Introduction
Examination of the physical and chemical evidence left at a fire scene may, provided that there has not been too much disturbance, provide evidence which may enable the cause of fire to be recognized, the point of ignition to be identified and the types of fuels first ignited to be established. Directional effects may be clearly evident and the causes for particular smoke staining patterns may often be deduced. The melting or discoloration of materials within the building will provide some evidence of the temperatures attained at particular locations within the fire. To some extent, fire scene patterns can assist in the recognition of the types of fire involved. Differences between slow burning fires and rapidly developing fires can normally be recognized and these may often be further categorized.
However, there is a considerable body of unsubstantiated mythology associated with the examination of fire scene patterns and some investigators make assertions about the behavior of charred wood, cracking glass and soot, which simple experimentation would have clearly shown to have been incorrect. For example, rules of thumb based on the depth of charring of wood should be treated with the deepest of suspicion. Some investigators still believe that all wood chars at a constant rate of 1/40 in min-1 (0.06 cm min-1). Simple tests demonstrate the intuitively obvious fact that the rate of charring depends on the amount of radiant heat acting on the piece of wood.
Preliminary Examination
It is a common experience for fire investigators, when first viewing a severely damaged building, to see the situation as hopeless and unlikely to yield any significant evidence. However, very considerable bodies of evidence may be recovered after even the most severe of fires.
The investigator will immediately upon arrival at the scene begin to form opinions as to the proportion of the building damaged and will be starting to try to assess which parts of the fire damage occurred during the early stages of the fire and which occurred later. Photographs taken during the fire-fighting by the brigade and by on-lookers may help in this assessment, although they are unlikely to be available during the initial stages of the investigation. Clearly, it is normal for fires to spread from low parts of the structure upwards through the building, and if there is any significant wind, it too is likely to have played its part in influencing the direction of spread. There is likely to be more severe damage on the downwind side of the seat of the fire, even to the extent that sometimes greater heat damage may be found on the downwind side of an item rather than on the side nearest the seat of the fire.
The fire-fighting, too, is likely to drive the flames away from the initial points of attack, whilst suppressing burning near to the regions of application of water. Positive pressure ventilation, a technique increasingly used by fire brigades, is also likely to result in unusual heat and smoke patterns. This technique involves the use of powerful blowers, which force fresh air into the building, displacing smoke. Some increased burning at high levels in regions remote from the seat of the fire may occur.
High-level investigations
At some stage in the investigation, it will be necessary to decide how much importance to attribute to evidence at a high level in the fire-damaged building. Although it is unlikely that the main point of ignition will be at a high level, in cases of deliberate ignition, there may be separate ignition points at different levels in the building. In addition, it is possible for fires to spread downwards, normally by the fall of burning material from a point of ignition at a higher level.
In general, however, the most easily investigated seat of fire is likely to be at a low-level within the burnt part of the building and it is possible that any evidence that had originally been present at a high level, will have been destroyed by the flames. Point of entry evidence is also unlikely to be evident at high levels in the building. It is, however, desirable for the investigator to have a pair of binoculars available, so that evidence visible in inaccessible parts of the building may be examined safely. If the building is unsafe to enter, some examination can still be carried out from a hydraulic platform using binoculars and a camera with a telephoto lens. It is also possible that the highest parts of the building were the places where victims had been trapped by the fire, and it may be necessary to recover bodies or to examine the places where they had been found.
Exterior walls
It is normally desirable to make an exterior examination of the building and of the smoke-staining which is often apparent on the walls above windows and doorways. This staining may provide evidence of forcible entry and of windows and doors which were the first to fail during the fire. The smoke-staining to the walls may indicate the direction of the wind prevailing at the time of the fire, although it should be borne in mind that the wind direction could have changed during the period that the building burned. Normal experience of bonfires and barbecues demonstrates that the direction of smoke travel is likely to change during relatively short periods of time.
