UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Exposure assessment of characteristic compounds at a fire & safety training center Hills, Dale 1998

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1998-0467.pdf [ 6.15MB ]
Metadata
JSON: 831-1.0088547.json
JSON-LD: 831-1.0088547-ld.json
RDF/XML (Pretty): 831-1.0088547-rdf.xml
RDF/JSON: 831-1.0088547-rdf.json
Turtle: 831-1.0088547-turtle.txt
N-Triples: 831-1.0088547-rdf-ntriples.txt
Original Record: 831-1.0088547-source.json
Full Text
831-1.0088547-fulltext.txt
Citation
831-1.0088547.ris

Full Text

Exposure Assessment of Characteristic Compounds at a Fire & Safety Training Center by Dale Hills B.Sc, The University of British Columbia, 1995 A THESIS SUBMITTED IN PARTIAL FULLFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Occupational Hygiene Programme) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA © Dale R. Hills 1998 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract Sampling for characteristic compounds emitted from fires at the Maple Ridge Fire and Safety Training Center was conducted to assess staff exposure to these compounds. Area samples were taken at outdoor burn sites to measure CO, CO2, S0 2, HCN, formaldehyde, particulate, and benzene, to determine if areas where staff technicians monitor emergency switches are free of exposure and to find if there are safe distances from these pits where harmful exposures may not occur. Area samples were taken inside the indoor bum structures to determine the time period to dilute these post-fire residual contaminants to levels where health effects would not be expected. All results were compared to the Workers' Compensation Board of BC exposure limits. Mean exposure at the propane pad fuel tank was below the detection limit, and at the large outdoor fuel and depth burn sites the mean exposure was below the exposure limits. At the extinguisher pad training site, mean levels of particulate and benzene exceeded the exposure limits (particulate 417 mg/m3, benzene 0.9 ppm). Mean exposure levels, following venting in the indoor burn sites, exceeded the exposure limit for CO (36 ppm), benzene (0.8 ppm), and formaldehyde (0.372 ppm), while C 0 2 (925 ppm), S02(1.5 ppm), particulate (0.2 mg/m3), and HCN (0.1 ppm) were less than the exposure limit. Formaldehyde (0.011 ppm) was detected in the burn building following the absence of any burning for up to one week, and was attributed to off-gassing of the chemical from the concrete walls. Control options to reduce exposure are proposed. They include: relocation of the extinguisher pad with future site expansion, altering work practices found to be predictors of exposure, and by avoiding entry into the ship mock-up for a minimum of 380 minutes, and the burn building for 140 minutes, following the last burning session. The results of the study suggest that for some burning sites, there are situations where exposure could occur, however, the risk of health effects may be slight if proper control of these exposures is exercised. Table of Contents Abstract 11 Table of Contents... 111 List of Tables vn List of Figures Acknowledgment. vni ix 1.0 Introduction 1 1.0.1 Objective 1 1.1 General overview of combustion and combustion byproducts 2 1.1.1 Previous study of byproducts of combustion 2 1.1.2 Concrete sink effect 4 1.2 Health effects of smoke exposure 5 1.2.1 Smoke inhalation 5 1.2.2 Regulations regarding smoke exposure 6 1.2.3 Controlling smoke exposure by use of SCBA 7 1.3 Studies involving fire fighting training centers 8 1.3.1 Prior study at the FSTC 9 1.4 Overview of the Fire and Safety Training Center 9 1.4.1 Site description 9 1.4.2 Detailed description of the sites with fire activity 12 1.4.2.1 Outdoor fuel and depth pits 12 1.4.2.1.1 T-pit 12 1.4.2.1.2 Round tank 12 1.4.2.1.3 Description of training activity at both the T-pit and round tank 12 1.4.2.1.4 Sampling rational at the outdoor pits 12 1.4.2.2 Extinguisher pad 15 1.4.2.2.1 Site description 15 1.4.2.2.2 Description of training activity 15 1.4.2.2.3 Sampling rational at the extinguisher pad.... 15 1.4.2.3 Propane pad 15 1.4.2.3.1 Site description 15 1.4.2.3.2 Description of training activity 16 1.4.2.3.3 Sampling rational at the propane pad 16 1.4.2.4. Indoor sites 16 1.4.2.4.1 Burn building 16 1.4.2.4.2 Ship mock-up 16 1.4.2.4.3 Description of training activity 17 1 .4.2 .4.4 Sampling rational 17 1.4.2.4.5 Propane side of the burn building_ 17 1.4.2.4.6 Sampling rational at the indoor burn sites 18 1.4.3 Fuel sources burned at the FSTC 18 1.4.3.1 Wood 19 1.4.3.2 Diesel 20 1.4.3.3 Gasoline 21 1.4.3.4 Propane 21 1 . 4 . 4 Potential byproducts of combustion at the FSTC 2 1 1.4.5 Use of SCBA at the FSTC and its effect on the study 22 1.4.6 Chemical combustion products sampled 23 2.0 Methods 2 5 2.1 Sample Collection 25 2.1.1 Observation of fire characteristics and weather conditions 25 2.1.2 The sampling stand 25 2.1.3 Outdoor burn sites: 26 2.1.3.1 Fuel and depth sites (T-pit and Round tank) sampling rational.. 26 2.1.3.2 Extinguisher pad sampling rational 26 2.1.3.3 Propane pad sampling rational 26 2.1.4 Indoor burn sites: Ship mock-up and Burn building (wood and propane sides) 27 2.1.4.1 Sampling Procedures 27 2.1.4.1.1 Background levels (environmental levels in the absence of burning) 27 2.1.4.1.2 Unvented Levels (worst case environmental levels after fires) 27 2.1.4.1.3 Environmental levels immediately following venting ... 28 2. /. 4.1.4 Assessment of time period for environmental levels (after venting) to dilute to insignificant levels 28 2.2 Sample collection and analysis 28 2.2.1 CO. CO7. SO? 28 2.2.2 HCN 29 2.2.3 Particulate 29 2.2.4 Formaldehyde 30 2.2.5 Benzene 30 2.3 Methods of data analysis 30 2.3.1 Frequency distributions 30 2.3.2 Censored Data 31 2.3.3 Data from direct reading instruments 31 2.3.5 Comparison of mean contaminant levels between sites 31 iv 3.0 Results 3 2 3.1 Outdoor Pits: sampling at Fuel and Depth fires (T-pit and Round Tank) 32 3.2 Outdoor Extinguisher Pad 33 3.3 Outdoor Propane Fires 34 3.4 Sampling Experiments at Indoor Sites: Burn building and ship mock-up 34 3.4.1 Background levels (environmental levels in the absence of burning) 34 3.4.2 Unvented Levels (worst case environmental levels after fires') 34 3.4.3 Environmental levels immediately following venting 35 3.4.4 Comparison of environmental levels immediately following venting and unvented levels (worst case environmental levels after fires) in both the ship mock-up and burn building 36 3.4.5 Assessment of time period for environmental levels (after venting) to dilute to insignificant levels 37 3.4.6 Comparison of Contaminant levels in the Burn building and Ship mock-up 43 3.4.7 Propane fires in the Burn Building 43 4.0 Discussion 4 4 4.1 T-pit and round tank 44 4.1.1 Micro-climate 4 1.2 CO and SO? exposure (examination of real time recordings) 47 4.1.3 HCN. formaldehyde, particulate and benzene exposure (examination of contaminants not measured as real time data) 47 4.2 Extinguisher Pad 49 4.2.1 Dry chemical clouds 50 4.3 Propane Pad 52 4.4 Indoor sites - Burn building and Ship mock-up 53 4.4.1 Background levels (environmental levels in the absence of burning) 53 4.4.2 Unvented Levels (worst case environmental levels after fires) 54 4.4.3 Environmental levels immediately following venting 56 4.4.3.1 Combined results (Burn building and Ship) 56 4.4.3.2 Specific observations - burn building 57 4.4.3.3 Specific observations - ship mock-up 61 4.4.4 Assessment of time period for environmental levels (after venting) to dilute to insignificant levels 63 4.4.4.1 Specific observations Burn building 64 4.4.4.2 Specific observations: Ship mock-up 64 4.5 Propane side of burn building 64 v 4.6 Off-gassing 65 4.7 Estimate of the risk of health effects to instructors at the facility 66 4.8 Generalizibility of the study results 68 4.9 Study limitations 69 5.0 Conclusions 7i 5.1 Outdoor fuel and depth pits (T-pit and round tank) 71 5.2 Extinguisher pad 71 5.3 Propane pad 71 5.4 Burn building and ship mock-up 71 Bibliography 73 Appendix 1A Description of byproducts of combustion measured and some general health effects 79 Appendix IB Occupational exposure limits for chemicals 83 Appendix II Individual measurements from sampling sessions 84 vi List of Tables Table 1: Byproducts of combustion, with highest published level of each contaminant detected and the percentage of structural fires at which they have been detected 4 Table 2: Byproducts of combustion and the common sources of materials that produce them 4 Table 3: General health effects of the combustion byproducts 6 Table 4: Concentration of emissions in diesel exhaust 20 Table 5: Comparison of diesel and gasoline exhaust emissions 21 Table 6: Summary of combustion gases and volatile materials at the T-pit and Round tank 32 Table 7: Summary of combustion gases and volatile materials at the extinguisher pad 33 Table 8: Summary of background contaminants at the burn building and ship mock-up 34 Table 9: Summary of unvented combustion gases and volatile materials measured in the burn building immediately fire suppression and before venting 35 Table 10: Summary of combustion gases and volatile materials measured immediately following venting at the Ship mock-up and Burn building 36 Table 1 1: Comparison of unvented levels and mean levels immediately following venting, along with the % decrease in contaminant levels, in the burn building and ship mock-up 36 Table 12: Summary of contaminants measured for 3 hours following venting in the ship mock-up and burn building 38 Table 13: Summary of benzene measured for 3 hours following venting in the ship-mock up 39 Table 14: Time period for contaminants following venting, in the burn building and ship mock-up, to