent. There is limited published information available on underhood surface
temperatures. Adding to this difficulty is the fact that exact operating conditions at the
time of fire that affect underhood temperatures are also difficult to determine. The
following information is based on available published tests, unpublished tests
conducted or seen by the authors and general principles of vehicle mechanics. [1-18]
6. Typically the exhaust system or the catalytic converter surfaces are the hottest
on a vehicle
7. In a normally operating vehicle, only exhaust system components are
hot enough to autoignite fuels in the motor vehicle. Surfaces for
potential ignition include exhaust manifolds, exhaust pipes nearest the
manifold, catalytic converters and turbochargers.
8. Exhaust manifold temperatures vary for different vehicles and operating
conditions
9. Manifolds and/or exhaust pipes on some vehicles can reach 1200
degrees F. It is rare to find temperatures this high in normal operation.
The hottest locations tend to be those for which there is some
constriction or impingement of exhaust gas. For example, a bend in an
exhaust manifold tube immediately outside the cylinder will have a
surface upon which the exhaust gas has a greater capacity to transfer
heat. Exhaust system temperatures will be reduced at any point of
contact with cooler components of large thermal mass, such as the
cylinder head.
10. Turbocharger system components in contact with exhaust gas are
frequently as hot or hotter than other exhaust system components.
11. A published study [10] of a 1996 Ford F150 V8 pickup truck showed
temperature of the exhaust pipe entering and exiting the first catalytic
converter on a level road and upgrade to be the following:
12. Exhaust System Temperatures at 60 mph (Degrees F.)
 Catalytic Catalytic
 
Converter Inlet Converter Outlet

Level road 757 770

7% grade 979 972

13.  
14. A published study [5] of four vehicles showed manifold temperatures in
the ranges provided in the following table:
15. Exhaust Manifold Temperatures (Degrees F.)
Qualification

  30 mph 70 mph

Level road 250 825

7% grade 360 1020

16.  
17. Catalytic converter temperatures also vary for different vehicles and operating
conditions
18. In a normally operating vehicle, the catalytic converter may be 750
degrees F, or more.  If one or more cylinders are not functioning and
unburned gasoline (or a richer fuel/air mixture) flows into a catalytic
converter, then temperatures can increase precipitously. The source
cited above [5] also contained the following catalytic converter
temperatures for the same four vehicles.
19. Catalytic Converter Temperatures (Degrees F.)
Qualification

  30 mph 70 mph

Level road 310 430

7% grade 640 735

20. Modeling and dynamometer data show higher catalytic converter

temperatures, but these sources did not incorporate representative
airflow [8,9].
21. Peak temperatures occur at road load, not idle
22. Peak temperature is related to engine speed (RPM), engine load, and
the volume of exhaust gas flowing through the exhaust system.
23. Higher temperatures occur in extreme and improper operating conditions
24. When an engine is operated at high RPM or load (upgrade, towing,
high wind, acceleration), exhaust system temperatures can increase
dramatically. Temperatures may also increase if the engine is not
functioning properly resulting in improper combustion and unburned
gasoline exhausting into the catalytic converter.
25. Cooling occurs within minutes of shutdown
26. Tests have shown hot surfaces can cool down as much as 400 degrees
F within 3 minutes of the vehicle being brought to a stop and the engine
being shut off [13]. The hotter the surface, the faster the expected rate
of cool down. The majority of data available to the authors have shown
that exhaust manifold and pipe temperatures fall immediately upon shut
down, and drop below fluid autoignition temperatures within 3 minutes
[14].
27. Catalytic converter temperatures may rise after the engine is shut down
due to continued chemical activity in the substrate material or heat
transfer from hotter internal components.
28. Further discussion of these data is available in the detail section for this
page.
29. Shielding may increase the likelihood of autoignition
30. Even when exhaust system shielding doesn't significantly increase the
temperature of a surface, the presence of a shield may increase the
chance of autoignition in at least two ways. First, by reducing air flow,
there is a greater opportunity to achieve air/fuel mixtures that support
combustion. Second, in stagnant air, the fuel may remain in the
presence of the heated surface longer and allow enough energy to be
transmitted to achieve combustion (higher "residence time"). A shield
may also allow an initial fire to burn and propagate when it would
otherwise be blown out. [13, 15]
31. Engine cylinder head and block temperatures are cooler than exhaust system
temperatures
32. The greater thermal mass of the engine and block, and the presence of
the cooling system, prevent the engine block and cylinder head from
achieving temperatures in the autoignition range for normally operating
vehicles. Engine surface temperatures increase for a period of time
after shutdown, but not to the threshold of autoignition.
33. A surface can be heated by dragging against the pavement
34. In addition to heating by exhaust gases, metal components dragging
against the pavement may become hot enough to support autoignition
of fluids. Steels may demonstrate their temperature history through
colors that remain after they are cooled. Such dragging may occur as a
result of collision or a component becoming loose enough to contact
the road.
35. Dragging metal also provides mechanical sparks as an
ignition source. You can click on the link below to review
an upcoming section on mechanical spark ignition now,
or continue studying autoignition.
36. Study Mechanical Sparks? Click here.
37. These are simple guidelines regarding the temperature of vehicle surfaces to support
ignition; if more definitive information is needed, tests should be performed. With
respect to exhaust systems, there are tests of surface temperature in published
literature related to a few vehicles and specific operating conditions. The variations
based on operating conditions (speed, grade) and vehicle model can be substantial.
Some conclusions can be drawn from this information, however. For example, it is fairly
certain that no surface in a normally operating vehicle will be of sufficient temperature
 after 5 minutes of idle operation for autoignition of coolant on an unshielded exhaust
manifold [3, 5, 13].
38. For references and more detailed temperature test data, click here.

Autoignition: Ignition Is a Probabilistic Function of Surface

Temperature
Hot surface ignition of fluids involves a complicated function of factors, surface
temperature being just one of them. Even at a constant surface temperature in the same
engine, it is possible to have ignition at some times and not others. For example, if wind
increases, the mixture may become too lean such that ignition is no longer supported.
Some tests have shown that ignition of an underhood fluid in a controlled laboratory
environment will have three ranges of surface temperature [1,2,3]:

1. No ignition is observed

2. Ignition occurs some of the time (becoming increasingly more likely as the
temperature increases)

3. Ignition occurs virtually all the time (under the same conditions)

The graph (below) showing this trend is included from one such study. For engine oil
ignition, the probability of ignition below about 570 degrees F (300 degrees C) is near
zero; it increases as shown to about 635 degrees F (335 degrees C), and ignition occurs
predictably above that temperature [1].

To understand the potential for autoignition, compare the autoignition temperatures of the
fluid in vehicle environments to the hypothesized surface temperature and determine
which of the three regions of ignition probability is likely to be appropriate to the case
being investigated.
References

1. Santrock, J. “Flammability Properties of Engine Compartment Fluids Other

than Gasoline,” NHTSA 98-3588-193, 1998.

2. LaPointe, R., et al., "Hot Surface Ignition of Gasoline on Engine Materials,"

SAE 2006-01-1013, 2006.

3. Davis, S., et al., "Hot Surface Ignition of Flammable and Combustible

Liquids," SAE 2006-01-1014, 2006

Autoignition: Videos of Ignition Testing

The previous pages in this section have introduced the concept of autoignition (hot
surface ignition) in vehicle fires and provided some information about the surface
temperatures available in vehicle environments under various operating conditions. The
page showing the probabilistic nature of the ignition events also demonstrates the
complicated nature of hot surface ignition. That complexity is also apparent in videos of
hot surface ignition tests.

The high speed video (immediately below) shows a controlled stream of 0.25ml diesel
fuel contacting a hot surface maintained at a temperature of 750 °C. The fuel vaporizes
on the hot surface and vapors are somewhat contained by the recessed fixture; ignition is
delayed until a combustible mixture is formed over an area of the plate. Combustion
continues until all liquid is vaporized. Note: The video is being played back in quarter
speed.

Video courtesy of Exponent, Inc.

 

The next video shows a similarly controlled stream of bio-diesel (B100) contacting the hot
surface maintained at 450 ⁰C in the same test fixture. This high speed video, shot at 500
frames per second and played at 1/8th speed, again shows the ignition delay and more
clearly the spread of combustion through the vapor cloud. Once the hot surface ignites
the vapor in one area, piloted ignition spreads rapidly through the rest of the vapor. It is
easy to see how introduction of air flow could disrupt the ignition during the delay.

Video courtesy of Exponent, Inc.

 

More information about the testing can be found in references 1 and 2.

The final video was taken by an infra-red video camera and also involved hot surface
ignition of bio-diesel on a steel surface. In this test, a single drop is released onto the
surface; the temperature distribution of the surface can be interpreted by comparing the
color of the image to the legend. Again the liquid vaporizes on the surface and ignition is
delayed until an appropriate mixture contacts an area of the surface of sufficient
temperature. The video also shows the motion of the droplet on the flat surface as it
vaporizes.

Video courtesy of Fire Research Group, University of Waterloo

 

More information about the testing can be found in reference 3.

References

1. Davis, S., Chavez, D., and Kyotma, H., “Hot Surface Ignition of Flammable
and Combustible Liquids,” SAE 2006-01-1014, 2006.

2. Somandepalli, V., Kelly, S., and Davis, S., “Hot Surface Ignition of Ethanol-
blended Fuels and Biodiesel,” SAE 2008-01-0402, 2008.

3. Byers, K., Epling, W., Cheuk, F., Kheireldin, M., and Weckman, B.,
“Evaluation of Automobile Fluid Ignition on Hot Surfaces,” SAE 2007-01-1394,
2007.

Autoignition Characteristics of Fuels

The autoignition and other characteristics of fuels available in the motor vehicle
environment are presented in the "Fuels" section. You can click on the link below to
review that information now, or continue with studying ignition.

Review Characteristics of Fuels? Click here.

Process of Investigation
Evaluation of Autoignition

To assess the potential for autoignition as an ignition source, the investigator must obtain
information about vehicle service condition prior to fire and vehicle operation at the time
of fire. With this information, the investigator can compare the potential surface
temperatures, conditions, and available fuels in order to evaluate the likelihood of
autoignition. To do so, consideration should include:

1. Vehicle speed at the time of the fire and for several minutes before the fire: If
the vehicle had been at idle and operating normally (at idle rpm, engine
operating in nominal condition), it is unlikely that surfaces are hot enough to
ignite any fluid. [1,2,3]
2. Vehicle operation: If the vehicle had been operating on a grade, fully loaded,
pulling a trailer, or passing, then the probability of high surface temperatures
increases. High load operation or high speed in a reduced gear translates to
higher surface temperatures.

3. Fluid vaporization: A stationary vehicle may have hot enough exhaust

surfaces to vaporize leaking fluids, but insufficient heat to ignite them. In
collisions, hot surface ignition may occur in the engine compartment of a
striking vehicle and then flash to the struck vehicle. Ignition of the flammable
vapors may also be caused by sources other than hot surfaces.

4. Air flow conditions: Non-collision fires frequently occur immediately after

vehicles come to a stop or slow from high speeds. Air flow at road speeds
may prevent the ignition of leaking fluids or the propagation beyond a
sheltered area. This is because the air flow may cool the hot surface, dilute
the mixture and or reduce the residence time.

5. Delays in identification of fire: Fires may initiate slowly, unobserved by

witnesses for several minutes. Lubricants in particular are able to sustain a
small flame (like an oil lamp). Such small fires may take time to become
apparent to witnesses who are often concerned with other things - such as
injury (in collisions) or their intended activities (in non-collision events). In
such cases, autoignition may occur while surfaces are still hot after operation,
even though witnesses did not observe it for some time. The appearance of
vapor after a collision may be only vapor - or the early evidence of fire
(smoke). If witnesses describe the sound of fire initiation (i.e., “I heard a big
'poof' sound”) or distinctive odors, the investigator may be able to isolate the
time of fire initiation more closely and therefore be able to evaluate the
potential for autoignition. Small flames may also be easier to observe at night
and in instances in which hood deformation allows greater visibility to all
areas of the engine compartment.

6. Vehicle maintenance condition: Obtaining a history of vehicle condition from

the operator may reveal engine or vehicle conditions related to surface
temperatures. Identification of a fouled spark plug may indicate improper
combustion and unburned gasoline released into the exhaust system with
resultant elevated catalytic converter temperatures.

7. Shielded manifolds, pockets on hot surfaces: Pockets in which combustible

vapors are held in proximity to hot surfaces lower the temperature at which
autoignition is possible (increasing the probability of autoignition) [3,4]. With
higher residence time, temperatures at which autoignition occurs in a motor
vehicle environment may approach (but can never be less than) ASTM E 659-
78 autoignition temperature measurements.

8. Coolant and gasoline: Both need unusually high surface temperatures (as
compared to other underhood fluids) for ignition on open (unshielded)
manifolds. Unless there is reason to expect temperatures of 900 degrees F or
more it is unlikely that coolant will ignite on an open hot surface. Similarly,
 unless surface temperatures are greater than 1100 degrees F, it is unlikely
that gasoline will ignite on an open hot surface [5].

9. Time: All surface temperatures are likely to be insufficient for autoignition after
3-5 minutes with the engine off.

References

1. Santrock, J., "Full Scale Vehicle Fire Tests of a Control Vehicle and a Test
Vehicle Containing an HVAC Module Made from Polymers Containing Flame
Retardant Chemicals," NHTSA docket number 98-3588-190, 1998.
2. Fournier, E., "Under Hood Temperature Measurements of Four Vehicles,"
MVFRI , http://www.mvfri.org/biokinetics-5.html, 2004.

3. Authors rely upon proprietary data from surface temperature tests they
conducted which are not available for publication.

4. Colwell, J. D., et al., “Hot Surface Ignition of Automotive and Aviation Fluids,”
Fire Technology, 41, pages 105-123, 2005.

5. Shields, L, Scheibe, R., “Computer-Based Training in Vehicle Fire

Investigation-Part 2: Fuel Sources and Burn Patterns,” SAE 2006-01-0548,
2006

6. Case Studies

1) Overheated wheel bearing ignition of bus tire

7. 2) Engine oil autoignition on exhaust manifold in crash

 Overheated bus rear wheel bearing and hub that ignited a tire.
8.  
9.

10. 1) During Hurricane Rita in May of 2005, a bus evacuating assisted living facility
residents caught fire 15 hours after travel began. Twenty three of the forty-four
passengers, many with limited mobility, were fatally injured. The United States National
Transportation Safety Board (NTSB) investigated the incident.  NTSB investigators
found that lack of lubrication in a rear axle wheel bearing led to an overheated hub and
ignition of the tire. For a pdf version of the NTSB report, click here. For a pdf version of
NTSB presentation slides with photos of this and other bus fires, click here.
11. 2) To continue and review the crash fire case study of engine oil autoignition on an
Escort exhaust manifold, click "next."

Case Study: 1995 Ford Escort

Overview

The fire vehicle had limited bun and collision damage, facilitating the
identification of origin and cause. (see larger photo)
 

Crash Information

 Car Front to Pickup Rear

 1995 Ford Escort (subject)

 1997 GMC Sierra crew cab pickup

 Car underrode pickup

 Delta V 6-10 mph

 Injuries: None

A 1995 Ford Escort four-door sedan traveling on a four-lane, rural highway at 65 mph
slowed and struck a 1997 GMC Sierra pickup truck from behind. The driver of the Escort
apparently failed to notice in time that the truck in front of him had slowed to make a left
turn. A small fire was noticed in and under the engine compartment within two minutes of
impact. Witnesses described the fire as small, "like an oil lamp," that dripped burning
material from the engine compartment to the ground below. The fire apparently burned
for five to ten minutes and remained quite localized (within a small area of the engine
compartment) before it was extinguished with a fire extinguisher. 

The fire was likely to have been fueled by engine oil or coolant that were released by
impact high in the engine compartment. The cam cover was fractured during impact and
had spilled oil onto the shielded exhaust manifold on the front of the transverse-mounted
engine. Coolant from the breached radiator was also available. Dark discoloration and
charring of the manifold and shield defined the most likely source and location of ignition.
The fire did not spread beyond the small zone at the right front of the engine; damage
was limited to local wiring and a plastic radiator tank.

Crush damage energy calculations showed that the Delta V for the Escort was 6 to 10
mph, and 2.5 to 4 mph for the GMC truck. The Escort sustained intrusion high in the
engine compartment from a trailer hitch on the GMC truck. Damage to the truck was
minimal and restricted to the rear bumper.

The driver of the Escort was not injured; his airbag deployed and he was belted. All four
occupants of the GMC truck were belted and none were injured.

Case Study: 1995 Ford Escort

Fuels Available
Collision damage to the Escort occurred high on the engine from a protruding trailer hitch
on the Chevrolet truck that was struck; the intrusion caused damage that allowed oil and
coolant to escape. The cam cover was fractured, spilling oil down the front of the engine
and onto the exhaust manifold. The passenger side radiator tank was broken near the
top, allowing coolant to reach the same area. The entire driver side radiator tank was
also broken. Collision damage fractured the base of the power steering reservoir,
emptying its contents on the right side of the engine compartment. Power steering fluid
leakage was well outboard or below heat damage; there is no evidence that the fluid
reached the vicinity of the fire so it was less likely to be a source of fuel for the fire.
 White arrows show the dislodged alternator (left arrow) and the damage caused
by the intrusion of the trailer hitch from the pickup (right arrow). (see larger
photo)

Case Study: 1995 Ford Escort

Ignition and Propagation

Ignition: The shielded manifold provided the most likely ignition source; it was contacted
by both oil and coolant. It is likely that the shielded exhaust manifold was hot enough to
ignite oil; heat damage was restricted to this region of the engine compartment. Electrical
arcing from the alternator, adjacent to the fire zone, was also possible. The alternator
was broken off its mounting bracket, and its case was cracked. Impact-damaged ignition
(spark plug) wires, in the vicinity of the fire, and at least two other broken or disconnected
wires in the right front of the engine compartment could also have provided an ignition
source through sparks or resistance heating.

In this instance, the burn damage was evident in the region of the shielded exhaust
manifold and autoignition was a continuously available source of ignition. The alternate
electrical sources of ignition would have sparked for short periods of time, intermittently
and in remote areas not necessarily in the region of burn damage; therefore they were
determined to be less likely sources of ignition than the hot exhaust manifold. It is also
possible that oil and coolant vaporized on the manifold, creating a vapor generally in the
engine compartment which was ignited by an electrical spark. If that was the case, the
fire would then have flashed to the manifold as the source of the vapor and continued to
burn there.
 Burn damage was limited to the area of the exhaust manifold under and around
the heat shield (white arrow). Some soot was present on the cam cover as
well. (see larger photo)
 

Propagation: The fire most likely originated on the right front of the engine, on or near
the exhaust manifold. Oil and/or coolant likely leaked between the exhaust manifold and
shield. The fire did not spread beyond the small zone at the right front of the engine.
Damage was limited to oxygen sensor wiring, the top of the right radiator tank (plastic)
and the spark plug wires for cylinders 1 and 2. Witnesses described the fire as small,
"like an oil lamp," that dripped burning material from the engine compartment to the
ground below. The fire was observed within 2 minutes of impact and apparently burned
for five to ten minutes before it was extinguished with a fire extinguisher.

Case Study: 1995 Ford Escort

Summary of Fire Investigation

 
   Escort: Investigator Assessments:

 Broken valve cover above shielded exhaust manifold

 Fuel likely: Engine oil, coolant

 Ignition:
 •Hot surface likely
 •Electrical spark possible
  Initiation: <2 minutes

 Propagation time: Extinguished within 5 min. w/o spread to

interior

 
Fire occurred after collision-related intrusion of a trailer hitch into the engine compartment
that caused fracture of the cam cover, radiator, alternator, and some wiring. Fire most
likely resulted from ignition of oil or coolant on the hot, shielded exhaust manifold. The
engine may have run for a short time after impact, causing additional oil leakage. Power
steering fluid, which was released  from the broken plastic reservoir during impact, was a
less likely source of fuel. Additional ignition sources include the alternator, spark plug
wires, or other electrical wires at the right front of the engine compartment.

This photograph of the pickup (struck vehicle) shows the aggressive hitch and
stiff bumper that contacted the underriding Escort.

Examples of Recalls - Autoigntion

ATF, Polymers, Brake Fluid, Gasoline, Coolant, Oil

2006 Mazda Mazda5
Exhaust system
NHTSA CAMPAIGN ID Number : 05V412000

On certain sport utility vehicles, a heat buildup in the exhaust system can occur.
Consequence:
This condition could result in a fire.

Remedy:
Mazda instructed owners (September 16, 2005) to bring in their vehicles in and will be
given a rental or loaner vehicle at no cost while Mazda repairs these vehicles. The recall
began October 26, 2005. Owners may contact Mazda at 1-800-222-5500.

Notes:
Mazda recall no. 3605J. Customers may also contact the National Highway Safety Traffic
Administrations Vehicle Safety Hotline at 1-888-327-4236; (TTY: 1-800-424-9153); or go
to http://www.safercar.gov
 

1996-2000 Chrysler Town and Country, Dodge Caravan, Dodge Grand Caravan,
Plymouth Voyager, Plymouth Grand Voyager
Gasoline fuel injection system
NHTSA Campaign ID number: 00V268000

Mini vans built with 3.3L and 3.8L engines have fuel rails with nitrole rubber o-ring seals
that can degrade over time. Fuel leakage from the underhood fuel injection fuel rail could
result, increasing the likelihood of vehicle fire.

Consequence:
Fuel leakage in the presence of an ignition source can result in fire.

