Click for next page ( 24


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 23
2 A PRIMER ON FIRE AND FIRE HAZARD * Combustion-product toxicology is becoming a technical specialty in its own right, but its regulatory utility is limited to the degree to which it aids the regulator in measuring and controlling the overall fire hazard of a material. Thus, it is important to develop understanding of the dynamics of fire to the point where the role of smoke and its toxic effects can be placed in perspective. That is the purpose of this chapter. (Two other sources that offer a more thorough treatment are the report of the Products Research Committeei82 and Drysdale.69) THE BURNING PROCESS The fuel for most unwanted fires is organic material, e.g., the wooden frame of a house, an item of furniture, or gas leaking from a heater. Except in smoldering fires, the combustion reaction itself occurs in the vapor phase, where fuel vapor and oxygen (O2) in the air can mix. The reaction is rapid, usually taking a few hundredths of a second. The speed of burning and hence the intensity of the fire are usually governed by the rate at which fuel vapor and air enter the flame, where temperatures are high enough to initiate their reaction. Most accidental fires involve such "diffusion" flames, as opposed to "premixed" flames. *Portions of this chapter have appeared in modified form in Clarke (copyright, 1986)5 and are published here with permission. 23

OCR for page 23
24 q" (external) .,. \ q" (external FIGURE 2-1 Schematic of burning fuel surface. vaporization The important transport processes in a diffusion flame are shown in Figure 2-1. Fuel vapor is produced when heat from the flame radiates back to the fuel surface. The hot vapor rises, mixes with air entering near the base of the flame, and ignites. This buoyant expansion in turn creates turbulence, which causes more air to be entrained. A sizable fraction of the heat produced by combustion appears as radiant energy, some of which is absorbed by the fuel surface beneath, so the evolution of fuel vapors continues. Adjacent surfaces are also heated until they are hot enough to evolve combustible amounts of vapors; this is how the flame spreads. Diffusion flames rarely produce totally oxidized products, such as carbon dioxide (CO2) and water in the case of hydro- carbons; fuel that is burned incompletely gives rise to visible smoke, as well as carbon monoxide (CO). The amount of visible smoke produced varies somewhat with the availability of 02; but the tendency to produce smoke also varies widely among materials.

OCR for page 23
25 If the fuel is a volatile liquid, very little energy is required to vaporize it. But the fuels in most accidental fires are polymeric, solid materials, and these generally must be thermally decomposed (pyrolyzed) to yield combustible vapors; because chemical bonds must be broken, this takes much more energy than does vola- tilization of flammable liquids. The surface of a burning solid usually either melts or chars. If it melts, the surface will remain well below the flame temperature, because much of the heat that the pool absorbs from the flame is carried off in the vola- tilization process. If the fuel bed chars, heat energy might have to penetrate into the interior of the sample to generate fuel vapors, and the surface will be cor- respondingly hotter, but still below flame temperature. Char-forming fuels often pyrolyze in two stages: most of the readily volatile material is driven off, and then the char left behind decomposes. A TYPICAL COMPARTMENT FIRE The following discussion is restricted to fires in compartments, e.g., an enclosed space. Indoor fires are by far the most important of these, with respect to safety. In addition, compartments in buildings, ships, and planes catch and hold heat and combustion products; this increases a fire's severity, both physically and in its impact on those exposed. To be life-threatening, a compartment fire must be of at least some minimal size. It usually will have begun small (e.g., with a dropped cigarette, a match in a wastebasket, or a frayed electric connection), but later spread to involve a major fuel source, such as an item of furniture. Such a fire will quickly exhaust the avail- able O2 in a normal room, and air for further burning will have to be supplied through a doorway or window. The hot combustion products rise from the fire, entraining additional air and forming a distinct, hot, smoky upper layer just below the ceiling, which will deepen as the fire continues to burn. When the hot layer extends down to the top of a doorway, open window, or other vent, smoke will begin to spill out of the room, some of it into the rest of the building. Doorways and windows provide both the air needed for continued combustion and a path for combustion products. Relatively cold air

