Determination of the origin of a fire is a necessary part of an investigation. Burn patterns at the loss site typically aid in the determination of the fire origin. The following are some thoughts to consider when assessing the origin of a building fire based on burn patterns.

The concept of low burn is often invoked as a way to determine the fire origin. The low burn area usually is the fire origin since the fire typically travels upward in a V shaped pattern. Figure 1 shows a low burn area where severe

Figure 1

damage has occurred, characterized also by deep charring of the wood. The well known phenomenon, that hot gases from a fire rise, accounts for the fire spreading towards the upper portion of the building (V pattern). In Figure 1 the pattern is typical of a single origin at the bottom of the V.

Figure 2 shows multiple V patterns in a building which is often construed to be multiple origins. Since multiple accidental fire origins are unusual, suspicion of arson is often opined. Other alternative possible causes should be investigated thoroughly. During one fire, the fire department began to fight a fire in a residential building until a substantial natural gas leak occurred. They were withdrawn from the scene until the gas could be turned off. As a result of holes chopped in walls and roofs by the fire fighters, the fire caused severe burnouts in several locations looking like a

Figure 2

multiple origin fire. The common characteristic of fire origins shown in Figures 1 and 2 are the lack of pattern distortion from other influences such as wind. Figure 3 shows burn pattern distortion from

Figure 3

wind. During windy conditions, a fire can easily spread downwind yielding a distorted V pattern. In some instances it is tempting to place the origin in the vicinity of the center of the burn (False origin Figure 3). If high winds were present, searching for an origin upwind from the center of damage may be fruitful.

Structural interactions with the fire can also result in a burn pattern distortion and false origin analysis. Figure 4a

Figure 4a

shows a fire starting above the doorway. As the fire progressed, it weakened the roof structure causing a collapse, resulting in the fire dropping to a lower level (drop down fire). This can result in a false origin determination as shown in Figure 4b.

Figure 4b

Flammable materials located in a building can cause distortion to a fire burn pattern. In Figure 5 a fire started to the right of the main entrance. However heat transfer from the fire ignited flammable materials nearby, causing a low burn area that can be mistaken as the origin.

Figure 5

In other instances, the fire may be so devastating that burn patterns may be inconclusive as far as the origin is concerned. Figure 6 shows a severe burnout of a building with no discernable pattern pointing to the origin. In these instances some other means may be required to find the origin.

Figure 6

Burn pattern recognition is quite useful in analyzing the origin of building fires. Thorough examination of burn patterns is necessary to avoid some of the pitfalls of false origin determinations.

 


 


Thermal Pattern Analysis

An Important Ingredient in Fire Investigation

 

 


After a fire, thermal patterns remaining at a fire scene are an important ingredient in fire investigation. Charred wood, soot deposition, melting, spalling and structural deformation are examples of thermal patterns caused by fires. When performing fire cause/origin analysis, fire investigators often seek out thermal patterns to aid in determination of the fire origin. The following article is a mini-encyclopedia of examples of thermal patterns along with what can and cannot be gleaned from the analysis of thermal patterns.

Figure 1a Classic V-pattern

Figure 1a depicts the classic V-pattern relied upon by many investigators. When hot gases from a fire rise in a natural convection mode, (as a result of heat related gas buoyancy) the gases tend to spread out as they rise, forming a V. Consequently, the fire investigator looks to the base of the V as the origin of the fire. In Figure 1a, the soot pattern, on the walls of the home, has spread outward near the eaves, suggesting that the fire origin is at grade level (arrow). Soot patterns inside the home were less severe, suggesting that the fire started on the

Figure 1b Melted gas meter and pressure regulator

outside of the wall. A closer look, at the base of the V, shows burned gas piping as shown in Figure 1b. Severe charring of wood and melted aluminum casings are indicative of a fire origin in this vicinity. It was determined that frost heaving caused movement of the natural gas piping, resulting in gas leakage through a compression fitting. The exact ignition source was undetermined. A possible ignition source could be a water heater, just inside the wall, igniting the natural gas before an explosive volume of gas could have accumulated. Air infiltration is a typical avenue for natural gas to

Figure 2 Kitchen fire

enter a home, especially in the winter. Figure 2 is a view of another classic V-pattern in a kitchen. The pattern is typical of an overheated appliance on the kitchen countertop. The arrow points to severely burned wallboard at the base of the V-pattern where a decomposed, plastic coffee pot housing was found, the apparent cause of the fire.

Propane gas is heavier than air and can cause random burn patterns throughout a building, provided there is sufficient venting of the deflagration so that an

Figure 3 Propane gas leakage causes a fire

explosion does not occur. In Figure 3, the propane cylinder on the fork truck (far left arrow) leaked, resulting in pooling of propane gas on the floor. A water heater pilot light is the likely ignition source of the propane. After ignition, random pools of propane ignited, causing multiple V-patterns as shown by the arrows on the right side of the photo. The amount of propane released before ignition was small, unable to sustain an explosion, but sufficient to cause random burning throughout the building, depending on the amount trapped in areas throughout the building. Multiple V-pattern observation is often a tenet of an opinion by an arson investigator that a fire of incendiary cause has occurred. Although multiple V-patterns have occurred in this example, the nature of the patterns is unique to propane gas accumulation and is not necessarily an arson.

Machine shops often use numerically controlled, electro discharge, milling machines. These machines use electrical arcs to erode away metal on the work piece to manufacture a finished part. The electrode and work piece are submerged in a dielectric oil for cooling and reduced oxidation of the electrode head. The machines run on a computer program and are often left running overnight, unattended. If the oil level drops below the electrical arc, ignition of the oil can occur, causing a smoky fire. A float switch is designed to shut off the machine if the oil level drops below a predetermined level. Figure 4 is a view

Figure 4 Electro discharge milling machine

of an EDM machine that was left operating overnight. A fire resulted, virtually destroying the building. Burn patterns on the side of the oil reservoir reflect a low oil level (arrow). The oil in the reservoir tends to cool the side of the tank, showing a different oxidation pattern than the upper side of the tank where high temperature oxidation is noted. A failure of the float switch is a possible cause of this fire.

Wood char depth is a burn-related pattern utilized by fire investigators. In Figure 5 the charring of wood joists, often

Figure 5 Char depth

called alligatoring, is a clue to the fire origin. The arrow shows the most severe char depth right above the furnace flue pipe, a typical consequence of flue pipes situated too close to combustible materials. Heating of the wood, over time, tends to lower its ignition temperature until it is ignited as a result of heat transfer from the flue pipe.