Remedy:
Dealers will install a seal on the vehicle fuel rails to prevent external leakage of fuel from
the fuel rail crossover tube. should the existing o-rings continue to degrade, owner
notification began on January 25, 2002. Due to the large quantity of vehicles involved in
this campaign, the owner notification will be phased over the next several months.
However, if a vehicle is leaking fuel from the o-rings, the vehicle should be taken into a
dealer to have this repaired as soon as possible. Owners who take their vehicles to an
authorized dealer on an agreed service date and do not receive the free remedy within a
reasonable time should contact Daimler-Chrysler at 1-800-8953-1403.

1991 Isuzu Rodeo

Automatic Transmission
NHTSA Campaign Number: 91V118000

The transmission fluid level gauges (dip sticks) installed in these vehicles are not to
designed specifications. The gauge indicates that transmission fluid should be added
when it is at the correct level.

Adding more transmission fluid than required will cause the fluid to leak through the fluid
level gauge tube onto the manifold as the engine temperature rises. This creates a
potential for a fire.

1988 Lincoln Continental

Wiring Harness: Front Underhood
NHTSA Campaign Number: 88V114000
Nylon shielded wiring harness, located in the engine compartment, is susceptible to heat
damage and melting of the nylon material in 100 degree plus weather.

Melted nylon shield could drip onto the exhaust manifold and result in an underhood fire.

Wrap wiring harness with a heat resistant fiberglass tape.

1992 Mercedes Benz 400E

Brake fluid hoses, crank case emissions control device
NHTSA Campaign Numbers: 92V083000, 94V004000

NHTSA Campaign Number: 92V083000: The plastic brake hydraulic hose from the

automatic slip control unit was misrouted too close to the pre-resistor for the electric
auxiliary engine cooling fan which becomes hot under certain conditions. This could
cause the hose to melt and leak brake fluid onto the hot pre-resistor.

Brake fluid can catch on fire when it contacts the hot pre-resistor, posing the risk of injury
to anyone in or near the vehicle.

NHTSA Campaign Number: 94V004000: High ambient engine compartment

temperatures are experienced after the vehicle is operated and then parked. If the
vehicle is restarted a short time later and the charcoal canister is saturated with fuel, it is
possible that fuel vapor may be expelled through the inlet vent line of the canister onto
the electric auxiliary radiator fan pre-resistor.

Under certain operating conditions the electric auxiliary radiator fan pre-resistor becomes
hot enough to ignite the fuel expelled from the charcoal canister and result [sic] in an
underhood fire. 

1987 Nissan Van

Engine, power steering hose, cooling system hoses
NHTSA Campaign Numbers: 87V109000, 90V136000, 91V211000, 93V036000,
94V031000

NHTSA CAMPAIGN ID Number: 87V109000: Valve cover gasket may leak oil.

Oil may drip on exhaust manifold and if engine is operated under severe condition or
operated at continuously high speeds. The manifold may become hot enough to ignite
the oil. This could result in an engine compartment fire.

NHTSA CAMPAIGN ID Number: 90V136000: Power steering hoses may leak fluid that
can spread to the exhaust manifold.

Under certain conditions, the manifold may become hot enough to ignite the power
steering fluid and result in an engine compartment fire.

NHTSA CAMPAIGN ID Number: 91V211000: The cooling system hoses deteriorate and

fail because of inadequate heat resistance. Failure first occurs at the clamp connections,
causing coolant fluid to be leaked or sprayed onto the engine compartment components.
If the vehicle is operated with a low coolant level, the cooling system temperature and
pressure would rise very quickly. Coolant reaching the hot exhaust manifold can ignite
and cause an engine compartment fire.

NHTSA CAMPAIGN ID Number: 93V036000: Nissan has made a preliminary

determination that a defect exists in the engine compartment component(s) that may fail
due to engine compartment heat.

The undetermined engine component(s) can cause an engine compartment fire, resulting
in possible injury to vehicle occupants.

NHTSA CAMPAIGN ID Number: 94V031000: Engine compartment fires and other

thermal incidents have occurred on vehicles which had received all of the remedies
provided by a previous safety-related recall (NHTSA Campaign No. 93V036000). The
fires are attributed to errors in installing the remedy specified in the recall.

If the remedy specified for Recall 93V036000 was not performed correctly or the vehicle
has not received the remedy, a fire could occur.

Nissan will offer to repurchase these vehicles from the owners. The vehicles will then be
destroyed. For owners that decline the repurchase offer Nissan will provide a
reinspection of their vehicle free of charge, to confirm the prior recall was completed
properly. Also, Nissan will perform any recall-related repairs that may be needed at no
charge to the owner.

Electrical Ignition
  There are three primary forms of ignition from electrical energy:

 Arcing

 Resistance heating

 Static electrical discharge

For ignition to occur, the electrical source must generate sufficient heat energy and
temperature in the presence of air, for a enough time, to bring a nearby combustible to
its ignition temperature. For fire to propagate, sufficient heat energy must be created by
the initial fire to sustain the fire and ignite other nearby combustibles.

Electrical Fires: Frequency

Experts demonstrate widely varying opinions regarding the potential and frequency of the
electrical system as a source of vehicle fires. Differences may be due to variation in
sample selection, category definition, or investigator judgment and bias.

Study Cause Percentage Notes Source

Ahrens, M., Short circuit (worn 4.3% Factors 1999-2001, NFIRS
 insulation)
Short circuit
(mechanical   2.4% contributing to
2005 [1] damage) and NFPA survey
ignition
Short (other)  15.9%
Other electrical 6.4%
Cars, UK, includes
Fray, M.,
Electrical 27% 1988 cases few percent collision
1991 [2]
fires
Powell, A.H.
Electrical 40% Non-deliberate UK Home Office
1989 [3]

References

1. Ahrens, M., “An Overview of the U.S. Highway Vehicle Fire

Problem,” SAE 2005-01-1420, 2005.

2. Fray, M., Southall, D., Galer, I.A.R., Reducing the Risk of Non-

Electrical Car Fires, SMMT, UMTRI - 82645, 1991.

3. Powell, A.H., "Review of the Electrical Causes of Fires in Cars," 

ERA Report 89-0047, February 1989.

Electrical Fires

Difficulty in Creating Electrical Fires

Many researchers and investigators have attempted to intentionally start electrical fires in
vehicles and found it difficult to accomplish. They have generally found that protective
systems (such as fuses) are effective, and circumstances that support fire and
propagation are very specific. At least 2 researchers have published reports of failed
attempts to create an electrical fire.

1.         Cole [1] made 4 attempts to cause a fire by electrical short in a 1989 Oldsmobile
Cutlass without success. Smoke and fumes were produced without flame.

2.         Hrynchuk [2] was unable to produce a fire from shorting of fused circuits, but did
produce a fire by shorting an unfused (bypassed) power line to ground under the
instrument panel of a 1980 GMC Sierra Grande.

Recall investigations have confirmed that electrical fires do occur.

References
 1. Cole, L.S., The Investigation of Motor Vehicle Fires, Third Edition,
1992.

2. Hrynchuk, R.J., "A Study of Vehicle Fires of Known Ignition

Source," International Association of Arson Investigators, Alberta
Chapter, Edmonton, Alberta, January, 1983 .

Arc vs. Spark: Definitions

The following definitions for an arc and an electrical spark are given in NFPA 921 as
follows:

 3.3.7       Arc. A high-temperature luminous electric discharge across a gap.

 3.3.42    Electric Spark. A small, incandescent particle created by some arcs.

This convention is not universal however; the term “spark” may be used by authoritative
sources to mean “arc” (as defined by NFPA 921). This is especially true in the context of
certain automotive components such as spark plugs were a “spark” may jump the gap or
an electric motor may be described as “sparking” during normal operation as motor
brushes break contact with the commutator. In NFPA 77, static electrical discharges are
referred to as “sparks,” inconsistent with the definitions in NFPA 921. As such, caution
must be used when interpreting or using these terms.

In this course, we will sometimes refer to electrical arcs and sparks interchangeably.

Arcing
Arcing results in high temperatures, sometimes in the range of several thousand
degrees, easily sufficient to ignite a flammable mixture under the right conditions.

A minimum of 350 volts is required for an arc to jump even the smallest gap [1]. Although
most vehicles have 12  or 24 volt electrical systems, which would be insufficient to cause
an arc to jump across a gap, ignition can occur from arcing in high energy electrical
components such as the secondary ignition wires (spark plug wires), ignition wire
connectors, coil(s), or distributor.

A parting arc is a brief electrical discharge that can occur when the current in an
energized circuit is interrupted by the separation of conductors. They can occur at lower
voltages, including in a 12 volt automobile electrical system. Parting arcs in vehicles can
occur from:

 Wiring that is severed or disconnected, such as during a collision

 Damaged insulation on wiring (worn, cracked, burned or missing), resulting in

unintended electrical contact and separation

 Switch contacts as the switch is turned on or off

 Brushes and commutators in electric motors

  Actuation of electromechanical control devices such as relays and solenoids

Parting arcs also occur when short circuited wires melt through, resulting in a brief arc as
the circuit is broken. Under some circumstances, small particles of hot, molten metal
(electrical sparks) can also be emitted from a short circuit when a parting arc occurs.

 Option: For more information about the potential for ignition by electric motor
sparks, click here.

Resistance Heating
Heat is created by the flow of electric current in a conductive material. According to
Ohm’s Law:

 Voltage=Current (measured in amperes) X Resistance (measured in

ohms)

 And because

 Power=(Current)2 X Resistance,

if there is too much current for the size of the wire and/or excessive resistance (such as
from a poor connection), undesired heating can occur. This heating can cause ignition of
nearby combustibles, including wiring insulation, vapors from flammable liquids, or other
solids. Light bulb filaments, typically tungsten, are designed to emit light through
resistance heating. Filaments operate at temperatures as high as 2550 degrees F. while
contained in a sealed, evacuated glass bulb [1] . Should the glass break from a collision,
the filament will oxidize and usually will fail in a short time, though exposure to flammable
vapors during this time can cause ignition. Other automotive electrical components,
including certain fan motor speed control devices, use resistors in the circuitry that may
generate sufficient heat to ignite combustibles should they come into contact with each
other.

 Automotive components most likely to cause ignition from resistance heating include:

 Batteries and battery cables

 Other high-current electrical cables, including those to the
alternator and starter motor
 Any wiring that does not have proper current overload protection
 Broken light bulb filaments
 Resistor packs used for motor controls
 Resistance heating elements
 Poor connections of conductors, including loose and oxidized
terminals
(For more detail regarding loose connections, click here)
  Overloaded or stalled electric motors
References

1.    NFPA 921, Guide for Fire and Explosion Investigations, 2004 Edition

Case Study: 1996 Dodge Caravan Crash Test Fire

 
 

This video shows a crash test between a stationary 1996 Dodge Caravan, with
underhood fluids on board, and a 3,611 lb deformable barrier traveling at 65 mph that
struck the van in the left front corner. The first segment shows the collision in real time;
the next shows it in slow motion. In the slow motion view, it is possible to see the fluids
relased and mixed during the collision. The next segment shows a another real time view
of the impact, during which no fire was immediately detected. The fire was first detected
approximately five minutes after impact, when blistering paint was seen on the upper
radiator tie bar near the front of the deformed battery. Flames were observed near the
battery approximately 30 seconds later. Approximately 10 minutes after impact (five
minutes after fire observation), the fire was extinguished. At that time, the fire was self-
propagating, but little propagation had occurred.

Case Study: 1993 Ford F150

Overview

Ford F150, showing burn damage concentrated at right rear of engine

 compartment (see larger photo)
 

Non-collision Fire Information

 1993 Ford F150 Pickup
 Operated normally prior to being parked
 High wind that day
 Driver smelled something when parking vehicle, saw nothing
 Engine compartment/interior fire discovered 10 to 15 minutes later

Burn patterns on this vehicle showed that the greatest fire damage was to the right
(passenger) side of the engine compartment, along the bulkhead. Fire damage also
extended into the interior. The right upper instrument panel had melted away from the
bulkhead, and fire damage extended across the headliner, which dropped burning
fragments onto the seats.

Engine compartment, showing damage near bulkhead where blower assembly

was located

Case Study: 1993 Ford F150

Blower Resistor and Burn Pattern

 
 Resistor pack for blower motor control, likely to have ignited combustible debris
such as leaves
Examination of the vehicle and review of documents and witness statements indicated
that the fire was located around the heater blower assembly located in the right rear of
the engine compartment. Potential scenarios included fire origination in the a) blower
assembly, b) wiring adjacent to the fan motor, or c) wiring between the heater blower and
the battery/relay. Only the first scenario was consistent with all evidence. It was
determined that the most likely cause of fire in the blower assembly was combustible
debris (such as leaves) in the blower assembly ignited by the blower resistor pack (motor
speed controller). It was also possible that faults in the wiring inside the blower fan or
leading into the fan assembly was the cause of fire.

For this vehicle, variable blower speed control is accomplished by switching resistors in
the circuit to drop voltage to the motor. When the fan is operating at full speed (“high”),
no resistors are in the circuit; at lower speeds, a series of external resistors, capable of
dropping voltage to the motor, are used to obtain the desired fan motor speed.
Resistance heating of combustible materials in contact with the resistor pack, such as
leaves or other debris, was a likely cause of fire.
 Wiring normally connected to the blower housing in the engine compartment
 

An exemplar F150 was inspected to confirm the location of wiring in pre-fire conditions. A
wire that had been routed inside the blower assembly and through a boot in the blower
housing showed more burn damage to the section that had been inside the housing than
outside the housing (and inside the engine compartment). This investigation utilized
large-scale features of burn pattern in identifying paint damage, rust, and damage to the
instrument panel that helped identify the region of origin. Small burn pattern features,
such as the wire indicated by the arrow in the photograph above, were equally helpful in
identifying the origin in a more precise way.

Resistance Heating

Lead/Acid Batteries
Lead/acid batteries have served as a mainstay in motor vehicles for many decades,
supplying electricity for starting the engine and running accessories. The electrochemical
system inside the battery consists of lead, lead dioxide and sulfuric acid. Over time, many
design improvements have been instituted to improve reliability, service life, and
durability.

For the purpose of fire investigation, there are several circumstances in which batteries
may be involved as an ignition source.

1)         Battery displacement or physical damage may result in a short circuit. A loose
battery may shift due to vibration or road bumps causing terminals to contact
conductive surfaces. If the contact is maintained, heating and fire may result.
Battery energy is high; it is generally expected that any short involving a
battery terminal or cable to ground will leave an easily visible mark. Collision
 damage may cause displacement of the battery, shorting of cables or damage
to the battery case that can result in fire.

An example of a fire caused by collision damage to a battery can be found in a

General Motors crash test report [1]. This test was conducted with standard fluids
and a charged battery. During the test, deformation caused a sharp screw to
penetrate the case of the battery, shorting against the internal plates, and
resulting in a fire.

2)         Batteries release hydrogen gas while charging. There have been cases of the
hydrogen gas igniting when subjected to sparks from static electricity or other
sources. This topic will be discussed in the Fuels Section.

3)         Batteries cases may also heat to the point of ignition. This does not appear to
be a frequent occurrence, but at least 2 failure modes have been
documented:

a)    Conductive contamination on the surface of the battery may create a high-
resistance short between battery terminals. For example, leakage of electrolyte
on the surface of the battery will create a circuit between terminals [4]. Most of
the time one would expect the heating itself to change the local conditions and
thus interrupt the circuit without heating to the point of ignition. However, in the
rare occasion when the short circuit is maintained, the battery may overheat and
ignite [4]. Vehicle operation with loose or missing hold downs may allow battery
vibration and leakage of electrolyte through vents.

b)    Battery abuse can also lead to overheating [2-6]. When batteries are
overcharged, heat is generated, hydrogen gas is released, and electrolyte levels
are reduced. In old, abused, or overcharged batteries, internal damage to the
battery may compound with use. This failure mode is expected to be
accompanied by a smell of sulfur (rotten eggs). A history of charging system
complaints or battery problems indicates increased likelihood that a fire was
caused by battery overheating.

c)    In recent tests of of battery abuse (in accordance with SAE J2464, "Electrical
Vehicle Battery Abuse Testing") 12 and 36 volt batteries were subjected to dead
shorts across positive and negative terminals. Shorts inside the battery stopped
the current before significant heating [7] Overheating would still be possible if the
short across terminals had enough resistance to heat but insufficient to cause an
early failure in the battery.

As in other factors of electrical causation, burn damage to batteries observed after the
fire may be a result of the fire and not because the battery was the source of ignition.
When investigating a battery fire in a fleet vehicle, battery conditions in non-fire fleet
vehicles of similar design, usage or maintenance may give information about the
probability of battery fire in the subject vehicle. This technique of examining non-fire
vehicles in a fleet can be applied to fire investigation in general.

To view references for this section before continuing, click here.

 Arc Tracking
Arc tracking is the process by which a plastic material is transformed from non-
conductive to conductive through a process of surface degradation. Arc tracking is a
potential ignition source in non-collision fires for systems with voltages greater than 14
volts only.

Click here to for more information about arc tracking.

Static Electrical Discharge

A buildup of static electricity results from the transfer of electrons between bodies,
causing one body to have an electrical potential when compared with the other. Static
electricity can be created from the movement of vehicles on the road, the movement of
clothing layers against each other or against upholstery, the movement of fluids
(including gasoline) in a pipe or hose, or it can be acquired from lightning. If the electrical
charge difference is neutralized by a sudden discharge of electrical energy in the form of
an arc in the presence of an ignitable atmosphere, there is the potential for fire. Such
forms of electrical ignition in automotive fires are rare, but should always be considered
during investigations if the circumstances are appropriate.

Static electricity has been the cause of fires during refueling of vehicles, though it is very
rare. The Petroluem Equipment Institute has published the results of their investigations
of this matter, as well as guidelines for prevention [1]. Also the PEI dispels the myth that
cell phones cause fires while refueling. It was dismissed as an email hoax.

References

1. Petroleum Equipment Institute website: http://www.pei.org/, January 2006

2. Design: Electrical System Protection

3. A variety of electrical system components are designed to protect against
fires. Overcurrent protection devices include fuses, fusible links, and circuit
breakers. Wiring design, insulation, termination, routing, and protection are also
crucial. Fuel pump shutoff systems are incorporated to limit fuel spillage in the
event of a collision. Numerous standards have been published to address some of
these issues.
4. Fuses
5. Fuses act as a safety valve to prevent current from exceeding the rating of wiring
and electrical circuit components. Overloads can occur through
either overcurrent or short circuit events. If the current exceeds the rating of the
fuse, the metallic fuse “element” overheats, melts, and opens the circuit. It is the
intent of the circuit designer to install a fuse that will open long before damage is
done to the circuit components. Once a fuse has opened or “blown,” it must be
replaced before the circuit can be re-energized. 
6. Fusible Links
7. A fusible link is a short section of wire (about 2 wire sizes smaller than the
surrounding wire, per SAE J156 Jun 94) that has a smaller diameter than the rest
of the circuit. It acts as a high-current fuse, typically with a capacity of 30 amps or
more, that protects circuits from excessive current draw. When current flow in the
circuit exceeds that of the fusible link, the wire melts and interrupts the circuit. This
type of fusible link is quickly becoming obsolete, as many carmakers have opted
for newer technologies such as the cartridge style fuse element or “Maxi” fuses.
The location of fusible links varies. The older, wire style of link can be hard to find.
It is usually located in a wiring harness near a main harness connection; the fusible
link may have a current capacity tag attached. There have been reports that fusible
links have initiated fires. 
8. Circuit Breakers
9. Circuit breakers are circuit protection devices that, like fuses, will open a circuit
when current exceeds the rating of the device. Unlike fuses, however, circuit
breakers can be reset without replacement. Some are designed to be self-
resetting; others require manual reset.
10. More information about automotive current protection devices can be found on the
Littlefuse website [1]: 
11. Wiring
12. Wire sizing, routing, insulation, termination and protection are important factors in fire
causation and investigation. [2, 3]
13. For more details on conductor size and wire gauging click here.
14. Fuel Pump Shutoff Devices
15. Some manufacturers (Ford, in particular) have historically used inertia switches to
interrupt power to the electric fuel pump circuit in the event of a collision. These
devices are often located in the trunk, and are resettable. Other manufacturers have a
different approach. Instead of inertia switches, they use an electrical signal from engine
rotation or crash sensors to disable power to the fuel pump. If the engine stops rotating,
as would be the case in most serious collisions, power to the fuel pump is interrupted.
 Photograph of fuel pump inertial cutoff switch in the trunk of a Ford Escort.
16.  
17. Battery Crash Protection
18. As described in the battery section of this course, vehicle batteries can be an ignition
source when short circuited during a crash. Insulating covers for the positive terminal
can reduce the chance of shorting of against other surfaces because of displacement
of the battery or deformation of metal.
19. Some vehicles have mechanisms to automatically disconnect the battery in a crash.
There are devices that use a pyrotechnic charge to disconnect the battery cable, and
others disconnect battery power electronically. Instrumented vehicle tests with crash-
induced short circuits have shown that power is frequently cut off  or stabilized early in
crash events, even in those vehicles without battery disconnect systems [6].
20. Standards Related to Electrical Systems
21. Numerous Society of Automotive Engineers (SAE) standards specify, among other
things, insulation type, adherence, wall thickness, dielectric testing, fluid compatibility,
pinch and abrasion testing, and tensile strength. [7]
22. Flammability of wiring insulation (and other materials) in horizontal and vertical
positions is covered by Underwriter’s Laboratories Standard UL 94 which is entitled
"Tests for Flammability of plastic Materials for Parts in Devices and Appliances." [8]
23. To view references for this section before continuing, click here
 Design: Failures Occur Despite Safeguards
Numerous electrical ignition sources for automotive fires are available, despite efforts
made in design to avoid them. Such ignition sources exist in properly maintained vehicles
in normal operation, as well as in vehicles compromised by damage, age, wear,
modification, or collision.