diffuse below the action limit 41 Table 15: Comparison of mean contaminant levels in the ship and burn building immediately following venting 43 Table 16. Formaldehyde levels in the burn building propane side 43 Table 17: Comparison of contaminants measured at the T-pit and Round tank with WCB exposure limits 44 Table 18: Comparison of contaminants measured at the extinguisher pad with WCB exposure limits 50 Table 19: Comparison of environmental levels immediately following venting at the Burn building and Ship mock-up with WCB exposure limits 56 List of Figures Figure 1: Site map of the FSTC 11 Figure 2: Locations of areas sampled at outdoor fuel and depth pits and the propane pad 14 Figure 3: Concentration of mean contaminant level vs. time since active venting in the ship and bum building 40 Figure 4: Concentration of formaldehyde and HCN in the burn building following active venting 42 Acknowledgment T would like to thank Dr. Chris van Netten for all the support and guidance throughout the entire project, without which it would not have been possible. Special thanks to Drs. Mike Brauer and Paul Demers for their advice and instruction. I am grateful to Victor Leung for his guidance and assistance in performing the laboratory analysis. Kristi Mclntyre and my parents deserve gratitude for their moral support and the encouragement they provided to complete this project. The staff and management at the Fire and Safety Training Center deserve special recognition for their cooperation, assistance, participation and patience while the project was completed. Finally, I need to thank the National Engineering and Science Research Council of Canada and the Justice Institute of B.C. for partial funding of this project. I X Chapter 1: Introduction Firefighters are routinely exposed to conditions that would be hazardous to unprotected individuals. Their working conditions are uncontrollable and the environment is proven to be deadly. Routinely, their call of duty requires them to accept the risk of hazardous exposure to toxic fumes, which can be harmful to their health. To help minimize the perils of this occupation, professional training is provided. Centers such as the Fire and Safety Training Center (FSTC) at Maple Ridge British Columbia were created to provide this training. The FSTC employs trained staff who are required to operate and assist in a variety of activities related to training students in the techniques of fire suppression. The staff are unique in that they are routinely exposed to fire and its combustion byproducts, often many times a week. Research has shown that fires produce many potentially hazardous chemical agents (Morse et. al. 1992). As a result, it has been said that the safety of firefighters is hard to guarantee (Selala et. al. 1993) and thus the health and safety of the FSTC staff may also be hard to guarantee. 1.0.1 Objective This study was undertaken to characterize the concentration of emissions of C0 2, CO, S02, HCN, particulate, formaldehyde, and benzene released from fires at the Fire and Safety Training Center, and to provide insight into their potential to pose a risk to the health and safety of the staff. Specific aims of the study included: 1. To evaluate the concentration of residual contaminants inside the indoor burning sites following fire suppression, and to determine the time for these contaminants to be diluted to insignificant concentrations. 2. To determine if the chemical contaminant concentrations at areas around the outdoor fire sites, where instructors are required to monitor emergency controls, exceed insignificant levels. 1 3. To determine if there exists a safe distance from the outdoor sites (T-pit, round tank, extinguisher pad and propane pad) where contaminant concentrations do not exceed insignificant levels. 4. To compare the levels of measured contaminants to the British Columbia WCB Permissible Occupational Exposure Limits, as listed in the Occupational Health and Safety Regulations effective April 15, 1998. For this project the levels of contaminants that are considered to be insignificant and not to pose a risk to staff health and safety will be arbitrarily set at Vi the WCB 8-hour time weighted average exposure limit the exposure level where employers must implement an exposure control plan (the action limit). 1.1 General Overview of Combustion and Combustion Byproducts 1.1.1 Previous study of byproducts of combustion The primary byproduct of combustion is smoke, which is a mixture of fumes, toxic gases, inert gases, particulate, organic acids, and many other compounds. Over 2000 individual chemicals belonging to 100 classes of chemicals have been identified in wood smoke alone (Burning issues - author anonymous). Smoke, the main hazard of fire fighting, is dealt with in the majority of current published literature (Committee on fire toxicology 1986), and is probably the most highly studied. However, other than the number of contaminants in smoke, little is known. The generation rate, release, and amount of most chemicals are poorly characterized, and may be impossible to fully describe, due to the uncontrollable nature of fire. Attempts have been made to characterize the emissions released from a fire, but little progress has been made. Previous study by Lowrey et. al. (1985) has concluded that there appears to be little correlation between the gases produced and the fuel source burning. They also reported that the same combusting material could produce different gases when burning conditions vary. Characterizing fire emissions is complicated by the fact that their generation and release differ at each fire (Prien 1988). Factors such as the ratio of oxygen to fuel determine the completeness of combustion. This is important as more toxic agents are formed by incomplete combustion of the fuel source (Guidotti and 2 Clough 1992). It has been found that the smoke released from synthetic materials such as plastics, polyesters, etc. is different from that released from natural products such as wood, cotton, or wool (Morse et. al. 1992). To further complicate the issue, some of the chemicals generated by fires are chemically reactive and continue to form other chemicals long after the fire has been suppressed (Decker and Garica-Cantu 1986). One factor influencing the types of combustion byproducts produced at a fire is the temperature of combustion (Murrell, no date). Murrell found that temperatures between 400-700°C produce the greatest number of byproducts. He also reported that at temperatures in excess of 700°C, complex organic molecules break down, which reduces the amount of these chemicals in the smoke. It has been found that increased temperature results in more complete combustion, and usually produces fewer toxic compounds (Bayer 1974, Guidotti and Clough 1992). However, higher temperatures, in excess of 1340°C, results in nitrogen fixation (Guidotti and Clough 1992). It has been found that HCN as well as oxides of nitrogen (NOx's) are formed in greater quantities as the temperature of combustion increases (Bayer 1974) (Committee on fire toxicology 1986). A summary of some of the levels of contaminants previously recorded at fires has been included as table 1, and a summary of the most common materials that produce these contaminants is included as table 2. Neither table 1 or 2 is an exhaustive list. Many other chemicals have been detected at fires, usually only in trace quantities. Numerous other materials may also form each chemical, mentioned in table 2, but are not as well recognized. 3 Table 1: Byproducts of combustion, with highest published level of each contaminant detected and the percentage of structural fires at which they have been detected Material Previously measured fire concentrations Carbon monoxide (CO) > 400 ppm in 28.5% of measured fires, >1500 ppm in 10.5%, measured in excess of 15, 000 ppm (*1) Carbon dioxide (C02) often exceeds 20,000 ppm, measured in excess of 60,000 ppm (*1) Hydrogen cyanide (HCN) detected at 12% of fires, measured in excess of 40 ppm (*2) Hydrogen chloride (HC1) detected at 9% of fires, measured in excess of 200 ppm (*1) Nitrogen dioxide (N02) detected at most fires, measured in excess of 10 ppm (*5) Sulfur dioxide found in 50% of fires in low concentration (*1) Hydrogen fluoride not often found but measured in excess of 6 ppm (*4) Hydrogen sulfide detected at 15% of fires in low quantities (*2) Acrolein detected at 56% of fires often exceeding 0.2 ppm (*1) Benzene detected at 20% of fires, and found in excess of 150 ppm (*2) Formaldehyde found at all most all fires, 2.5% exceeded 3 ppm (*1) Particulate found at all fires, measured in excess of 560 mg/m3 (* 1) *1 -Treitman^. al. 1980 *2 - Lowrey et. al. 1985 *3 - Gold et. al. 1978 *4 - US dept. of Health 1974 *5 - Guidotti et. al. 1992 Table 2: Byproducts of combustion and the common sources of materials that produce them Material Source Carbon monoxide (CO) incomplete combustion of organic compounds (*1) Carbon dioxide (C02) incomplete combustion of organic compounds (*1) Hydrogen cyanide (HCN) wool, silk, nylon, paper, polyurethane (*1) Hydrogen chloride (HC1) poly vinyl chloride (PVC), plastics (*2) Nitrogen dioxide (N02) cellulose and nitrate containing compound (*1) Sulfur dioxide (S02) sulfur containing compounds (* 1) Hydrogen fluoride (HF) synthetic polymers (*2) Hydrogen sulfide (H2S) sulfur containing compounds such as wool, rubber, diesel (*1) Acrolein cellulose, wood, polyoleifins (*1) Benzene combustion of hydrocarbons and organic compounds (*1) Formaldehyde combustion of wood, cellulose containing materials (*1) Particulate incomplete combustion of organic materials (* 1) *1 - WCB report 1994 *2 - Guidotti et. al. 1992 1.1.2 Concrete sink effect The concrete sink effect is not a byproduct of combustion but a special hazard. Concrete retains heat very efficiently, and due to its porous nature may entrap gases in the pores. These pores expand as they warm and absorb chemicals, similar to a sponge absorbing water. Later, as they cool, they slowly release any gases absorbed into the pores (Guidotti and Clough 1992, WCB report 1992, Dyer et. al. 1976). This 4 effect is common where concrete is exposed to combustion byproducts and has been termed off-gassing. This is a hazard for firefighters, as chemical release can occur long after the fire is extinguished, resulting in potential chemical exposure after the fire is suppressed and the visible smoke has cleared. At these times firefighters may incorrectly believe there are no chemicals remaining. Firefighters may deem the area relatively free of exposure and remove their SCBA, which is heavy, hot, cumbersome, and restrictive. This can result is significant chemical exposure (WCB report 1992). 