Point of entry
In many deliberately ignited fires, the premises will have been forcibly entered by an intruder, and such evidence may be clearly visible to fire-fighters and subsequent investigators. If the intruder ignited the fire from within the building, it is probable that he would have started the fire at some point remote from his point of entry, in order to escape safely as the fire developed. For this reason, it is not uncommon to find significant evidence at points of forcible entry even after the fire and the fire-fighting, provided that reasonable care has been taken by the fire brigade to preserve this valuable evidence.
Work has been carried out on techniques of restoring fingerprints left on glass and other smooth materials, subsequently affected by smoke, heat and water. In addition there may be footprints on window sills, tables and other articles of furniture near to points of entry. Other evidential materials, such as glass, blood and fibers, may also have survived the fire and fire-fighting at the point of entry. In many cases of arson by intruders, far greater implicative evidence is likely to be found at the point of entry than at the point of ignition. It should also be borne in mind that many of the more experienced criminals tend to ensure a point of exit for themselves and this region is even more likely to be free from fire spread damage.
Location of Point of Ignition
Fire scene patterns are among the methods which may be used to locate the point or points of ignition. In general, these techniques depend on the assumption that where the fire has been ignited, it is likely to burn for the longest time, producing the greatest amount of damage. Although in simple fires this may be a valid assumption, in complex fires involving difficult fire-fighting, unequal distribution of fuels or variable ventilation throughout the building, then fire scene patterns can produce misleading evidence for fire seat location and more reliable and impartial techniques should be used. Pattern-based techniques which have been commonly used include the following.
Seeking the region of maximum damage within the building Although this technique may be fairly reliable in the case of small fires, in large fires the maximum damage can be very much influenced by ventilation, the available fuel, and the fire-fighting techniques used.
Depth of charring to wood The rate of charring of wood will depend on the type of wood involved and the amount of radiant heat flux to which it has been subjected. The greater the heat flux the more rapid the rate of charring. This effect can be used to establish where the greatest heat has been involved, by comparing charred wood in different parts of the building. However, in all fires there is likely to be a considerable thermal gradient, with higher temperatures being attained at high levels within rooms. For this reason, the most meaningful comparison will normally be made at a low level where differences in temperature are more significant. The depth of charring does not provide any absolute measurement of the time for which the timber in question had been subjected to heat.
Spalling of plaster When plaster is subjected to heat it may spall from the underlying brickwork of the building (Fig. 1), providing some evidence which can be used in fire seat location. However, the sudden cooling effect due to fire-fighting water jets may also spall plaster from brickwork, although in this case there is unlikely to be any smoke deposition on the surface of the bricks.
Distortion to glass and plastic Many plastics soften at temperatures of 130-400°C. The manner of their distortion can give indications as to the direction and magnitude of the heat acting on the original item.
Figure 1 Plaster spalled from a wall as a result of cooling by fire fighting jets. The direction of the spray can be seen from displaced soot on the ceiling.
Glass tends to soften at 750-900°C. Above 900°C it is sufficiently mobile to fall to a lower, cooler position.
Distortion of metalwork Structural steelwork in buildings distorts in severe fires and the degree of distortion may indicate the regions where the greatest heat had been released.
Melted metal Commonly encountered metals melt at temperatures between 326°C for lead and over 1000° C for metals such as copper and steel. Evidence of significant quantities of melted metal may indicate that very high temperatures had been achieved in certain parts of the building.
Damage to concrete floors Concrete spalls at elevated temperatures, and high temperatures at a low level within the building may be indicated by spalling to concrete floors. Unfortunately, because of the variation in quality of concrete likely to be encountered in many buildings, confusing damage may occur. Contrary to general belief, spalling to a concrete floor does not necessarily indicate that a flammable liquid had been burnt on the surface and in fact during the burning of a flammable liquid the wetted surface tends to remain cool until all of the liquid has evaporated.