Figure 6a Before the fire

Figure 6a is a view of a propane gas fired heater in a produce truck, used to keep produce from freezing in the winter. The arrow points to pyrophoric decomposition of the wood near the heater exhaust. The discoloration of the wood is an indication that ignition of the plywood sheathing is eminent. Figure 6b is a view of a heater in an identical truck that caught fire, destroying the produce and part of the van. The arrow points to the area where the plywood sheathing had been installed. In this vehicle, pyrophoric decomposition of the wood has reduced the ignition temperature to nearly that of the exhaust pipe, resulting in the fire.

Figure 6b After the fire

Figure 7 Passenger seat auto fire

Figure 7 is a view of the passenger seat in a 4-door sedan. Burn patterns tend to focus on the seat as an origin of the fire. Melted polymers, such as polyurethane, are evident on the seat and dash. The damage to the dash is most likely a result of a fire on the seat. The vehicle was found in this condition, with the fire self-extinguished from lack of oxygen (doors closed). Arrows point to remains of emergency flares near the center of the seat. There are no vehicle related ignition sources in that area. Is arson ruled out in this case? Probably not, since spontaneous ignition of road flares on seating in automobiles is unlikely.

Figure 8 Fuel line repair

Burn patterns may not always point to the actual square inch where an ignition occurred. Fuel leakage in a vehicle, typically sprays fuel throughout the engine compartment, causing a diffuse burn pattern. Inspection of the fuel lines may show deficiencies that are causative to the fire, as shown by the improper fuel tube repair in Figure 8.

Figure 9 Fire remains in a dishwasher

Smoke damage occurred to a home as a result of soot evolution from a dishwasher. The dishwasher was in the drying cycle at the time of the fire. A view inside the washing compartment yields thermal patterns that are clues to the fire cause. The most severely burned area in the compartment is shown near a collection of utensils. The upper arrow, in Figure 9, shows a steak knife with the wood handle burned away. The lower arrow points to the heating coil used in the drying cycle. Thermal patterns are consistent with one of the steak knife handles being ignited by the hot drying coil in the washer.

Figure 10 Extension cord coil

Figure 10 shows a 100-foot extension cord wrapped in a neat coil and placed on a rug. The homeowner had connected a new dehumidifier to the cord in a finished basement. Thermal pattern analysis indicated melting of insulating material on the extension cord. There was smoke damage to the home, but no other areas suffered the degree of melted polymer than the electrical cord. Apparently, the extension cord was too long for the appliance, causing a voltage drop and resistive heating. The coil itself limited heat transfer from the wiring. The consequences were increased heating and increased electrical resistance, resulting in ignition of the insulation material and carpeting.

Burn patterns on water heaters are dramatic, in that one views a gradation of thermal patterns from the bright white of the original painted surface to the severely corroded, oxidized, sheet metal. Figure 11 depicts a typical V pattern with the base of the V near the gas valve. Caution should be exercised when interpreting such a thermal pattern. Possibilities of the fire cause are gas valve failure, gas pipe leakage, combustible materials placed too close to the heater, and

Figure 11 Typical gas fired water heater

flame rollout. In the case of Figure 11, flame rollout ignited nearby garbage bags.

One large loss involving a 2-story home occurred during the winter, while the owners were on vacation. Approximately 6 inches of water was found in the basement as well as water vapor throughout the home. Wallboard had deteriorated and fallen from wall studs exposing several burned areas, like that shown in Figure 12. Some investigators were interpreting the burned areas at the pipe joints to be a result of a malfunction of the water heater. What had actually occurred was a freeze failure of a few wall-mounted water pipes during unusually cold weather. This caused hot water to run through the pipe fractures into the basement, generating the large amount

Figure 12 Burn pattern near pipe joint

of water vapor that damaged the drywall. The burn patterns at the pipe joints were formed during construction of the building when plumbers soldered the pipes with torches and were unrelated to the loss.

Figure 13 The grounder

Some thermal patterns yield inconclusive results, such as the “grounder’ shown in Figure 13. The destruction is so widespread that pinpointing the fire origin, based on thermal patterns, is difficult. Other thermal patterns can be more definitive, as evidenced by previous examples. Thermal pattern recognition and interpretation will most likely remain a significant ingredient in fire cause/origin analysis.


 



During the heating season, many fire losses will occur with origins in building mechanical equipment. A thorough investigation requires information on the probable cause of the fire. The discovery of defective mechanical equipment design or installation will certainly enhance the chances of monetary recovery. The following are examples of fire causation often found in mechanical equipment in buildings.

Boilers

Figure 1 shows a badly overheated boiler that was the cause of a fire in a small apartment building. The building was nearly destroyed. The cause of the fire was the overheated boiler that ignited

Figure 1

floor joists in the vicinity of the flue pipe. A clogged low water cut-off control did not sense the drop in water level in the boiler, resulting in the overheating. The maintenance service hired by the insured did not follow appropriate blow-down procedures that would have prevented the malfunctioning.

Boiler control malfunctions can cause fires. The controls should be carefully removed and tested. Boiler installation should be inspected for proper clearances to combustible materials. Since boiler flue temperatures often reach 500 to 600 degrees Fahrenheit, a properly operating boiler can cause a fire if the flue pipe is near combustible materials such as wood structural components. Finally, operational procedures should be reviewed to ascertain the effect of the behavior of boiler operating personnel on the probable cause of the loss.

Furnaces

A significant number of furnace malfunctions occur in the flue pipe system. Flue blockage, insufficient flue distance from combustibles and improper flue assembly usage are typical causes. Control malfunctions may result from either gas valves or limit switches. In some instances, the limit switch failed causing overheating of the furnaces. Electrical faulting in wiring connected to furnace controls has been known to occur, thereby initiating a fire.

Water Heaters

Figure 2 shows a burned water heater that was the origin of a fire. Food grease from an overflowing trash container near the water heater had spilled onto the heater and drained down the side into the

Figure 2

burner area, resulting in the fire. Careless placement of combustible debris in the vicinity of water heaters is a common cause of fires. Other causes include malfunctioning gas valves and flue blockage.

Kerosene Heaters

Figure 3 shows a burned kerosene heater at the origin of a fire. Typical causes of kerosene heater fires are usage of gasoline as a fuel, spillage of fuel from the heater while ignited, placing the heater near combustible materials, and improper adjustment of the heater. Improper adjustment of the heater can result in a sooty flue product from the heater causing significant smoke damage without a fire. In such an instance, a defective wick adjustment mechanism was discovered which did not allow the insured to properly adjust the heater.