1.2 Health Effects of smoke exposure 1.2.1 Smoke inhalation Examination of firefighters injured in the line of duty shows that over 80% of injuries are from smoke inhalation, and of all firefighter deaths that occurred on the job over 50% were caused by smoke inhalation (WCB report 1994, Terrill et. al. 1978, Utech 1975). Thus smoke inhalation is the primary hazard of firefighting. Smoke inhalation results in exposure to a multitude of combustion byproducts which can result in a variety of pathological conditions in the human body. The location and extent of the injury will depend on the size of the particles in the smoke, the solubility of the gases, and the duration of exposure (Rinke 1987). Each single contaminant in smoke by itself may be associated with a specific form of injury or multiple physiological effects. The fact that smoke is a chemical mixture, and that synergism and interactions between each chemical in the mixture may occur, compounds the problem even further (Guidotti and Clough 1992). Combinations of effects between two or more agents can results in unique and unexplained effects at concentrations lower than necessary for either constituent acting alone (Committee on fire toxicology 1986). Furthermore, uncertainty exists about the role of toxic gases in fire-related injury and death. A study by Burgess et. al. (1995) reported that the four most serious combustion byproducts, representing the most serious hazard were CO, HCN, HC1, and particulate. CO has long been implicated as the main cause of fire-related deaths because study has shown most other gases are often not present in sufficient quantities to 5 be life ttireatening (Prien 1988). Guidotti and Clough (1992) have also reported that gases other than CO and C0 2 are rarely produced in concentrations that are considered to be lethal. However, in contrast, some studies have measured other in excess of levels immediately dangerous to life and health; under these circumstances these gases are present in sufficient quantity to cause death. Bayer (1974) and Treitman (1980) have both hypothesized that N0 2 may play a key role in fire-related deaths. Decker and Garcia-Cantu (1986) have proposed that it is the synergistic effect of a variety of chemicals that is responsible for fire-related deaths. Finally, Becker (1985) has proposed that cyanide from HCN may play a vital role in fire-related deaths. There may be no way of predicting possible health outcomes following exposure or predicting at what level of exposure these effects may occur. To answer the questions about possible health effects from smoke exposure, it has been argued that exposure levels must first be quantified. Only then can these levels be compared to reported exposures at which previous health effects have occurred (U.S. Dept. of Health and Welfare symposium publication 1985). The most notable general health effects are summarized in table 3. A more detailed description of these health effects has been included in appendix IA. Table 3: General health effects of some combustion byproducts Benzene Carbon Dioxide Carbon Monoxide Formaldehyde Hydrogen Cyanide Particulate Sulfur Dioxide asphyxiant x ••carcinogen X X cardiovascular X cellular poison X hematological X X immune effects x X X irritants X X X mutagenic reproductive X respiratory - X X X X sensitizer X ••Defined by the International Agency for Research on Cancer (IARC) 1.2.2 Regulations regarding smoke exposure Regulating bodies, such as the Workers' Compensation Board of British Columbia, set limits regarding exposure to chemical substances to protect the health and safety of workers. These limits define the 6 maximum level to which most workers may be exposed, without any undue health effects. The limits set by various government bodies in British Columbia and the U.S. have been included as appendix IB. Unfortunately, the nature of firefighters' jobs, suppressing uncontrolled fires, makes adherence to these limits impossible without personal protective equipment. Smoke is a complex mixture of chemicals, particulate, gases, and other compounds. To address the problem of regulating exposure to mixtures, the Workers' Compensation Board of B.C. considers the effects of a mixture to be additive, unless it is known otherwise. To calculate the additive exposure of a mixture the following equation is used: AE=%EL(A)+%EL(B)...+%EL(n) (WCB 1998) where 'AE' is the calculated additive exposure, 'EL(A)' is the measured exposure to component A, expressed as a percent of its exposure limit, and 'EL(B)' is the next component in the mixture, etc. 1.2.3 Controlling smoke exposure by use of SCBA Recognition of smoke inhalation as the primary concern to a firefighters' health has been the main impetus for the development of better personal safety practices, better self contained breathing apparatus (SCBA), and binding safety regulations. Regulations (NFPA) require that SCBA is to be used at all fires, hazardous material incidents, overhaul operations if a hazardous atmosphere is present, if the atmosphere is suspected of being hazardous, or if the atmosphere may rapidly become hazardous. All SCBA must maintain positive pressure in the facepiece and meet the requirements of the National Fire Protection Association (NFPA) standard 1981. Under ideal circumstances SCBA will provide a lab tested protection factor of 10,000, which is the maximum protection provided by any form of breathing apparatus. This means that if the concentration of a gas outside a mask is 10,000 parts per million (ppm) then the concentration inside the mask will be only 1 ppm. The protection factor is derived from laboratory testing and one never achieves maximal protection in 7 the field (Utech 1978). Problems such as the ability to fit one generic model of mask to everyone's face, the presence of facial hair, or glasses compromise the fit and contribute to leakage (Utech 1975). Regulations requiring the mandatory use of SCBA have reduced smoke inhalation injury. Presently, most smoke inhalation cases are now believed due to non-use of SCBA use, and early SCBA removal (Heinemann et. al. 1987). Studies have reported that air samples taken inside the mask have shown low level exposure to some chemicals and is believed to be due to leakage (Jankovic et. al. 1991). However, follow-up study has shown that chemical exposure from the effects of mask leakage is insignificant compared to early mask removal (Jankovic et. al. 1991, Levine 1978). Early mask removal occurs when firefighters remove their SCBA, often during the overhaul phase of fire suppression. The "overhaul" phase is when the main body of flame is suppressed and little fire or smoke is visible, however, the fire may still be smoldering and emitting gases (Morse et. al 1992, Heinemann et. al. 1987). Studies have shown that there is no correlation between the amount of gases such as CO, C0 2 , HC1, or benzene in the air and the amount of visible smoke present (Guidotti and Clough 1992, Dyer et. al. 1976, Bendix 1978). This may lead to exposure, as undetectable toxic gases may be present, and firefighters remove their masks (Guidotti and Clough 1992). 1.3 Studies involving fire fighting training centers There exists very little literature specifically pertairiing to fire fighting trairiing centers. Booher and Janke (1997) evaluated plume and ground level concentrations of contaminants from crude and fuel oil fires at a fire training facility in Calgary Alberta. They reported that measurable quantities of hazardous emissions are produced, that the majority rises into the air, and is significantly diluted before again settling to the ground. They also reported that measured amounts of PAH's, well below occupational exposure limits, were the only quantifiable ground level exposure. A similar study by Leahey et. al. (1993) studied emissions from naval training fires and concluded that the air quality impact of the fire fighting training is much lower than the relevant time weighted averages established to protect workers' health. Neither study 8 by Janke (1997) or Leahey et. al. (1993) reported if they examined the possibility of additive effect of contaminants measured in their studies. In contrast, a study by Feunekes et. al. (1997) recently reported that uptake of PAH occurred amoung fire training instructors at the Royal Netherlands fire training school, despite short term smoke exposure. They also suggested appropriate control measures against PAH uptake may be necessary. A related study by Atlas et. al. (1985) measured emission of chemicals from fires at a fire training center. Unfortunately, they only measured combustion byproducts directly in the smoke plumes and did not evaluate personal exposure of staff at the facility. Stevenson (1985) studied the effect of radiant heat in fire training instructors but did not assess ground level smoke exposure. 