Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
CHAPTE R 6 SMOKE AND TOX ICITY 6. 1 1 ntroduction The toxicity problems encountered in a mine or bunker fi re tend to ch ange as the fi re progresses. At the time of ignition, the atmosphere is oxygen-rich and the toxicity is simi l ar to that of the co mbustion products of the m aterials involved; l ater, the atmosphere is oxygen-lean and the tox icity is similar to th at of pyrolysis of the same materials. Toxi city arises from the presence of com bustion products in a given location. The major factors that i nfl uence human survival du ring a fi re in a confi ned s pace are : 1 . Thermal ene rgy - The hazard to hu man health is either that of local therm al damage, usually of the skin or respi ratory ai rway, or genera l i zed thermal shock. 2. Decreased atmospheric oxygen concentrati on - Th is factor is associated with a l l materi als undergoing ox i dation at a rapid rate . The health hazard is th at of deprivation of an essential amou nt of avai l a ble respi ratory oxygen. In fi res occu rri ng in mines or bu nkers, decreased oxygen m ay rapi dly oecome a signif· icant factor. 3. Carbon monoxi de and other nox ious gases and aeroso ls - Like other organic materi als, any pol ymer undergoing less than com plete com bustion i n the presence of oxygen yields carbon monox i de. The tox icol ogical consequences of h u m an ex posure to carbon monoxide h ave been extensively rev iewed ( N a tional Advisory Center on Toxicol ogy 1 973, Stewart 1 976). Carbon mon oxide is the major single cause of fatal ities in most fire situations and is especially significa nt in the oxygen-lean atmosphere that devel ops in mine and bu n ker fi res. De pendi n g on the types of organic materials present, the py rolysis or combustion of pol ymers may produce, in addition to carbon monox i de, com bi nations of tox icants incl u ding hydrogen cyan ide, sulfur com pounds, nitrogen oxides, hal ogen acid gases, vapori zed fu mes of oth er hal i des such as ethyl bromide, organic gases such as ethyl brom ide, al deh ydes l i ke acrolein, and such exotic co mpou nds as the h igh ly tox ic bicycl ic phos phates from some plasticizers ( Petajan et a l . 1 975). 4. Smoke - Massive exposu re to smoke may result i n di rect mechanical plugging of ai rways or smoke particles may carry adsorbed gases or l i q u i ds or resi dual heat to various pa rts of the respiratory tract and produce injury by direct contact. Smoke also obscu res vision, impedes escape, and h i n ders fi refighti ng efforts. ( See Appendix C). 121
M I NES A N D B U N K E R S 5. Fear, pan ic, or i ncidental trauma - These factors are recogn i zed causes of i ncapacitation and death in fi res. 6. Pre-ex isti ng or con cu rrent psychophysi ological disease or impai rment - Pre· existing heart disease has been established as a significant factor in fire deaths. Acute alcohol i ntoxication also may be a factor, and the disabled obvi ously are at a great disadvantage. These are on ly a few of the more apparent diseases and impai rments affecting an i n divi du a l 's abi l ity to escape and su r vive. M i n es and bun kers a re examples of confi ned spaces in wh ich fire represe nts a serious h aza rd. They may be remote from outside firefighti ng assistance and avenues of egress m ay be l i mited and difficult. The materials i n mi nes and bun kers that are most l i kely to undergo com bustion or py rol ysis are natural and synthetic polymers ( for the purpose of th i s chapter, fue ls, l u bricants, and hy drau lic fl u ids are excluded ) i ncluding wood, wool, cotton, paper, other cel l u losic materi als, and a wide range of synthetic pol ymeric m ateria ls. Polymeric materials cu rrently i n use i n m i nes and bu n kers are descri bed i n Chapter 4. There general ly are no speci fic requi rements regu lating nox i ous gas generation for m ateri als used in m i n es and bu nkers under fi re conditi ons. 6.2 Perspective on Experi mental Data The n u m ber of stu dies concerned with nox i ous gases and smoke resu lting from thermal degradation of polymeric m ateri a l s has increased during the past 1 0 years. Pyrolysis or com bustion produ cts of the poly mers used in m i nes and bu nkers h ave bee n found to include car bon monox i de ( CO ) , carbon dioxide ( C02 I. hydrogen cyan ide ( H C N ) , oxi des of nitrogen ( NO x l ⢠ammon ia ( N H 3 ) , hydrogen su lfide ( H2 S) , phosgene ( COCI 2 I. and many other com pounds. The physiol ogical hazards of these gases are descri bed in Appendix D and Kimmerle provi des a more detai led compi l ati on of effects at various atmospheric concentrati ons ( 1 974 ). F rom fires i n confi ned spaces, the predom inant tox ic therm al degradati on prod· uct is CO. I ncapaci tati ng or lethal amou nts of CO can devel op with i n m i n utes. Appendix E presents a brief review of CO toxicity . The net physi ological response from CO and other thermal degradation gases is fa r from clear although awareness of th is pro bl em and the difficu lties in assessing it are increasi ng. F i re si tuations can have an extre mely co mp lex toxicology . I f a lethal CO at· mosphere is not reached, other lethal or disa bl ing factors sti l l may be present. For example, i n the series of experi ments reported by Corn ish and Abar ( 1 969) pu l· mon ary injury fro m HCI developed i n the a bsence of leth al effects from CO. A more subtle effect noted by Effen berger ( 1 972) is th at bu rn ing po l ysty re ne does n ot cause rats to die or devel op sign ificant amou nts of carboxyhemoglobin, but rath er the styrene it yields appa rentl y has an immobi l izing effect on rats. If this i nte rpretation may be extended to fi res i nvolv i ng hu mans, death cou l d resu lt be· cau se a bi l ity to escape from the fi re is i m paired. 1 22
SMO K E A N D TOX I CITY Smo ke presents a nu mber of hazards and occu rs in a wide variety of com posi ti ons. Smoke basical ly is a mixture of carbon parti cles and materials evolved from com busti on. It may conta i n i rritants adsorbed on the particles and be m ixed with thermal decomposition gases. The hazards of smoke m ay be physical ( blocking vision or airways) , phys iol ogical ( l ocal or systemic chem ical i rritation, tox icity, or heat i nju ry ), or psychologica l (fear and pan ic). Appendix C su rveys the hazards from smoke and describes cu rrent measurement techni ques. 6. 2. 1 Experi mental Work U l tra Systems, I nc., under a U. S. Bure au of M i nes contract, is investigati ng 3 2 materials com m on l y carried i nto m i nes or bunkers b y workers or occu pants . The behavior of these materials under control led pyro l ysis conditions is being stu died and the m ateri als are ass igned rati ngs based upon various factors, incl u d i n g tox icity. 6.2.2 Clinical Data Based on Ai rcraft Fires Qu antitative tox icol ogical data on mine and bun ker fi re victi ms are nonex istent; however, l i m ited data are ava i l able on ai rcraft fi re victims and are perti nent in that the synthetic polymers involved are s i m i l ar. ( See Volume 6, Aircraft : Civil & M i l i tary ). Sm ith and Associates have descri bed the resu l ts of the i r fore nsic investiga tions of several ai rcraft acci dents as fol l ows : Two com mercial ai rcraft acci dents i n the Un ited States during the 1 960's ( Denver, Col orado, 1 96 1 , and Salt Lake City, Utah, 1 965) con tri buted greatly to the i n iti ation of the present concern over the tox ic hazard of the gases generated i n ai rcraft fi res. These accidents we re of special significance because carefu l anal ysis indicated that few, if any, of the occu pants wou ld have suffered significant physical inju ry from the rel ativel y m i l d im pacts i nvolved; yet, a total of 60 persons perished as a resu l t of thermal and chem ical i n ju ries su sta i ned i n the ensuing fires. Carboxyhemogl obi n measu rements on 16 victi ms of the Denver crash revealed CO saturations rangi n g from 30 to 85 percent with a mean of 63. 3 percent. Si milar anal yses on 36 victi ms of the Salt Lake City accident yielded CO saturations ra ngi ng fro m 1 3 to 82 percent, the mea n bei ng 36. 9. The l ower carboxyhemoglobin val ues found in the second accident have been attri buted to the fact that fi re was present with i n the a i rcraft before evacuation cou l d be atte mpted and that the survival ti me of many victi ms must have bee n shortened by di rect ther mal effects. It also has been assumed that gases other th an carbon monox ide must have contri buted to the tox icity of the cabin envi ron ment, but there is no supporti ng evidence for the assumption. I n 1 970, bl ood sa mples from victi ms of an ai rcraft crash followed by fire ( Anchorage, Alas ka, Novem ber 1 970) were analyzed fo r the pres- 1 23