OCR for page 23
26 flows in through the bottom part of the vent and hot fire gases flow out through the top part. Assuming that available fuel is sufficient to consume all the available air, a steady state will eventually be reached in which the burning rate is limited by the rate at which new air is supplied. For a fire whose air is supplied by a normal doorway, 80 in. (203 cm) high by 36 in. (91 cm) wide, the maximal fire intensity is about 2-3 MW.i 16 The availability of air influences the products of combustion, as well as the intensity of a fire. When a fire is relatively small, and excess air is available, relatively little CO is formed. As the fire grows, it becomes more difficult for air to reach all parts of the flame while the vaporizing fuel and partially oxidized products are still hot enough for further reaction. As the fire approaches its maximal size, O2 depletion becomes pronounced, the fraction of CO in the smoke increases appreciably, and complex pyrolysis products are likely to appear (in particular, products that would be oxidized further if more O2 were available). For this reason, the toxicity of smoke from a fire usually depends on the intensity of the fire and certainly on the avail- ability of air. A small fire might produce mostly CO2 and water vapor and little else; smoke from the same material burning near flashover conditions (see below) can contain large quantities of CO and unoxidized pyrolysis products. The amount of ventilation, not the size of the compartment, controls the fire's eventual rate of energy output. Compartment size does, however, influence the rate at which the fire grows and the likelihood that it will spread beyond the compartment. As the upper part of a room becomes filled with very hot combustion products, this hot layer, like the flame itself, radiates energy to the fuel bed. The extra radiant heat makes the fuel burn faster than it would otherwise. Combustible items some distance from the original fire are also exposed to the radiation from the hot layer, so they will be ignited sooner than they would otherwise. This phenomenon constitutes a major threat to anyone still in the room; one need not be close to the original fire source to be severely burned by radiation from the hot layer. Figure 2-2 is a schematic of a room fire, showing the development of the hot upper layer and the flow of hot and cold gases through a vent. Much of the heat

OCR for page 23
27 cold air radiation from hot layer Target 1 J ~ ~ A, . :. I) flame /:: or radiation Fire r- Plume } \ :[ fuel ~ bed L ~ FIGURE 2-2 Two-zone schematic of fire burning in enclosure; doorway at left. generated is carried out the doorway by the hot smoke. The rest goes either into the ceiling and the portions of the walls in contact with the hot layer or into the room and its contents by radiation (from both the flame and the hot layer). As the room's surfaces become hotter, they themselves begin to radiate heat back into the room. The net result is that all combustible materials in the room are heated. If their ignition temperatures are reached before the initial fuel supply is exhausted or the fire is extinguished, burning will no longer be confined to one item; the whole room will become involved in flames. This phenomenon, called flashover, is the typical result of an unchecked fire in a residence or a commercial occupancy that contains an abundance of combustible materials. At flashover, more combustible fuel vaDor is being Produced than can be consumed by the ~ ~ _ _ _ _ ~ ~ : An ~~~ ~ he ~~~w=~~ air coming in, so not vapors are carrlea cur Alla UWLW= where they burn as they encounter more air. Combustible materials in adjacent spaces can then be ignited by flames emerging from the original fire compartment. Even where such additional combustible material is not available, the production of heat will increase dramatically, because additional air is available. Obviously, a flashed-over

OCR for page 23
28 room is difficult for firefighters to approach, so there is often little opportunity to apply water to the burning fuel in the room of origin. The consequences of any large room fire are poten- tially serious. The temperature of the hot gases coming out of the room as the fire approaches flashover typi- cally exceeds 700C; fuel is consumed at rates of around 0.5 kg/s; CO content of the smoke might be 5%--high enough for a few breaths to be disabling or lethal. Such a fire produces hot gases at several cubic meters per second, so an entire floor of a building can be filled with smoke within a few minutes. In such a situation, the magnitude of the hazard is dominated by the size of the fire. No matter what materials are burning, the threat is acute; no big fires are safe. FIRE HAZARD ASSESSMENT DEFINITIONS: RISK AND HAZARD Risk and hazard are defined in various ways. AS used in this document, fire hazard is the potential for exposure to a fire or its products. Thus, the relative hazard posed by two materials is the relative potential for exposure they offer. Fire risk is the probability that a given fire outcome will occur. A discussion of fire risk, like that of any class of risk, needs to include both the likelihood and the severity of the event. 5 7 ~ 9 S The Committee does not concur with the American Society for Testing and Materials, which blurs the terms "hazard" and "risk" and defines "fire hazard" as a fire risk greater than acceptable." 3 It is impossible to discuss the fire hazard associated with a product without knowing the circumstances in which it is used and the fire conditions to which it will be exposed. These circumstances together constitute the fire "scenario"s 2 for which hazard is to be assessed. The simplest scenario would be a fire involving a small quantity of combustible material in an essentially closed compartment and no fire spread to neighboring objects. Suppose an occupant is sleeping soundly in the compartment. In a fire, toxic smoke accumulates and mixes roughly uniformly with the air in the compartment.