Electrically protected circuits can still have ignition potential. For instance, if a stranded
wire is abraded such that only one strand remains, even if it is operating at rated current,
the damaged wire may heat to the point of igniting nearby combustibles, or it may burn
through and cause a parting arc. Localized heating of conductors from grounding,
shorting, or poor or intermittent connections can cause sufficient heat for ignition while
staying below the current level necessary to open the fuse. In rare circumstances, normal
arcing within an electric motor (e.g. fan motor, electric pump, wiper motor) can cause
ignition of flammable vapors.

Some high current electrical cabling has no circuit protection by design (starter,
alternator, positive cable from battery). If a short to ground occurs, arcing and sparking is
likely to occur.

Loose connections can be the source of high-resistance heating. For more

information, click here.

Process: Investigative Considerations

It is advised that fuses be inspected during the course of most automotive fire
investigations. This requires locating the fuses, which can be challenging in some
vehicles. It is not uncommon for fuses to be located in more than one location –
fuseboxes are usually found in the engine compartment and/or interior. In-line fuses may
be located anywhere in the vehicle. Fusible links can be particularly difficult to find,
especially after a fire. Access to a shop manual (or even the owner’s manual) for the
vehicle will help the investigator learn fuse locations, intended current ratings, and the
circuits controlled by each fuse.

Fuses should be inspected to determine if they have the correct current rating for the
circuit and if they are open or intact. Obviously, severe heat damage from fire may melt
the fusible elements and preclude any meaningful interpretation of fire cause. Also, short
circuiting often occurs during a fire, so caution in the interpretation of blown fuses must
be exercised. Nevertheless, fuse condition can give clues as to the fire origin or cause
and it can sometimes help the investigator determine the time history of electrical power
on the vehicle as the fire progressed. For example, for an advanced fire, a lack of blown
fuses could indicate that system power was interrupted before the fire progressed very
far.

Aftermarket wiring is always suspect (due to uncertain quality control) and should be
carefully inspected. Wiring associated with sound systems, alarms, remote starters,
engine block heaters, and other add-on accessories may not be correctly fused, and may
not be properly insulated or protected from unwanted electrical contact or abrasion. It is
also possible that factory wiring may be damaged by installation of other accessories
when holes are drilled in the vehicle or sharp screws penetrate the insulation of a
harness.

Process: Evidence of Electrical Causation

Automotive fires are relatively infrequent events; hence those of electrical origin are even
more rare. As such, it is prudent to carefully investigate the history of electrical problems
with the vehicle. If a fire was caused by electrical means, there is some likelihood that the
cause was related to a traceable history. Specific evidence of electrical causation may
include:

History of fuse failure

Interviews with witnesses (the vehicle owner/driver or the person most responsible
for maintaining the vehicle) may reveal a history of fuse failure and replacement that
may give credence to a theory of electrical ignition source.

Wiring failure consistent with operating conditions and current

The investigator should attempt to determine if there was an electrical component in

the car that reportedly malfunctioned, had been operating erratically, or had an odor
prior to the fire. Was the electrical system energized? Could the available current
from the system have caused the observed damage? Are there alternative
explanations for the observed damage? Sometimes fires occur in parked vehicles
after a significant period of time. In such cases, there is little chance that a hot
surface on the engine could provide an ignition source. Electrical malfunctions or
arson are the only likely possibilities.

Burn pattern: fuels and ignition sources

Sometimes the physical fire evidence suggests a possibility, or even likelihood, of

electrical ignition. This could be in the form of arced connections, electrical overheating,
beaded wires, internal heating, or other indication. Other times, by elimination, there is
little possibility of any other form of ignition. For example, after a vehicle is parked for an
hour, electrical malfunction and arson may be the only possible ignition sources. In any
case, the investigator must strive to find consistency in all the evidence, including witness
statements, to identify the ignition source. In an accidental fire, if there is no hot surface
in the vicinity of the fire origin (and the fire origin is definitively located), then an electrical
ignition source becomes much more likely. 

Arc damage

Arc damage or beading, may suggest a possible electrical origin. Arc damage may
involve a transfer or loss of material. The investigator should look particularly
carefully at the wiring in systems with a history of fuse failure, recent maintenance or
problems in operation (e.g., reported intermittent function of electrical accessories or
components).
Evidence of overheated wiring

Overheated wire should be carefully examined to determine (if possible) whether the
heating was internal or external. If a wire has been overheated internally, its
insulation will be affected along the entire length of the wire, from the current source
to ground. If the wire was externally heated as a result of fire, there may be telltale
areas where the wire was shielded from heat (such as behind a clamp). Sometimes
disassembly of a wiring harness will give clues as to specific wires that may have
overheated. If an overheated wire is found, the investigator should look for
information about the related circuit. Is there any evidence of a problem with this
circuit before the fire (owner report, maintenance, fuse replacement)? Was the circuit
powered prior to the fire?

Process: Evidence Can Be Difficult to Find

Beware: Short circuits, arcs, and overheated wires also occur after fire initiation

There may be many arcs or shorts that occur as a result of a fire, depending on the
extent of the fire. Evidence of arcing or wire beading may have occurred after the fire
was initiated. Unfortunately, when there has been a complete burn-down of a vehicle,
where no combustibles remain, there is no reliable means to distinguish between an
internally and externally heated wire. There is no discernable metallurgical difference
in a copper wire that has been exposed to internal heat as compared with external
heat. There are some controversial methods (that have been proposed and
published) to determine whether an arc was the cause or result of a fire, but none
have been proven to be reliable. To learn more about this topic, click here.

Photograph of battery cable short against housing as a result of burn damage to

insulation after fire initiation.

Arcing is difficult to detect, particularly in small wires

There are many hundreds of feet of wiring in an automobile. Small arcs may be difficult to
find after fire damage, collision damage, or in large wire bundles that are difficult to
access. Also, the arc evidence can be subtle and nearly microscopic. Other evidence,
including witness statements, may be necessary to corroborate a theory of electrical
ignition.

In this 1993 Honda Prelude collision fire, the

energized alternator terminal was forced through a
steel shroud. (Arrow shows the area of the alternator;
front of vehicle is to the right in the photograph.)
 

The alternator positive terminal stud, located at

the front of the engine compartment, must have
been powered at the time of contact with the
steel shroud, yet neither the stud nor the shroud
showed any evidence of arcing or material
transfer. It may be that the contact was so fast
and secure that there was no opportunity to form
a parting arc. This is an example in which there was a known short and no clear evidence
of arcing remained.

Poor connections and deficient insulation may not be evident after the fire

Fire often destroys or consumes the very evidence of its cause. Threaded connections
can loosen (or appear to loosen) in a fire from expansion of the metal, or from heat
consumption of washers or other components in the connection. Also, loose connections
can corrode after a fire, giving them the appearance of being tight.

Process: Complex Electrical Activity in Collisions

The response of an automotive electrical system following a collision and during a fire
can be very complex. Intermittent shorts have been identified in crash tests. In this
example (below), involving a crash test on a 1997 Ford Explorer, internal alternator
shorting was identified, which may have been the cause of the system voltage dropping
between 40 and 110 milliseconds. There was no evidence of burning and charring in or
around the alternator [1]. In this case, evidence of a short was found after a crash, but it
did not result in fire. This underscores the complexity of electrical activity in crash and fire
events, and the difficulty in assigning causality even to the evidence found.

In general, after a fire has started and regardless of its ignition source, the compromise of
wiring and other electrical components during the fire can cause errant operation of any
of the electrical systems on the car. It is not unusual for the starter motor to operate, the
headlights to turn on or off, the wipers to actuate and the horn to honk. When such
systems are energized due to fire-compromised insulation, evidence of short circuits
unrelated to fire causation is expected. This should be kept in mind during the
investigation phase when speaking with witnesses about their observations of the fire.

 
References

1. Jensen, J. et al., "Evaluation of Motor Vehicle Fire Initiation and Propagation;

Part 8: Crash Tests on a Sport-Utility Vehicle", NHTSA docket number 98-
3588-139, 1998.

Recalls: Electrical System Examples

Fused, Unfused, Low Current, Breaker

1994-1996 Ford / Bronco 1994-1996
1997-2002FORD / EXPEDITION
1994-2002 Ford / F150
1994-1999 Ford / F250
2002 Lincoln / Blackwood
1998-2002 Lincoln / Navigator
Vehicle Speed Control
NHTSA CAMPAIGN ID Number : 05V388000

Summary:
On certain pickup trucks and sport utility vehicles equipped with speed control, the speed
control deactivation switch may overheat, smoke, or burn.

Consequence:
A fire at the switch could occur.

Remedy:
By letter dated September 12, 2005, owners were instructed to return their vehicles to
their dealers to have the speed control deactivation switch disconnected. Owners who
have had their speed control deactivated are being notified that parts will be available
and advised to make an appointment to reconnect the speed control beginning in
February 2006. Owners who did not have their speed control deactivated are being
notified to have their systems remedied beginning February 2006. Owners are urged to
avail themselves of the free disconnect service as soon as possible because of the
significant risk of fire. Owners may contact Ford at 1-800-392-3673.

2002-2006 International/ 7600, 7700, 8600

International Truck & Engine Corporation
Engine and engine cooling
NHTSA CAMPAIGN ID Number: 05V567000

On certain heavy duty trucks equipped with Caterpillar diesel engines, the engine
electrical ground cable that connects the engine block with the starter may be routed in a
way that could cause chafing against the starters positive battery cables.
Consequence: If chafing occurs between the positive and negative cables, an electrical
short may result and may cause a vehicle fire, possibly resulting in property damage,
personal injury or death.

Remedy:
Dealers will inspect the engine ground cables for proper routing. If any chafing is found,
the engine block ground cable will be replaced with a shorter (280 mm) cable and any
other damaged cables will be replaced and routed correctly free of charge. The recall
began on December 21, 2005. Owners may contact International at 1-800-448-7825.

2002-2004 Kia Sedona EX

Seats
NHTSA CAMPAIGN ID Number: 05V463000

Summary:
On certain passenger vehicles equipped with power seats, the wiring under the front
seats could be misrouted, allowing contact with metal components.

Consequence:
This contact could eventually result in a short circuit condition that may result in a fire in
the junction box area of the vehicle.

Remedy:
Dealers will inspect and re-route the wires under the power seats, if necessary, as well
as install protective seaming welts around those wires to better insulate them. The recall
is expected to begin during November 2005. Owners should contact Kia at 1-800-333-
4542.

1987 Saab 900

Electrical System: Wiring Under Dash
NHTSA Campaign Number: 88V173000

Wiring harness may chafe against sharp edges under the dashboard and against a
support under the rear seat (1988 models only).

This could cause a short circuit, the resulting heat buildup could cause smoke or a fire in
the passenger compartment.

1985 Chevrolet Caprice

Electrical System: Battery cable
NHTSA Campaign Number: 85V153000

The positive battery cable may contact the upper control arm or exhaust manifold shield
and eventually wear through the insulation and ground the cable. Consequence of defect:
This could result in an underhood fire which could cause burn injuries to occupants.

1990 Chevrolet Celebrity

Interior Systems: Bench backrest
NHTSA Campaign Number: 92V124000
The power seat or recliner wiring harness can be pinched or can chafe on the seat frame
or spring support rod, wearing through the wire insulation and causing an electrical short.

The high current caused by the electrical short welds the circuit breaker elements
together in the closed position, leaving the circuit unprotected and generating sufficient
heat to burn the wire insulation and ignite the seat cushion material.

Mechanical Spark Ignition

Mechanical sparks are another ignition source in vehicle fires. There is a limited amount
of information to help the investigator determine the likelihood of mechanical sparks as
an ignition source in specific cases.

The following definition for a spark was given in NFPA 921:

  1.3.113 Spark: A small, incandescent particle.

 Such sparks can occur through the following means:

 Collision (metal to metal or metal on pavement).

 Dragging metal object on pavement.

Test Results

Drag tests of metals on concrete have shown the following characteristics for mechanical
spark ignition [1-3]:

Steels

 Consistent ignition of gasoline when drag speeds were 10 mph and

contact pressures were 20 psi and higher.

 Inconsistent ignition of kerosene, though it becomes more consistent

at drag speeds of 20 mph and contact pressures of 50 psi.

Aluminum

 No ignition of gasoline or kerosene at drag speeds of up to 45 mph

and contact pressures of greater than 1455 psi.

Pavement factors

 There was a lower ignition potential on asphalt than concrete.

When mechanical sparking occurs, minimum ignition energy of the fuel must be achieved
for the spark to be considered a valid ignition source.

In general, mechanical sparks are rather poor ignition sources because of their small size
in relation to their temperature, which gives them little energy. Titanium and magnesium
have higher energy sparks than steel or aluminum and have a higher propensity for
ignition of a fire. However, both titanium and magnesium are rarely used in any quantity
on modern road vehicles. A dragging metal component can heat to high temperatures
which also raises the potential for autoignition of flammables.
  

References

1. Mowrer, F.W., "Heavy Truck Fuel System Safety Study: Final

Report," DOT HS 807 484, 1989.

2. Johnson, N., “Spilled Fuel Ignition Sources and

Countermeasures," DOT HS S01 727, 1975.

3. Arndt, S., "Motor Vehicle in the Post-Crash Environment, An

Understanding of Ignition Properties of Spilled Fuels," SAE 1999-
01-0086, 1999.

Case Study: 1991 Toyota Previa

Overview

1991 Toyota Previa van, showing fire damage caused by undercarriage collision
with towing dolly that penetrated the gas tank

 
Minivan Undercarriage Crash Information
 1991 Toyota Previa
 Override of tow dolly
 Delta V minor
 2 occupants, burn injuries

A 1991 Toyota Previa van was traveling 50 to 55 mph on a multi-lane divided highway in
an urban setting when it struck a tow dolly in the roadway. The tow dolly had apparently
fallen off another vehicle or come loose from a vehicle being towed. The underside of the
Previa struck a rod on the tow dolly, puncturing the fuel tank of the Previa; the driver was
able to bring the vehicle to a stop in her lane of the highway. The vehicle was surrounded
by flames immediately. The driver and her son were able to exit the vehicle without
assistance, but both had to cross through flames. 
The only available liquid fuel was gasoline, and it was the most likely primary fuel source
for the fire. Contact with the rod from the tow dolly created a 3-4 in. long by 0.5 in. wide
breach in the lower forward portion of the steel fuel tank, which was on the
undercarriage. The breach was close to the driver door. Ignition was immediate and most
likely from mechanical sparks created by abrasion between the tow dolly and the
pavement; several abraded areas on the tow dolly were consistent with this scenario.
A pool fire resulted, causing an intense fire that blocked the occupants' exit paths.

Propagation to the interior was within seconds of the doors being opened. Each occupant
exited through their respective front doors, and the driver door was left open. The
passenger door was likely open during a portion of the fire. Heat damage was most
intense on the undercarriage in the vicinity of the tank breach, but heat damage to the
interior was more extreme to the passenger side interior. This was reconciled by an
assessment of vehicle position (during the fire) relative to wind direction, which revealed
that the wind was blowing from the driver side toward the passenger side.

Both occupants sustained severe bum injuries. The driver suffered burns to her hands,
feet, and face, and was hospitalized for two and a half months. Her son sustained burns
to his face and was hospitalized for three days. The driver has had multiple surgeries for
her burns.

Case Study: 1991 Toyota Previa

Fuels Available
 

Undercarriage view. Gas tank in center of photo, near top. Front

to the left. (see larger photo)

 
The only source of fuel was gasoline from a breach in the tank. This
was caused by the undercarriage of the van striking a protruding rod
on a tow dolly in the road.  The breach was 3-4 in. long and
approximately 1/2 in. wide (the diameter of the rod).  The breach was
through the bottom of the tank; the rod also punctured a 3/4" diameter
hole in the top of the tank.  No other systems were disturbed.

Penetration in gas tank from tow dolly

Case Study: 1991 Toyota Previa
Ignition and Propagation
 
 Tow dolly that was stuck by Toyota Previa
Ignition:  The primary source of ignition was likely to be from
mechanical sparks created by the tow dolly rubbing against the
pavement, as evidenced by score marks on the tow dolly.  Less likely
is the possibility of auto ignition on the exhaust manifold or braided
metal exhaust pipe covering.  These exhaust system components
were fairly distant from the fuel source and not likely to be hot
enough.

 Propagation:  The fire likely originated close to the hole at the front

outboard corner of the tank, adjacent to the driver's door.  The fire
propagated to the interior through the open driver's door.  The
passenger door was also likely open at some point during the fire,
then closed or nearly closed as evidenced by the burn pattern.  The
undercarriage fire was hottest near the leak (left side), but the interior
fire was hottest on the right side (likely due to wind direction). Wind
direction was confirmed through local weather services. Fire likely
initiated immediately, with propagation to the interior almost
immediately after the doors were opened.

Case Study: 1991 Toyota Previa

Summary of Fire Investigation
  

The tow dolly penetrated through the top and bottom of the tank
Gasoline was released from a large hole in gasoline tank caused by
collision with a tow dolly.  Ignition was likely from mechanical sparks
caused by abrasion between the tow dolly and the ground.  Fire
propagation was through open door(s).

End of Ignition Section

Congratulations!

You have completed review of the section covering ignition sources in

motor vehicles. Clicking the "next" buttons on this page will take you
to the section called "Fire Pattern."

Fire Pattern
Investigation of automotive fires often involves the interpretation of
burn or heat damage patterns on the vehicle body, in the engine
compartment, passenger compartment (interior), or cargo space.
Such patterns can provide clues as to the origin, cause or
propagation path of the fire. However, the use of such patterns can
be misleading and their interpretation must be used with care.

High-heat areas, such as portions of the body where all traces of

paint have been burned off (and metal has often rusted) can provide
clues as to the fire origin. This is especially true if a vehicle has not
been burned extensively. High heat areas may also overpressurize
sealed chambers, such as hood struts, hollow driveshafts, and other
vessels that show distortion. The more limited the fire damage (from
either self or external extinguishment), the greater the potential for
identifying the fire origin based on fire patterns. Obviously, if fire is
restricted to a small area, the fire origin is most likely within that
region. The interpretation of fire patterns for purposes of identifying
origin and cause must always be considered along with other
evidence including observations, witness statements, analytical test
results and the history of vehicle operation and repair.

This section will cover the interpretation of fire patterns and show
examples of various features.

Toyota Camry collision fire. Note that fire spread from

engine to interior, and heat radiated outward from
passenger windows. Also, there was less exterior burn
damage to lower area of rear doors.

Fire origin in the engine compartment can be identified even in some

events with extensive burn damage. The following picture shows the
engine compartment of a Toyota Camry with almost all combustibles
consumed in a post-collision fire. With collision and fire damage this
extensive, sometimes the origin of the fire can only be broadly
identified.

Extensive engine compartment burn

damage in a post-collision fire of a
Toyota Camry

A reliable identification of the engine compartment as the origin was

possible in this case based on the following:

 Collision damage was limited to the engine compartment. It is

likely that fire initiated in and around a damaged system.

 The burn damage in the engine compartment was near total;

the burn damage in the passenger compartment was
progressive from front to rear. If the fire had initiated in the
passenger compartment it would be likely to have appeared to
progress forward and rearward from the origin.

 Witness observations were consistent with fire initiating in the

engine compartment. Even without the witness statements,
identification of an engine compartment origin was justified.

 Fire Pattern
  Evidence of Origin
 Particularly in cases of limited burn damage, the pattern of fire
damage can be indicative of the fire origin. The more extensive the
damage, the less precise the information is likely to be and the larger
the identified area of possible origin. The photographs that follow on
this page and others in this section will give examples of the
interpretation of overall fire pattern ("macro" features) and of fire
damage to individual components ("micro" features).

Burn pattern on underside of BMW hood, showing heat

concentration near rear of engine and on right side. Forward
on vehicle is up and left in photograph.

  
 The lighter area in the burn pattern on
this Ford F150 correlates to the likely
location of the origin of the fire. For more
details about this case, click here.

  

A collision damaged the valve cover of

the Ford Escort pictured above, allowing
oil to leak onto the hot exhaust manifold
and ignite. Burn damage was limited
such that the only evidence of fire was in
the region of the exhaust manifold. For
more about this fire, click here.

 The investigator should also be aware that in instances of significant

burn damage, and even in some cases of limited burn damage, the
“hot spot” is not a function of the point of fire origin, but can instead
be due to the location with the greatest fuel load. If the fire begins in
a region with relatively little consumable material and then
propagates to an area with a large amount of flammable liquid or
combustible polymer, the secondary area will appear to be the hot
spot.
 
 This 1994 Saturn SL2 was subject to rear-end impact, fuel tank
damage and immediate fire. The most likely location of fire origin was
adjacent to tank damage under the passenger compartment, but the
fire quickly propagated through splits in the body sheet metal and
burned the interior.

Close-up view of gas tank through split body

panels.

 The hottest fire burned in the interior, which would mislead the
investigator as to the origin unless taken in context of all information.
For more on this fire click here.
 This metal tank in the following picture was punctured when the
driver drove over a tow dolly, spilling gasoline on the pavement and
initiating a pool fire under the vehicle. In this case, the origin would
be under the vehicle.