 Fireplaces

Fireplace fires are know to be caused by defective flue liners, improperly installed flue liners, creosote build-up in the chimney, blocked flue pipes (bird nests), insufficient screening of the fireplace, gas igniter malfunction, and insufficient hearth area in front of the fireplace. Check local building codes for installation deficiencies that could have caused the fire. Check to see if chimney sweeps had cleaned the fireplace flue. In one instance, a chimney sweep reassembled a flue liner improperly resulting in a gap that allowed hot debris to enter the attic.

Several causes of fires due to mechanical equipment have been reviewed. Many more can occur, depending on the complexity of the mechanical equipment.

Figure 3

 




Fires started by furnaces have been a major factor in losses associated with residential and commercial property. The beginning of the heating season usually brings a rash of fire losses as marginal equipment that survived the previous season fails, causing a fire. During the heating season as cold spells occur, marginal equipment is often over stressed, resulting in a fire. After a thorough burn pattern analysis is performed on the building and a preliminary assessment indicates a furnace related fire, a closer look at the furnace installation may be fruitful. The following is a listing of some typical fire causation modes associated with furnace (and boiler) fire losses that may be helpful in a causation analysis:

Combustible Material near the Furnace

Some furnace rooms act as a depository for cardboard boxes, garbage and other combustible materials. Placing these items close to a furnace can cause a fire as a result of heat transfer from the furnace or flame rollout from the draft diverter. Check for the remains of unusual combustible material near the furnace. Inquire as to what was in the furnace room at the time of the fire.

Improper Furnace Installation

Furnaces and furnace flues are often required to be installed with certain clearances from combustible building materials. Figure 1 shows the remains of a single wall flue vent installed near a floor joist causing the fire. The flue vent should have been a double wall design that reduces heat transfer from the hot flue gases to the floor framing.

Figure 1

Control Malfunction

Several controls are installed in a furnace to reduce the chance of a fire. Such controls are the high temperature limit switch, thermocouple pilot flame failure sensor and flue spillage sensor. Failure of any of these could result in a fire. Figure 2 shows a typical furnace burner

Figure 2

with the gas control valve indicated by the arrow. The gas control valve houses several safety features that should be checked during examination. Figure 3 shows a typical high temperature limit switch that is designed to turn off the furnace when plenum overheating occurs. If a control malfunction is suspected, laboratory testing should be performed on the suspected control mechanism if possible. Handle the controls with care. Jarring a sticky gas valve could result in loss of evidence.

Figure 3

Figure 4

Cracked Heat Exchanger

Figure 4 shows a cracked heat exchanger brought on by cyclic thermal stresses. Figure 5 is a close-up of the crack. As the crack develops, hot combustion products can impinge on combustible materials causing a fire. Heat exchanger cracks can also cause indoor air pollution with sometimes serious health effects.

Figure 5

Improper Wiring

Although rare, improper wiring of furnace control systems can cause a fire. One recent loss was a result of an improperly connected wire that bypassed overheating safety devices. When the fan motor failed, the furnace overheated since the high temperature safety devices were ineffective, resulting in severe damage to the building.

Figure 6

Sludge

Accumulation of debris or sludge in boilers can cause over- heating and fire. Figure 6 is a section of a boiler heat transfer tube that is partially blocked with calcium deposits as indicated by the arrow. Note the burned heat transfer fins. This is usually a result of poor maintenance on the boiler system.



 



Furnace and water heater related fires in residential buildings continue to be a cause of loss to many insurers. The previous article: “Furnace Related Fires,” reviewed several examples of furnace related fires. The following are some additional case studies. Figure 1 shows a furnace that sustained a relatively minor

Figure 1

fire that initiated in the vicinity of the high voltage module that drives the spark ignition system. The arrow in Figure 2 points to the badly damaged module which was the ignition source of the fire. Although the direct damage to the home from the fire was minimal, the induced damage was substantial. The home was unoccupied while the insured was on vacation during the winter. The

Figure 2

furnace caught fire but fortuitously self extinguished from lack of fuel in the vicinity of the fire origin. This caused the furnace to be inoperative, resulting in severe water damage to the home once the frozen water pipes thawed. The exact cause of the failure of the ignition module could not be determined because of severe damage to the unit from the fire. Figure 3 is a view of an old forced air furnace in a home that was badly damaged by a fire. Figure 4 is a view of a typical floor heating duct register that is badly burned. Figure 5 is a view of another register that is connected to the forced air system and is in good condition. Did the furnace overheat and cause a fire? The controls and burner which were located at the lower portion of the furnace were in good condition. Testing showed no malfunction in the burner gas valve or limit switches. The most

Figure 3

Figure 4

severely burned areas were near a few registers on the forced air ducting that was far away from the furnace. If the furnace was the heat source causing severe pyrolysis at the registers, how come the wood near the burners or the wood near some other registers is not pyrolyzed. This is a violation of the principles of heat transfer since heat does not naturally flow from a cold source to a hot source. Chemical analysis in the vicinity of the badly burned registers

Figure 5

Figure 6

confirmed the existence of an accelerant. Figure 6 is a view of a home after a natural gas explosion had occurred. A fire resulted but was quickly extinguished by

Figure 7

the fire department. The furnace and water heater were suspected as having malfunctioned but no apparent deficiencies were found. Within a week, two more homes on the same block suffered explosions. Suddenly the focus of the investigation was directed toward the gas utility piping. The gas main was excavated and several cracks were found in the piping like that shown in Figure 7. Figure 8 is a view of a typical part of the gas main showing markings indicative of having been struck by an excavator.

Figure 8

About 10 to 15 years before the loss, water pipes had been placed in the ground next to the gas main. Apparently the excavator had struck the gas piping during installation causing random cracks throughout the piping system. Seasonal soil movements enlarged the cracks which began to leak natural gas more profusely. Some of the residents noticed brown areas on their lawns despite every effort to properly care for the lawn during the summer. This resulted from natural gas filtering through the soil to the atmosphere. However, when the winter frost layer developed, the avenue of escape for the natural gas was blocked. The natural gas then moved horizontally, entering voids near sewer lines and other utilities, causing gas accumulation in some homes. The gas was most likely ignited by pilot systems of gas fired equipment.

In another case, a water heater was involved in an explosion when the insured attempted to light the pilot after the propane fuel cylinders were replaced. The insured indicated that he did not smell any gas. Chemical analysis of gas samples from the cylinders indicated very little or no odorant was added to the gas. The gas valve on the water heater was apparently on recall as a result of a propensity to remain open when the pilot light extinguished. The insured typically would let the gas run out before replacement of the fuel cylinders, relying on the gas valve safety system to close the valve when the pilot flame was extinguished. When the new propane cylinders were connected, gas accumulated in the basement that could not be smelled by the insured because of lack of odorant. The interaction of a sticky gas valve with the lack of odorant in the gas precipitated a hazardous environment, resulting in the loss when the insured attempted to light the pilot.