Residual heat The region where the greatest burning had occurred is likely to have been heated more than any other part of the building. It has been argued that, after the fire has been extinguished, the regions where the greatest residual heat remains may represent the seats of the fire. One complicating factor is that during fire-fighting water applied to suppress the fire will not necessarily cool the whole building uniformly and so the residual hot areas may simply represent areas where less fire-fighting water was deployed.
Burning of soot Although soot deposits in different ways (to be discussed later) one indication which has been used for fire seat location purposes, is evidence that soot has been burnt from surfaces after first being deposited. This indicates an elevated temperature and an oxidizing atmosphere and may be of significance in the enquiry.
Low burning Since fires spread predominantly upwards, the lowest region of burning might reasonably be supposed to be the seat of the fire. However, burning materials can drop down and start other fires and, in addition, a well-developed fire may possibly burn through a floor, causing damage at a lower level.
Holes in the floor When there is a single hole in the floor of a building, this may well be the seat of the fire. However, there are other possible explanations, for example it might be a region of particular fuel loading, a region of particularly good ventilation, a region where the floor had been weaker than normal or a region last to be extinguished by the fire brigade. If liquid fire accelerants have been used, they may leak through joints between floorboards causing characteristic burning patterns. However, burning beneath and between floorboards can be produced by normal fire spread. This damage cannot be regarded as conclusive evidence of the use of fire accelerants unless supported by analysis information.
Funnel patterns A seat of fire may produce a characteristic funnel pattern either in smoke or heat damage in the room or building involved (Fig. 2). However, funnel patterns are not necessarily categorical evidence of seats of fire because regions of particular fuel loading or regions where burning materials have dropped can also produce similar patterns.
Characteristic structural collapse Buildings may collapse in a manner characteristic of a fire having been started in a particular region. Interpretation of such evidence should be carried out with considerable caution because widespread travel of hot air and smoke may produce conflicting effects, as also may unequal distribution of forces within the building.
Thermal direction indicators If the furniture and contents of the building have been largely left in position, then charring and heat damage to particular faces of the furniture may provide an indication of the direction of the seat of the fire. It is important that the original positions of items should be known categorically for this type of evidence to be used.
Figure 2 A characteristic funnel pattern indicating an armchair as the point of ignition. The fire had, in fact, been ignited approximately 500 mm from the chair, but had spread to it at an early stage in the fire.
Ceiling damage Particularly intense damage to one region of the ceiling may indicate a seat of fire beneath it. However, it must be recognized that a fire at a low level in a room will cause widespread and general heating to a ceiling and the damage at a high level is less likely to be diagnostic than intense damage at a low level.
When the fire-spread pattern information and the other more objective techniques of ignition point location have been employed, it is likely that the investigator will be able to define the probable seat of fire to within certain limits. It should be possible to draw a line defining the outer limits for the possible position of the point of ignition, the confidence perimeter or radius of error. Excavation within the area defined by the confidence perimeter must include the seat of the fire. It is clear, therefore, that the whole of this area must be excavated during the investigation if any physical evidence of the cause of the fire is to be found. In general terms the more rapidly the fire has developed, the larger the confidence perimeter, because a rapidly developing fire tends to produce relatively uniform damage.
If a fire started slowly, for example by smoldering, then very localized damage may be found. It is, therefore, possible by examination and comparison of the damage to form some opinion as to how rapidly the fire has developed. In a fire which has started by smoldering and burnt slowly until it was extinguished, characteristic localized deep charring may be found. The smoke-staining, characteristic of a slow-burning fire, may show evidence of nonbuoyant pyrolysis products causing staining on the horizontal upper surfaces of objects in the vicinity.
A rapidly developing fire, on the other hand, can cause relatively uniform heat damage, with smoke which is likely to be buoyant and probably more carbonaceous than that from a smoldering fire (Fig. 3). A complication may arise because a fire which started in the smoldering form may develop into a free-burning flaming fire at some stage, possibly when the ventilation is increased, causing confusing evidence. In the most extreme case, the transition from smoldering to a free-burning fire may take the form of a ‘back draught’. This phenomenon occurs when flammable gaseous pyrolysis products accumulate within the building under conditions of limited ventilation. The opening of a door or window may allow the flammable gases to ignite explosively.