Gas tank punctured by tow dolly which led to a

pool fire. For more about this fire, click here.

ase Study: 1999 Pontiac Grand Am

Overview

Pontiac Grand Am non-collision engine fire. Fire originated

between engine and bulkhead. Note damage to right front
 fender.

 1999 Pontiac Grand Am non-collision fire

 Witness reports driving 5-10 mph on
residential street
 First noticed power steering assist was lost
 Smoke observed by 2 occupants
 Parked immediately
 Fire developed within 1-3 minutes

A 1999 Pontiac Grand Am was traveling on residential streets in an urban setting; the
vehicle had been running for 20-25 minutes. The vehicle reportedly had 8,000 miles on
the odometer. It was raining and daytime. The driver slowed to 5-10 miles per hour as he
approached an intersection with a small traffic circle. His intention was to continue
straight on the same road. To navigate around the turning circle he released the brake
and steered right, left, and right again to straighten the vehicle on the street. As he turned
left and right to straighten, he perceived the steering was "stiffening up". The driver
recalled rubbing or buzzing sounds while the passenger described clicking. As the
vehicle passed the traffic circle and after the steering stiffened, thick gray smoke was
visible coming from the engine compartment. The driver attempted to make a hard right
turn into a driveway to take the car off the street but the steering didn’t respond as he
expected. He was able to park.

When smoke was first visible, both witnesses described it as emanating from both right
and left sides of the engine compartment and toward the rear. The driver, front seat
passenger, and child restrained in a safety seat located in the back of the vehicle were all
able to exit without injury. The driver escorted the passengers 15-20 feet away and then
turned back to the vehicle. Flames were first observed at that time in the rear of the
passenger side front wheel well. Approximately 1-3 minutes transpired between steering
stiffness and the first observation of flames. There was no burning material evident on
the ground; flames were located on the bottom of the undercarriage. After the
passengers were clear, the driver returned to the vehicle and turned off the engine.

The resident of the adjacent home met the vehicle occupants as they got 15-20 feet from
the vehicle and offered to call 911. She walked back into her home and did so. The fire
department arrived 6 minutes after the call was placed and extinguished the fire. The fire
was extinguished 9 to 12 minutes after smoke was first seen and before reaching the
passenger compartment.

1999 Pontiac Grand Am Case Study Fuels and Ignition

Fuels: Possible initiating fuel sources included power steering fluid, coolant and
gasoline. There was no evidence of any breaches though all these fluids were in the area
of fire origin as indicated by the burn pattern. Witness statements suggest power steering
failure just prior to observed smoke and while the engine was still running. Power
steering lines and fittings were in close proximity to the center of the burn pattern and the
shielded exhaust manifold. While polymeric fuels were secondarily involved, the fire
scenario suggests that they were not involved in fire initiation. Based on the incident
description, burn pattern and location of potential fuels, power steering fluid was
determined to be the most likely initiating fuel.

Passenger side rear of engine compartment; bulkhead is

toward top in photo. Power steering lines are shown by arrows.

Ignition : Power steering fluid most likely autoignited on the shielded exhaust manifold. It
showed dark discoloration on its underside, toward the passenger side in the area of
power steering lines. Electrical wires were not a likely ignition source because they were
not located in the area identified as the fire origin. Given the description of power steering
failure, which was most likely a pre-fire event, it is unlikely that ignition was caused by
electrical system malfunction because this would require two simultaneous failures.
1999 Pontiac Grand Am Case Study Origin and Propagation
Fire originated near the rear of the engine, low in the engine compartment and somewhat
toward the right side. The burn pattern on the hood and behind the engine confirmed this.
Fire spread upward in the engine compartment and remained high. The discoloration on
the underside of the shielded rear exhaust manifold was concentrated on the passenger
side only. Fire was extinguished prior to entry into the passenger compartment. The
windshield showed heat damage but no breach.

Hood folded in half during salvage. Burn pattern toward rear,

which is on left side in photo.

Power steering lines and fittings were in close proximity to the shielded exhaust manifold
and to the center of the burn pattern. Although higher on the engine, gasoline and
coolant lines were located within the heat damaged area as well and could not be
eliminated as possible sources of originating fuel. Given the driver’s observations of
steering system malfunction just prior to seeing gray smoke, it was concluded that most
likely the fire originated due to a power steering fluid leak ignited on the exhaust
manifold.

 
 Engine compartment, with bulkhead toward right in photo. Fire
is believed to have originated on rear of engine, near bulkhead,
low in engine compartment.

Smoke and heat damage on inside of windshield and on dash

cover. Fire had just begun to propagate to interior when it was
extinguished.

Case Study: 1999 Pontiac Grand Am

Summary

 
  Fuel: Power steering fluid likely
• Gasoline, coolant, electrical possible
 Ignition: Hot exhaust manifold likely
 Propagation time: Extinguished within 9-12
minutes without spread to the interior
 
Witness statements suggest a power steering failure occurred; smoke was noticed
immediately after power steering malfunction and fire was observed within three minutes
thereafter. Physical evidence showed the fire was concentrated behind the engine and
that it initiated low and likely at the shielded exhaust manifold. This was consistent with a
breach in a power steering hose, though fire damage to the hoses prevented
confirmation. Other possible fuels included coolant and gasoline, though they were
considered to be less likely because they were higher in the engine compartment and no
driveability problems were noticed other than the loss of power steering.

Fire Pattern Myths

It was once thought that a complete “melt down” of a vehicle was a certain indication of
arson. That is no longer the case. Vehicles are now constructed with a larger quantity of
flammable materials, plastics in particular, that can burn rapidly and completely. It is not
unusual for accidental fires, with no accelerant, to consume all vehicle combustibles if
there is no intervention. It is also not unusual for the interior of a vehicle to be gutted,
even though the fire started in the engine compartment.

Myths: Confounding Evidence of Hot Spots

Fire patterns show gradations in heat damage that may or may not be relevant to the fire
origin. For structural fire investigations, the area of lowest burn (or of highest heat), may
have great significance to the origin determination. Because automotive systems involve
multiple locations where quantities of flammable liquids are found (under the hood, for
instance), such “hot spots” may indicate locations of high fuel load and not be indicative
of origin. For example, the origin may be quite subtle, such as at an electrical connection,
but nearby flammables may propagate the fire and ignite a cache of brake fluid in the
master cylinder reservoir and leave evidence of a local hot spot by the reservoir.
Likewise, other flammable solids, including plastics, interior components, and tires can
provide distinctive burn patterns that may or may not point to the fire origin.

References

1. Hrynchuk, R.J., “A Study of Vehicle Fires of Known Ignition Source,”

International Association of Arson Investigators, Alberta Chapter, Edmonton
Alberta, January, 1983.

 
 Wind and Vehicle Attitude
Other factors must be considered in the interpretation of fire patterns. Fire generally
burns upward but wind can alter the fire pattern significantly. The windward side of a
vehicle can be nearly devoid of damage, while the lee side may be completely burned,
regardless of the origin. If a vehicle is tilted or on its side or roof during the fire, fire
patterns will generally reflect this. As with structural fires, molten or burning material,
such as in the interior of a vehicle can drop down and propagate fire to lower parts of the
interior.

Toyota Previa, showing extensive burn from a pool fire caused

by a ruptured fuel tank. Though rupture was actually on the
passenger side, fire was hotter on the driver side because of
prevailing winds from the direction of the passenger side.
To see a case study of this vehicle, click here.

 
 Plymouth Sundance collision fire, showing burn pattern that
occurred while vehicle was inverted. Paint remains on the hood
and roof where the vehicle rested on the ground.

Sundance fire pattern shows greatest heat low on the car and
near the rear. It was this part of the vehicle that was highest while
it burned in the inverted position.

Propagation through Openings

Fire propagates through openings that are available and often creates those openings as
it progresses. Openings in the metal bulkhead allow fire to pass from the engine
compartment to interior, or vice versa. Bulkhead openings are often found for the heating,
ventilating and air conditioning (HVAC) system, the steering shaft, wiring harness, throttle
cable, brake and transmission controls, and other accessories. Such openings may be
surrounded by a grommet or seal made of flammable material. Some HVAC penetrations
are covered by polymeric components that are completely consumed in a major fire. Fire
patterns at penetrations can sometimes be interpreted to determine the direction of fire
travel (e.g. from the interior to the engine compartment) as in the Dodge van photos
below. Openings in the cowl, just below the windshield can be interpreted similarly.

Ford Explorer engine compartment bulkhead, showing HVAC

and wiring openings though which fire propagated to interior.

The bulkhead of this heavily fire-damaged engine

compartment of a 1995 Toyota Camry is at the top of
the photograph. Several penetrations are visible
showing potential paths for fire.

 
Dodge van non-collision fire engine compartment (after being
cut and partially disassembled during salvage after the fire).
showing burn pattern that indicates fire initiated behind the
bulkhead and propagated through an opening in the bulkhead.
Arrow shows direction of propagation. Rusted areas typically
indicate regions of high heat where paint is thoroughly
consumed by the fire.

Dodge van non collision fire, showing damage to steering shaft

boot on underside of vehicle. Heat damage to the boot clearly
 shows fire propagated from left to right through steering shaft
hole in bulkhead.

The arrows show the interior appearance of the

grommets sealing the floor pan openings for the
steering column (lower arrow) and brake pedal of a
1999 Ford Windstar.

Burn damage to wiring insulation, carpet, or other polymers immediately on the non-fire

side of metal bulkheads or panels can sometimes be caused by transmission of heat
through metal panels even though there is no penetration. If the carpeting has been
charred, melted or otherwise deformed, then the location can be considered as involved
in the fire.

The undercarriage has polymeric plugs (fully covering holes) and grommets (surrounding
wiring and controls) that penetrate the passenger compartment and trunk. These plugs
and grommets would be some of the earliest materials expected to demonstrate the
transmission of fire to or from the passenger compartment.
 The arrow indicates the location of an unburned plug in the floor
pan of a 1995 Ford Escort.
Note that carpeting on the floor and lower areas of the seats are not frequently consumed
even in extensive vehicle burns.

Even though the upper half of the passenger compartment had

little remaining combustible material, the flooring and lower
portion of the driver’s seat foam remained in this 1994 Saturn
SL2. More photos of this vehicle are presented in a case study
within this section.

Propagation through Openings: Burn Tests

As part of research sponsored by General Motors, instrumented vehicles were subject to
front or rear crash tests. The

  Vehicle Crash/Ignition Flame Penetration into

 the Passenger
Model
Compartment

Through broken
windshield, AC
1996 Front crash; electrical
evaporator and
1 Dodge ignition around the battery
condenser-line pass-
Caravan in the engine compartment
through and HVAC air
intake

1996 Rear crash; ignition of

Through split weld
2 Plymouth gasoline pool under the
seams
Voyager vehicle in the rear

Through crashed
induced seam openings,
1997 Rear crash; ignition of
gap between the driver’s
3 Chevrolet gasoline pool under the
door and door frame,
Camaro vehicle in the rear
and drain hole in the
floor panel

Front crash; ignition of

sprays and pools of Through windshield and
1997 mixtures of hot engine HVAC module in the
4 Chevrolet compartment fluids by a dash panel, both of
Camaro propane torch in and which were broken in the
below the engine crash test
compartment

Through crashed
induced seam opening,
1998 Rear crash; ignition of
ignition of quarter trim
5 Ford gasoline pool under the
panel by fire plume and
Explorer vehicle in the rear
heat conduction across
the floor pan.

Front crash; ignition of

1998
gasoline pool Through openings in the
6 Ford
under the vehicle in the floor panel
Explorer
rear

7 1998 Front crash; ignition of Through windshield and

Honda sprays and pools of pass-through openings
Accord mixtures of hot engine
 compartment fluids by a
propane torch in and
in the dash panel
  below the engine
compartment

1998 Rear crash; ignition of

Through crashed
8 Honda gasoline pool under the
induced seam openings
Accord vehicle in the rear

Effectiveness of Fire Retardant Treatment of HVAC Unit

1999
Chevrolet
Camaro,
9
Non-FR
HVAC Front crash; ignition by
Through windshield and
(control) electrical igniter installed in
HVAC module in the
the air cleaner housing in
dash panel
the engine compartment
1999
Chevrolet
10
Camaro,
FR HVAC

Effectiveness of Intumescent coating of the Underbody of the Vehicle

1999
Ford Rear crash; ignition of Through electrical pass-
Explorer, gasoline pool through openings in the
11
Test #6 is under the vehicle in the floor panel similar to test
the rear #6 (control)
control

Effectiveness of Fire Suppression System for Engine Fires

1999  Front crash, ignition of Test terminated well

Honda fluids in the engine before flame penetration
12
Accord compartment in the crash into the passenger
  test compartment

References:
 1. Tewarson, A., et al., "Post Collision Motor Vehicle Fires," Prepared
for Motor Vehicle Fire Research Institute, October 2005.

2. Windows and Doors

3. The position of windows and doors can often be determined by studying fire patterns.
Open windows and doors create large propagation paths and often leave distinctive
heat patterns around the openings. It can be a challenge to determine whether the
opening from the door or window existed prior to the fire.
4. Discontinuities in heat pattern between the trailing edge of a door and the adjoining
body panel may indicate whether the door was open during the fire. Doors may change
position during the course of the fire; they may have been opened by the occupants in
an effort to escape the fire or by emergency personnel during a rescue.

Toyota Previa collision fire, showing discontinuous heat

damage at rear of driver’s door, indicating door was open
during fire.

5.  
6. As with structural fires, windows often break from heat; it is sometimes possible to
ascertain whether the broken window pre-existed the fire or was the result of fire. If
there are window fragments inside the car (or at the fire site) that are not heat
damaged or covered with soot, then there is a higher possibility that the broken glass
preceded the fire. Glass fragments and loose components found on the ground at the
scene are often swept up and put in the vehicle during salvage.
7. In some cases the position of the openable windows may be determined by noting the
position of the window lift mechanism inside the door. In extreme fires, where lift
mechanisms distort or few consumables remain, it is often impossible to make a
reliable determination of window position based on this technique.
8. Fire can penetrate around the polymeric materials surrounding windows, or break
windows propagating fire across the boundary.
 
 An engine compartment fire propagated through the
base of the windshield of this Taurus.

9.  
10. The response of the vehicle windshield can be indicative of fire origin. A fire originating
rear of the instrument panel in the passenger compartment will often damage the top of
the windshield first. Fire patterns may then radiate outward (on the hood) from the
windshield. Engine compartment fires tend to cause damage to the base of the
windshield, near the cowling.
11.  

Chevrolet Cavalier collision-fire, showing how heat from engine

fire damaged lower portion of windshield.

12.  
13. For another example, you can revisit the Grand Am case study by clicking here.
14.  
 Plymouth Voyager collision-fire, showing collision to front end
and heat damage to hood from engine fire.

15.  
16. Contrary to this observation, fires that occur in the instrument panel (vehicle interior)
may also tend to damage the lower part of the windshield first. Crush damage to the
hood may provide a path for fire either toward or away from the windshield which will
alter the propagation path.
17. It should also be noted that the body panels of some vehicles are composed of
aluminum or fiberglass. Propagation dynamics will inexorably change (when compared
to steel) if the panels deform during the fire.

Gas filled hood strut that burst from internal pressure during a
vehicle fire.

Collision Fires
 Collision fires have additional fire paths that may be readable with burn pattern
interpretation. Broken welds, torn sheet metal, broken windows, deformed or open doors
and dislodged components can allow propagation of fire into the vehicle spaces adjoining
those where the fire originated, or can allow flame passage from the ground in the event
of a pool fire of gasoline. Of course, collision fires also raise the likelihood that spilled
automotive fluids will cause or propagate a fire. More information about collision fires can
be found in the Collision Fire section of this course.

Case Study: 1994 Saturn SL2

Overview

Extensive crush to Saturn from high-speed rear impact.

Driver’s door likely removed by emergency services. Burn
damage significant inside passenger compartment.
 
 1994 Saturn SL2 (fire vehicle)
 Rear impact by front of 1995 Ford Taurus wagon
 Delta V 42-48 mph
 1 occupant, fatal injuries
 
Overview: A police officer was attempting to stop a 1994 Saturn SL2 for speeding on an
interstate highway. The Saturn driver inexplicably stopped in a lane of travel rather than
pulling off the road and onto a shoulder. The front of a 1995 Taurus wagon that had been
behind the Saturn and traveling in the same direction, struck the rear of the Saturn
almost immediately after the Saturn came to rest. Sparks were observed underneath the
Saturn and it caught fire immediately. The road was straight and dry and it was cloudy
and dark with no streetlights.

Collision and Fire Damage: Crush-related intrusion extended forward to the rear of the
driver's seat. Within 1-3 minutes of impact, a witness approached the driver's side of the
Saturn. There were no flames visible on the ground or in the rear of the vehicle, but
flames were visible in the passenger compartment behind the driver's seat and
approaching the back of the head of the motionless driver. Dry chemical fire
extinguishers reduced the visible flames briefly, but they resumed within 30 seconds. The
passenger compartment was fully involved within 3-5 minutes of impact. Logs showed
fire department personnel arriving within 7 minutes of impact. After the fire was
extinguished, burn damage was largely limited to the passenger compartment region,
with little extending to the undercarriage or engine compartment. The undercarriage
surrounding the plastic gasoline tank showed very little evidence of burn damage.

Delta V of the Saturn was estimated to be 42-48 mph by momentum and crush energy
methods. Delta V of the Taurus was estimated to be 32-38 mph.

Injuries: The first witness approaching the Saturn, a trained emergency medical

technician, observed that the only occupant was motionless and upright in the driver's
seat, with eyes open and pupils fixed. It was reported that both front doors were jammed
shut and the occupant could not be removed before the flames were too intense. The
Saturn driver died, most likely from traumatic injuries from impact. The Taurus had two
occupants that had been belted and protected by air bags. The only injuries to either
were neck and shoulder bums from seat belts.

Taurus, showing front end damage from collision with rear of

Saturn.

1994 Saturn SL2: Case Study Fuels and Ignition

 
 Underside shows reduction in wheel base to rear. Fuel tank at
arrow is in crush zone forward of rear axle.
 
Fuels: All impact damage was to the rear of the vehicle. Only gasoline and brake fluid
were in the crush zone, and there was no indication that brake fluid was in the region of
burn damage. The Taurus came to rest remote from the Saturn and was not involved in
fire. Consequently, gasoline from the Saturn gas tank was the only possible liquid fuel
source. The rapid rate of fire initiation and propagation was consistent with a fire fed by
liquid fuel. Also, during the inspection almost three months after the incident, gasoline
leaked from the tank to the ground when the vehicle was rolled onto the passenger side.
The leak could have been from a dislodged sending unit, breached fuel lines, filler, or
tank. No leaks were observable in situ and the tank could not be removed for further
inspection.

Undercarriage of Saturn; rear is down in photograph. Arrow

pointing from bottom to top identifies the plastic fuel tank and
associated shielding. Arrows from top to bottom point to
underbody scrapes.
 
Ignition: No definitive source of ignition was identified. A witness observed sparks under
the vehicle during or immediately after impact. The road was dark with no artificial lighting
that would have obscured the visibility of sparks. By the witness description, the volume
of sparks was most likely to have been produced mechanically. Several undercarriage
surfaces, including the exhaust pipe, showed evidence of scraping against the pavement.
There may also have been mechanical sparks from metal-to-metal contact between
vehicles and from crushing metal surfaces within the Saturn. Electrical sparks and
exposed hot filaments were present from energized lights on both vehicles.

View looking down behind driver’s seat (frame of seat in

foreground), through torn sheet metal. Dark, burn damaged
region of plastic fuel tank visible through tear.

Case Study: 1994 Saturn SL2

Origin and Propagation

 Passenger compartment burn damage down to seat cushions.
Frame of driver’s seat visible on right. Arrow shows burned
edge of fuel tank exposed through gap in torn sheet metal.

Origin and Propagation: Fire damage to the undercarriage was limited to a small area
forward of and above the left rear wheel well. Consumable materials remained in that
region and showed little heat damage. The upper corner of the fuel tank (in damaged,
post-impact position) showed heat exposure but no breach. Fire damage was extensive
in the passenger compartment with little in the trunk. In the passenger compartment,
most consumables remained at the level of the seat cushions and below. Some foam
seat material remained as well. Carpeting showed little burn damage anywhere. The
instrument panel was burn damaged on the upper left, upper right and lower right. No
burn damage was evident in the engine compartment.

With extensive crush, the surface of the gasoline tank adjacent to the floor pan could not
be examined in its entirety. However, inspection from above revealed no openings in the
floor pan in the region of the center of the tank. Passenger compartment integrity was
breached by seam openings of the left and right rear fender wells and floor pans; neither
of these areas was adjacent to any observable fuel leaks. The area of suspected leakage
was approximately centered in the vehicle; likely propagation paths were located along
the sides of the vehicle, through the torn fenderwell-to-body spot-welds. There was
minimal observable burn damage between these two points under the vehicle. There was
no breach of the trunk floor pan from the front of the spare tire rearward and no burn
damage in this area. Since burn damage was predominantly to the passenger
compartment with limited incursion to the undercarriage, a continuous tank leak was
discounted. Most likely, gasoline was released during impact compression of the tank
and was sprayed into the passenger compartment.