OCR for page 23
29 The toxic smoke approaches a maximal concentration as the combustible material is consumed. The occupant, who is presumed to be unable to escape, might or might not be able to survive the exposure. Assessing smoke hazard in this scenario involves, for most purposes, only elementary calculations; required inputs are the mass of combustible material involved, the volume of the room, and the lethal concentration of the combustion products, assuming a substantial exposure time. The simple refinement of considering the ventilation rate through the compartment would provide a dilution factor. as well as a limit to the duration of exposure. Given valid smoke toxicity data, such calculations are simple--no computer is needed. This simple fire scenario represents an extremely common type of fatal fire. In fact, most residential- sized rooms contain many times the amount of combustible material that, if burned, would produce a lethal amount of smoke. In other words, to avoid death in most poten- tiallY lethal fires, it is necessary either to suppress the fire or to escape from it. The remaining categories of fire scenarios deal with the possibility of escape. (The detailed ramifications of fire suppression are not considered in this report.) QUANTIFYING HAZARD Fire behavior is a time-dependent process, but, even after a fire itself has reached a steady state, the concentrations of smoke in most of the building will continue to change. Hence, it is natural to use time as a basis for evaluating the relative hazard of different fires and of a given fire in different locations. This is generally done by identifying some temperature or smoke concentration that is unacceptable for safety and determining how long the fire in question takes to reach those points. 2 0 6 - Figure 2-3 shows a generalized fire growth curve, where the ordinate is a measure of the intensity, or size, of the fire. The figure could represent the upper room temperature in a room in which an item of fuel was ignited with a match. Little energy is generated at the outset, but eventually the fire becomes large enough to begin heating the room. The average room temperature, which reflects the size of the fire, begins to increase

OCR for page 23
30 / / ~ - - - - Fire Size - T. Ime FIGURE 2-3 Typical growth of room fire. rapidly. This corresponds to the steep middle portion of the curve. Finally, the fire will reach its maximal size. This limit will be reached either because the entire surface of the item is involved or because air cannot enter the doorway any faster. In either case, the temperature will approach a limiting value, governed by the relative size of the fire and the rate at which hot gases escape from the doorway. If Figure 2-3 were drawn for a longer period, the temperature would eventually decrease as the item of fuel burned itself out. Figure 2-4 shows fire growth curves from different fuel packages--say, two sofas. As the temperature increases, it reaches the upper limit of possible escape, shown here arbitrarily as 100C. The sofa represented by curve 2 produces this temperature at time t2, and the slower- burning sofa represented by curve 1 somewhat later, at t1. The difference between these two times is a measure of the relative hazard posed by these two sofas in this particular room environment and for this ignition scenario. Choosing a critical temperature much higher than 100C (or much lower) would change the difference between t1 and t2. This illustrates that the perceived performance of materials can depend heavily on the chosen criterion of hazard.

OCR for page 23
31 4J At Q E a a Q Q / t2 t1 . Time _ 100 c FIGURE 2-4 Comparative growth of two fires burning in room, showing time taken in each case to reach 100C. The situation is different when one wishes to evaluate hazards associated with the smoke. Figure 2-5 shows the same fire growth curve as Figure 2-3, with a dotted curve added to show smoke production. Smoke production con- tinues to increase very steeply even after fire tempera- ture has reached a constant value. In effect, the amount of smoke produced is proportional to the integral of the fire-size curve with respect to time. Unlike thermal hazard, the amount of smoke that represents untenable conditions is different for each material. The toxicity of the smoke must be measured by some appropriate method, and the "toxic" concentration for a given material must be determined. Dose can generally be related to concentration and time, so it is possible in principle to identify the point on the smoke concentration-time curve that corresponds to the arrival of unacceptably toxic conditions. Then the two materials can be compared as they were above (see Figure 2-6). Smoke also interferes with visibility. It is pos- sible to estimate relative smokiness of materials by measuring the light attenuation produced by the smoke from a known mass of sample--the so-called mass optical density. In a manner analogous to determining the onset