1.3.1 Prior Study at the FSTC A previous evaluation by Lockhart Risk Management Limited (1992) reviewed potential hazards associated with the open air burning of diesel fuel. They reported that the staff at the Maple Ridge Fire and Safety Training Center are at higher risk than students, due to their ongoing work at the site. Though Lockhart Risk Management reported that there are risks, they could not quantify the risks, they did not focus on any other hazards associated with burning at the facility, nor did they address the individual exposure of the instructors at the facility. A more recent investigation by van Netten (1996) measured worst case exposure at the facility and recommended that a more detailed study of the potential exposure of the employees at the center be done. Considering the recommendations of previous study, this thesis was conducted to assess the staff exposure to ground level smoke during their work at the institute. 1.4 Overview of the Fire and Safety Training Center 1.4.1 Site description The institute is located on an isolated five-hectare site, at Maple Ridge British Columbia. The institute employs 12 people; 5 are instructors who are involved in field exercise demonstrations, 3 are technicians who maintain equipment and take care of the facility, 2 are managers, and 2 are clerical workers. Adjunct 9 instructors periodically help with training. The center offers various courses designed to teach safe fire suppression. There is a 12 week new recruit program where newly hired recruits from Lower Mainland fire departments are trained. All other courses are shorter in duration, and include a week long training course on industrial fire fighting, a week long course on marine firefighting, courses on fire extinguishers that may last one to three days, and a few other short courses lasting only a day or two. All training programs employ classroom and field exercises as part of their content. Classroom session cover theory on fire behavior, burning, and suppression. Field exercises consist of hands-on training in the use of equipment, fire fighting techniques, and active fire suppression. All fires are conducted in a controlled manner under the supervision of highly trained fire instructors who have been, or still are, fire fighters. The FSTC has eight training sites, six of which are designed to allow fire demonstration. They are: the burn building, ship mock-up, propane and gas pad, extinguisher pad, round tank, and T-pit. The other two sites are used to instruct search and rescue techniques and confined space entry, they are the smoke-house and rail cars, and not employ any burning. Ancillary to the center are sites where SCBA tanks are refilled, SCBA tanks are stored, fire extinguishers are refilled, a maintenance area which contains washing facilities for personnel and for equipment, a fire truck storage garage, an office complex, two pump houses, and a number of indoor classrooms. A site map is provided as figure 1 (following page) and shows the layout of the facility. 10 Figure 1: Site Map 11 1.4.2 Detailed description of the sites with fire activity 1.4.2.1 Outdoor fuel and depth pits 1.4.2.1.1 T-pit This site offers trairiing in suppressing a 'fuel and depth' fire, a with liquid hydrocarbons floating on top of a layer of water. The fuel floats on top a pit of water that is approximately 0.65 meters deep, is shaped similar to the letter "T", and is sunk about 1 meter below grade. Training involves instruction on hose handling techniques as well as foam applications, which are necessary to suppress these fires. 1.4.2.1.2 Round tank This structure is used to simulate a leaking pipeline fire. The structure is a circular tank partially filled with water, creating a fuel and depth situation (as the T-pit), however, the sides of the tank extend about 1.5 meters above grade. There is a raised platform at one end where pipes are located and stairs leading up to the platform. The actual fire is generated from the pipes on the raised platform, as fuel leaking from the pipes is ignited. The structure is designed so that there is little danger of the pipeline exploding. Extmguishing the fire involves the use of foam and dry chemicals (fire extinguishers). 1.4.2.1.3 Description of training activity at both the T-pit and round tank Burning at both the T-pit and round tank produces large fires that release copious quantities of black smoke into the air. Both sites generate similar fires and are expected to produce similar contaminants in comparable quantities. Trairiing procedures at these pits consist of lighting a fire, allowing it to briefly "preburn", have crews quickly suppress the fire, and then restarting the fire for another practice suppression. Actual fire suppression usually lasts 2-5 minutes and rarely exceeds 10-15 minutes. In total 3-7 fires are lit during a trairiing period, generally one after the other, unless there is need for a brief discussion. 1.4.2.1.4 Sampling rationale at the outdoor pits At the round tank a technician is required to monitor a fuel pump and emergency shut off switch, at a location about 12 meters North-North East of the middle of the pit. Course participants often observe the 12 burning unprotected by SCBA at an area between the T-pit and round tank. At both locations there is the potential for smoke and contaminant exposure to various chemical combustion products. Sampling at the outdoor pits was to assess contaminant levels at the two areas where SCBA is not used, and at locations deemed to be at the downwind side of the pits, which may represent worst case exposure areas. The objective was to verify that chemical exposures do not exceed recommended levels, confirm that SCBA will not be required, and help alleviate health concerns about exposure at these sites. The areas sampled are indicated on figure 2 (following page), which is a site map of the outdoor pits, extinguisher pad, and propane pad. Area 1 is where the emergency shut off switch is located, area 2 is where people may observe without SCBA, areas 3 and 4 both represent areas in the path of the predominant wind direction at the facility, and are hoped to represent worst case exposure areas. 13 Figure 2 Site sampling location diagram Extinguisher pad Propane and gas pad 14 1.4.2.2 Extinguisher pad 1.4.2.2.1 Site description This area is used to train students in the proper use of fire extinguishers. There are several fire pits similar to the round tank and T-pit except that they are much smaller. Fuel and depth fires are lit in these pits and then suppressed with 10 to 20 pound hand held fire extinguishers. 1.4.2.2.2 Description of training activity Training consists of lighting a fire and having course participants extinguish the fire with the fire extinguisher. SCBA is seldom used as fires are small and suppressed within seconds. However, the use of the fire extinguishers creates large clouds of dry chemical powder (sodium bicarbonate) suspended in the air, which may drift with the wind currents before settling to the ground. It is suspected that these clouds contain contaminants formed as a result of burning, dry chemical particles, and C0 2, which is used as the propellant in the extinguishers. 1.4.2.2.3 Sampling rationale at the extinguisher pad Sampling at the extinguisher pad was to conducted to assess contaminant levels in the floating clouds of dry chemical powder. The objective was to verify that chemical exposures do not exceed hazardous levels, confirm that SCBA is not required, and help alleviate health concerns about exposure from these clouds. 1.4.2.3 Propane pad 1.4.2.3.1 Site description This site employs a "controlled" propane leak in a large tank that could hold 150 kilograms of propane. The tank has been modified to ensure there is no boiling liquid expanding vapour explosion (BLEVE) potential (a hazard associated with propane canister fires). This has been achieved by use of safety valves and by cutting a hole into the backside of the tank. The propane fires are lit, and then suppressed by turning off the gas supply valve, which is 1 meter from the flame source. Once the propane fuel supply is turned off the fire is immediately extinguished. 15 1.4.2.3.2 Description of training activity Fires at this site are impressively intense and loud, yet burn cleanly with almost no visible smoke emission. Trairiing consists of lighting the fire, allowing the fire fighters to extinguish the fire, and then restarting the fire for another practice suppression. During most practice sessions 3 to 5 fires will be lit and suppressed. A technician is required to be present near the site holding an emergency shut offline in case of an emergency. The emergency line is located at a station about seven meters west of the fire site. Traditionally no SCBA is worn, as the staff at the site deems the fires clean burning with no potential for exposure. 1.4.2.3.3 Sampling rationale at the propane pad Sampling at the propane pad was to assess contaminant levels at the site where the technician stands, that is indicated as area 5 on figure 2. The objective was to verify that chemical exposures do not exceed hazardous levels, confirm that SCBA will not be required, and help alleviate health concerns about exposure at this site. 1.4.2.4 Indoor sites 1.4.2.4.1 Burn building This large concrete three story building was made to simulate house fires. All fires occur in rooms inside the building and are thus enclosed indoor fires. The structure is divided into two sides, one side burning propane and the other side burning wood pallets, which are initially ignited with a propane torch. There is a local exhaust ventilation system at one end of the building, which consists of two wall fans placed on each floor. 