Propagation time was established from several witness statements. The police officers,
parked on the shoulder to the right of the Saturn point of impact, reported seeing flames
immediately. An off-duty emergency medical technician (EMT) was passed by the police
 vehicle, then came upon the scene even before the police officers exited their stopped
vehicle. He stopped immediately, identified himself to the officers, and approached the
Saturn. As he approached the Saturn he observed flames behind the driver’s seat but
none under the vehicle. His estimates and estimates from the activities described were
consistent with his approaching the vehicle within 1-3 minutes of impact. He attempted to
open the driver’s door, the passenger’s door, and the driver’s door again before the
flames prevented his further attempts. The EMT walked quickly to the Taurus to assist
occupants there and found it empty. Looking back at the Saturn he observed the
passenger compartment was fully involved. It was estimated that the fire was fully
involved within 3-5 minutes of impact. Fire service logs indicated their arrival 7 minutes
after impact.

Case Study: 1994 Saturn SL2

Summary

 
 Fuel:
• Gasoline
 Ignition:
• Mechanical spark
• Electrical spark
 Initiation: Immediate
 Propagation time: 1-3 minutes
 
The rear end impact caused seam openings on right and left sides of the passenger
compartment. Gasoline was most likely sprayed through these openings during the
impact, resulting in fire largely limited to the passenger compartment. Fire began
immediately from mechanical or electrical sparks, or hot filaments from lights. Mechanical
sparks were considered to be the most likely source due to their observed volume.
Propagation to the passenger compartment was within 1-3 minutes of impact.

Roadway: Burn Patterns

Sometimes it is possible for the investigator to visit the scene or site of the fire. If that is
the case, burn and debris patterns can provide clues as to fire origin and cause.
Sometimes, in vehicle collisions, there will be evidence of fire onboard another vehicle,
which will raise questions as to which vehicle was the cause. The attitude of the vehicle
can often be established, and it may make it easier to assess the effect of wind.
Depending on the incident location, the National Oceanic and Atmospheric
Administration (NOAA) may be contacted to provide information as to the prevailing wind
direction at the time of the fire. Burn patterns on roadways can be especially useful if
scene photographs exist that can be used to further identify the location of the fire and
other details of interest, such as the attitude of the fire vehicle, fluid spillage stains
leading to the fire vehicle, or the position of other vehicles that may have relevance to the
event.

Burn pattern on roadway, showing heat-damaged pavement,

burned vegetation, and debris within two months of the fire.

External Heat Patterns

Sometimes gasoline or other flammable liquids are sprayed on the body of a car, causing
that paint surface to become heat damaged. It is also possible for radiant heat from
a pool fire or another source (such as another fire burning nearby) to cause such
damage. In the event of multiple areas of heat damage that are not contiguous, there is a
greater possibility of an intentional fire.

Mitsubishi Expo, showing significant fire damage pattern over

engine compartment, but also surface heat damage to right
front fender and bumper. Close inspection showed that the two
 areas of heat damage were not contiguous. Arrow indicates
undamaged paint between two areas of burn damage.
 
 

Expo, showing closer view of bubbled paint and bumper, most

likely from an accelerant that had been applied there. The lack
of burn damage inside the engine compartment and on the
undercarriage indicated that the fender and bumper burn
damage was not caused by fire in either of those locations.
Fire was determined to be arson.

The Toyota Corolla above shows external front end burn

damage. The Corolla struck a Ford Explorer that also burned.
After collision, both vehicles came to rest adjacent to each other.
Witness statements corroborated physical evidence that the fire
started on the Explorer and propagated to the Corolla.
 Heat Shadows
Burn patterns can be used to pinpoint fire origin by observing where the shadowing of
heat has occurred. “Shadows” of relatively undamaged material may be visible in the
area surrounding the heat source if those materials were blocked from the heat. An
electrical conductor overheated by internal resistance will exhibit damage along its entire
length from the power source to ground. Alternatively, an electrical conductor damaged
by a fire may have insulation remaining in areas protected from the heat.

Engine compartment of Jeep Cherokee after non-collision fire;

the bulkhead is to the left. An aftermarket alarm wire (at upper
arrow) showed considerable heat damage. At the lower arrow,
there was less damage to the alarm wire insulation because of
heat shadowing from the larger wiring harness adjacent to it. This
ruled out internal overheat of the wire as a potential fire cause.

Metals: Melting Temperatures

The temperature history of various parts can sometimes be determined by observation of
molten metals. If aluminum in the vicinity has melted but copper has not, then it is
possible to put upper and lower bounds on the highest temperature reached at a
particular location. If there is evidence of molten steel, there is a strong indication of
some type of accelerant and a higher likelihood of arson. Published melting temperatures
of various metals are shown below.

Melting Temperatures of Metals

Material Melting
Point (°F)

Aluminum [1] 1220

 Brass [2] 1700-1850

Bronze [2] 1615-1940

Cast Iron [3] 2150-2300

Copper [1] 1985

Lead [1] 622

Nickel [1] 2651

Steel [4] 2600-2750

Tin [5] 450

786
Zinc [5]

References

1. Shackelford, J., Introduction to Materials Science for Engineers, Prentice Hall,

2000.

2. ASM Handbook, Volume 2, ASM International 1990.

3. Properties of Some Metals and Alloys, International Nickel Company, Inc.,
1968.

4. Avallone, E.A. et al., Marks’ Standard Handbook for Mechanical Engineers,

McGraw-Hill, 1986.

5. Shackelford, J.F., et al., The CRC Materials Science and Engineering

Handbook, CRC Press, 1992.

Combustible Metals
Motor vehicles are incorporating a growing volume of combustible metals (such as
Magnesium and Aluminum) for weight reduction. The following table lists properties of
some of these metals in their pure form. The metals used in vehicles, likely to be alloys
developed to improve properties related to corrosion and high temperature resistance,
would have different properties than those listed.  
 Properties of Combustible Metals

Material Melting Solid Metal

Point Ignition
(°F) Temperature
(°F)

Aluminum [1, 2] 1220 1030

Magnesium [2, 1200 885-1150

3]

Titanium [2] 3140 2900

1652
Zinc [2, 4] 786

Magnesium is light in color, silvery white, and when ignited burns with a "dazzling white
flame." It is used in flares and pyrotechnics. [5]

References

1. Shackelford, J., Introduction to Materials Science for Engineers, Prentice Hall,

2000.

2. NFPA 484 Standard for Combustible Metals, Metal Powders, and Metal
Dusts, 2002.

3. Mallinckrodt Chemicals, MSDS "Magnesium Metal, Powder," 1979.

4. Shackelford, J.F., et al., The CRC Materials Science and Engineering

Handbook, CRC Press, 1992.

5. Weast, R., CRC Handbook of Chemistry and Physics, 49th Edition, 1968.

Propagation Time
Knowledge of expected propagation times may assist the fire investigator in evaluating
hypotheses as to the initiating fuel and ignition source of a fire. For example, in one
series of tests of non-gasoline fed engine compartment fires, propagation to the
passenger compartment took from approximately 5 minutes to almost 25 minutes. In
contrast, gasoline pool fires propagated to the passenger compartment in approximately
1-4 minutes [1].
The next pages will show video clips of crash tests that resulted in fire and examples of
propagation times measured in full-scale vehicle fire tests. Such testing is helpful in
establishing sample propagation times. It should also be understood that test results are
highly dependent on the conditions of the tests and may vary substantially. Some factors
of importance that may be difficult to quantify in field fire investigations are wind speed
and direction, leak rates (in fluid fires) and electrical resistance (in fires resultant from
damaged and shorted wires).

In two unusual frontal crash tests, vehicles caught fire spontaneously (with no external
ignition source or artificial fuel application). The tests were performed with all the fluid
reservoirs filled, operating electrical systems and engines running. Instrumentation
recorded temperatures, the presence of vapors in different regions of the engine
compartment, and the fire was captured by high-speed film coverage. 

In a frontal crash test of a 1998 Honda Accord, it was struck by a 3,613 lb. deformable
moving barrier in a frontal impact at 65 mph; the impact was offset 50% to the driver's
side and at a 21 degree angle such that the principal direction of force was through the
center of gravity of the Accord [2]. The engine of the Accord was operated at high idle for
55 minutes prior to impact to simulate on-road exhaust system operating temperatures.

The high-speed film of the crash tests shows a flame in the engine compartment from
approximately 40 milliseconds to 400 milliseconds after impact. Visible flames were
observed again in the region of the windshield washer fluid reservoir approximately 5
minutes after impact until the fire was extinguished after an additional 5 minutes.
Instrumentation and testing indicated that the first flames were likely due to power
steering fluid ignition on the hot exhaust manifold. It should be noted that mechanical and
electrical sparks may also have been present and other fluids were released in the crash.

Further testing indicated that the methanol vapors in the crash-compromised windshield
washer fluid reservoir could sustain a non-luminous flame for 5 minutes before spreading
to other materials and becoming luminous. These results were consistent with the initial
flame causing ignition of the windshield washer fluid vapor in the reservoir, an incubation
period, and then fire propagation.

References

1. Tewarson, A., et al., "Post Collision Motor Vehicle Fires," Prepared

for Motor Vehicle Fire Research Institute, October 2005.
2. Jensen, J., et al., "Evaluation of Motor Vehicle Fire Initiation and
Propagation Part 11: Crash Tests on a Front-Wheel Drive
Passenger Vehicle." NHTSA docket number 98-3588-179, 1998.

Propagation Time

Dodge Caravan Crash Test and Fire

The next crash test and fire involved a 1996 Dodge Caravan Sport struck by a 3,611 lb.
deformable moving barrier in a frontal impact at 65 mph; the impact was offset 47% to
the driver's side and at a 25 degree angle such that the principal direction of force was
through the center of gravity of the Caravan [1]. The engine of the Caravan was operated
at high idle for one hour and 46 minutes prior to the crash to simulate on-road exhaust
system operating temperatures.

According to the test report:

"There were three separate electrical occurrences caused by the collision, which in
combination, caused the post-impact fire.

One event was the pinching and partially severing of a wiring harness between the front
edge of the left frame rail and the transaxle housing...Several wires in this harness,
including the starter cable, were shorted to chassis. The insulation on the starter cable
showed signs of melting and there was evidence of corrosion caused by electrical arcing
at the tips of the copper wire. Also, electrical arcing eroded a small area of the frame
rail...

A second electrical event noted during the test was the disconnection of the negative
terminal from the internal plates of the battery (which was identified during the post-test
inspection of the battery). The disconnection was caused by the motion of the battery top
and terminals upwards during the impact (which was observed in the high speed film).
The entire system will not operate and current will not flow without a battery ground...

A third electrical event that was noted was the penetration of a self-tapping screw into the
battery contacting the negative (and possibly positive) plates inside the forward cell of the
battery. The screw was identified during the disassembly of the battery and the
penetration was noted from inside the battery housing. The screw initially mounted the
metal bracket holding the power distribution center housing.

The penetration of the screw into the battery resulted in local resistive heating of both the
screw and the surrounding steel structure to which it was mounted...

The shorting of the starter cable, the disconnection of the negative battery terminal, and
the penetration of the screw into the battery, in combination, were the causes of the
electrical fire. However, in drawing conclusions on how post-impact fires might start, the
specific outcome of this test should not be regarded as determinative. It is impossible to
determine the repeatability of these specific occurrences, or whether they were affected
by the modifications made to the production vehicle in this specific test. As an example,
the removal of the hood could have increased the vertical displacement of the battery
contributing (to) the separation of the negative battery cable."

 
 Photograph of the burn-damaged battery after the crash and fire.

A video clip of the Caravan crash test and fire can be reviewed by clicking the play button
below.

References

1. Jensen, J., et al., "Evaluation of Motor Vehicle Fire Initiation and

Propagation Part 2: Crash Tests on a Passenger Van." NHTSA
docket number 98-3588-30, 1998.

Propagation Time

Test Measurements
Published research contains reports of controlled tests in which the time for fire
propagation is recorded. Each provides an example of fire propagation under the given
conditions. Frequently, the methods of ignition used are not representative, and thus the
results may be difficult to generalize. Even when representative ignition is achieved,
there are a numerous factors that can influence the time to propagate that are difficult to
identify in real-world events (such as wind speed and direction, quantity of material
leaking in fluid events, and resistance in electrical shorts). Still the published tests
provide insight and information as to what is possible under the conditions shown.

In the following test sequence, the vehicles were first crash tested at a General Motors
facility, and then taken to FM Global for instrumented burn tests. In the crash tests,
vehicles were impacted with deformable moving barriers at approximately 50 mph and 65
mph for rear and front, respectively [1]. A more detailed description of each of the burn
tests can be found on a table previously shown in this section by clicking here.

These tests show a clear distinction in propagation time between gasoline pool fires and
engine compartment fires; pool fires propagate to the interior much faster. “Flashover
time” is the time from fire ignition to flashover in the passenger compartment.

 
Flashover
Test #
Vehicle Time
 
(min)

Rear Crashed Vehicles, Gasoline Fires Started Under the Vehicle in

the Rear, leak rates shown

98 Honda Accord
8 #201 1.6
400 mL/min

97 Ford Explorer
5 #188 2.6
750 mL/min

97 Chevy Camaro
3 #158 3.3
515 mL/min

96 Plymouth
2 Voyager #143 3.3
250 mL/min

97 Ford Explorer
6a #18-9 4.0
350 mL/min

Front Crashed Vehicles, Fires Started in the Engine Compartment

 96 Dodge Caravan
1 #119 10.5
Battery fire

99 Chevy Camaro
9 #190 11.5
Std HVAC

99 Chevy Camaro
10 #190 11.5
FR HVAC

97 Chevy Camaro
4 #178 15.5
HVAC

98 Honda Accord
7 #203 26.5
washer fluid

 
This bar graph shows the same series of tests tabulated above [1]. The vehicles
subject to rear crash tests and pool fires are shown below the dividing line.

 
 The next burn tests shown (in the following table) involved fires initiated by igniting pans
of heptane (a major component of gasoline) in the engine compartment. A small
difference in time to propagate was observed between the tests, one of which had the
HVAC fan on [2].

Smoke to Fire to
Initiation Pass. Passenger Condition
Vehicl
Location Compartme Compartme s
e
s nt nt  
(minutes) (minutes)

 2/3 from
front to
bulkhead
Saab (400 ml Ventilation
8:45 10:05
9000 heptane fan off
each
side)  
 

2/3 from
front to
Ventilation
bulkhead
fan on.
Saab (400 ml
8:00   Fan
9000 heptane
shorted at
each
5:30
side)
 

In the following test, a pan of heptane was ignited in the passenger compartment, with
observations as to the spread to other areas of the vehicle [3].

Vehicle Initiation Location Propagation

1.5 liters heptane in open Immediate temperature rise in

Ford tray under LF seat, ignited. passenger compartment, increase in
Taurus LF door ajar, windows engine compartment ~15 minutes,
partially open. increase in trunk ~10 min.

 
Another set of tests were performed to quantify the rate of flame spread in buses. For
these, full-size seats were installed in enclosed models of regions of a bus. The seats
were then ignited in the manners identified below. One of the researchers [4] noted that
the results were dependent on the size and shape of the compartments used for testing.
Without full-scale vehicle tests, the results could still be used to make comparisons
between seat materials as to their relative tendency to propagate fire.

Braun & Klote, 1991 [4]

 Seats installed in model enclosures

 Ignition to seats: Gas burner, 100 KW flame

 Incapacitation in 2-3 minutes with some seat assemblies

Nelson & Bridgman, 1977 [5]

 Enclosure with seats, comparison of materials

Ignition of newspaper

 Flashover in 6-7 minutes

 Consideration of more than one material necessary

References

1. Tewarson, A., et al., "Post Collision Motor Vehicle Fires," Prepared

for Motor Vehicle Fire Research Institute, October 2005.

2. Johansson, E., “Investigation of Fire in Cars,” Saab-Scania AB,

Sweden, International Technical Conference on Experimental
Safety Vehicles, 12th, volume I, Washington, D.C., NHTSA, 1989.

3. Mangs, J., Keski-Rahkonen, O., “Characterization of the Fire

Behavior of a Burning Passenger Car. Part I: Car Fire
Experiments”, Fire Safety Journal, 23, 17-35, 1994.

4. Braun, E., et al., "An Assessment Methodology for the Fire

Performance of School Bus Interior Components, Fire Safety
Science, Third International Symposium, 1991.

5. Nelson, G., et al., "Material Performance in Transportation Vehicle

Interiors,"  Journal of Fire and Flammability, Vol. 8, July, 1977.

Case Study: Propagation Time

Freightliner Non-collision Fire

An investigator encountered a fire in a Freightliner tractor by the side of a highway. He
documented the propagation until the fire department arrived and extinguished it. [1]
The truck driver reported that the engine ran roughly, stalled, and came to a stop on an
eastbound road. The driver observed fire in the engine compartment and attempted to
extinguish it with a 5-pound fire extinguisher. The investigator arrived about 3 minutes
later and began to photograph the fire. He reported the following events:

The first photo was taken at approximately 6:10 PM and note that there
was a strong constant wind (estimated at 15 to 20 mph) from the
northeast. The greatest concentration of fire was on the lower left side of
the diesel engine (also confirmed by the truck driver) where the engine's
fuel system components and piping are located. Cause of the fire was not
known by the truck driver.

Fire was observed in the engine compartment. Wind directed propagation

rearward and to the curb-side. [1]

Subsequent digital still and video photography was sequential starting

from 6:10 PM and concluded at approximately 6:40 PM after fire
department had extinguished most of the tractor fire and was
concentrating on the enclosed aluminum refrigerated trailer containing
frozen potatoes.
The progression of the fire was documented. [1]

There were several truck drivers who came to the aid of this fire with their
respective dry chemical extinguishers that they carried on their semi-
tractors. From approximately 6:12 PM to approximately 6:16 PM there
were four dry chemical extinguishers applied to the engine compartment
through various openings around the hood but to no avail. Intense heat in
the engine compartment proved the extinguishing agent from various
extinguishers ineffective. There were two 8 lbs. and two 5 lbs. the
extinguishers used.

During the fire growth stage and after consuming most of the fiberglass
cab of the tractor, this investigator heard approximately five mild
explosions within the engine compartment and under the cab. Although
none of the mild explosions caused shrapnel to travel beyond two foot
from the tractor, one explosion near the left saddle tank lifted the tractor
several inches.

During this same period, this investigator heard a "pressure release

sound" from the front engine compartment area; most likely was from the
air-conditioning system; this investigator could not determine if there was
a corresponding intensity of the fire due to the strong wind.

First Mongo Fire Department quick response fire unit arrived on scene at
 approximately 6:28 PM and within two minutes began applying foam
solution through a single 1 ½” handline on the tractor. Mongo tanker with
two firefighters arrived on scene two minutes later and supplied water to
the attack unit. Fire had involved the semi-trailer before the arrival of the
fire department,

A video clip of the still photographs was produced. The progression shows propagation
being forced rearward more than would otherwise be expected. At times during the fire
wind direction shifted, as evident by the smoke. Fire propagated forward as well, in spite
of the wind. The tractor fiberglass body panels became heavily involved, as did burning
material on the ground.

References

1. Photographs and case information are provided courtesy of Robert Marihugh,

Defiance, OH.

End of Fire Pattern Section

Congratulations!

You have completed review of the section covering Fire Patterns in motor vehicles.

For now, clicking "next" will take you to the next section: Arson

Arson Investigation
 

Arson, or the deliberate setting of fire, is a crime that is difficult to prove or detect. Even
when arson is confirmed, it can be difficult to prosecute the offender unless there is an
admission by the suspect, reliable eyewitness testimony, and such testimony is
accompanied by consistent physical evidence. The arsonist will usually commit the crime
without witnesses and will often go to great lengths to conceal the cause of fire.
Determination of arson is further complicated by the very nature of fire, which tends to
consume the physical evidence of its cause, as would be the desire of the arsonist. But
often the fire investigator is interested only in classifying a fire as accidental, intentional,
or undetermined.

This section covers the tools used to help determine whether arson should be considered
as a possible fire cause, and presents useful techniques for identifying and reducing
arson.

Sections will cover:

 Scope of Problem
  Motivation

 Physical Evidence

 Risk Factors

 Consistency of Evidence

Scope of Problem
 

The problem of arson is significant. In a study of national fire statistics [1], the estimated
annual average number of fires in highway vehicles was 266,000 which resulted in 350
civilian deaths and 960 million dollars in direct property damage. Of these, fires described
as “intentional” accounted for 15.6% of the total, 9% of the deaths, and 25% of the direct
property damage. Of the 45,100 estimated annual number of fires reported to originate in
the operator or passenger area, 53% were attributed to an intentional cause.

References

1. Ahrens, M., “An Overview of the U.S. Highway Vehicle Fire Problem.” SAE
2005-01-1420, 2005.

Motivation for Arson

 

There are many motivations for arson. They include:

 Fraud (insurance-related)

 Personal profit
 Concealment of another crime

 Vandalism

 Excitement (including pyromania)

 Social protest, riot, and terrorism

 Revenge or retaliation

 Jealousy

Motivation alone, of course is not sufficient to classify fire cause as arson. The arsonist
must also have the opportunity or access to the vehicle, and the means, which includes
the requisite knowledge, skill, and tools to set the fire. As a fire investigator, it is important
to be aware of these motivations and the various risk factors that may be indicative of
arson.

Physical Evidence of Possible Arson

 

There are many physical clues that may suggest arson is a fire cause. They include:

 Burn pattern: multiple origins or unusual origin location

 Presence/absence, and/or condition of ignition key

 Ignition lock/switch condition

 Accelerants in unexpected location(s)

 Missing accessories, components, or personal property

 Components that have been tampered with

 Evidence of forced entry

 Metal content in engine or transmission lubricants

 Physical damage or abuse to vehicle

For example, if a burned vehicle, reported stolen by the owner, is found in a deserted
area with the key in the ignition switch, then it would obviously raise questions as to how
the key was obtained by whoever “stole” the vehicle. In such a case, the owner may have
been motivated to burn his or her vehicle to defraud the insurance company and get out
of making payments on the vehicle. Likewise, if there had been no key and there was
evidence of forced entry, then that would also be of interest. In this case, someone may
have stolen the vehicle and then burned it to destroy evidence of theft.