OCR for page 23
32 o . _ o c: o o a' as Q E Temperature Smoke Concentration _ _ - Time FIGURE 2-5 Growth of temperature and smoke concentration as functions of time. of toxic conditions, one can then estimate when, in a developing fire, the smoke will be dense enough to block sight-directed escape." 89 In reality, however, smoke might effectively impede vision at concentrations below those at which it blocks light transmission, if it irri- tates the eyes; no adequate biologic model is available for assessing such properties. The time available for escape or rescue begins when the fire is detected. For those in the room of origin of the fire, this detection might occur as soon as the fire starts. In many cases, however, exposure does not begin until later, e.g., when the hot upper layer descends far enough from the ceiling to be breathed. Smoke is usually the first sign of fire detected, either by those exposed or by a smoke detector. The properties of the smoke, and hence of whatever material is producing it, influence how readily it can be detected. Obviously, the sooner the fire is detected, the greater the fraction of the available time that can be used for escape. Three kinds of information must be available for time available for escape (TAE) to be determined:

OCR for page 23
33 o . _ Cal 4 - a) c' Toxic Dose (2) o <~ Toxic Dose (1 o in / / ~~/ / / 71 t1 Time t2 FIGURE 2-6 Comparison of smoke production and development of toxic smoke dose for two different materials. . toxicity. The unacceptable temperature or amount of smoke The temperature, amount of smoke, or time at which the fire is detected. . The fire or smoke growth curve. The first is derived (directly or indirectly) from exposure of a test animal to hostile conditions of heat or smoke. The second depends on the scenario. The third can be obtained either from full-scale burn experiments or, in many cases, analytically from small-scale data, laboratory data, and knowledge of the fire scenario. Calculations of fire growth (fire modeling) are discussed in more detail in Chapter 3. Computation of TAE permits one to compare the relative hazard posed by products in the same application; a ranking of TAEs is a ranking of hazard for the scenario under consideration. TAE alone, however, does not

OCR for page 23
34 determine whether a difference in hazard is significant and does not determine the degree of ~safety" of a given product. For example, if product A offers 2 min of available escape time more than product B. but product B provides 50 min of escape time when 10 are required, then the difference will be relatively unimportant. If product A offers 5 min and product B 3 min. but escape requires 10 min. then neither product is acceptable. TIME NEEDED FOR ESCAPE Determining whether a fire situation is survivable requires knowing the time needed for escape (TNE), as well as TAE. The escape margin, TAE - TNE, is a measure of safety. 6 7 A negative escape margin is inherently bad; the larger the margin, the better, although, as discussed above, it is possible to reach a point of diminished significance. Although the effect of smoke on visibility (discussed in the next section) plays a role, TNE depends almost entirely on the nature of the structure, the capabilities of those exposed to the fire, and human behavior. In short, it is not a function of a material's fire or smoke properties, so it is of only indirect interest to this study, although it is critical to the proper use of the results of fire hazard assessment. Figure 2-7 illustrates the conceptual framework of fire hazard assessment modeling. The five components labeled N1-N4 and As' permit computation of the time needed (N) for escape; the seven labeled A1-A7 permit computation of the time available (A). N1, occupant location and condition, includes such information as how far occupants are from exits or refuge areas, whether they can escape unaided, and their expected ages. N2, the decision/behavior model, permits prediction of how the occupants will behave in a fire emergency once alerted to the fire; how and when they are alerted depends on the protection system, N3. The evacuation model, N4, predicts how long the building population will take to reach safety in a given layout (A5'). The output of N4 is thus TNE. The "exposure response evaluation" compares TNE with TAE, which is computed from the "A" components (discussed in detail in the next section) by determining when fire