It should be noted that in two of the mentioned case studies the furnace or water heater did not cause the loss. It is not unusual for the furnace or water heater to be blamed for a loss, despite a lack of supporting evidence. Each of the case studies, however, had sufficient information to evaluate the role that the furnace or water heater played in the event.





As the heating season arrives, so do the fire losses originating in or near chimneys, fireplaces and flues. Typical causes include deficiencies in installation, deficiencies in manufacture, wear-out, biological (flue blockage from bird nests), lack of maintenance and improper usage. The following case studies serve to illustrate typical fire origin scenarios along with hints on what to look for when performing a cause and origin analysis.

Figure 1

Figure 1 is a view of a new wood stove in the middle of a family room, which was being used for the first time when a fire developed at the ceiling joist near the flue. Figure 2 is a view of burned ceiling joists where a single wall flue had been installed with little clearance between it and ceiling framing. Wood stove flue gases can be quite warm when compared to fuels such as natural gas or oil. Consequently, multiple wall flues are required or significant clearance between the flue and combustible material must be maintained. Since the fire occurred on the first ever usage and the owner had fueled the stove properly, an installation deficiency was suspected as born out by the discovery of a single wall flue with insufficient clearance.

Figure 2

Figure 3

Figure 3 is a view of a badly burned condominium unit with a fire origin near the fireplace. The owner indicated that he had smelled natural gas odor near the fireplace for several weeks prior to the fire. Why he did not attempt to correct the problem prior to the loss remains a mystery. Figure 4 shows a gas shut-off valve removed from the gas pipe that supplied natural gas to the fireplace burner. The gas valve had been installed backwards in the gas line. On certain gas valves an arrow indicates the direction of flow so that when the gas is shut off the

Figure 4

valve stem seal (indicated by the arrow in Figure 4) is not pressurized. Testing indicated that the valve stem seal leaked significantly. If the valve is installed backwards, the stem seal is then pressurized when shut, which can result in gas leakage and a possible fire or explosion. During operation of the burner, the pressure drops significantly in the valve resulting in virtually no leakage through the valve stem. In this case an installation deficiency caused the fire.

Figure 5

Figure 5 is a view of a metal flue pipe showing holes that have been drilled by the installer to secure a bracket for mounting in a wood frame chase. Apparently these holes were not in the proper location and the bracket was secured several inches above. Burn patterns suggest that the origin of the fire was at wood framing that was about 1 inch from these holes. The drilled holes penetrated both walls of the double wall flue. The fire occurred on a very cold day when the gas fired furnace was working at or near capacity. The holes offered an avenue for hot gases to exit the flue and come in contact with wood framing; setting in motion pyrolysis of the wood and eventual ignition. The penetration of the flue liner is an obvious construction deficiency.

Figure 6

On a cold autumn evening a homeowner noticed a sudden roaring sound coming from the fireplace after a fire was started in the fireplace. Shortly afterward, severe cracking was noticed in the chimney tile as shown by the lower arrow in Figure 6. The upper arrow points to severe sooting and creosote buildup in the chimney. The owners burned soft wood regularly and had not cleaned the chimney in 2 years. This is a classical chimney fire where creosote deposits, built up over the years, ignited, causing severe thermal damage to the chimney. This usually manifests itself in cracking of chimney tile and brickwork. Here lack of maintenance is the deficiency that caused the fire.

Figure 7

Figure 7 is a view of a fireplace showing the origin in the vicinity of the ash pit. Figure 8 is a view of the ash pit, which was improperly constructed since the hot

Figure 8

ash pile contacted combustible floor framing. This is a construction related deficiency.

Bird nests sometimes cause havoc with chimneys either by blocking the flue or by contacting hot metallic components.

In some instances owners use fireplaces as an incinerator, burning garbage and flammable waste.

A statement from the owner is often helpful in establishing the nature of the loss such as misuse, improper maintenance or a construction deficiency. Inspection and documentation of the scene is necessary especially in those matters where subrogation is contemplated. Even though the building may have been constructed several years before a fire occurred, an installation or design related deficiency would not necessarily cause a fire immediately. The fire depicted in Figure 5 did not manifest itself until one of the coldest days since the home was built was experienced. Finally, keep in touch regarding the latest developments in spoliation of scene evidence. Damaged evidence or failure to notify potential litigants regarding inspections and testing can limit monetary recovery potential.




Fires Resulting from Low Water Cutoff Control Malfunctions

 

 


Good design practice and local codes require that water heating boilers be equipped with a low water cutoff control. This device signals the burner to shut down when the water level in a boiler decreases to a predetermined level. An alarm then sounds for the boiler attendant to provide manual feed water (more water), or an automatic feed water control activates, bringing the water level to the appropriate height. The water level may decrease, due to small leaks in the system, causing overheating of the boiler if the low water cutoff malfunctions. The boiler can be damaged and the heat from the upper portion

Figure 1

of the boiler can ignite combustible materials such as wood joists in the boiler room. Figure 1 shows a boiler that has experienced overheating. Burn patterns are severe on the top side of the boiler

Figure 2

while the lower portion is unscathed. Figure 2 shows the deepest char depth, which is just above the boiler toward the flue end, a place expected to be quite hot from an overheating boiler. This particular boiler had a float type low water cutoff control. Electrical measurements were performed on all controls. Testing indicated that the low water cutoff control was still signaling the burner that the water level was within limits yet there was no water in the boiler. This testing

Figure 3

was performed with the control on the boiler. Figure 3 shows the low water cutoff after removal from the boiler. Notice the severe sludge deposits inside the control as shown by the arrow. This prevented the float from descending and opening the circuit to the burner. After the control was cleaned it worked properly. The probable cause of this fire was lack of maintenance causing a low water cutoff control to malfunction. The manufacturer of this control installed a warning tag on the unit stating that the operator should blow down (clean out) the control every week. It appeared that this had not been done for several months. The system had no water treatment apparatus installed.

While analyzing a fire in a building housing a boiler, the following characteristics may indicate a low water cutoff control related fire:

a.       The fire started in the boiler room or along a passage near the boiler flue.

b.      The boiler is badly damaged especially on the upper portion near the breeching.

c.       The boiler tubes on top are severely distorted while those on the bottom are not.

d.      There is no water treatment on the system to reduce the accumulation of contaminants entering the boiler water circuit.

e.       There appears to be generally poor housekeeping and poor maintenance in the boiler room.

f.        The boiler fluid delivery piping was known to have leaked extensively.