The presence of accelerants, such as gasoline, which can sometimes be confirmed with
analytical tests, is not always conclusive and must be used with caution. Most vehicles
are powered by gasoline, and it is not necessarily unusual for someone to carry gasoline
in a container in their car. It is also possible for inadvertent contamination to occur at the
fire scene before a valid sample can me made. However, the use of an accelerant such
as gasoline to start a fire can cause intense burns in carpeting and other combustible
materials that can leave telltale patterns that should be considered when making the
determination of cause.

In order for analytical tests for accelerants in carpet or other vehicle materials to be
accurate, evidence handling is extremely important. This includes not only using the
proper sampling and storage technique, but also the ability to gather samples and
conduct testing in a timely fashion. Any material that may contain a flammable or
ignitable liquid loses a quantity of the accelerant every day as it is exposed to weather.
One source indicated that using an analytical technique called gas chromatography
coupled with tandem mass spectrometry required that the proper sampling technique to
be used within 17 days in order for the testing to be valid. Lesser techniques required
even more prompt sampling, often within a few days [5].

Additional details on arson inspection techniques can be found in references 1-8.

References

1. Carter, R. E., Arson Investigation, Glencoe Press, 1978.

2. Cole, L., Investigation of Motor Vehicle Fires, Lee Books, 1992.

3. DeHaan, J., Kirk’s Fire Investigation, Prentice Hall, 1997.

4. ATF Arson Investigation Guide, Bureau of Alcohol, Tobacco, and Firearms,

The US Treasury Department, (no date).

5. Sutherland, D., et al., “GC/MS/MS An Important Development in Fire Debris

Analysis,” Fire and Arson Investigator, October 2000.

6. Seufert, F., “Steal and Burn,” Fire and Arson Investigator, December 1993.

7. Herndon, W., “Vehicle Arson/Fraud,” Fire and Arson Investigator, January

2000.

8. Hrynchuk, R., “A Study of Vehicle Fires of Known Ignition Source,” IAAI

Alberta Chapter, January 1983.

Risk Factors for Automotive Arson

 

There are a number of risk factors that may be indicative of an increased likelihood of
vehicular arson. These include:

 Event occurs in summer months between 8PM and 2AM

 Often set out of view

 Owner behind on payments

 Major repairs needed to vehicle

 Evidence that car did not operate before fire

 Reported stolen, but sound system, wheels, and other items of value present and
intact

 Reported stolen but key found in car

 Not reported stolen but stripped

 Tires/wheels not appropriate for car and/or don’t match owner’s statement

 Lug nuts not tight

  No evidence of personal property claimed to be in the car at time of fire

 Evidence of accelerant containers in vehicle or area of fire

 Damage to vehicle body, perhaps from hit and run collision

These risk factors alone obviously do not constitute sufficient reason to declare arson as
the cause. The investigator must consider the consistency of circumstantial and physical
evidence, as well as witness statements, before making the determination of arson.
Unlike accidental fires, the difficulty with arson is that the perpetrator will often attempt to
obscure the cause of fire, sometimes with sophisticated means.

Consistency of Evidence
 

Unless there is a direct admission of guilt, a definitive photograph or video tape showing
the crime, or irrefutable eyewitness testimony, it is difficult to convict an arsonist. More
traditionally, it is necessary to make the determination of arson by considering the
consistency of the physical evidence, witness statements, and factors that support the
motivation, opportunity, and means for a potential criminal. As fire investigators,
particularly in the private sector, it may not be necessary to convict or even identify the
potential arsonist. It may be more important to simply classify the fire as accidental or
intentional, and to have sufficient evidence to do so on a “more likely than not” basis.
There are occasions in which evidence is insufficient to meet the higher standard for
criminal conviction (beyond a reasonable doubt), but sufficient in civil proceedings to
deny insurance claims because of arson and intent to defraud.

If arson is suspected, the consistency of evidence related to the fire characteristics

should be evaluated for a number of issues including:

 Burn pattern compared to reported fire characteristics and progression

 Accelerants in unexpected locations

 Ignition and propagation times compared to witness reports

 Evidence of fluid leakage from vehicle in its usual parking area

Similarly, the consistency of evidence related to vehicle characteristics should be

evaluated for a number of issues including:

 Operational condition of ignition/steering wheel lock

 Missing accessories

 Evidence of forced entry

 Report of damaged bodywork and vehicle defects

  Inventory of ignition keys

 Remnants of key evident after fire

 Service records indicating that operating condition was consistent with owner
statement

 Vehicle service history relative to fire origin

 Condition of engine internals, including lubricants

Technology can be used for differentiation of potential fire causes. Some of the analytical
tools used for fire and arson investigation include

 On-board computer records, which are stored in the memory of modern vehicles
can sometimes be used to assess the operational condition of the vehicle prior to
an incident, including evidence of recent crash impulses or fault codes that may
indicate engine operational problems or trauma. Extracting and interpreting these
data may require specialized equipment and expertise. Of course, if the fire was
extensive, such records may have been lost.

 Gas chromatography can be used for chemical identification of possible

accelerants that may remain in trace amounts after the fire. Proper sample
gathering and handling must be followed, with analysis by a reputable laboratory.
As previously mentioned, caution must be exercised in the interpretation of results;
a positive test for accelerants, such as hydrocarbon fuels, may not indicate a
definitive cause of fire [1].

 Oil analysis can be conducted by a laboratory to evaluate the metal content in a

sample of engine or transmission oil. Such analysis may also be used evaluate the
oxidation level of the oil, that may be indicative of internal overheat from poor
operating condition. Note that full scale vehicle burn tests have shown that even
extensive fires are not likely to deteriorate oil and transmission fluid [2].
 Material and metallurgical testing and analysis can be conducted to determine the
cause of failed material (for example that may be indicative of forced entry or pre-
fire steering column lock damage), the temperature that a metal or polymer
attained, or other qualities that may be important to fire cause determination.
To learn more about a technique some suggest can be identify the last
key used in a lock (key pathway analysis), click here.

References

1. Sutherland, D., et al., “GC/MS/MS An Important Development in

Fire Debris Analysis,” Fire and Arson Investigator, October 2000.

2. Hrynchuk, R., “A Study of Vehicle Fires of Known Ignition Source,”

IAAI Alberta Chapter, January 1983.
 
 se Study: Mitsubishi Expo LRV

Overview

Mitsubishi Expo showing external heat damage to hood and

right front fender and bumper
 

Non-collision Fire Information

 1993 Mitsubishi Expo LRV

 Reported stolen

 No witnesses to fire, fire extinguished

without fire department involvement

 Fire damage limited to engine

compartment

The owner of a 1993 Mitsubishi Expo LRV minivan discovered it was missing and
reported it stolen. Two days later, another homeowner called the police to report that the
same vehicle had been parked by his house; the homeowner had been away from home
for an unknown duration. The vehicle had been parked down a steep grass driveway
where it was not be easily visible from the street. While no one reported seeing the fire,
police found there had been fire in the engine compartment. The fire was not burning
when the police arrived and they were not aware of any witnesses to the fire. Although
the ignition lock was not broken, the circumstances were consistent with theft. Police
reported having arrested individuals for car theft in the region and finding sets of master
keys in their possession.

Burn damage was limited to the upper, right rear area of the engine compartment. There
was no propagation to the interior; polymeric grommets in the bulkhead remained intact.
A yellow-white powder was present in the engine compartment throughout the burn-
damaged area. The powder had the appearance of fire suppressant.

Limited burn pattern to the rear of the engine compartment and hood indicated fire
predominantly rear of the center of the engine compartment, high, and toward the
passenger side. One flexible gasoline line was the only breached fluid source in the burn-
damaged area of the engine compartment. The brake fluid reservoir, also on the right
side, was burn damaged with fluid remaining. The charred remnants of a nylon garment
were found on the right rear of the valve cover. It is possible that the material was used
during an arson attempt on the vehicle.

Underhood view, showing heat damage concentrated near

bulkhead and toward passenger side.
 

View looking rearward toward bulkhead on passenger side.

Brake master cylinder reservoir is in the center. Green-colored
fabric, just behind valve cover is visible in the lower portion of
the photo.
  

In addition to the fire in the engine compartment, the right-front fender and adjacent
bumper also showed scorching from heat. This region was not directly associated with
other areas of burn damage.

Mitsubishi Expo LRV: Case Study

Fuels and Ignition

Fuels: This vehicle was stolen and reported to have been subject to engine compartment
fire damage during the two days it was missing. It was clear by the absence of impact
damage that this was a non-collision fire. Gasoline was the only liquid fuel system
breached during the fire. Although the engine could not be started to confirm it, there was
no other liquid fuel leakage evident. The only gasoline line breached was the hose
between the filter and fuel rail. The damage evident may have been due to heat (post-
fire) or from leakage leading to the fire. The brake fluid reservoir was also in the region of
heat damage; however, fluid was still present in the reservoir. It was considered a less
likely source of originating fuel; the lines to the master cylinder were not disturbed, the
reservoir itself was not pressurized, and the reservoir was rearward and outboard of the
region of predominant heat damage. There were electrical wires with heat damage as
well, but all appeared to have been subject to external heating.

The possibility of arson was also considered. A heavily burn-damaged piece of nylon
cloth-like material was found on the right rear of the valve cover. This material was
reasonably central to the heat damage. In addition to the fire in the engine compartment,
the right-front fender and adjacent bumper also showed scorching from heat. This region
was not directly associated with other areas of burn damage. Other combustibles located
between the right-front fender and engine fire did not demonstrate evidence of burn
damage. It was considered possible that, during an arson attempt, gasoline or other
accelerants were poured in the rear right of the engine compartment. If accelerants were
spilled on the right-front fender the burn damage observed would be possible.

 
 Scorched green fabric material, previously located behind valve
cover was moved to top of valve cover for a photograph
 

Scorched right front fender and bumper, showing evidence of

external heat, likely from liquid accelerant burning on the
surface. This area was not contiguous with other areas of heat
damage on hood.
 

Ignition: No physical evidence of ignition was found in the engine compartment. The
shielded exhaust manifold was in front of the engine and lacked evidence of scorching or
fire damage. Sparking from normally operating electrical devices or shorting of wires later
disturbed by fire could not be ruled out. No evidence of pre-fire electrical system
malfunction was identified. Matches, a butane lighter and cigarette butts were found
inside the vehicle and may have been involved in arson.
 Mitsubishi Expo LRV: Case Study

Origin, Propagation and Summary

Origin and Propagation: The fire was predominantly high in the right rear of the engine
compartment. There was no propagation to the interior; polymeric grommets remained
intact. A yellow-white powder was present throughout the area of burn damage indicative
of the likelihood that the fire was extinguished. The right front fender was heat damaged
in a location disconnected from other heat damage; there was undisturbed combustible
material between the two areas.

Summary

 Fire damage in the right-rear of engine

compartment

 Fire damage in two non-contiguous

areas

 Remnant of charred garment in area of

fire damage

 Ignition lock appeared to be

undamaged

 Arson likely

The incident was a non-collision fire involving a car that was reported stolen. Little
information about the fire and recovery was available. While gasoline could not be
eliminated as an initiating fuel source, there was otherwise no cause to suspect it. Given
the reported theft, arson was also evaluated. Remnants of a nylon cloth-like material
were found on top of the engine on the right side. It was possible this material was
soaked with accelerant and ignited. The burn damage to the right front fender could have
been caused by accelerant spilled and ignited and would be consistent with arson as
well. Although less likely, the heat damage to the right-front fender could also have been
from heat reflected from an externally burning object.

End of Arson Section

Congratulations!

You have completed review of the section covering arson in motor vehicles.

Collision Fires: Introduction and Purpose

Collision fire events, though relatively infrequent when compared with non-collision fires,
are of significant importance because of their high injury potential. A collision carries the
risk of incapacitation and entrapment that has grave consequences both physically and
emotionally should a fire erupt, hence they continue to receive a great deal of attention
from investigators, vehicle designers, lawyers, insurance companies, and the general
public.

Previous sections of this course presented fundamental principles of vehicle fire

investigation; in this section, considerations specific to collision fires will be covered.
 

 Incide
nce of post-collision fires shows that they are relatively rare events [1]
 
 Fatal crashes with fire represent 3% of all fatal collisions; demonstrating a
disproportionately high injury rate due to either the high-energy collisions
that lead to both fatality and fire or fire as a cause of fatality.
 

Because of the complexity of collision fire events, it is difficult to relate the results of
laboratory and statistical studies to actual field incidents. Motor vehicle accident
databases, though very helpful for general statistics, lack the necessary detail to
understand the specific causes of fire, fuels involved, fire propagation factors, and injury
mechanisms. For greater detail, in-depth field studies of individual incidents have been
performed that provide insight into some of these issues [2, 3].

  

References

1. Parsons, G., “Motor Vehicle Fires in Traffic Crashes and the Effects of the
Fuel Integrity Standard,” DOT HS 807 675, NHTSA, 1990.

2. Scheibe, Robert R., Shields, Leland E., and Angelos, Timothy E., “Field
Investigation of Motor Vehicle Collision-Fires,” SAE 1999-01-0088, 1999

3. Ragland, C., et al., “A Case Study of 214 Fatal Crashes Involving Fire,”
NHTSA 98-S4-O-08, 16th ESV Conference, 1998.

Investigation of Collision Fires

The investigation of collision fires involves the same principles as those that apply to the
investigation of non-collision fires, and each can be used to help study the other. All the
same vehicle systems must be inspected and considered during an investigation. But
when a collision is involved, on-board systems that were disturbed by collision provide
clues to fire causation. In particular:
  Fluid systems can be breached and disrupted

 Electrical systems severed and circuits opened or shorted

 Flammable materials moved into contact with hot exhaust components

 High-impact or rollover collisions can disperse and/or spill flammable liquid fuels
over a broad area

 Mechanical sparks from metal-to-metal or metal-to-pavement contact can provide

additional ignition sources

If there is another vehicle involved, the fluids, electrical system, and exhaust components
from that vehicle can become involved as well. It is not uncommon in a car-to-car
collision fire that the first fuel ignited may be contributed by one vehicle while the source
of ignition may be contributed by the other.

Compared to non-collision fires, there is obviously a lower likelihood of fraud or arson,

especially for those cases involving substantial impact.

Collision Type and Severity

Collisions can be of several types:

 Frontal

 Rear

 Side

 Rollover

 Undercarriage or other

Collision severity is often characterized by parameters that can be helpful to the fire
investigator. Frequently, an accident reconstructionist makes judgments about collision
damage according to the seven-digit SAE Collision Deformation Classification [1]. This
classification is used to characterize the location and extent of damage in an abbreviated
format. Some examples of the coding for this classification include:

Direction of Force During Impact: The Principal Direction of Force (PDOF) is the

direction from which impact force is applied to the vehicle. A clock face is assigned to the
vehicle with 12:00 being at the front. Hence, a frontal collision into a wall would have an
impact location of "Front" and a PDOF (on the subject vehicle) of 12:00. A sideswipe from
an overtaking car on the driver’s side of the subject vehicle would have an impact
location of "Left" but a PDOF of 6:00. In the SAE Collision Damage Classification system,
the first two digits are the PDOF clock position.

Area of Deformation: The area of deformation broadly defines the area of the subject
vehicle that was damaged. It is given a letter designation, such as “F” for front, or “U” for
underside.
Other codes exist for

 Specific Longitudinal or Lateral Area

 Specific Vertical or Lateral Area

 Type of Damage Distribution

 Extent of crush damage

These factors, as well as measurements of crush damage and other vehicle data are
analyzed by the accident reconstructionist to determine a measure for the severity of
impact. This is often characterized by the “Delta V” or speed change of the vehicle as a
result of impact. Delta V is used to standardize the assignment of impact severity by
normalizing the result for the weight of both the vehicle, and the object with which it
collides. It is typically calculated by analytical techniques, and sometimes confirmed by
crash tests.

For statistical information relating fire location to collision type, click here.

References

1. Collision Deformation Classification – SAE J224, March 1980.

Process of Investigation
1. Collision fires are a small percent of all accidents and occur over a wide range
of circumstances and severity. Be aware that collisions with negligible Delta
V, under certain circumstances, may be sufficient to cause a fire.

2. Pay close attention to the disruption of vehicle systems, especially those

containing flammable fluids. Trace each and every fluid system to see which
of them can be ruled in as a possibility or ruled out. Because so many fluids
exist in the engine compartment, it can be very challenging to make this
determination when there is substantial collision damage to the underhood
area.

3. Examine electrical systems thoroughly, again focusing on any area of

disruption or deformation. Look for evidence of arcing, shorting, or
overheating that may provide clues as to electrical causes or involvement.
Recognize that shorted wires are not necessarily a cause of fire; they often
result from fire as well.

4. Look for deformation that may have forced flammable materials into contact
with a hot surface. For example, polymeric components normally separated
by space may have been crushed into contact with hot exhaust components.

5. During the collision, dynamic crush can be significantly greater than static
crush visible during a post-impact inspection. Note possible areas of
disturbance that may have taken place during the impact. For example, the
 engine may be pushed rearward during the impact in a manner that deforms
or breaches a fuel rail between the engine and bulkhead. The engine position
may then restore when collision forces are removed leaving little direct
evidence as to the source of the damage to the fuel rail.

6. Consider possible ignition sources such as mechanical sparks caused by

collision with other vehicles or objects, and the possibility of metal dragging
on the pavement. Also, consider that the ignition source may have been from
another vehicle involved in the collision (light bulb filaments, parting arcs from
broken electrical wiring) that may not be present at the inspection.

7. Be aware that the first material to ignite may have been splashed or sprayed
from its normal location to a distant location, vaporized by a hot surface, and
ignited by yet another source distant from the hot surface. Such fires can be
very complex and difficult to track.

8. Interview witnesses, if possible, to learn details about exactly where the fire
was first observed. Ask how much time elapsed between impact and ignition,
whether there were any smells or sounds that preceded the eruption of fire
(e.g., “poof”sound or explosion). Also inquire about the characteristics of the
flames first observed (robustness, color, size), the nature of the how long it
took for the fire to propagate from one place to another, what path the fire
took, how long it took to be extinguished, and how it was extinguished.

9. Examples of Collision Fire Investigations

10.
11. During a side swipe event, the right-rear quarter panel of a 1995 Pathfinder was torn
free, severing the fuel filler. Gasoline available from breaches to the filler and vent
hoses was distributed during the subsequent rollover of at least two rotations. Ignition
was likely caused by mechanical sparks during the rollover event. Other ignition
sources were possible. Reportedly, fire initiated immediately, consuming virtually all
combustibles.
12.  
 Close up of Pathfinder fuel filler and vent hose
remnants still attached to the body panel that
was torn free.

13.  
14.  
 15.
16. 1994 Toyota Camry impacted a narrow object to the front, off center to the left, with a
Delta V of 31-39 mph. Investigation revealed that coolant and brake fluid were released
in the collision, and that ignition of those fluids was likely through a hot surface (such
as the exhaust manifold) or possibly an electrical spark. Fire initiated within two
minutes of impact and propagated to the interior within five minutes.
17.  

1994 Camry, showing frontal damage

1990 Lincoln Town Car: Case Study Overview

  

 
Lincoln Town Car, showing extensive crash and fire damage

Collision Fire Information

 1990 Lincoln Town Car (subject)

 Rear impact by front of Chevy C20 Van

 Impact offset right

 Delta V 50-60 mph

 2 occupants, both fatalities

A 1990 Lincoln Town Car was being driven on a straight, two-lane road in the rain when
the driver lost control. The Town Car spun, crossed the centerline and was struck in the
rear right by the front-right of a Chevy C20 van. The two vehicles were originally traveling
in opposite directions; the road had a speed limit of 65 mph. Delta V of the Lincoln was
estimated to be 50-60 mph by the crush energy method.

 
 Lincoln Town car, showing total burn damage

A witness reported that fire began in the Town Car immediately after impact. By the time
the fire department arrived approximately nine minutes later the vehicle was fully
engulfed. All four tires had blown by that time. The fuel tank was ruptured in several
locations, the largest of which was a 4 inch long tear in the bottom of the tank caused by
a tire lug wrench, Collision forces pushed the wrench through the floor of the trunk and
the tank, puncturing the top and bottom of the tank. Other punctures from trunk items, as
well as tears at folds and seams were evident. Ignition of gasoline was likely from
mechanical or electrical sparks during collision.

Both occupants of the Town Car were fatally injured. The driver died from multiple blunt
force injures. The passenger was partially ejected; records were not available to
determine the cause of death of the passenger. Police reported that the driver of the van
was not injured, two adults and one child had incapacitating injuries, and two children
had non-incapacitating injuries. One of the adults, sitting in the right front seat of the van,
had a dislocated hip and shoulder.

1990 Lincoln Town Car: Case Study Fuels and Ignition

 

Fuels: Impact to the right-rear resulted in several holes and tears in the gasoline tank.
Breaches in the tank were caused by at least two rods, each puncturing both the top and
the bottom of the tank. One of the rods was a tire lug wrench still in position through the
tank at the time of inspection. The lug wrench induced a 4" tear in the bottom of the tank.
There were also tears at folds and seams.