OCR for page 23
- - - ll l l i l 1 1 1 Iz 1 ~ 35 -1 o ._ A Q E E Woo l . l l ,. l , I< . _ ~ .~ ' E ~ ._ l __j Cal A ~_: 1 ~ . l li i l 1 l _J _ Z ~ ~ ~ o ~ . ~ A 4, ._ O ~ ~ ~ 1 ~ . C ~ o A= o ~ I I in o . ~ ~ ~ ~ At, - ~1 | m~ ~ ~ .0 A o l -- <1: O ~~ Z=~ i C ~ 1 ~ lot ~ j O at rt~ O it i i. ~ .~., 1 ~ o~= 0 =5 ~ . ~ O n.~ .- cn ~ lL 11_~___. N a ,] o tn o t) o ~, ~: 1 3

OCR for page 23
36 conditions become untenable (A7). The escape margin might differ for occupants in different locations, and the overall effect of a given fire is determined by computing its impact on occupants of specific locations. It seems likely that the behavior of those exposed to a fire, and hence the time required to vacate a fire zone, would be influenced by the presence of smoke. Even rela- tively little smoke is enough to impair visibility, as discussed below, so one can usually expect that visibility will be partly or fully blocked before lethal temperatures or toxic conditions are reached. However, information on the sublethal effects of smoke is not generally available in a form that permits it to be used in detailed predic- tions of occupant behavior (N2) and the corresponding evacuation times (N4). It seems reasonable to assume that evacuation will be slowed if visibility is impaired or effectively blocked and that the TNE should be increased over what it would be in clear air. The simplest way to do this is to multiply the estimated TNE by some appropriate "safety factor." Visibility is usually blocked early in a fire, so the most conservative assumption is that the safety factor must always be applied; hence, the detailed smoke properties of a given set of burning fuels do not have a strong influence on TNE. In other words, it is assumed that evacuation will always be accompanied by poor visibility. T IME AVAILABLE FOR ESCAPE As discussed above, the sublethal effects of smoke from various materials are not understood in sufficient detail for their influence on TNE to be predicted with confidence, and it has been suggested that those effects and reduced visibility be approximated by applying an appropriate multiplier, a 'safety factor," to a computed TNE. If it is assumed that TNE is a constant for a given scenario, usually independent of the burning material, and that, in contrast, the fire and smoke properties of a given material influence TAE, then, of a series of materials postulated to be burning in a given scenario, that with the largest TAE offers the greatest opportunity of escape. Hence, TAE becomes a surrogate for the relative degree of hazard that a material offers in the scenario under study.

OCR for page 23
37 Without a knowledge of THE, it is not possible to say whether a material provides acceptable fire performance; therefore, in using TAE as a measure of relative hazard, one should not infer that it represents a degree of safety. However, computing TAE does constitute a method for assessing the overall fire hazard of a material or product in a given scenario and comparing it with that of others intended for the same use. The details of fire hazard depend on the scenario. For example, a rapidly developing flaming fire whose products accumulate in a relatively confined space, such as a small apartment, produces a well-defined hot gas layer that descends rapidly. There is little opportunity for the smoke to lose energy to the walls and ceiling, so the layer is very hot. Once the hot layer has descended to a few feet above the floor, it is difficult to escape without coming into direct contact with it. Regardless of its chemical composition, this layer poses an immediate threat because of its temperature. In a larger space, or if the fire is burning more slbwly or perhaps smoldering, the layer is cooler. In fact, a distinct upper layer might not be apparent, because high temperature is what gives the layer its buoyancy and results in stratifica- tion. In such a case, the toxic properties of the smoke become important, because they, not the temperature, determine the tenability of the compartment. In larger structures, it is common practice to provide barriers to the free passage of smoke or fire between floors, into exitways, and often between groups of rooms. (Modern apartment buildings, for example, have fire walls between apartments, but not as interior partitions.) The smoke from a fire might leak into spaces far from the fire, but rarely as a hot buoyant stream, the form in which it exists when it is nearer the fire. Therefore, the primary threat away from the fire is usually smoke toxicity, not heat. One uses scenarios involving multiple burning items when one wishes to examine the hazard attributable to an item that can burn only after exposure to a fairly large ignition source. Examples of single-item and multiple- item scenarios are described below.