M I N ES A N D B U N K ERS ence of cyan i de (apparently the fi rst ti me such anal yses were m ade on ai rcraft fi re victi ms) . Measurable amounts of cyanide were fou nd in 1 8 of the 1 9 speci mens submitted, and were accompan ied by carbon mon oxide saturations rangi ng from 1 7 to 70 percent. I n the one sample in which cyanide cou l d not be detected, the carboxyhemogl obin concen tration, 4.9 percent, did not exceed that wh ich cou l d resu lt from smoki ng, indicating the pro ba bi l ity of death on i mpact. Bl ood cyanide l evels in these victi ms corresponded cl osely with those reported in the l i te ratu re for victi ms of structu ral and vehicu lar fires rangi ng from the lower detection limit (ci rca 0.01 J,Lg/ml ) up to 2.26 J,Lg/ml . The rel ation· s h i p between cyan i de levels and carboxyhemoglobin content varied in random fash ion, perhaps representi n g rel ative prox i mity of the vi cti ms to cyan ide-producing materials. Alternatively the vary i ng cyani de levels reported may be due to u ncontrol led auto-production of cyanide i n and from the tissues. Nothing i n these fi ndi ngs permitted specu l ation concerning the rel ative contri bu tion of the two gases to lethal ity. In add ition, there was no way of assessing the possi bl e contri bution of other gases that must have bee n present i n the pyrolysis mixture to which these victi ms were ex posed. 6.3 Evaluating the Hazards of Toxic Fumes and Smoke 6.3.1 Comparison of Materials Valid com parisons of different mate rials must be based on s i m i l a r or reasonably standard conditions. The desirabi l ity for a standardized test procedure pertinent to proposed use appl i cation is obv ious. However, real fi res possess two essenti a l l y u ncontrol lable variables - oxygen supply a n d tem peratu re - that m ake selection of such test procedu res i nherentl y difficu lt. Genera l l y , laboratory thermal degradati o n tests i n a n oxygen- lean atmosphere are descri bed a s py rolysis tests whereas com bu s tion with actua l flame indicates an oxygen-ri ch atm osphere . Si nce either pyrolysis or com busti on can be the more hazardous depending on the n atu re of the material bei ng consumed, a standard procedure shou l d take both catego ries into accou n t, e ither separately or together. ( See Appendix E for gu ide l i nes for suggested screen ing tests. ) In the case of fi res in mines and bu nkers, as previously noted, the atmos phere tends to ra pi dly become oxygen-lean and h igher carbon monox i de leve l s wou ld be ex pected earl ier than usual . 6.3.2 Thermal Decomposition Temperatures F i re te mperatu re depends on the pyrolysis and/or com bu stion processes and al so on the ca loric value of the product or products consu med by fire . For the therm al decom position of wood, seve ral anal ysts h ave fol l owed a class ification of fou r distinct tem peratu re zones (see Appendix F ) . H oweve r, the committee i s u naware of any attem pt at such a cl ass ification for the multitude of synthetic polymers th at 1 24
SMO K E A N D TOX I CITY exist today. 6.3.3 Method of Study Recent literature descri bes the fo u r ty pes of methods discussed below (see Ap pendix G ) . 6.3.3.1 Analysis Testing to identify the chemical components involved can help in understanding the effect of altering variables such as temperature and oxygen . The rel ative hazard or lethal ity of the produ ct can be esti mated with reasonable confi dence if a single com ponent, such as CO or H CI , is clearly predom inant and no other significant sou rce of stress is present. If thermal de�adation generates a significant quantity of m iscel l aneous gases, heat or smoke , the net physiologica l response is difficu lt to esti mate; however, anal ysis of such m i xtu res or the i r degradation products is be· com i ng less difficu lt with the devel opment of more soph isticated (and expensive) analytica l tools. To com pound problems, fi nished products m ay be composed of basic elements, antioxi dants, fil lers , additives, and finishes, and even major per· centage components ofte n are not stated. 6.3.3.2 Biological Testing D ata obta ined from tests on l a boratory animals can be expressed as the LC7b5 0 or concentration expected to produce death in 50 percent of the ex posed animals and the EC5 0 or conce ntrati on ex pected to produce a specified effect, such as i ncapacitation, in 50 percent of the exposed ani m als. It is in th is latter area of biol ogical test for toxicity that special attention is requ i red. A safer product from the standpoint of a flammabi l ity test does not al ways result in a more desirable product i n that : 1 . Structure modification or additives may resu lt n ot on ly in retarded com bus tion of the pol ymer but also i n increased tox icity of decom position gases and/or dense smoke from the smoldering of the polymer when it is ex posed to external heat. 2. Many of the more common fire retardants contain halogens, such as ch lorine and brom i ne, that theoretical l y make poss i bl e the production of therm al decompos ition products such as hydrogen ch l oride ( HCI ) , phosgene ( COCI 2 ), and hydrogen brom ide ( H Br). 3. The presence of nitrogen atoms, either in the pol ym er or in the additives, ⢠i ntroduces the possibil ity of thermal degradation to HCN or NOx. 4. Pol y mers based on propoxyl ated tri methy lopropane polyols and fi re-retarded with phos phorous-coating retardants m ay yield h igh ly toxic bicycl ic phos· phorous esters whe n therma l l y degraded ( Petajan et al. 1 975). 6.3.3.3 Extensions of Analysis and Biological Testing 1 25