 
 Tire iron pierced through trunk floor and breached gasoline tank

Ignition: No definitive physical evidence was found identifying ignition sources. The
collision occurred at night and presumably lights of both vehicles were on. Wiring was in
the crush zone of the rear of the Town Car and in proximity to the gasoline tank. Since
witnesses reported that the fire was immediate, it is also possible that the C20 van was
the source of ignition. The van had substantial front-end damage that may have provided
electrical sparks, resistance heating, or mechanical sparks from metal-to-metal contact.
The Town Car also had fluid leakage from a cracked transmission housing in the forward
end of the vehicle. Exhaust system components were located on both sides of the
transmission housing. If transmission fluid reached the exhaust system, it may have
penetrated gaps in shielding and ignited on the hot surfaces.

 
 Chevrolet C-20 van, which collided with Lincoln, may have provided a source of
ignition

1990 Lincoln Town Car: Case Study Origin, Propagation and

Summary
 

Origin and Propagation: While there were some differences in heat and extent of
consumption, all areas of the vehicle were substantially burned and most combustibles
were consumed. No definitive evidence of propagation direction remained. The likely
path is from gasoline underneath, burning hottest on the left, then propagating forward.
Even if transmission fluid was the ignition source, it was likely to have leaked out
immediately upon impact leaving little fuel for continued combustion. Sheet metal was
torn and spot welds broken in several locations. The trunk floor was torn adjacent to the
gas tank. Sheet metal was separated at the floor-to-sill junction by both sides of the rear
seats and forward to the 'B' pillar. The back light and rear side windows also broke during
impact. Right side doors were open during the fire due to impact damage (this was
confirmed by the investigating police officer). The area behind the rear seat, largely free
of a metal barrier, was also a potential propagation path.

 
 Torn sheetmetal in trunk and elsewhere provided propagation paths for the fire

A witness reported that fire began in the Town Car immediately after impact. The vehicle
was fully engulfed by the time the fire department arrived 11 to 12 minutes after impact.
At their time of arrival, the vehicle was in flames from front to back, all four tires were
involved, and metal was bending from heat.

Summary

 Breached tank by trunk contents and

other sources

 Fuel:

 Gasoline most likely

 Automatic transmission fluid

possible

 Ignition:

 Mechanical spark likely

 Electrical spark possible

 Initiation: Immediate

 Propagation time: Fully engulfed within

9 minutes
 End of Collision Section
 

Congratulations!

You have reached the end of the Collision section. Alternative and Hybrid is next.

 Alternative and Hybrid Vehicles

Traditionally, road vehicles have been powered by an internal combustion engine fueled
by spark-ignition gasoline or compression-ignition (diesel) engines. When we refer to
“Alternative Vehicles” in this course we address all vehicles with drive systems that
incorporate fuels other than gasoline or diesel, and use motive power from some means
other than (or in addition to) internal combustion engines.

Passenger, commercial and utility vehicles powered by propane (liquefied propane gas
or LPG) or compressed natural gas (CNG) have been retrofitted for use with conventional
engines for decades. Still, the presence of manufacturer-produced vehicles with
alternative drive systems and fuels is relatively new in the United States. The
technologies for these vehicles are evolving quickly and the numbers on the road still
relatively small. We have little real-world experience to discuss the fire investigation of
alternative vehicles, but reports on the subject are now being published.

Hybrid cars and SUVs are currently manufactured and sold by Toyota, Honda, Ford,
General Motors, Daimler-Chrysler and others; with US sales of about 175,000 vehicles in
2005 (1% of total sales) [1,2].

The designs are diverse as companies innovate to find the most efficient effective
designs. To support the increasing popularity of these vehicles, there is a growing body
of literature regarding the fire safety of vehicle occupants, fire service personnel, and
occupants of residential and commercial structures in which alternative vehicles are
parked.

This section covers types of alternative vehicle fuel and drive systems.

References

1. http://www.eere.energy.gov/afdc/pdfs/my2006_afv_atv.pdf , February 2006.

2. Wernle, B., Automotive News, January 9, 2006.

Types of Alternative Vehicles and Fuels

Alternative Drive Systems
      Electric
      Hybrid
Alternative Fuel
      Design for One or More Fuels
            Single fuel vehicles
            Flex-fuel vehicles (FFV)
            Bi-fuel vehicles
      Fuel Types
            Alcohol
            Biodiesel
            Compressed natural gas (CNG)
            Liquefied petroleum gas (LPG)
            Hydrogen

Alternative Drive Systems

Vehicles have been produced solely with electric drive motors for limited applications and
for trials of performance and customer acceptance (e.g., California and New York).
Currently, the most popular alternative vehicles are “hybrids” that combine both the
internal combustion engine (gasoline powered) and electric motors, utilizing the best
features of each.

The course material to follow will briefly review the components of electric vehicles as
they relate to fire investigation, and then address the components used in hybrid
vehicles.

Electric Vehicles
An electric vehicle (EV) is driven by an electric motor with power supplied by on-board
batteries. In the 1990s California environmental policy and regulation encouraged the
sale and lease of electric cars and supported recharge facilities. In early 2006, a small
proportion of those vehicles remained on the road [1]. There have also been
experimental programs for electric vehicles in the New York City area [2]. Manufacturers
have R&D programs for future production of EVs, but current production favors hybrid
vehicles combining gasoline and electric drive.

Electric vehicles have been built with AC and DC motors, operating in the range of 300
volts. Power is supplied most commonly by either nickel-metal-hydride or lead-acid
batteries. Packs of individual batteries (often 12 volts each) are generally located forward
of the rear axle in closed, vented compartments. The batteries in series supply the high
voltages required by the motors.

Vehicles sold in the United States operating with more than 48 volts are required to
comply with Society of Automotive Engineering "Recommended Practice for Electric and
Hybrid Electric Vehicle Battery Systems Crash Integrity Testing" (SAE J1766), and US
Federal Motor Vehicle Safety Standard 305, “Electric Powered Vehicles, Electrolyte
Spillage and Electrical Shock Protection.” These requirements limit the amount of
electrolyte spillage and electrical insulation degradation in frontal (FMVSS 208), side
(FMVSS 214) and rear (FMVSS 301) crash tests. Even if a vehicle complies with these
standards, impacts of differing configurations or energy levels may still compromise the
high-voltage systems resulting in fire or electric shock.

References

1. EV World, http://www.evworld.com/evguide.cfm?
section=evguide&evtype=production, January 12, 2006.

2. Sharke, P., Mechanical Engineering, January, 2002, p 37-41.

Process of Investigation
EVs contain some fuel and ignition sources common to traditional vehicles, and some
systems with unique fire risk. The key factors are listed below:

1. There is significant variation in layout and equipment used in EV design;

owner’s manuals and service manuals should be consulted for information
relevant to fluid and electrical systems for fire risk.

2. EVs generally have coolant for engine cooling and interior heating, hydraulic
brakes, windshield washer fluid and may have automatic transmission fluid
and other oils. As in traditional vehicle fire investigation, each fluid system
should be checked for involvement in fire.
3. EVs operate at voltages well above traditional vehicles increasing the risk of
fire due to electrical ignition. Higher voltages result in greater tendency for
arcing, higher energy arcs when they occur, and the potential for arc tracking
to create conductive pathways in polymers. (For more information about arc
tracking, click here.)

4. EV battery packs, with multiple batteries and high voltages, may overcharge
individual batteries if the batteries within a pack are not uniformly balanced.
Overcharging may overheat batteries or release hydrogen, either condition
increasing risk of fire.

5. High-voltage wiring insulation (as well as standard wiring insulation) may

become compromised as a result of collision or fire. High energy wiring would
be even more likely than 12- or 24-volt wiring to leave observable evidence of
a short after the fire.

Hybrid Vehicles
Hybrid vehicles are characterized by having more than one drive system. Typically,
gasoline engines and electric motors are employed to use the most effective operation of
 each. The gasoline engine is relatively efficient when operated at a steady speed for
which it was optimally geared. Thus, at a constant highway speed gasoline engines
provide good gas mileage. Electric motors are more efficient at lower and changing
speeds and are therefore optimal for stop-and-go city driving. Hybrids are designed to
automatically determine which drive system to utilize and to what degree for a given
driving condition.

Hybrids are inherently more complicated and expensive due to having multiple drive
systems, the necessary gearing to connect them to the drive wheels and processing to
determine how each motor is used.

Photograph under the hood of a 2005 Toyota Prius. The arrow

to the left indicates the top of the internal combustion engine.
The arrow to the right indicates the top of the inverter/converter;
the electric motor and transaxle are below. Fluid reservoirs are
also evident.
 

 
 Photograph of the same Toyota Prius looking in from the driver’s
side.
 

Hybrid vehicles may connect drive motors in series, parallel, or a combination of both
systems. The Toyota Prius includes two electric drive motors as well, one on each axle.
Power assisted steering is frequently applied electrically (as in the 2005 Toyota Prius and
Honda hybrids), eliminating the need for hydraulic fluid. Gasoline, coolant, engine oil,
brake, automatic transmission and windshield wiper fluids are generally present as liquid
fuels.

In addition to traditional sources of fuel and ignition, hybrid vehicles also have high
voltage wiring and battery supplies. A 12-volt lead-acid battery is often mounted in an
easy-to-find location, with either a Nickel Metal Hydride (NiMH) or lead-acid high-voltage
battery pack hidden beneath panels in the trunk or passenger compartment. The battery
packs have individual cells wired in series to develop high voltage, which is carried in
large orange cables from the battery compartment to the inverter/converter under the
hood. The inverter/converter takes the DC voltage from the batteries, converts it to the
voltages needed for various vehicle systems, and inverts it to AC for the electric drive
motors. AC motors are used for their greater efficiency

The Chevrolet manufactures a version of the Silverado that is an example of another

variation of called a “mild hybrid” vehicle. The gasoline engine of the Silverado shuts off
when braking and at idle. A 42-volt electric motor is used to generate electrical energy
during braking and for smooth starts when the brake pedal is released in order to begin
acceleration. Three lead-acid batteries provide the voltage for the electric motor. The fuel
savings is generated through the regenerative braking and engine shut-down when
power is not needed.

For more information about voltage used in some hybrids, click here.

 Process of Investigation
The key factors are listed below: .

1. There is significant variation in layout and equipment used in hybrid design;

owner’s manuals and service manuals should be consulted for information
relevant to fluid and electrical systems for fire risk.

Manufacturers provide emergency response guides that have some

information about hybrid and electric vehicle configurations. Some can be
found at manufacturers' websites such as the following:

Ford

https://canada.fleet.ford.com/pdfs2006/EscapeHybridEmergencyRespo
nse.pdf

 Honda (Select "Hybrid/Emergency Response")

https://techinfo.honda.com/rjanisis/logon.asp

Toyota and Lexus (Select "Emergency Response and Hybrid Information")

http://techinfo.lexus.com/

2. Hybrids generally contain most traditional fluids and often have an additional
cooling system for the inverter/converter. As in traditional vehicle fire
investigation, each fluid system should be checked for involvement in fire.
3. Hybrid electric motors operate at voltages well above traditional vehicles
increasing the risk of fire due to electrical ignition. Higher voltages result in
greater tendency for arcing, higher energy arcs when they occur, and the
potential for arc tracking to create conductive pathways in polymers.

For more information about arc tracking, click here.

4. Hybrid battery packs, with multiple batteries and high voltages, may
overcharge individual batteries if the batteries within a pack are not uniformly
balanced. Overcharging may overheat batteries or release hydrogen, either
condition bringing risk of fire.

5. As in electric vehicles, high-voltage wiring insulation may become

compromised during collision or as a result of fire.

6. Exhaust temps may be lower on hybrid vehicles (relative to traditional

vehicles) for low speed operation during which the hybrid electric motor
provides most of the drive energy.

7. Hybrid manufacturers frequently develop documents for emergency response

personnel that specify battery and high-voltage wiring placement. These
documents, available through manufacturer websites or customer service,
can be helpful during an investigation as well.
  Alternative Fuels
Traditional vehicles have a single drive system with an internal combustion engine driven
by a single fuel, gasoline or diesel. Some alternative vehicles are also built with a single
drive system but with a capacity to utilize more than one fuel as a power source.

This section will describe alternative types of engines and fuels not already covered in
the electric and hybrid vehicle sections.

 Design for One or More Fuels

            Single fuel vehicles
            Flex-fuel vehicles (FFV)
            Bi-fuel vehicles
 Fuel Types
            Alcohol
            Biodiesel
            Compressed natural gas (CNG)
            Liquefied petroleum gas (LPG)
            Hydrogen

Alternative Fuels: Design for one or more fuels

Single fuel vehicles (SFV):.Alternative SFVs are designed to operate with one fuel that
is other than gasoline or diesel.

Flex-fuel vehicles (FFV): FFVs are constructed with a single drive system and single
fuel tank, but the engine may operate on different fuels. FFVs provide options for
operation with a choice of fuels, depending on what is available. Some examples of pairs
of fuels currently used in FFVs are [1-2]:

1. Pure unleaded gasoline or a mixture of unleaded gasoline and alcohol: E85

(85% ethanol with gasoline) or M85 (85% methanol with gasoline).

2. Diesel and bio-diesel: Current research is exploring 5-20% bio-diesel mixed

with standard diesel [3].

FFVs are now common in Brazil after decades of government subsidy for the
development of ethanol as a vehicle fuel. Last year over one million FFVs were sold in
Brazil (70% of the market) that were designed to run on a blend of 22% ethanol/78%
gasoline, or nearly pure ethanol [4].

Bi-fuel vehicles (BFV): BFVs have two separate fuel systems (storage tanks) with the
capability of drawing from one or the other. Some examples of fuel combinations in
current BFVs are:

 Gasoline and CNG (compressed natural gas)

 Gasoline and LPG (propane, or liquid petroleum gas)

References

1. Alternative Fuels Data Center, US Department of Energy,

http://www.eere.energy.gov/afdc/, January 2006.
2. Department of Energy Vehicle Buyer’s Guide for Consumers,
http://www.eere.energy.gov/cleancities/vbg/consumers/, January 2006.

3. Alexander, D., “Fuel for Thought,” Automotive Engineering, Society of

Automotive Engineers, September 2005.

4. Press Release, Bosch, January 2006.

Alternative Fuels: Fuel Types

Alternative fuels, their characteristics and use in vehicles will be covered on this page.
Additional information and resources are available on the US Department of Energy
website [1].

Alcohol and Biodiesel: Alcohol and biodiesel are the alternative fuels that can most
efficiently utilize existing infrastructure for delivery and existing vehicle technology with
relatively few modifications. Mixtures of alcohol and gasoline are in common use, with
some designs utilizing mixtures of 85% alcohol (ethanol) and gasoline. There are also
initiatives to produce biodiesel from farm products and reduce dependence on oil
resources.

Natural gas (compressed natural gas, or CNG): Vehicles may have CNG fuel storage
alone, or CNG and a gasoline tank. According to the Natural Gas Vehicle coalition, there
are 130,000 CNG vehicles on the road in the United States and 2 million world-wide [2].
CNG tanks may operate at pressures above 2,400-3,600 psi [2,3]. Properly designed and
assembled systems will include a pressure relief device on the tank to release pressure
and fuel in a controlled manner if the vehicle burns. Without the pressure relief device,
overheating of the tank may result in a catastrophic failure of the tank, releasing
tremendous amounts of stored mechanical energy due to compression, as well as the
flammable gas. Even without the chemical energy from gas flammability, the mechanical
release of pressure can cause the tank to break its restraints, pierce any containment
provided by the vehicle body and fly great distances. With the pressure relief devices, the
controlled release may result in an intense column of flame for a short time (until
pressure in the tank is relieved), but the overall risk is likely to be reduced.

All vehicles powered by CNG, gasoline or other fuel have an inherent risk due to the
flammable material carried; some rate of fire is expected for all designs. In one survey of
over 8,000 CNG vehicles that traveled almost 180 million miles, there were 7 reported
fires, one of which was related to the CNG fuel system [4].

Video clip of a Southwest Research Institute fire test of a

CNG tank during which a PRD failed to open. [5]
 If you would like more information about CNG tank types and inspection
of CNG tanks for structural integrity, click here.

Propane, or liquid petroleum gas (LPG): LPG is currently used to power 350,000
single-fuel or bi-fuel vehicles in the United States  (4 million world wide)[2,6]. Most are
aftermarket conversions, but some are made by original equipment manufacturers.

LPG is 90% propane, with the balance made up of butane, propylene and other gases.
On-board propane tanks are generally at moderate pressure (160-200 psi) as compared
to hydrogen and compressed natural gas. Vehicles may have storage tanks only for
propane or they may have tanks for propane and gasoline. Otherwise, these vehicles
generally have traditional fuel and ignition sources.

Because LPG tanks are at a lower pressure than CNG (and hydrogen) tanks, there is
less stored mechanical energy to be released if there is a tank structural failure or over-
pressure due to a fire. Systems are still constructed with pressure relief devices to
prevent catastrophic failure of the tank structure in a fire.

Hydrogen and hydrogen fuel cells: Prototypes of hydrogen fuel cells are currently used
to power transit buses [7], and passenger cars [8].

The fuel for current hydrogen vehicles is stored as a compressed gas or in a cooled,
liquefied form. Research is also being conducted on reforming hydrogen from other
onboard fuels, such as ethanol, methane, and gasoline. Even more advanced methods
are being explored for storage of hydrogen at lower pressures within containers that
absorb and release hydrogen as a function of pressure and temperature.

In another Southwest Research Institute test of a

compressed hydrogen, the PRD was disabled so that a
catastrophic failure would ensue during a fire test.

Designs using compressed and liquefied hydrogen have some of the same design and
fire considerations related to fuel storage as the CNG vehicles already described.
Systems using reformers will include risks associated with storage of the relevant fuel.
Including fuel storage (or reforming) the fuel cell requires:

The part of the system referred to as the “fuel cell” takes hydrogen, either from stored
tanks or reformed from another fuel, chemically combines it with oxygen, producing
electrical power to drive a motor and water as a byproduct. The most promising current
fuel cell technologies for vehicles are:

 Alkaline fuel cells (AFC): Using an alkaline electrolyte, AFCs are

efficient and adaptable to the low-temperature vehicle environment, but
require pure hydrogen.

 Proton exchange membranes (PEM): PEMs use a solid electrolyte, are

also adaptable for low temperature environments, are less efficient than
AFCs but do not require a pure hydrogen source.

This video clip from UTC

Power (the manufacturer of
fuel cells), provides
 Information about how fuel
cells work. [9]

Additional information is available on the website of Ballard Power Systems (another fuel
cell manufacturer):
http://www.ballard.com/be_informed/fuel_cell_technology/how_the_technology_works#.
The Ballard Power Systems video also includes information about the fuel cell vehicle
and hydrogen distribution system.

The components of a fuel cell work together in the following way [10]:

Hydgrogen Fuel Cell Schematic

  

 Hydrogen is either stored in tanks for reformed (split from) other fuels.

 A controller senses the need for power and modifies the amount of fuel directed to
the fuel cell stack.
  The fuel cell stack takes hydrogen and air as inputs, and electrochemically
combines them to produce electricity to drive the motor, and water and heat as
byproducts.

 A cooling system removes the waste heat.

 A DC/AC converter increases the voltage (to the range of 300 V) and decreases
the current from the fuel cells.

 Electric power is directed to:

 An inverter to produce AC electric power (A/C motors are generally more

efficient for vehicle use).

 For battery charging.

 Power from the inverter drives the motor.

 The motor drives the axle and wheels.

A great deal of work on hydrogen vehicle safety is ongoing [11-13].

References

1. Alternative Fuels Data Center, US Department of

Energy, http://www.eere.energy.gov/afdc/, January 2006.

2. Department of Energy Vehicle Buyer’s Guide for Consumers, CNG and

propane, January 2006.

3. Webster, C. et al., "Experience Using Pressure Relief Devices in Compressed

Natural Gas Vehicles and Fill Station Service," Canada. International
Symposium on Protection of Dangerous Goods Tanks and Cylinders in Fire,
October 2002.
4. Clean Vehicle Education Foundation, Technology Committee Bulletin,
September 1999.

5. Stephenson, R., "Fire Safety of Hydrogen-Fueled Vehicles: Systems-Level

Bonfire Test," International Conference on Hydrogen Safety, 2005.
6. US DOE http://www.eere.energy.gov/afdc/afv/prop_vehicles.html%20,
January 2006.

7. Santa Clara Valley Transportation Authority, California, January 2006.

8. For example, see Phelan, M., “Honda’s FCX Performs Like a Conventional
Car, But Hydrogen Storage, Availability Are Hurdles,” Detroit Free Press,
March 2005.
9. Video included courtesy of UTC Power. It is also available (along with other
information about fuel cells) on their
 website. http://www.utcpower.com/fs/com/bin/fs_com_Page/0,5433,03540,00.
html, March 2006.

10. CRC, Electric and Hybrid Vehicles: Design Fundamentals, 2003.

11. Stephens, D., et al., “Survey of Potential Safety Issues with Hydrogen-
Powered Vehicles,” SAE 2006-01-0327, 2006.

12. Ohi, J., et al., “The Department of Energy’s Hydrogen Safety, Codes, and
Standards Program: Status Report on the National Template,” SAE 2006-01-
0325, 2006.

13. Scheffler, G., et al., “Developing Safety Standards for FCVs and Hydrogen
Vehicles,” SAE 2006-01-0326, 2006.

Compressed Gas Cylinders and Fire Safety

Reports have shown that gas cylinders used in CNG service have performed safely over
many years and miles [1]. Review of cylinder accidents indicates failures do occur from
service/maintenance-related issues (tanks damaged in impacts, corrosion) and
component failures. In addition, vehicle burn tests of hydrogen and CNG cylinders with
proper pressure relief devices have shown safe behavior in the event of fire [2].