The reverse of this process may occur if a large fast-burning fire is imperfectly extinguished and an area of smoldering remains unnoticed by the firefighters. Under these circumstances, a smoldering fire can occur causing intense burning in one region. Such regions of burning are likely to be in inaccessible places, in the floor boards, at the base of the wall, at the sides of cupboards and under or behind furniture. Such regions of damage, which are known as ‘bulls’ eyes’, are relatively easy to recognize.
Figure 3 Typical smoke-logging in a passageway. Hot, buoyant smoke produces staining at a high level.
Information from Smoke Staining
The appearance of deposited smoke at fire scenes can provide valuable evidence, giving information relating to the position of items, their temperature, the behavior of the fire and the direction of smoke travel. Although the composition of smoke can vary considerably, the buoyant black smoke produced in a free-burning fire tends to deposit preferentially on cold items rather than hot, and rough surfaces rather than smooth and it deposits better when flowing in a turbulent manner, rather than laminar.
The tendency for smoke to deposit on cool surfaces is often clearly apparent, resulting in effects of selective deposition related to the underlying materials beneath the surface in question. Electrical wiring, structural steelwork and even heads of nails may act as preferential areas for smoke deposition under certain circumstances (Fig. 4). Surfaces coated with soot may suffer a rise in temperature during the development of the fire and particularly if this rise in temperature is associated with increased availability of oxygen; then areas of previously deposited soot may be burnt away producing a clean area in the middle of a blackened region.
Deposits caused by smoke from smoldering fires tend to be different in appearance, consisting of sticky, tarry pyrolysates, which tend to deposit very much more evenly, although there is a pronounced tendency for the smoke to fall, lacking as it does the buoyancy of the carbonaceous smoke from a free-burning fire.
Recognition of Types of Fire
Fire scene patterns may provide evidence to indicate the type of fire involved. Fires accelerated by the use of flammable liquids may show characteristic pool burns, particularly if they are extinguished at a relatively early stage. A pool burn shows a hard-edged demarcation line, between the irregularly shaped liquid pool and the remaining unburnt floor covering. Because of the lowering of the surface tension of most commonly encountered flammable liquids when they are heated, liquid accelerant tends to be driven under the unburnt region of floor covering. After a destructive liquid fire, samples taken from this transition zone are more likely to contain recognizable accelerant residues than samples from other regions.
Although a liquid-accelerated fire may well be evidence of arson, it is possible for flammable liquids legitimately present on the premises to cause similar effects. In addition, certain synthetic carpeting mate-rials may burn in a manner which mimics the effect of a pool burn, even if no liquid accelerant had been present. For this reason, it is normally essential that samples should be analyzed to demonstrate the presence of a liquid fire accelerant
Figure 4 Smoke tends to deposit preferentially on cooler surfaces and may reveal positions of nails and joints normally hidden by the plaster..
Fires caused by smoking materials
Although smoking materials are commonly blamed for many fires, it is difficult to demonstrate that a fire has been caused in this way. Carelessly discarded smoking materials include not only discarded lighted cigarettes, but also discarded lighted matches and tapers. Although a discarded lighted cigarette can normally only initiate smoldering, a carelessly discarded lighted match or taper can start a fire which shows all the characteristics of a free-burning flaming fire from its inception. If a lighted cigarette is suspected of having caused the fire, very localized damage is likely to be found at the point of ignition. The position of this localized damage must have been within material which was capable of smoldering and which could have been ignited by a small ignition source. Bedding, soft furnishings and piles of rubbish are the materials most commonly ignited by cigarettes. Normally, smoldering can only occur in cellu-losic materials capable of forming a rigid char. Upholstery foams are not easily ignited in this way unless a cellulosic fabric is also involved. In addition, evidence that the occupants of the building had been careless with smoking materials is likely to be found, with cigarettes, spent matches or spent tapers scattered carelessly in the unburned parts of the building.