1.4.2.4.2 Ship mock-up Similar to the burn building, this site also has fires that are enclosed inside the building. This site is a large metal structure designed similarly to a ship, and training in marine fire suppression occurs inside. The fuel source is primarily wood with gas and diesel used to initiate combustion. In one room, a simulated engine room, the fire is fueled with diesel. 16 1.4.2.4.3 Description of training activity All fires occur in rooms inside these structures, resulting in smoke accumulating inside the buildings, with little escaping to the surrounding outside area. Rooms inside the buildings are connected by means of hallways and doors with vents in them. The floors are connected by stairwells, which also have doors at the top and bottom landings of the stairwell. When closed these doors prevent most smoke from escaping into other sections or floors of the building, but not all. To ensure safety at these structures it is mandatory that anyone inside or entering, during active burning, must be wearing SCBA. Fire training consists of lighting the fire and then having the firefighters enter the structure, locate the fire, and then suppress it. After suppressing the fire the structure is 'vented' to remove residual smoke quickly. Venting is accomplished by opening all windows and doors, ttirning on portable venting fans, and by spraying a stream of water out a window or door. This has the effect of pulling smoke out of the structure. In most circumstances venting is performed until the structure is relatively clear of residual smoke, and generally lasts 10 minutes, but may take longer in some instances. Venting in the ship is performed in nearly the same manner as the burn building, except that the burn building has local exhaust fans built into the building, and the ship does not. These fans can be turned on to help flush smoke from the building. Following venting, the firefighters clean up fire hoses and gear and then let the building air out. 1.4.2.4.4 Sampling rationale After venting, not all the smoke may have been cleared from the structure. If so, maintenance, cleanup, or post fire discussions in the structure may be unsafe due to residual smoke exposure. If smoke exposure may occur, then the structures should only be entered after a sufficient amount of time has elapsed, to ensure that contarninant levels have diluted to levels that would not jeopardize health. 1.4.2.4.5 Propane side of the burn building This side of the burn building is similar to the other side, except that is uses propane fuel. The training procedures are similar as those in the other side, except, this side of the building is equipped with exhaust ventilation fans. Another notable exception is that theatrical smoke (Rosco smoke fluid) is used with the 17 propane fires to help produce smoky fire conditions. The supplier of the theatrical smoke, and material safety data sheet (MSDS) claims it is non-toxic and non-irritating and poses no health risk. 1.4.2.4.6 Sampling rationale at the indoor burn sites Unvented levels (worst case environmental levels after fires): sampling was conducted to determine the levels of gases, residual after the fires had been suppressed, but before the structures have been vented. This was to determine if contaminant levels would be dangerous to health, verify that SCBA is necessary to enter the structure, and give baseline levels of contaminants before venting the structure. Sampling of unvented levels was only conducted in the wood side of the burn building and not in the propane side or ship mock-up. Environmental levels immediately following venting: At both the wood and propane side of the burn building and ship mock-up, sampling was conducted to determine levels of contaminants residual immediately after venting, and to assess whether SCBA would be necessary to enter the structure following venting. Assessment of the time for environmental levels (after venting) to dilute to insignificant levels: If levels remained elevated after venting it is assumed that they would eventually dilute to insignificant levels. If so, after an undetermined period of dilution one could safely enter the structure without SCBA. Thus sampling was conducted to determine the time periods for gases to dilute to insignificant levels. Sampling was done in the burn building (wood side) and ship mock-up. Background levels (environmental levels in the absence of burning): Sampling was conducted to determine if background levels of contaminants may be present in the indoor sites, even after long absences of any burning in the structures. 1.4.3 Fuel sources burned at the FSTC The use of live fire is required to allow hands on experience in fire suppression. Due to ongoing training, the instructors are exposed to byproducts of combustion on almost a daily basis. This differs considerably from firefighters at municipal departments who may wait weeks or longer between fire calls. Another 18 consideration is that all fires at the facility are controlled and employ limited types of fuel. The limited fuel sources may decrease the variability of the byproducts of combustion to which FSTC staff are exposed. The main fuels burned at the site consist of diesel mixed with 7% gasoline, wooden pallets, and propane. The wood pallets are used at the indoor burn sites, propane is used at the propane pad, and the gas diesel mixture is used in the fuel and depth pits. The fuel and depth pits require a mixture because gasoline is required to initially start the fire burning, and diesel is used to allow prolonged burning. Following is a brief discussion of each fuel source and potential byproducts of combustion. 1.4.3.1 Wood Materials located on the low end of the energy scale, such as wood or charcoal, generate more emissions than high energy products, such as natural gas or propane, which tend to burn cleanly (Anon. Burning issues, Larson and Koenig 1994). This is supported by research that has found that combustion of wood releases higher levels of particulate and organic carbon than oil or gas (Committee on fire toxicology 1986, Westbrook and William 1983). Not all chemicals generated from wood fires have been identified and their release is only partially clarified (Ware 1991). To date over 2000 chemicals have been identified in wood smoke, of which over 100 are known carcinogens. Due to the high number of carcinogens present in wood smoke, the US Environmental Protection Agency has estimated that the risk of cancer from wood smoke is 12 times greater than cigarette smoke. Particulate emissions predominately consists of particles less than 1 um in size, allowing them to remain airborne for up to three weeks (Committee on fire toxicology 1986). Particles of this size may also penetrate deeply into the lung, to the alveoli, where they may cause considerable harm (Burning issues). Wood combustion is not usually a source of NO x or SOx production due to lower combustion temperatures (Vedal 1993). However, at the lower temperatures more organic acids and aldehydes are produced, with higher levels of formaldehyde, acrolein, benzene, toluene, etc. being associated with wood combustion (Ware 1991). 19 1.4.3.2 Diesel Diesel is a less volatile petroleum distillate than gasoline and consists of long chain olefinic, paraffinic, and aromatic hydrocarbons, and generally contains less than 0.5 % sulfur. Few studies have identified the common products of combustion of diesel in open air burning, and the literature, focuses on diesel exhaust emissions from internal combustion engines, van Netten (1994) reported that open air combustion, which will be less complete, may produce increased levels of organic emissions, including poly aromatic hydrocarbons (PAH's). Bayer (1974) has reported that products of incomplete combustion tend to be more harmful than those from complete combustion. Known products of combustion include unburned fuel, CO, C0 2, SOx, particulate, soot, aldehydes, benzene, aliphatic hydrocarbons, and PAH's, many of which are carcinogenic (Atlas et. al. 1985). It is also known the products of combustion can vary considerably during burning (Atlas et. al. 1985). Little literature exists concerning open-air combustion of diesel fuel. A reference about the adverse health effect of diesel fuel combustion in automobile engines, which is well studied and documented (Steenland 1986, Minty 1985, Gharibeh 1988, Sampara 1987) is included here. Previous studies of internal combustion engine emissions have looked at possible links with cancer, as diesel emissions have been found to be mutagenic and carcinogenic in bioassay (Gharibeh 1988, Sampara 1987). However, these studies have reported that at occupational exposure levels, vehicular diesel emissions are unlikely to be carcinogenic in humans (Gharibeh 1988, Sampara 1987). Table 4, below, summarizes the most commonly measured components of diesel exhaust fume from machinery in underground mines. Table 4: Concentration of emissions in diesel exhaust Emission Concentration (ppm) CO 0-20 co 2 400-1500 NO 1-20 N0 2 0.2-0.4 so x 0.5-0.7 Diesel particulate (PAH's) 0.1-2.0 mg/m3 *From Sampara - Canadian center for occupational health and safety. 