There are four types of compressed gas cylinders used in vehicles. Each of these has
characteristic failure modes [3]:

Type 1: All-metal cylinders made of steel or aluminum.

Steel or aluminum liners with a hoop-wound composite
Type
wrapping. The metal liner is designed to withstand pressure
2:  
25% above the service pressure.
Type 3:  Thin metal liner and composite overwrap.
Type 4: A non-metallic liner with a composite overwrap.

Metal cylinders have high impact and fire resistance but are subject to corrosion.
Composite materials are lighter weight, often have good fatigue resistance and
resistance to environmental damage (corrosion) but are at higher risk of impact and fire
damage.

When inspecting a gas cylinder for damage, the factors to consider include:

1. Corrosion and pitting of metal, paying particular attention to surfaces under

attaching brackets.
 Surface pitting and corrosion of steel cylinders: Corrosion and pitting can reduce
surface thickness leading to loss of integrity. [3]

2. Corrosion of composite from electrolyte leakage or other strong chemicals.

Fiberglass wrap with stress corrosion cracking, likely due to exposure to

solvents. [3]

3. Scuffing, cuts or abrasion of metal or composite.

 Structural integrity of a fully wrapped cylinder is particularly sensitive to surface
cuts. [3]

4. Fiber cracking of composite indicative of impact. Damage may be greater

under the surface and not visible.

Impact and scuff damage to the surface of a cylinder with carbon composite
wrap construction may indicate greater damage to the cylinder wall. [3]

5. Bracket integrity and damage to tank from bracket. Brackets should have
appropriate liner for the cylinder.
 Brackets that are loose or missing gaskets can lead to wear or abrasion, as on
this composite wrap cylinder. The cylinder was also subject to heat damage
during the removal of a strap.  [3]

For more detailed information about visual inspection and care of CNG cylinders, see [3].

References

1. Clean Vehicle Education Foundation, Technology Committee Bulletin,

September 1999.

2. Yasumasa, M., “Test of Vehicle Ignition Due to Hydrogen Gas Leakage,” SAE
2006-01-0126, 2006.
3. Natural Gas Vehicle Business Unit, “Compressed Natural Gas (CNG)
Cylinders: Tips for General Visual Inspection and Care,” Gas Research
Institute, 1996. http://www.gastechnology.org/webroot/app/xn/xd.aspx?
xd=10AbstractPage\10682.xml

Photographs on page courtesy of Gas Technology Institute.

Case Studies

Three Cases with Natural Gas Fuel Cylinder Failures

 
 Fire and explosion damage to a Honda Civic CNG vehicle after tank rupture and
explosion.

1)    Honda Civic non-collision fire and CNG tank explosion: The Seattle Fire Department
was called to an arson fire in a parking lot. Fire fighters were on site when the fire, which
had involved a Honda Civic CNG vehicle, had a tank rupture and explosion. To review 54
presentation slides prepared by Randy Hansen of the Seattle Fire Department
summarizing the investigation, click here (12.5 MB).

To view case studies of two additional CNG tank failures described below, click "next".

2) Fueling rupture of a CNG tank without fire.

3) Pressure relief device venting of a transit bus CNG tank during a non-collision fire.

Case Study

Natural Gas Fuel Cylinder Failure

 
 Surface pitting and corrosion of steel cylinders: Corrosion and pitting can reduce
surface thickness leading to loss of integrity. [3]

The following description of a CNG cylinder rupture is quoted directly from the fact sheet
prepared by Charonic Canada Inc. in conjunction with Powertech Labs Inc. and the
Ontario Technical Standards and Safety Authority. [1] The case demonstrates the
importance of regular cylinder inspections.

At approximately 21:00 hours on Wednesday, September 3rd, 2003 a

natural gas cylinder installed on a vehicle ruptured during fuelling at a
Shell Canada station located at 1201 Markham Road, Scarborough.

The cylinder was one of three that were installed on a 1994 Dodge Ram
B250 Van when it was converted to bi-fuel gasoline/natural gas operation
in April 1994. The three cylinders were 70-litre Faber, 15-year life, all-
steel cylinders manufactured in accordance with CSA B51 Appendix G,
with Canadian Registration Number K4928.5. This registration requires a
visual inspection at least every three years.

The two cylinders mounted inside the vehicle were undamaged and had
retained the original paint coat applied before installation. The failed
cylinder had been mounted under the vehicle parallel to the exhaust
system and had suffered severe localized corrosion. At the time of failure
the vehicle was fitted with what was said to be a “fairly new” replacement
exhaust system.
Prior to the incident, the driver of the van had reported on the mornings of
September 2nd and 3rd that there was a smell of natural gas in the van.

The driver reported that he believed the cylinders were empty when the
vehicle arrived at the station and that only $10 worth of fuel had been
dispensed into the vehicle at the time the cylinder failed. He stated that
usually it took $25 to fill the vehicle from empty. A full fill on the day of the
incident would have resulted in a cylinder pressure of approximately 3400
psi, so that it is estimated that the failure occurred at less than 1700 psi.

Following the incident, the vehicle was inspected and the failed cylinder
was sent to Powertech Labs Inc. for evaluation. Powertech conducted a
failure analysis and reported that:

 The materials properties of the ruptured cylinder were consistent

with the specifications and test results provided by the cylinder
manufacturer.

 An area of significant corrosion measuring approximately 10 inches

in length and 4 inches in width was visible on the exterior surface of
the cylinder.

 Large-scale pits at depths of up to 66% of the cylinder wall

thickness were observed in the heavily corroded area.

 The accelerated external corrosion is consistent with what would

be expected from exposure to heat and exhaust emissions
associated with the cylinder’s proximity to the vehicle’s muffler.

 The cylinder failure mechanism is consistent with what would be

expected from the reduced wall thickness associated with the
observed external corrosion.

Powertech did not observe any retest or inspection markings on the failed
cylinder and no documented or anecdotal evidence has been found
indicating that it had been inspected by a qualified person.

Conclusions:

1. There is no evidence that this cylinder type or model is unsafe

if properly installed and maintained.

2. The cylinder failed through weakening caused by external

corrosion.

3. The corrosion was probably caused by exposure to hot

exhaust gases from a leaking exhaust system.

4. The corrosion would have been readily apparent on visual

inspection.

5. The cylinder would have been condemned if inspected by a

qualified person.
Investigators noted that the first owner of the vehicle had been instructed as to its use
and the need for inspection, and that the owner at the time of the rupture was the third
owner, a warning sign was proposed for refueling stations.

Fueling station warning sign recommended after this incident.[1]

References

1. Case description and photographs are courtesy of Nick White, Charonic

Canada, Inc.

Case Study

Fire in a CNG Transit Bus

 
 Fire on MARTA bus; it appears that water has not yet been applied. [1]

A fire originated in the rear-tire region of a Metropolitan Atlanta Rapid Transit Authority 
(MARTA) CNG bus during normal operation. The fire spread throughout the bus,
including to the roof-top CNG tanks. The PRDs vented prior to overpressurization of the
tanks, releasing gas which burned in the air above the bus. There were no injuries.
 A stream of gas burns above the bus, likely due to PRD
release. [1]

 
 Fire continued to propagate and consume side body panels. [1]

Smoke tended to rise to the street-side, with little fore-aft bias. The relatively lower
intensity of burn damage to the rear of the bus could not be attributed to wind direction
and was consistent with fire that initiated forward of the engine compartment.

 
 Engine compartment after the fire was extinguished. [1]

 
 The heavily damaged side of the bus is visible through the steam.
[1]

References

1. Photographs are courtesy of Peter Jensen, Atlanta, Georgia.

2. Case description courtesy of Hank Seiff, Clean Vehicle Education Foundation.

Process of Investigation
The key factors are listed below:

1. There is significant variation in layout and equipment based on the fuel and
drive system used. Owner’s manuals and service manuals should be
consulted for information relevant to the specific vehicle. It is important to
research or see an exemplar especially when inspecting vehicles of unfamiliar
design.
 2. As in traditional vehicle fire investigation, each fluid system should be
checked for involvement in fire.

3. FFVs have generally similar configurations and flammable materials as in

traditional vehicles. Investigators may need to consider the characteristics of
the alternative fuels used. (Specifications of both alternative and traditional
fuels are included in the Fuels section.)

4. For BFVs, fire investigators have two separate fuel containment systems and
two fuel delivery systems to inspect.
5. Electric drive motors operate at voltages well above the electrical systems in
traditional vehicles thereby increasing the risk of fire due to electrical ignition.
Higher voltages result in greater tendency for arcing, higher energy arcs when
they occur, and the potential for arc tracking to create conductive pathways
in polymers.

6. Compressed and cryogenic gas cylinders have additional risk factors for fire
and explosion.

7. Hydrogen and methanol may burn with a colorless flame. Witnesses may not
accurately detect when fire began. The presence of contaminants in the flame
(as would occur if vehicle components become involved) is likely to produce
observable color.

8. Propane is heavier than air; leaking fuel will pool in low, un-vented areas.
Hydrogen and CNG are lighter than air; leaking gas will gather in high un-
vented areas.

9. Location of pressure relief devices (PRDs) may be important prior to and

during a fire. If a PRD vents due to overpressurization of the system, it would
be desirable to have gas released in a vented region and in a direction that
venting during a fire would minimize exposure to those in the area.

10. Catastrophic overpressurization of a tank, during fire or otherwise, may

indicate improper functioning of the PRD.
 

References

1. www.fueleconomy.gov for DOE government site, January 2006.

End of Section
You have reached the end of the Alternative and Hybrid Vehicle section.

Clicking the "Next" button will take you to the Fire Suppression section.
 

Fire Suppression
Fire suppression systems, as of October 2005, are not common in passenger vehicle and
light truck applications and are seldom a factor in passenger vehicle fire investigation.
However, there are commercially available fire suppression systems used in some police
vehicles, heavy trucks, and transit buses, so it would be useful for fire investigators to
have familiarity with suppression systems at least in order to recognize them when
present.

This section provides information about fire detection and suppression systems:

 Applications currently in use

 Types of systems

 Detection

 Suppression

 Manual

 Passive

 Active

 Investigative considerations

Fire Suppression: Applications

Police cruisers which frequently stop on the shoulder of high-speed roadways are at
particularly high risk of severe rear-end collision, fuel system compromise and fire.
Transit buses, operating continuously in stop-and-go operation day after day, driven by
non-owners, and subject to various levels of maintenance attention are also at high risk
for mechanical failure and fire. Some heavy truck applications at higher risk for fire, such
as waste haulers, also install fire detection and suppression systems upon occasion. This
has led some fleets to purchase and install optional fire detection and suppression
systems. When studying fires in these industries the investigator should be especially
attentive to the presence of such systems.

Detection: Purpose
Detection systems are designed to a) activate an alarm, b) adjust vehicle operation, or c)
activate fire suppression. Once a fire is detected, some systems notify the driver by
activating an alarm bell and/or a light on the instrument panel. The alarm signal may also
alter the function of various systems to enhance fire safety. For example, in some
vehicles the same signal that activates the alarm shuts down the cooling and ventilation
fans (to reduce spread of fire and prevent dispersal of suppression agents), and may
alter engine operation to encourage the driver to get off the road quickly. Early
notification may also give the driver an opportunity to intervene with a manually activated
fire extinguisher before the fire is beyond the scope of the extinguisher.

The fire detection signal is also used to activate automatic suppression systems.
Whether a manual or active system is used, the increased airflow due to vehicle motion
decreases the effectiveness of the suppression agent. There is a trade-off in suppression
system design; waiting to activate the system until the vehicle slows or comes to a stop
prevents the agent from being blown away by the additional airflow, but allows the fire to
grow more and possibly exceed the systems ability to suppress it.

etection: Methods
There are sensors based on heat, radiation and fumes that are designed to detect fire.
Available sensors include:

Thermal sensors: Heat is detected directly in specific locations (spot detectors) and
along the length of wires designed to activate when exposed to heat (linear detectors) [1].
Spot detectors are sensitive to fire in a small area; many are typically installed to
sufficiently protect an engine compartment. Linear detectors can be many feet in length
and strung throughout the engine compartment to detect fire around components with
complicated geometry. Both spot and linear sensors require cleaning to prevent
contamination and insulation from mud and debris.

An example of a thermal spot detector. [2]

 
 An example of thermal linear detector cable.

A linear detector (arrow) mounted on the ceiling of

an engine compartment. The wire is routed
throughout the engine compartment to enhance
early detection.

Radiation sensors: Radiation sensors are generally designed to detect specific

frequencies of light characteristic of fire; they are effective at locations remote from the
fire if the sensing element is in the field of view of the fire. Various sensors use infrared,
ultraviolet visible or a combination of spectra. If not cleaned during maintenance, the
sensitivity can be muted, preventing detection or increasing the time to detection.
 An optical infrared sensor
used in transit bus fire
detection systems.

The same optical detector mounted in a CNG

engine compartment.

References

1. NFPA Fire Protection Handbook, 17th Edition, 1991.

2. Photographs on page courtesy of Kidde Dual Spectrum.

Suppression Systems
There are two general categories of suppression systems; those inherent in design
requiring no moving parts (passive), and those involving a forced dispersal of a
suppressant agent (active). The following topics related to active and passive systems
will covered.
 Passive
  Crash release of dry chemical
  Fire blankets fall during fire
     
Active
  Manual activation
  Automatic activation
    Dry chemical or foam agent dispersal [1]
    Clean agent dispersal (halon gas and halon replacements)

References

1. Dierker, J., "Development of Ford Fire Suppression System," SAE 2005-01-

1791, 2005.

Suppression Systems: Passive

Some vehicles have fire retardant material in the engine compartment sound insulation
mounted on the underside of the hood. Plastic mounting clips melt in a fire and drop the
blanket to inhibit fire propagation.

A polymeric panel containing dry powder suppressant installed behind fuel systems is an
example of a passive design to protect against fuel system crash fires. When a vehicle
with the panel installed is struck and crushed in the region of the fuel tank, the panel
ruptures and releases the suppression agent to inhibit combustion of spilled gasoline.
These panels are currently installed on some police and race vehicles [1].

Limitations of fire retardant polymers

Ignition and propagation can be inhibited by use of fire retardant materials. However, it
should be understood that flammability is often not a binary (flammable/ non-flammable)
question. Many materials used in vehicle manufacture can be chosen for their
flammability properties, as characterized by standard tests [2] that measure degree of
resistance to burning.

Full-scale burn tests were used to evaluate the effectiveness of using fire retardant
polymers in an A/C system housing [3]. The housing was installed in a test vehicle. The
test vehicle and a control vehicle with OEM components were both crashed, and then
burned. The fire retardant component did not improve the propagation properties of the
test vehicle. For this test, other underhood polymers were of standard composition.

 
References

1. Bennett, J. M., “Principles, Testing and In-Field Experience for the Fire
Panel,” SAE 2005-01-1790, 2005.

2. FMVSS 302 (horizontal samples), UL 94 (vertical and horizontal samples).

3. Santrock, J., Hodges, S., "Evaluation of Automatic Fire Suppression Systems

in Full Scale Vehicle Fire Tests and Static Vehicle Fire Tests," SAE 2005-01-
1788, 2005.

Suppression Systems: Active

Manual activation

The manually activated extinguishment systems may be a hand-held tank-type

extinguisher, or an extinguisher mounted in the engine compartment but activated by the
operator in the driver’s seat. For systems activated remotely by the driver, there will be a
switch or button actuator in on the instrument panel a suppressant agent storage
container in the vehicle, possibly piping to direct the agent where needed and one or
more dispersal nozzles. The installed systems are analogous to the automatically
activated system, differing only by their means of activation. The components will be
further described in the automatic activation section.

Automatic Activation

Once the sensors signal for suppressant system activation, one of several strategies may
be initiated.

1. Immediate release of the suppression agent. For this approach, no

processing component is necessary for the system.

2. Delay in activation for an amount of time (for the vehicle operator to bring the
vehicle to rest) or until vehicle sensors indicate the vehicle is below a preset
speed. By delaying for a slower or stopped vehicle, the airflow is likely
reduced and a greater amount of the suppressant agent remains in the area
of need.
 The control module pictured above is a sample
of the type used in transit bus applications and
generally located in the operator’s area. [1]

When suppression systems are fixed in the vehicle, components remaining after the fire
may include:

1. Nozzles for dispersal of dry chemical or clean (gaseous) agents. Often the
nozzles have a similar appearance to water-based systems found in
buildings.

2. Suppressant agent reservoirs. The reservoirs often have appearances similar

to small fire extinguisher cylinders.

3. Valves on reservoirs that may be activated by solenoid (electric control) or

pyrotechnic detonators (squib).

4. Less commonly, there may be an un-pressurized agent reservoir and a

pressurized cylinder of gas to expel the agent when needed.

5. Piping between the nozzle(s) and reservoirs and wiring to detectors or alarm
signal processor.
 An extinguisher bottle as used in commercial vehicle suppression systems.

Sample nozzles used for the distribution of dry chemical suppression agent.

Example of a System in a Transit Bus

This diagram shows a fire suppression system used in a transit bus. A detection system
and both manual and active actuation systems are described [2].

References

1. Photographs on page courtesy of Kidde Dual Spectrum.

2. Reproduced with permission of Port Authority of Allegheny County,

Pennsylvania.
 Process

Investigative Considerations
If a fire suppression system was fully effective in stopping a fire the burn pattern will
reasonably represent the conditions at the time the fire was suppressed. If the
suppression system was not fully effective, however, the pattern may have been altered
by the agent; the fire may have been unabated in some areas and slowed or temporarily
stopped in others. This effect should be considered when analyzing the burn pattern.

For systems with fire detection and an alarm signal, the operator and other witnesses can
provide information as to the conditions of operation at the time of fire. Examination of the
positions of sensors can then be considered with respect to the likely origin of the fire.
For example, if the likely origin of the fire was hidden from optical detectors, the fire
would have burned for some time prior to the alarm signal.

Fire suppression system function may also alter the investigator's evaluation of
probabilities of fire causation. For example, if witnesses report evidence of smoke or fire
for a significant period of time prior to the alarm signal, then one may conclude that the
fire originated in an area remote from the detector.

In cases where the effectiveness of the fire suppression system is one of the questions
being studied, then the investigator may need to evaluate:

 When the fire was detected with respect to time of ignition

 Where the fire started with reference to detection sensors

 The intensity of the fire at the time of suppression system activation

 Vehicle operating conditions at the time of ignition, detection, and suppression

system activation

 Evidence of detection and suppression system condition before the fire

(maintenance, pressurization)

 Evidence suppression effectiveness or suppression with re-ignition

End of Fire Suppression Section

Congratulations!

You have completed review of the section covering Fire Suppression in motor vehicles.
Clicking the "next" buttons on this page will take you to the section called "Resources"
 Resources
Resources included with this course:

Inspection Data Collection Forms

     List of all forms
     Checklist (pdf)
     Detailed Forms (pdf)
 

Reference Data

These links will take you to data within the course material.

 Fuel characteristics
   ·Fluids
   ·Polymers
   ·Gas

 Hot surface temperatures

 Metal Melting points

Bibliography
The bibliography is a comprehensive list of publications related to vehicle fire safety,
design and investigation. Some but not all of the publications have been referenced
within the course.

Glossary
For definitions of words common in vehicle fire investigation and research, click here to
see the glossary.

Standards
For a listing of standards relevant to vehicle design with respect to fire investigation, click
here.

Help

   System Compatibility
   Navigation
   Contact

Resources available on the internet:

 

National Highway Traffic Safety Administration (NHTSA)

The NHTSA website has crash tests, a search engine to check for vehicle recalls, and
access to databases of crash events.

http://www.nhtsa.dot.gov/portal/site/nhtsa/menuitem.bead436724af02e770f6df1020008a
0c/

Motor Vehicle Fire Research Institute (MVFRI)

Through the MVFRI website, there is access to reports of funded research on many
topics related to vehicle fire.

http://www.mvfri.org

George Washington University National Crash Analysis Center (NCAC)

The NCAC Library provides crash test reports, a portal to the Fatal Analysis Reporting
System (FARS) and access to a collection of papers related to vehicle safety issues in
general.

http://www.ncac.gwu.edu/filmlibrary/index.html

Society of Automotive Engineers (SAE)

The principal international organization for technical paper presentation and publication in
the ground vehicle industry. The website has a search engine for technical papers.

www.sae.org

National Fire Protection Association (NFPA)

NFPA publishes "NFPA 921: A Guide for Fire and Explosion Investigation," that includes
a substantial section for vehicle fires, "NFPA 556: Guide on Methods for Evaluating Fire
Hazard to Occupants of Passenger Road Vehicles" (in draft form as of 2008), and other
documents with material flammability properties.

http://www.nfpa.org/index.asp,

Neptune Engineering, Inc. (Vehicle data)

This commercial site provides vehicle data (Motor Vehicles Manufacturer data), crash
test data and reports, and a Vehicle Interchange List (a list of vehicles and their “sisters
and clones”; similar vehicles that may have different Makes and model years).

http://www.neptuneeng.com/products.cfm
 
Motor Vehicle Fire Research Institute (MVFRI) (Additional vehicle data)
This resource is a database of vehicle data (2003 model year) related to fuel system
design and containing vehicle photographs. It was developed by Biokinetics & Associates
Ltd under a grant from MVFRI.

Contact
If you would like to report a problem with program functionality, or have suggestions for
additional resources to include, please send email to: vehfire@u.washington.edu.

About

Course version, authors and acknowledgements