M I N ES A N D B U N K E RS Present models incl u de co mbined testi ng, predictive testi ng, and "room " or l arge-scale fi re tests. 6.3. 3.4 Epidemiological Studies Critical epidemiol ogical anal yses h ave been applied to fire tox icity only recently. The re is a need for evaluated reports on mi ne and bun ker fires that provide com pre hensive casu alty data, i ncluding a qu antitative tox icological and pathologica l eva l u a tion of the victims. To consider only one decom position product, a recent d is cuss ion of HCN as a major lethal factor i n aircraft fi res indicates the cu rrent interest and concem in this area and also bri ngs atte ntion to the difficu lties that may be associated with such anal yses. 6.4 Special Considerations Obviously, products or formulations requ i re more stringent evaluation when i ntended for use i n m i nes and bu n kers than for many othe r appl i cations. The long and tedious escape routes from u nderground spaces and the rel ative unavai labil ity of firefighti ng equ i pment and person nel inherently expose the occupants to a great er fi re and toxi city hazard than ex ists in more ordin ary occupancies. If any new products or formu l ations are intended for use i n m i nes and bu nkers in appreciable quantity, their potential toxicity when bumed must be evaluated with i n experi mental gu ide l i nes and reviewed in the context of ava i lable epide m i ological data. 6.5 Conclusions and Recommendations Conclusion: Pu bl ished data re l ating to the tox icity of th e com bustion and py rol ysis products of polymeric materi als, indiv i du al l y and i n com bin ation with other pol y mers, are sparse. Recommendation: A central agency shou l d be establ ished to col lect and anal yze data from and promul gate information regarding the tox icity of the combustion and pyrolysis of pol ymeric materials in test and actual fi res. Conclusion: Most tox icity stu dies relating to the products of com bustion a n d pyrolysis of polymeric materi als are oriented toward establ ishing lethal levels and incapacitation is rel ative ly ignored Recommendation: Stu dies of the tox icity of pol y meric mate rial combustion and pyrolysis products shou l d address incapacita tion as we l l as lethal ity. Conclusion: Little is known a bou t the additive or synergistic effects of the tox ic agents produced from the com bustion and pyrolysis of polymeric m aterials. Rec ommendation: Methods shou l d be esta bl ished for studying the tox icity not onl y o f i ndividual toxic agents release d du ring the com bustion an d pyro lysis of pol y meric materi als but also the toxicity of combi nations of such tox icants l i kely to be encountered in varying fire situations. Conclusion: Cl i n ical tox icity data on the com bustion and pyro l ysis products o f pol ymers in actual fi res i s very sparse. Recommendation: G u idel i nes shou l d be establ ished and pu bl ished for the col lection of tox icity data on f i re victi ms by 1 26
SMO K E A N D TOX I CITY paramedical rescue personnel and medical centers. These gu i de l i nes shou ld specify data col lecti on techniques that wil l not interfere with fi rst aid or intermediate or defi nitive medical care. 6.6 References H. H. Corn ish and E. L. Abar, "To xicity of Pyrolysis Products of Vinyl Plastics," ArchiiHI$ of Environmental Heslth 1 9 : 1 - 1 5, 1 969. E. Effen berger, "Toxische Wirku ngen der Verbren nugsprodu kte 110n K unstoffen," Stadtehygiene 1 2; 275-80, 1 972. G. Kimmerle, "Aspects and Methodology for the Evaluation of Toxicological Parameters Dur ing Fire E xposure," Journal of Fire and Flammability/Combustion Toxicology 1 :4-5 1 , 1 974. National Advisory Center on Toxico logy . National Academy of Sciences-National Research Counci l , Committee on Toxicology . Guidt18 for Short-term Exposurt18 of the Public to Air Pollutants, Vol . V I , Guide for Carbon Monoxide, Washington, D.C., 1 973. J. H. Petajan, et al., "Extreme Toxicity from Combusti on Products of a Fire-Retarded Poly u rethane Foam," Science, 1 87 74244, 1 975. R. D. Stewart, "The Effect of Carbon Monoxide on Humans," Journal of Occupational Medi cine 1 8 : 304-309, 1 976. 1 27