20 1.4.3.3 Gasoline Gasoline is a clear, volatile liquid that is exceptionally flammable. It is a complex mixture of aromatic, olefinic, and paraffinic hydrocarbons, primarily ranging from C 3 to Cn compounds, with as many as 200 hydrocarbons identified in gasoline (Biological exposure index for gasoline ACGIH). Typically the composition consists of 80% paraffins, 14% aromatics, 6% olefins, with the content of benzene approximately 1%. Combustion of gasoline can result in a multitude of byproducts, however, as with diesel fuel, the majority of research is based on gasoline combustion engines and little exists that concerns open air combustion. For comparison, diesel exhaust emissions are compared with gasoline engine emissions in table 5. Table 5: Comparison of diesel and gasoline exhaust emissions Emission Gasoline compared to diesel Hydrocarbons higher levels CO higher levels NOx same SOx same Particulate 50-100 lower *From Sampara - Canadian center for occupational health and safety. 1.4.3.4 Propane Propane is a colorless gas, odorless when pure, but has an added odorerant. Natural propane is a mixture containing 96% propane and 4% or less butane (Biological exposure index ACGIH). Combustion of propane is much cleaner than other hydrocarbon fuel (Westbrook and William 1983, Allara and Edelson 1975), with far fewer emissions released and generally at much lower concentrations. Combustion products likely to be detected at the site are CO, C0 2, and possibly NOx. Other byproducts may be present but are thought to be undetectable in open air burning situations. 1.4.4 Potential byproducts of combustion at the FSTC Due to the fuel used (diesel, gasoline, wood, and propane) there should be limited production, if any, of chemicals that have been associated with combustion of synthetic materials. Consequently HC1, HF, polyvinyl chloride, etc. are not expected to be detectable at the facility. Government regulations require the 21 sulfur content of diesel to be < 0.5%, which limits the amount S0 2 and H2S produced. Additionally, due to the low odor threshold of H2S, one would be able to smell it, but this has not been reported, or observed in personal visits to the site. Actual burn periods, during training exercises are brief, rarely more than 30 minutes and the fuel supply is limited. This limits the temperature of combustion since the fires will not have time or fuel supply to reach high temperatures. Previous measurements of large liquid pool fires by Koseki (1989) reported that maximum temperatures reached in gasoline fires were 1360°C, and averaged about 1100°C. Kerosene had a maximum temperature of 1380°C. Diesel fuel fire temperatures were not reported. The maximum temperature of the gasoline and kerosene fires were slightly above temperatures known to allow nitrogen fixation (1340°C), but were only achieved after allowing the full combustion potential of the fire to be reached. Koseki (1989) stated that fire intensity and burning rates increased with tank size, and also estimated that maximum temperatures would reach 1300-1400°C in 30-50 meter diameter tanks. Fires at the FSTC are in smaller tanks (the largest tank the round tank is about 7 meters in diameter) and are suppressed quickly, probably before maximal temperature is reached. As a result, fixation of nitrogen should be limited, reducing the production of oxides of nitrogen. Subsequently, limited formation of N0 2 is expected. NO and HCN may be present in small quantities but may be difficult to detect during the brief burning periods. It is believed that the common byproducts of combustion may be present at the facility, they are: CO, C0 2 , benzene, particulate, formaldehyde, thus they are the main chemicals studied. Additionally, S0 2 and HCN were measured, which would allow investigation of the amount of production of these chemicals. 1.4.5 Use of SCBA at the FSTC and its effect on the study At the FSTC, exposure during training sessions is minimized, as the institute supplies turnout gear and SCBA to all participants in all courses. Proper use and care of the equipment is demonstrated, and use of the equipment is required during all training sessions. Instructors have their own equipment present during 22 the field exercises and use of their equipment is also required of them. Additionally, the institute has technicians who maintain and service this equipment on site. With the high level of protection that SCBA provides it is recognized that each instance of smoke exposure may not result in harmful exposure. Previous study by Levine and Hothman (1978) concluded that SCBA protects firefighters from smoke exposure. This is supported by a risk assessment conducted by Burgess and Crutchfield (1995), who concluded that the NIOSH recommended protection factor of 10,000 was adequate respiratory protection for firefighters. The majority of smoke exposure at the facility will occur while not using SCBA. To make the project feasible, in a reasonable period of time, the assumption that SCBA provides adequate respiratory protection was made. No fit testing was conducted to determine the protection factor that the SCBA provided, or to validate the previous assumption. Thus one can not be confident in the fit and seal of the equipment and protection it offers. It is recommended if further study is to be conducted that this assumption be verified. Conforming to the previous assumption, no evaluation of exposure was conducted for instructors or course participants in situations where they use SCBA. 1.4.6 Chemical combustion products sampled Sampling focused on specific compounds chosen as indicators of exposure to specific classes of contaminants. This was done to decrease sampling time and laboratory analysis. Benzene and formaldehyde are believed to be in the smoke emissions and were measured as representative compounds of total volatile hydrocarbons and aldehydes. Formaldehyde was chosen because it is reportedly found in about 10 times as great a concentration as other aldehydes, such as acrolein (Ware 1991). Additionally, due to the extremely irritating nature of acrolein one may be able to detect emissions, but this has not been reported, nor observed in personal visits to the site. Benzene was chosen as it has been found in detectable amounts at many fires (WCB report 1992), and is a volatile organic chemical with welMmown health effects. Total particulate was measured to allow investigation of potential levels of soot and particles formed from burning. It is believed that the main gases formed during combustion would be CO, C0 2, and S0 2 so they 23 were measured to determine the levels present. HCN was measured as a study by Becker (1985) has reported it may play a key role in smoke related injury. 24 Chapter 2: Methods Sample Collection 2.1.1 Observation of fire characteristics and weather conditions All sampling was conducted during course training sessions to determine exposure during these activities. Factors, such as size of the fire and duration of suppression that may add to the variability and quantity of contaminants were recorded. Fire sizes were recorded at the outdoor pits (T-pit and round tank) and were categorized as large, medium or small. Fire size was determined by approximating the height of flames above the surface of the pits. Fires with flames less than 2 meters in height were recorded as small, flames greater than 2 meters but less than 4 meters was recorded as medium, and flames greater than 4 meters were considered large. All fire sizes were assessed by personal observation. Weather conditions such as wind direction, temperature, precipitation, and humidity was measured daily prior to sampling. Wind direction was determined by noting the direction of a blowing flag. Relative humidity and temperature were measured using a sling psychrometer, psychometric chart, and thermometer. All readings were conducted outside the East Side of the main office. Sampling pumps were turned on prior to use and flow rates were measured, and rechecked after sampling, with a calibrated rotameter. Real time electrochemical gas detectors were started prior to sampling, to briefly monitor outdoor ambient gas levels, in an area deemed contaminant free. 2.1.2 The sampling stand All sampling for the project was performed using a sampling stand. This stand consisted of light aluminum shelving (purchased at Lumberland Building Materials) on which sampling apparatus could be placed. Sampling for CO, CO2, S0 2 was conducted by placing the gas detectors on the top shelf. Sampling for formaldehyde, HCN, and benzene was done by taping the sampling tubes, cassettes, or impingers to the top shelf. All sampling was conducted on the top shelf, to simulate breathing zone levels, as this shelf was 160 cm above the ground. 25 2.1.3 Outdoor burn sites: 2.1.3.1 Fuel and depth sites (T-pit and Round tank) sampling rationale Over a period of six months (June to December 1996) sampling was conducted. Sampling stations were located at the areas marked 1, 2, 3, and 4 on the site map provided (figure 2). Location 1 is at the site where the technician must stand to monitor an emergency switch for the round tank (about 12 meters North east of the center of the round tank). Location 2 is between the T-pit and round tank, but is located South West of the round tank (about 10 meters from the round tank and 12 meters from the T-pit) at the area where people often observe without SCBA. Location 3 is between the T-pit and round tank, but is located South East of the round tank (about 10 meters from the round tank and 12 meters from the T-pit). Location 3 represents the predominate natural direction of wind from the round tank. Location 4 is southeast of the T-pit and represents the predominate natural direction of wind from the T-pit tank (about 10 meters from the T-pit). Sampling was started immediately before ignition of the first fire, was conducted over the entire duration of fire activity, and was halted immediately following suppression of the final fire. During the sampling period the time of ignition and suppression of each fire was recorded, along with the predominant direction of smoke plume travel. 