If a malfunction of the boiler is suspected, DO NOT remove the boiler controls until they are tested in place. A slight movement of a sticky low water control can cause it to return to an operational condition resulting in loss of important evidence. The remaining controls should also be tested to rule out any additional or aggravating causes such as a malfunctioning gas valve or temperature limit device.




Gas valves are installed in many appliances such as dryers, water heaters, furnaces, and industrial machinery. The purpose of the gas valve is to reduce the gas line pressure to a level required by the using appliance. The gas valve also acts as a safety device to shut off the gas automatically in case of flame failure. Modern gas valves incorporate pressure regulation for reliable burner operation. A diaphragm connected to a valve controls gas flow and pressure. A thermocouple senses heat from a pilot flame and generates a current that holds open a solenoid valve. If the flame near the thermocouple is extinguished, the current from the thermocouple stops and the solenoid valve closes off gas flow.

Although relatively rare, gas valves can fail causing a fire or explosion. Gas valves can fail from a variety of causes including contamination, product deficiency, and misuse.

CONTAMINATION

Most building mechanical codes require a sediment trap in the gas line, often called a drip leg. Figure 1 is a view

Figure 1

Figure 2

of a typical drip leg. Theoretically any debris that flows toward the gas valve will be snared by the drip leg. Figure 2 is a view of contaminants in a water heater gas valve. These contaminants accumulated on valve seats allowing gas flow when the heater was in the off position. This resulted in gas leakage and an eventual explosion. The contaminants were analyzed and found to be copper sulfide, a byproduct of the reaction between odorant in the gas and copper. It was found that a copper tube was used to connect the gas service to the water heater without using a drip leg. In this case, the two deficiencies were the use

Figure 3

of a copper line in gas service with a sulfur based odorant and the absence of a drip leg. Figure 3 is the radiograph (x-ray) of the water heater gas valve showing internal components. An x-ray is a valuable tool in analyzing the condition of a gas valve prior to disassembly since it is a nondestructive testing method.

Figure 4

PRODUCT DEFICIENCY

Figure 4 is a view of a home that sustained an explosion in the basement. The arrow points to structural damage at the wall. Figure 5 is a view of the water heater which was being lit by the home owner when the explosion occurred, causing serious burn injury. Apparently the home owner was in the habit of allowing the propane gas tanks to become

Figure 5

empty before calling for a refill. The gas valve on the water heater was an old design that had been recalled because of intermittent operation of the flame failure safety shut off. With this particular valve, one could turn the gas valve control knob from pilot to on when the pilot light was out. This caused damage to the valve which did not close during flame failure. When the homeowner went into the basement to light the water heater, he was unaware that gas was flowing through the gas valve due to an open main valve. An unfortunate synergism had occurred in that the propane gas that had just been delivered was not supplied with odorant. Consequently, he did not smell the leaking propane. Since propane is heavier than air, a pool of gas had accumulated on the floor which initiated an explosion when a flame was struck to light the water heater.

Figure 6

JURY-RIGS

Some gas valve related failures are a result of field expedients often called "jury-rigs.” Figure 6 shows a flattened tin can being used as a baffle to prevent the pilot flame from being extinguished by down drafts. The problem with this "new design" is that when the thermostat calls for heat and the gas valve opens, it becomes very difficult for the pilot flame to communicate with the raw natural gas at the burners. Consequently a large amount of natural gas flows from the burners before ignition occurs. This results in several large flash ignitions of the natural gas, with one such ignition finally igniting nearby combustible materials.

Gas valve malfunctions can cause property loss and personal injury. Statements as to activities by people near the appliance can be helpful such as the information on the habit of the homeowner who allows the gas to run out before reordering. Testing for the existence of odorant in the gas may explain why the leakage may not have been detected. Review of product recalls may show a deficiency in the gas valve. Finally, testing of the valve may be required for further insight as to the cause of the accident.





Hydrogen peroxide (H202) is a widely used chemical in the minerals, food processing, paper pulp, cosmetic, pharmaceutical and textile industries. Hydrogen peroxide is typically provided in a solution with water ranging from 2-70% concentration, depending on application. Hydrogen peroxide is a strong oxidizer and will initiate or sustain a combustion process quite readily. (Oxidizers are chemicals that release oxygen during a reaction.) If a hydrogen peroxide solution evaporates on a combustible material such as clothing, a fire may result spontaneously, without the need for an ignition source. Hydrogen peroxide was also used as an early rocket fuel oxidizer.

The following case study is illustrative of an oxidizer related fire. Figure 1 depicts the roof of a warehouse, the arrow pointing to a severely burned area of the metal deck. Looking below this burn pattern yields the photo shown in Figure 2. The arrow points to a badly decomposed mass of plastic bottles, packaging

Figure 1

Figure 2

material and wood pallets. The fire occurred early in the morning, before the dayshift arrived. The apparent source of the fire was several polyethylene bottles of hydroxide bleach, an example of which is shown in Figure 3. Conversations with the manufacturer of the bleach

Figure 3

product indicated that the concentration of peroxide in water was approximately 25%. The US Department of Transportation requires a DOT Oxidizer label on a product container for concentrations of

8 to 52% peroxide. (49CFR 172.101) The product involved in the fire did not have such a hazard label. Prudent packaging engineering dictates that the bottle for the 25% peroxide solution be opaque and have a vented cap. Sunlight can cause decomposition of the peroxide solution, resulting in oxygen evolution and

Figure 4

high pressures. A vented bottle cap is required to relieve pressures in the storage containers to prevent a catastrophic failure of the container. Figure 4 shows the type of opaque bottle that should have been used. A venting cap is not shown but is also required.

A reconstruction of the fire scenario is as follows: Peroxide solution (25% in water) was being stored in clear polyethylene bottles. Either as a result of agitation, sunlight, or contamination, the pressure increased in the peroxide laden bottles. Since there was no means of pressure release, the pressure in the bottles continued to build until they failed, releasing the peroxide. As the solution evaporated, the peroxide reacted with nearby combustible materials causing a fire. Spontaneous ignition of combustible materials near the oxidizer suggested leakage of the solution.

Storage and handling of hydrogen peroxide solutions is well regulated with special safety features mandated for containers. Product chemical analysis acts as a starting point to determine the cause of a fire whose characteristics mimic those found in this case study.