OCR for page 23
38 BURNING OF A SINGLE ITEM Consider a fire ignited in a compartment of 250 m3 (which corresponds roughly to 1,000 ft2 of floor area and a normal 8-ft ceiling) and restricted to one rela- tively large item, such as a heavy upholstered chair or loveseat. Because the fire is restricted to an item of furniture, the fire properties of the rest of the room are unimportant, although the thickness and thermal properties of the walls and ceiling should be known, if one is to determine how much of the heat energy of the fire is lost to these surfaces. As for the furniture itself, its burning rate (heat release and mass loss rate) must be measured or calculated from small-scale test results. These data constitute the input to As in Figure 2-7. Because the scenario envisions monitoring conditions in the same room as the fire, the layout of the building (A5') is not very important in this cal- culation. Neither are construction material properties (A4), other than those already mentioned for A2. Assume that an air temperature of 100C is the upper limit for human escape from the compartment. Finally, suppose that smoke toxicity data (A6) available on the furniture material fairly reflect the toxicity under actual burning conditions. The most useful measurement is the L(Ct)50, the concentration-time product required for death to occur in 50% of animals exposed to the smoke. In a smoke toxicity test, this product is obtained by continuously monitoring the smoke concentration to which the animals are exposed and reporting the time integral of this quantity when the animals die. (This takes no cognizance of the possibility that animals die after exposure.) The tenability limits for temperature and toxicity are deter- mined in A6 and constitute A7. The simplest burning scenario is one in which a moder- ate fire begins on the furniture and does not spread appreciably. If the fire size is 100 kW--i.e., about 0.6 m in diameter--it can be shown that the hot smoke will have filled the room to a depth of 1 m from the floor in about 6 mini the temperature of the hot layer will have reached 100C after 11 min.sS Hence, by the temperature criterion mentioned above, the environment will have become lethal in 11 min.

OCR for page 23
39 Whether smoke toxicity becomes a problem sooner depends on whether the occupants have been exposed to the smoke throughout the course of the fire and on how toxic the smoke is. For smoke toxicity to be the immediate threat in this scenario, the atmosphere must become lethally toxic before it becomes lethally hot. When the burning rate of the fire and the associated mass loss rate are known, it is simple to compute the average smoke concentration in the hot layer. Assuming that the occupants have been exposed to the smoke from the time when the hot layer was at the 1-m level, the time to receive a lethal dose of smoke, TAE, is given by the integral over time, aft: MAE L(ct)50 = / CS(t)3t, tl where L(Ct)50 is the lethal dose determined from a laboratory toxicity measurement, t1 is the time at which the smoke reaches the 1-m predetermined level, and Cs(t) is the smoke concentration, expressed as a function of time. The smoke concentration, Cs(t), is a function of the mass loss rate, m(t): rt Cs (t)V J m(t) At (1) (2) TAE is plotted as a function of L(Ct)50 for this scenario in Figure 2-8. If the furnishing material has a smoke L(Ct)50 below about 200 g~min/m3, death from smoke toxicity could be expected to occur before con- ditions were thermally untenable; otherwise, in this scenario, the thermal hazard is more immediate. It is also instructive to compute the smoke density, and hence the visibility, in the upper layer when it has descended to 1 m above the floor. Depending on the mass optical density of the fuel, visibility in the upper layer after 6 min of burning will be no more than about 2 m, and more typically about 0.8 m. If visibility is

OCR for page 23
40 400 lo 300 200 . a, In o ~ 100 o In o . _ _ / / / / 0 2 4 6 8 10 12 14 16 Time Available for Escape (TAE), min FIGURE 2-8 Buildup of lethal smoke dose with time single-item scenario. ., restricted to 0.8 m, a person cannot see any farther than an outstretched arm, and even 2 m of visibility is likely to be of little real help, in that typical room dimensions are at least twice as large. Studies by Jinxes in Japan have shown that, when smoke optical densities are above 0.25 m~l (i.e., when visibility is less than about 5 m), movement by those exposed is slowed appreciably. For a fire to produce so little smoke in the foregoing scenario, it would have to have a mass optical density of 20 m2/kg, which is about one-tenth the smoke-producing potential of a typical furnishing material, such as polyurethane. If, instead of a flaming fire, the fire on the furni- ture is smoldering, too little heat will probably be generated to maintain a stable upper layer, and the smoke will disperse generally uniformly through the compartment volume. TAE (including time needed to detect the fire) is given by: ^TAE L(Ct) 50 = ~ Cs (t) aft. o (3)