2.1.3.2 Extinguisher pad sampling rationale Over a period of three months (Nov 96 to Jan 97) sampling was conducted at the extinguisher pad. Sampling was performed to measure the level of particulate and possible levels of contaminants trapped in the dry chemical clouds. Sampling stations were placed on the downstream side of the pit, in the direction of which the dry chemical was sprayed. In this manner it would measure levels of contaminants contained within the cloud of dry chemical powder. Samples were started immediately before initial fire ignition and continued until the last fire was suppressed. 2.1.3.3 Propane pad sampling rationale Over a period of four months (Sept 96 — Dec 96) sampling was conducted at the propane pad. 26 Sampling was done for CO, C0 2, and S0 2 and no other contaminants because propane is reported to burn much more cleanly than wood or diesel (Westbrook and William 1983, Allara and Edelson 1975, Anon, from Burning Issues). A sampling station was located at the area where the technician must stand to monitor the emergency shutoffline. Samples were started immediately before initial fire ignition and continued until the last fire was suppressed. During the sampling period the time of ignition and suppression of each fire was recorded, along with the direction of smoke plume travel. 2.1.4 Indoor burn sites: Ship mock-up and Burn building (wood and propane sides) 2.1.4.1 Sampling Procedures 2.1.4.1.1 Background levels (environmental levels in the absence of burning) Background levels were measured only if no fire activity occurred for approximately 1 week (7 days). In the burn building the sampling stand was placed in the center of the room in the North west corner of level 1, the room where the majority of fire activity occurs. Sampling for CO, C0 2, and S02, HCN, formaldehyde, particulate, and benzene was conducted. All formaldehyde, benzene, particulate, and HCN samples were conducted for 50 minutes and 4 consecutive samples were collected. Real time analysis of C0 2, CO, and S0 2 was conducted for the entire duration. In the ship mock up the sampling stand was placed in the center of the north end of the ship, in the corridor directly beside the simulated engine room. This room was chosen due to the proximity to the engine room, where the majority of fire activity occurs. Sampling was not conducted in the engine room due to the steep narrow access stairs, which would make entry into the room with equipment dangerous. 2.1.4.1.2 Unvented Levels (worst case environmental levels after fires) Sampling was done by placing the sampling stand inside a hallway in the middle of the structure next to the room where a fire had previously been extinguished, but not vented. Due to anticipated levels of gases SCBA was necessary for entry into the structure to place the equipment, and was done with a FSTC staff technician. Sampling was conducted for 30 minutes only, long enough to get samples, yet short enough to not damage the equipment due to the heat residual in the structure. 27 2.1.4.1.3 Environmental levels immediately following venting The sampling stand was placed in the center of the room where fires had been lit and suppressed. Placement occurred after the firefighters had completed venting and had cleared their equipment from the building. In most cases, sampling was conducted within 10 minutes of final fire suppression. Placement of the stand could not be accomplished before the firefighters were finished venting, as the structure is obscured by smoke, visibility is reduced, and equipment could be damaged by firefighting activities. Sampling was conducted for 30 minutes. 2.1.4.1.4 Assessment of time period for environmental levels (after venting) to dilute to insignificant levels These samples were identical to the sampling conducted immediately following venting. Sampling was conducted by taking four consecutive samples of HCN, formaldehyde, benzene, and particulate. Sampling was started immediately following venting, and the first long term samples were the samples used to determine the contaminant levels immediately following venting. For HCN, formaldehyde, particulate, and benzene the first two samples were 30 minutes in duration, and then the following two were collected as 1 hour samples. In this manner 4 samples were collected over a three hour period. For CO, CO2, and SO2 real time data collection was done with the toxic gas detectors. 2.2 Sample collection and analysis All samples of HCN, particulate, formaldehyde, and benzene collected at the FSTC were transported back to the UBC Occupational Hygiene Laboratory and analyzed following WCB standard methods. All data for S02, CO, C0 2 was downloaded, at UBC, onto a Pentium computer. 2.2.1 CO. CQ 7, SOi Measurement of these gases was done with datalogging electrochemical gas detectors, calibrated every two weeks. All monitors were set to record the time weighted average concentration of gas over a 60-second period. This allowed real time data collection that measured change in gas concentration (rate of increase or decrease). CO and S0 2 were measured using a TMX 410 passive electrochemical gas analyzer made by 28 Industrial Scientific Corporation. CO2 was measured using a Young Environmental Systems-203 (YES) toxic gas detector. Limits of detection for CO and CO2 were 1 ppm and 0.1 ppm for SO2. 2.2.2 HCN HCN was collected using midget impingers following the British Columbia WCB method 0700. This method entailed drawing a known volume of air through an impinger, filled with an ammonium solution of nickel (II) chloride. Air was drawn through the impinger at a flow rate of approximately 2.0 L/min, using a Aircheck Sampler Model 224-PCXR3 pump (SKC Inc. Eighty Four, PA.). Pumps were calibrated on site with a Matheson Gas Products Rotameter that had been previously calibrated with a soap bubble meter. The formation of a tetracyanonickelate ion was spectrophotometrically measured at 267 r\m with a LKB Biochrom Ultrospec II spectrophotometer (Cambridge, England). The samples were quantified by comparison with standard cyanide solutions made at the University of BC occupational hygiene lab. The detectable mass of HCN was 5 u.g, which corresponded to a concentration limit of detection of about 0.01 ppm, depending on the volume of air sampled. 2.2.3 Particulate Total particulate was measured gravimetrically following the WCB method 1150. A known volume of air was pulled through a two section closed face filter cassette loaded with 37-mm diameter, 5 micron pore size polyvinyl chloride filters (Costar Nucleopore). Air samples were drawn with an Aircheck Sampler Model 224-PCXR3 pump (SKC Inc. Eighty Four, PA.) calibrated on site at a flow rate of approximately 2 liters per minute. Before sampling, filters were preweighed 3 times to within 0.01 mg using a Sartorius Micro balance (Sartorius GMPH Gottingen, Germany). Post-sampling filters were weighed 3 times, on the same balance at the UBC Occupational Hygiene Laboratory. One blank sample was collected with every five samples taken, in an manner identical to other samples, except that no air was drawn through the cassette. The limit of detection for particulate was approximately 0.05 mg/m3, depending on the volume of air collected. 29 2.2.4 Formaldehyde Formaldehyde samples were collected following the WCB method 5270. A known volume of air was drawn through sorbent tubes (SKC Eighty Four, PA.), packed with two sections of impregnated silica gel, at a flow rate of 0.5 L/rninute. Sampling used an Aircheck Sampler Model 224-PCXR3 pump (by SKC Inc. Eighty Four, PA.), calibrated on site with a calibrated rotameter. Samples were extracted with acetonitrile and aliquots were injected into a Dionex high pressure liquid chromatography (HPLC) instrument equipped with a Dionex variable wavelength UV-visible photometric detector set at 353 run. The samples were quantified with formaldehyde standards prepared at the UBC Occupational Hygiene lab. The detectable mass of formaldehyde was 0.15 ug, which corresponded to a limit of detection of about 0.005 ppm, depending on the volume of air sampled. 2.2.5 Benzene Benzene was sampled following the WCB method 3301. A known volume of air was drawn through Supelco ORBO 32 standard charcoal tubes (Supelco Inc. Bellefonte, PA.) packed with two sections of activated charcoal. Sampling used a flow rate of 0.2 L/min, using Dupont UL model P30A pumps (Dupont Inc.), calibrated on site with a calibrated rotameter. Samples were extracted with the desorbant CS2/undecane. Aliquots were injected into a Varian 3400 Gas Chromatograph (Walnut Creek, CA.) equipped with a flame ionization detector, quantified by comparing the peak area of the sample to a benzene standard prepared at the UBC Occupational Hygiene lab, and identified by the relative retention time. The detectable mass of benzene was 10 u.g, which corresponded to a limit of detection of about 0.5 ppm, depending on the volume of air sampled. 2.3 Methods of data analysis 2.3.1 Frequency distributions For each set of measurements, for each contaminant, the shape of the distribution curve was determined as either lognormal or normal. The choice of distribution was made by treating the data alternatively as normal or lognormal and by inspection of the histograms. The histogram, that better approximated the 3 0 distribution of the data, was chosen as the appropriate distribution. If the distribution could not be readily determined by inspection of the histograms then the distributions were examined using the Z-score histogram option in Statview (Abacus Concepts, Berkley, CA.) and the most appropriate distribution was chosen. 