 



Flammable Liquids in Plastic Containers

Pose Fire Control Problem

 

 


A technical report entitled "Exploratory Testing of Flammable Liquids in Plastic Containers" (FMRCJ.10NOE4.RR) was published by Factory Mutual Research Corp., Norwood, Mass. on April 6, 1986. The article is somewhat technical, requiring a BS degree in science or engineering to totally understand the intricacies of the testing. The report is a preliminary study of the hazards of combustible and flammable liquids stored in plastic containers. Although many flammable liquids are stored in plastic containers, the study was limited to three typical liquids: mineral spirits, alcohol and corn oil. Polyethylene containers were used for testing because of their widespread use in packaging flammable liquids. Three sets of tests were performed. The first test dealt with characterizing the flammable materials from product data sheets and laboratory bench tests. Technical data such as specific gravity, heat of combustion, flash point, fire point, etc. was obtained and reported. The next two tests dealt with measurements of heat release rate, flame height, thermal radiation, temperature and visual observations of the flammable liquids in their plastic containers after being subjected to a heat load.

After this preliminary testing, Factory Mutual Research concludes that fires involving mineral spirits and alcohol products in their plastic containers will be difficult to control using conventional sprinkler technology. Corn oil testing indicated a better chance of control using conventional sprinkler systems. It appears that a way to analyze the flammability of liquids in plastic containers can be developed with additional testing. Implications are that flammable liquids in plastic containers pose a different type of fire hazard that may be difficult to control. The report may be useful to the underwriting department to make sure that they are aware of additional hazards of insureds utilizing flammable liquids in plastic containers.

While adjusting a loss, analyze the involvement of flammable liquids in plastic containers. They may have contributed to the loss, which could play a significant role in subrogation efforts.


 



 



While investigating a fire loss, the possibility of an electrical malfunction causing the fire is often at the top of the list of most analysts. It is not uncommon for an investigator to blame a fire on an electrical deficiency after ruling out other causes, even though there is insufficient evidence to formulate an opinion on an electrical origin. Care should be taken when analyzing electrical equipment to avoid the pitfalls and embarrassment of the "false origin." The following are some case studies that give insight into analyzing electrical malfunctions as a fire cause. Electrical short

Figure 1

circuits or faults typically generate heat and high temperatures that tend to melt wiring. The molten wire often forms spherical droplets before solidification occurs. This is typically called beading or balling. Figure 1 is a photograph of a typical electrical fault in a wire resulting in beading. A fault has occurred between two wires generating significant heat and melting of the wires. Does beading necessarily mean that one has found the cause of the fire. Not necessarily. Figure 2 shows several areas of beading on an

Figure 2

electrical conductor found in a fire. In many instances the beading may be a result of a fire that burns through electrical insulation causing a fault. At this point burn pattern recognition in the vicinity of the beading is important. If there is uniform destruction throughout the building with evidence of beading throughout the building, the beading is probably a result of the fire. However, if a precise area of fire origin is identified from burn patterns and a single wire is beaded in this area where no other ignition sources exist, it is likely that the electrical fault is a cause of the fire.

Electrical faults can also exist without beading. Figure 3 is a view of a wire that

Figure 3

short-circuited with insufficient resistance to cause arcing or beading. The fault developed and melted a portion of the wire without telltale beading. This does not necessarily mean that beading did not exist in the primary event. It means that the increased heating has caused the liquid copper droplets to coalesce into the wire strands causing a smooth solidified surface at the end of the wire. This is typical of smoldering faults which slowly heat and cause ignition.

In some instances the arcing generated from a fault can cause arc erosion. The discharge of electricity through the wires can erode away metal with disastrous consequences. Figure 4 is a view of a wire that was resting against a fuel line in a large tractor. During operation of the tractor, the wire insulation rubbed against the fuel line causing breakdown of insulation. As the electrical wire faulted, it generated a hole in the fuel line causing the fuel to be ignited by the arc. Severe damage was sustained by the tractor as a result of the fire.

Figure 4

Faulting in electrical wiring may or may not be a cause of a fire. Care should be taken in analyzing burn patterns and other ignition sources at the origin of the fire.

 


 



Chafing is the failure of wire insulation, usually as a result of mechanical means. Figure 1 shows typical examples of wire chafing. Figure 1A depicts the common "cut through" of the insulation from a sharp object. Wiring is often erroneously

Figure 1

routed over sharp edges that cut the insulation, causing a short-circuit and possible fire. Figure 1B shows the classical "wear through" configuration, a condition where wiring is in contact with an oscillating or vibrating part. The mechanical motion wears the wire insulation until a short circuit occurs along with a possible fire. In Figure 1C, foot traffic has caused a crush like failure to wiring routed under a carpet. This failure mode is known in the motor vehicle industry. Figure 1D is the pinch failure mode where the wire insulation is damaged by two mechanical parts acting like shears. A related failure mode is chafing as a result of rodent chewing.

Figure 2 is a view of a heating unit in a high rise building that suddenly started generating smoke. After the fire department had left, claims investigation began with the heating unit and associated wiring. There was severe faulting of the

Figure 2

Figure 3

power cable in the vicinity of a metallic enclosure inside the unit. Figure 3 is a view of the faulted power wire that had been removed from a flexible metallic conduit. Burn patterns confirmed that the electrical fault was the likely cause of the fire. In the apartment there were two additional identical units, certainly a fortuitous find since comparison with an exemplar usually aids in the investigation. Figure 4 is a view of the power wiring of the exemplar showing the flexible

Figure 4

metal conduit with the power wiring entering a metallic enclosure. Surprisingly, none of the other units in the apartment were equipped with a connector to connect the flexible metal conduit to the enclosure. The flexible metal conduit was unrestrained, relying on the wire to support it at the entry into the enclosure. The enclosure was constructed of a thin gauge of galvanized steel with sharp edges at the knockouts. The power wire rested against the sharp edge supporting the flexible cable. During heater operation, fan vibration caused the flexible metal conduit to oscillate, resulting in severe chafing of the wiring. Wires in the exemplar units were in various stages of chafing and on their way to eventual failure. This installation was deficient in that a connector should have been installed to stabilize the wiring and act as a protector from vibration and chafing.

Identification of the chafing failure mode obviously depends on the quality of the evidentiary remains. Usually the cause of failure is installation related such as improper securing of wires or lack of hardware. Sometimes a design deficiency arises through improper selection of wire insulation. There are various grades of wire insulation, some of which are chafing resistant. A review of the quality of wiring used in an installation may demonstrate that substandard wire insulation was the ultimate cause of the failure.

 

 


 



Fire development in clothes dryers causes a range of losses from severe smoke damage to the building interior to disruption of the entire building. The typical clothes dryer is composed of a large rotating drum which provides an environment of warm air and moisture removal. The drum is supported on bearings and usually powered by an electric motor. The heat source is usually an electric heating coil or a gas fired burner. Temperature limit switches are placed at strategic locations throughout the dryer to provide input to temperature control devices. There are a variety of causes of dryer fires with the following being typical examples.