OCR for page 23
41 For a smoldering fire that does not change in constant), Equations 2 and 3 can be solved to yield: TAE = [2~(Ct) 50 (V/m)] 2 ~ ze (m is (4) For one that grows linearly with time (m = kt), the same equations yield: TAE = [6L (Ct) 50 (V/k) ] 3 ( 5 ) The higher order of time dependence on mass loss rate, the less sensitive TAE is to changes in L(Ct)50. This is the case for flaming fires, as well as smoldering fires, although the former are complicated by the hot- layer descent discussed above. Most fire scenarios are too complex for calculations of TAE to be expressed in analytic form; computer-based numerical solutions are usually required. The logic of the procedure, however, is the same. The material properties that control burning rate and fire growth are at least as important as are smoke toxicity characteris- - ~ . ~ tics In determining --the rare properties alone control the hazards associated with thermal exposure and the rate at which smoke is produced. BURNING OF MULTIPLE ITEMS Most real fires involve several items. In some cases, the sequence of ignitions seems idiosyncratic--it is as easy to envision drapes igniting from a burning chair as the reverse. In other cases, however, the sequence is likely to be predictable--combustible materials, such as plastic pipe or wiring, behind a wall are much more likely to be exposed to heat from a fire in the room than to be ignited directly by a small ignition source. Figure 2-9 shows the buildup of temperature in a room as the result of a known fire, the standard time-tempera- ture curve for the ASTM E119-83 fire endurance test.) 2

OCR for page 23
42 800 700 800 500 o o ~ 400 it' 300 200 100 o ~1 19 exposure _ / 7 Air in shaft with 5~" gypsum wallboard 1 1 1 1 0 5 10 15 20 . . [ It mln # FIGURE 2-9 Time-temperature profile of fire simulating ASTM E119 fire endurance test and of cavity behind gypsum wallboard.

OCR for page 23
43 It also shows the buildup of temperature behind a wall of Run-in. gypsum wallboard. These two curves are inputs Al and A2, respectively, for determining the smoke production rate of the room fire and the concealed combustible materials. It should be clear that the temperature in the room becomes untenable long before the region behind the wall warms appreciably. At some distance from the fire, the sole hazard of concern is the toxicity of the smoke. For the first 15 min or so, smoke issues only from the room fire. When the temperature behind the wall is high enough, the combustible materials will begin to decompose then to ignite and burn. The data necessary to character- ize this process include ignition temperature, mass loss vs. temperature, heat release rate, and amount of hidden fuel. These data constitute part of A2. The curves in Figure 2-10 show the contribution of the in-room and behind-wall fuel packages to the total smoke produced. One must know the toxicity of the smoke from both fuel sources to predict the effect of the behind- wall material on TAE. Formal expressions have been proposed by Bukowski42 and others to compute TAE associated with fires involving multiple components. TAE can be computed from the equation: rTAE 1 rTAE 1 ~ ( ) L(Ct)so(2) Jo ) (6) 1 L(ct)50(l) Jo where (1) and (2) refer to the smoke generated from the in-room and behind-wall fuels, respectively. In practice, the smoke production curves are followed, integrated, and normalized with respect to the toxic dose (determined in advance by small-scale tests). When the normalized contributors sum to unity, TAE is deemed to have been reached. The smoke concentration depends both on the mass loss rate of the fuel and on the point in the building at which the hazard is being assessed. Hence, knowledge of the building layout (A5') and the construc- tion material properties (A4) might also be needed for this computation.

OCR for page 23
44 HI Smoke 7 = o . e Room FUJI 8 II 0 5 10 15 20 Tim~,m~ FIGURE 2-10 Smoke production for fire burning in room and igniting combustible materials behind wall. In comparing two alternative materials for the same use behind the wall, it is possible to compute a difference in TAE associated with the change from one material to another. How this is accomplished is the subject of Chapter 6.