Recognition of spontaneous combustion
In most cases of spontaneous combustion, the development of heat is likely to have been slow, and the maximum temperatures are likely to be attained within the bulk of the material which has self-heated. Although there will evidently have been disruption of the material during fire-fighting, there may still be evidence that the interior of the pile of material has reached higher temperatures than the exterior. The heating will occur in a susceptible substance, such as hay, cotton, fishmeal, sawdust and coal. There is also likely to have been a background history of heating prior to the fire.
More reactive substances such as oil-soaked rags, car body filler, quick-drying paint and certain metals are likely to self-heat more rapidly and can do so in smaller quantities. With the more reactive substances, the evidence of internal heating may not be so clearly defined as in the less reactive substance, where a larger quantity must necessarily be present for ignition to occur.
Recognition of electrical fires
The fire scene pattern evidence for electrical fires is confined to the recognition that the seat of the fire is in, or associated with, an electrical appliance or system. For an electrical appliance to cause a fire it must necessarily either emit sparks or flames, or reach a temperature such that its exterior surface is hot enough to ignite materials in the vicinity. Under both of these circumstances it is to be suspected that the interior of the appliance will have been significantly hotter than its surroundings. This is normally apparent on examination, but the increasing use of beryllium in electronic components makes the examination of electronic devices which have achieved very high temperatures extremely dangerous, because of the possible risk of inhaling beryllium oxide dust. Such examinations should only be carried out under laboratory conditions, where stringent precautions can be taken against this serious danger.
Evidence of electrical arcing in a fire scene does not, in itself, provide evidence that the fire was electrical in origin, although it may be taken as evidence that the conductors in question were live at some time during or after the fire. However, an electrical appliance suspected of having caused the fire must firstly have been live and must secondly have been in the region where the fire started. For this reason, damage to the insulation of wires within or supplying the appliance is likely to result in electrical arcing, which may have caused the operation of a fuse or miniature circuit breaker. The significance of arcing within the appliance can therefore be assessed accordingly.
Recognition of fires caused by lightning
When a fire occurs in a building during an electrical storm, the suspicion may arise that it had been caused by a lightning strike. In most countries records are kept of individual lightning strikes and it is possible to establish from a meteorological department whether such a strike had occurred in a particular area, at a particular time. In addition, examination of the fire scene patterns may demonstrate damage to roofing tiles and electrical arcing effects to vertical wiring. The effects may be particularly characteristic in wires leading between telephone inputs and receivers and vertical wires to television aerials. In addition, plug sockets may have been disrupted explosively by the lightening strike and there may be particular damage to television sets. Such evidence may well survive after relatively destructive fires.
Recognition of arson
A high percentage of fires in many countries is attributed to deliberate ignition. Although it is possible to cause a serious fire by the application of a flame to a single artifact within a building, it is more common in cases of deliberate ignition for there to be several points of ignition and for liquid fire accelerants, such as petrol or paraffin to be used. The presence of extraneous liquid fire accelerants provides strong evidence that the fire has been ignited deliberately and their presence may be suspected if liquid pool burns are found or traces of unburned liquid are smelt (Fig. 5).
Multiple seats of fire must be shown to be independent of one another if found and in a large fire it may be difficult to establish whether apparently separated seats of fire were simply the result of differing firefighting effort or different amounts of fuel present in the regions of the building. If it can be conclusively demonstrated that there had been several separated seats of fire then, provided that accidental reasons for multiple seats can be excluded, then the fire can be concluded to have been deliberately ignited. Accidental reasons for multiple seats include secondary electrical faults, disturbance of burning material, interrupted critical processes, and abnormal fire spread.
Figure 5 A typical petrol vapor explosion, which caused almost complete demolition of the building, followed by a widespread and destructive fire.