Figure 1 is a view of an electric dryer

Figure 1

Figure 2

that sustained a fire in the drum. The origin of the fire appeared to be near electric heating coils as shown in Figure 2. Combustible debris such as floor dust appeared to be contacting the heating coil. A look under and around the dryer revealed a large accumulation of dust and lint. Ingestion of combustible debris into the dryer which contacts heating coils is a common cause of fires.

Figure 3

Figure 3 is a view of a gas dryer with a relatively low burn pattern in the vicinity

Figure 4

of the burner. Figure 4 is a view of one of the temperature limit switches that had parted from its mount. The small plunger that was a part of the limit

Figure 5

switch had fallen out (Figure 5), rendering the high temperature limit function ineffective. This resulted in overheating, causing the fire. This appears to be a manufacturing related deficiency. Figure 6 is a view of a gas dryer that caught fire

Figure 6

shortly after the owner heard a scraping sound. Figure 7 is a view of a failed support bearing which caused the drum to wobble and scrape the back of the dryer. This opened a gap between the drum and

Figure 7

dryer housing, resulting in clothes falling out of the drum onto the bottom of the dryer.

Figure 8

Figure 8 shows clothing tangled in the other support roller as shown by the right arrow and clothing contacting the burner shroud as shown by the left arrow. The bearing failure appears to be an end of useful life failure, as the dryer is old.

Figure 9

Figure 9 is a view of a large industrial dryer that sustained a fire in the vicinity of the air circulating impeller, which was constructed of a magnesium alloy (Figure 10). The magnesium alloy had

Figure 10

ignited and had to be extinguished with a chemical agent. The impeller was partially consumed, yet one interesting

Figure 11

piece remained. In Figure 11 we see a lump of magnesium with a bent paper clip lodged inside. Scrape marks on the paper clip suggest that it became engaged in the gap between the rotating impeller and shroud, causing frictional heating that ignited the magnesium.

Some explosive events occur in dryers, usually as a result of volatile cleaning chemicals entering the dryer or, in some instances, the ignition of butane cigarette lighters in the dryer drum.

Some items to consider when attempting to determine the cause of a dryer fire are statements from witnesses, the physical handling of the dryer after the fire, storage of the dryer and the fire origin. In some instances, the burned dryer may be a "victim" of the fire and not a cause. The typical burn pattern analysis and witness information will aid in this determination. Burned dryer components are often delicate and rough handling, as a result of moving, can cause a loss of evidence. Finally, analyze the dryer as soon as possible since environmental influences such as corrosion and animal infestation can cause evidence deterioration.

 


 


 



Electric blankets are a useful product and have been well accepted nationally, by consumers. Electric blanket manufacturers and Underwriters Laboratories have worked together to improve product safety. Despite these efforts fires do occur in relatively few of the blankets in use. Typical failure modes that cause electric blanket fires are often difficult to analyze due to the inherent destruction of the blanket once a fire has started. Nevertheless, current thought suggests the following causes of electric blanket fires:

1.      Thermostat malfunction

2.      Control unit malfunction

3.      Wiring faults

4.      Product abuse (improper laundering, leaving the blanket heating in a bunched condition, smoking)

All manufacturers provide some thermostatic mechanism in the electric blanket wiring as mandated by UL standards. Figure 1 is a photograph of a typical thermostat used in electric blankets. This

Figure 1

unit, which is one of several, interrupts the power to the blanket if overheating results when the blanket cannot dissipate the heat that is being generated.

When investigating a fire where a thermostat malfunction is suspected, it is useful to delicately recover the electric blanket and have the thermostats radiographed to ascertain their positions. Radiography is also helpful in analyzing malfunctions of control units. Figure 2 shows a typical control unit. Wiring

Figure 2

faults due to insulation failure can be detected by visual examination of the wiring. However a wiring fault can be a result or a cause, so interaction with the environment around the blanket must be studied carefully. Product abuse is an important contributor to blanket malfunctions. Improper laundering at high temperatures often violates manufacturer's instructions, causing thermostat failure. Check for fabric shrinkage or water inside the thermostat housings as evidence of having been damaged during laundering. Check out smoking habits of the blanket user. It is thought that leaving the electric blanket on in a bunched condition could cause a fire. Although unlikely (thermostats should protect the blanket from overheating) check the burn patterns on the blanket to determine if multiple burns are found at several positions on the blanket. Figure 3 shows multiple burn areas which suggest a bunching related fire. In this case the overheating occurred as a result of a damaged thermostat. Finally, your investigation of an electric blanket fire may result in a conclusion of 'undetermined' due to the all too often occurrence of severe damage that masks the evidence of fire causation.

Figure 3

 


 



Figure 1 is a view of a combination excavator loader that was part of a rental fleet. Typically these vehicles remain idle through the winter but with the onset of spring, rental demand returns. The vehicle was started and warmed up by the renter at a construction site when a fire developed in the engine compartment according to witnesses. Attempts to extinguish the fire were unsuccessful and a near total loss of the vehicle resulted. The vehicle was new with no service history problems and no recalls.

Figure 1

Figure 2

Burn patterns in Figure 2 tend to confirm a fire origin in the engine compartment. The burn was low with classic V-pattern. Opening the engine compartment cowling yielded the photo show in

Figure 3

Figure 3. The corroded turbocharger housing (left arrow) is quite evident having achieved normally operating temperatures in excess of 500F as a result of hot exhaust gasses that drive the unit.

Figure 4

The right arrow points to combustible debris on top of the engine near the turbocharger.  No electrical or mechanical deficiencies were found at the origin that could have caused a fire. Figure 4 is another view of the debris pattern on top of the engine. Figure 5 is a close-up of an eggshell in the middle of the debris which was mostly straw and twigs.  Skilled analysts would opine that a bird (North American Robin, Turdus Migratorius) had made a nest in the engine compartment. The spring had been late that year and cover for birds was at a premium since the trees had not yet

Figure 5

leafed out. There was easy access to the rain free engine compartment through the exhaust stack clearance hole shown in Figure 6. The massive steel engine

Figure 6

provides heat capacity and a stable temperature in the engine compartment.   After the birds parted, combustible nesting materials remained and were ignited by the hot turbocharger surface once the vehicle warmed up. Turbocharger surface temperatures are typically high enough to easily ignite typical nest materials.

Bird problems are nothing new to the claims industry. Bird nests in chimney flues which offer a warm environment have resulted in flue fires and heating equipment malfunctions as illustrated by the bird remains causing overheat damage to the boiler in Figure 7. Equipment that is not being used during the spring is a prime candidate for a bird nest site which can be constructed in days. Screen guarding of access openings is a typical preventive measure for bird entry.

Origin of fire: engine compartment, cause of fire: biologic intervention.

Figure 7

 


 

 



Figure 1 is a photograph of the remains of a building after an explosion has occurred. Explosions often result in severe damage to a building from the initial blast as well as the ensuing fire. Losses can reach several millions of dollars.

Figure 1

The adjuster may be one of the first of a long line of personnel involved in the analysis of the explosion and resulting subrogation that may last for several years. Quick action on collecting data concerning the explosion results in better documentation of the accident scene for future analysis. Here are some thoughts that may help in the collection of data.

Explosions often result in debris and damage patterns around the building that are easily identifiable. A debris pattern reflects the rest positions of objects that were hurled away from the explosion initiation area. A damage pattern is the configuration of the building that exploded as well as neighboring buildings that were damaged by the explosion. This information should be photographed immediately from the ground and from the air if the patterns are widespread. Explosive debris and minor damage to other buildings is often cleaned away or repaired quickly, resulting in a loss of valuable information to the explosion analyst. The explosion analyst uses this information in an attempt to identify the initiation point of the explosion - the point at which the explosion started. After an assessment of the seriousness of the loss, explosion analysts may be required. If so, get the analysts and experts in right away while the scene is relatively undisturbed. The experts will usually try to identify the initiation and ignition points along with the location of explosive fuel. The ignition point is the location where explosive fuel was ignited and may differ from the initiation point. In vapor cloud explosions, the initiation and ignition points are often different locations. Photographic documentation of the interior of the building is a must. The building should be secured so that the scene is undisturbed. Analysts should be brought in to identify the ignition, initiation and fuel source locations. This may mean stationing a full time guard at the scene to prevent loss of valuable evidence. Having established the initiation, ignition and fuel source locations, the next area of inquiry is why the explosion occurred. Was the release of an explosive mixture the result of carelessness of the insured or was this a result of a component or system defect. The identification of a component defect that caused the explosion significantly affects subrogation interests. Analysis and inspection of building systems including electrical gear, process systems and environmental systems should be performed to isolate the probable cause. After determining the probable cause of the explosion, artifacts that were removed should be identified and documented so that the chain of evidence is preserved. Since the building is likely to be torn down and rebuilt, it is necessary to remove certain critical pieces of hardware to the evidence storage locker.

The following sums up some ideas on how the adjuster can effectively influence the analysis of an explosion. Photograph debris and damage patterns, secure the scene, call in the experts immediately if required, and document the removal of evidence for possible future litigation.


 


Explosions Caused by Leaks in Flexible Gas Connectors

 

 


When analyzing the cause of a fuel gas related fire or explosion, examination of flexible gas connections in the building is advisable. Leaks in the flexible gas connector can be caused by improper installation, excessive bending from movement of the appliance, defective design or manufacture and environmentally assisted cracking. Improper installation such as cross threading of connectors and insufficient torquing of the connection can cause leakage. Inexperienced installers are likely to be at the root of poor installations. Excessive bending of the gas line can occur by forcing the appliance against an immobile object, such as a wall, causing crimping of the tube. It does not take too many cycles of excessive bending to cause partial or total failure of the tube. Older homes and installations are candidates for this type of failure since remodeling and other rearrangements of appliances over the years have taken their toll.

Figure 1

A typical manufacturing or design defect is insufficient soldering of the tube assembly to the end ferrule. Explosions have occurred from leaks caused when the solder connection failed allowing a large crack near the end of the connector. Utility companies often compile lists of flexible connectors with poor performance histories.

Environmentally assisted cracking (stress corrosion cracking) occurs in certain types of brass used in flexible connectors. Figure 1 is a view of a typical flexible gas connector. Figure 2 is a view of a crack in a flexible connector caused

Figure 2

by environmentally assisted cracking. The crack is in a stressed section of the tube. Chemical analysis of the brass showed it to be Cartridge Brass (nominally 70% copper and 30% zinc). Brasses containing more than 20% zinc are highly susceptible to stress corrosion cracking. Figure 3 is a photomicrograph of a cross-section of the tube showing the grain structure of the brass and severe intergranular corrosion in the tube, characteristic of stress corrosion cracking. The corrosion is aggravated by household cleaning chemicals such as

Figure 3

ammonia. The cracking can occur with very little movement of the tube and is often a subtle failure.

These are a few examples of deficiencies that can occur in flexible brass connector installations which can result in substantial property or casualty loss. Clues to causation of the loss may lie in the detailed examination of flexible brass connectors attached to gas appliances.


 



Periodically, an explosion loss occurs that is related to the flexible gas connector used to transport fuel gas to various appliances. A unique failure of the brass flexible connector has been observed

Figure 1

involving the solder joint near the flare/threaded nut end. Figure 1 is a cross-sectional view of an old flexible connector design where the corrugated flexible brass tubing is soldered to the end flare. The end flare is the mating surface with the appliance fitting and is secured by the threaded nut. Apparently, over time, the solder joint fails and a crack forms at the solder/brass interface. Newer connectors do not have this reliability problem.

Figure 2

Figure 2 is a cross-section of the more modern connector with a flare that is formed as an integral part of the connector. The absence of the solder joint enhances flexible connector reliability.

Figure 3

Figure 4

Figures 3 and 4 show two different solder joint connectors that failed causing an accumulation of natural gas in a home and an explosion. The characteristic solder fracture surface is apparent in each of the connectors. The flexible connectors showed little sign of excessive bending or strain. In each case the connectors were over 30 years old.

Figure 5

Figure 5 is a photograph of a modern flexible connector with a protective

Figure 6

coating. Figure 6 is an end view of the integral flare. Figure 7 is a view of the manufacturer’s identification band. The apparent cause of the failures is an aging effect, characteristic of solders. Solders are multi-phase lead/tin alloys, which tend to precipitate over time, causing a reduction in fatigue strength. As a result of cyclic loading brought on by periodic forces such as thermal expansion and contraction of the solder joint, metal fatigue initiates a crack that eventually propagates through the solder joint causing the gas leak and possible fire or explosion.

Figure 7

When analyzing an explosion or fire, which is a result of a gas leak in a flexible connector, determine if the manufacturing method involves a solder joint. If a fractured solder joint is found, then metallurgical analyses may be required to verify the failure mode. The age related failure of the old solder joint technology is tantamount to a "time bomb" in that it can occur without warning and at virtually any time.