2.3.2 Censored Data For each data set the number of samples above the limit of detection was determined. Depending on whether f the data was lognormal or normally distributed, censored values were treated differently. For data sets distributed log normally, calculations of summary statistics was done assuming data below the LOD had value of LOD/V2 (Hornung and Reed 1990) and for normally distributed data, calculations were performed by assigning a value of LOD/2 to values below the limit of detection. 2.3.3 Data from direct reading instruments Particulate, HCN, benzene, and formaldehyde sampling methods measured the average concentration of contaminant over the sampling period. Direct reading instruments yield continuous data over the entire sampling period. Therefore, for CO2, CO, and S02, peak or maximum values were determined for each sampling period. The peak values from each sampling period were tabulated and summary statistics were performed on the peak values recorded over the sampling period. 2.3.5 Comparison of mean contaminant levels between sites Comparison of mean contaminant levels was conducted on several occasions, by using the non-parametric Mann-Whitney test. This test was used as the equivalent parametric test (T-test for comparison of two means) required the assumptions that the data were normally distributed and were from data sets roughly the same size. Though the T-test is robust, the last two assumptions (equal variance, and similar size data sets) could not be met, and the non-parametric test was used. 31 Chapter 3: Results The results are divided into sections to differentiate between each site sampled. In each section the results are tabulated to show the number of samples above the limit of detection, mimmum and maximum values, and the mean and standard deviation of the data sets. For lognormal distributions the geometric mean and geometric standard deviations are reported and for normal distributions the mean and standard deviations are reported. Individual values for each section can be found tabulated in appendix II. 3.1 Outdoor Pits: sampling at Fuel and Depth fires (T-pit and Round Tank) Observations of daily weather and burning conditions were recorded during each of the 9 burn session sampled. These observations can be found listed in appendix II. For formaldehyde, benzene, HCN, CO, and S0 2 less than 50% of the samples were above the limit of detection. Only particulate was recorded above the detection limits in > 50% of the samples. All data sets were lognormalfy distributed. These data are summarized in table 6. Table 6: Summary of combustion gases and volatile materials at the T-pit and Round tank Contaminant LOD # samples #of minimum maximum Geo. Geo. (ppm) >LOD samples value (ppm) value (ppm) Mean Stdev Formaldehyde 0.005 9 19 <0.005 ppm 0.13 0.02 1.9 Benzene 0.2 1 15 <0.2 ppm 0.24 0.15 1.2 Particulate (mg/m3) 0.03 11 15 <0.03 18.1 0.29 6.3 HCN 0.01 3 19 <0.02 ppm 1.0 0.01 2.2 CO 1 5 11 <1.0 ppm 19 2 3.4 S0 2 0.1 2 11 <0.1ppm 0.7 0.1 2.1 The average levels of contaminants at the outdoor fuel and depth fires were found to be much lower than the relevant time weighted exposure limits, established by the WCB. Of all samples at these sites, only particulate was detected at levels in excess of its WCB exposure limit of 10 mg/m3. Particulate was detected at a peak level of 18.1 mg/m3, but had a mean concentration of 0.29 mg/m3. In all instances S0 2 was detected, on three of the five instances CO was detected, they were brief peaks recorded soon after ignition of the fire. These chemicals quickly dissipated, and were not detected in the next 1 minute sampling interval. On two occasions (Oct 24 and Dec 5) CO was detected, the values 32 remained elevated for more than 1 minute, but on both occasions, dissipated below the LOD within 3 minutes. Recorded C0 2 levels fluctuated, however, they were always within 60 ppm of ambient levels (recorded earlier the same day) and never exceeded 450 ppm (data not included). 3.2 Outdoor Extinguisher Pad Observations of daily weather and burning conditions were recorded during each of the five burn session sampled. These observations can be found listed in appendix 2. Of all contaminants, only benzene and S0 2 had < 50% of the samples above the limit of detection. All other contaminants had > 50% of the samples above the limit of detection. Particulate and C0 2 were found to approximate normal distributions while formaldehyde, HCN and CO were lognormally distributed. These data are summarized in table 7. Table 7: Summary of combustion gases and volatile materials at the extinguisher pad Contaminant LOD # samples #of minimum maximum Geo. Geo. (ppm) >LOD samples value (ppm) value (ppm) Mean Stdev Formaldehyde 0.005 5 10 <0.005 0.082 0.011 3:8 Benzene 1.0 2 10 <0.2 1.3 0.9 1.2 Particulate (mg/m3) 0.03 10 10 <0.03 1169.7 *417 *325 HCN 0.01 8 10 <0.01 1.03 0.42 1.9 CO 1.0 8 10 <1.0 81 4 5.2 so 2 0.1 4 10 <0.1 0.7 0.1 2.5 C0 2 1 6 6 652 3190 *I873 *1147 *These distributions were normal and are expressed as the arithmetic mean and arithmetic stdev. Similar to the outdoor fuel and depth pits, except for particulate and benzene, the average levels of contaminants are well below the WCB exposure limits. As shown in table 7, the mean particulate level was about 40 times higher than its respective exposure limit and the mean benzene concentration was about twice as high as its exposure limit. In all instances that S0 2 was detected, it was a brief peak recorded after ignition of the fire and dissipated within 1 minute. CO on 4 of 8 occasions dissipated within one minute, but was also detected for up to 7 minutes on the longest occasion. C0 2 was found on all occasions sampled. 33 3.3 Outdoor Propane Fires Three sampling sessions were conducted at this site between September to December 1996. On no occasion were any detectable levels of any of the sampled contaminants measured. Limits of detection: CO 1 ppm, S0 2 0.1 ppm, C0 2 lppm, HCN 0.02 ppm. 3.4 Sampling Experiments at Indoor Sites: Burn building and ship mock-up 3.4.1 Background levels (environmental levels in the absence of burning) Over the course of the sampling regime HCN, benzene, particulate, formaldehyde, CO, C0 2, and S0 2 were recorded. All data sets were lognormally distributed. Of the contaminants measured only particulate and C0 2 were found at detectable levels in > 50% of the samples. These data are summarized in table 8. Table 8: Summary of background contaminants at the burn building and ship mock-up Contaminant LOD # samples #of minimum maximum Geo. Geo. (ppm) >LOD samples value (ppm) value (ppm) Mean Stdev Formaldehyde 0.005 4 10 <0.005 0.12 0.011 5.6 Benzene 0.2 0 10 <0.2 <0.2 0 0 Particulate (mg/m3) 0.09 10 15 <0.09 0.33 0.10 1.6 HCN 0.01 0 10 <0.01 <0.01 0 0 *co 2 1 0 4 343 367 0 0 CO 1 0 4 <1 <1 0 0 S0 2 0.1 0 4 <0.1 <0.1 0 0 •Natural background levels for C02 at the site was between 343-367 ppm. As can be observed from table 8, particulate was detected on 67% of occasions sampled, and of all other contaminants, only formaldehyde was also detected, and was detected on 40% of occasions. C0 2 levels measured during these periods were found to vary 10-15 ppm, over the 4 hour sampling period, but were considered normal fluctuations in ambient levels (data not included). 3.4.2 Unvented levels (worst case environmental levels after fires) On July 24 and October 10 sampling was carried out to determine unvented levels of levels of contaminants produced within the burn building immediately following a burning exercise. Because of anticipated high contaminant levels, temperature, dangerous conditions, and the need to use SCBA only two sampling sessions were conducted. Sampling was conducted only in the burn building and not ship mock-up. Over 34 the course of the sampling regime HCN, benzene, particulate, formaldehyde, CO, C0 2, and S0 2 was recorded. These data are summarized in table 9. Table 9: Summary of unvented combustion gases and volatile materials measured in the burn building immediately after fire suppression and before venting. Contaminant LOD # samples > # of samples minimum value maximum value LOD (ppm) (ppm) Formaldehyde 0.005 ppm 1 1 2.99 2.99 Benzene 0.3 ppm 1 1 0.4 0.4 HCN 0.01 ppm 2 2 2.04 3.5 Particulate 0.08 mg/m3 1 1 1.39 mg/m3 1.39 mg/m3 C0 2 1 ppm 2 2 3190 3456 CO lppm 2 2 544 859 so2 0.1 ppm 2 2 9.6 52.1 No summary statistics are presented because only a maximum of two measurements were made for any contaminant. Unfortunately, the second samples for formaldehyde, benzene, and particulate were lost or damaged. For combustion gases (C02, CO, and S02) the levels stated are the maximum value recorded over the 30 minute sampling duration. Measurements of 0 2 depletion during the non vented sampling sessions was also done. These data, though not tabulated or reported, showed that oxygen levels remained high when the building was not vented. The levels dropped from 21.8% to 19.6% (the lowest value of oxygen measured) during the July period and the oxygen levels dropped from 21.9% to 20.0% during the October period. 3.4.3 Environmental levels immediately following venting Sampling conducted in the ship mock-up and burn building immediately following venting after a burning exercise revealed that all contaminants except benzene were detected in >50% of occasions sampled. Formaldehyde, particulate, benzene, and HCN data sets were lognormally distributed whereas the CO, C0 2, and S0 2 data sets were found to be normally distributed. These data are summarized in table 10. 3 5 Table 10: Summary of combustion gases and volatile materials measured immediately following venting at the Ship mock-up and Burn building Contaminant LOD (ppm) # samples >LOD #of samples minimum value (ppm) maximum value (ppm) Geo. Mean Geo. Stdev Formaldehyde 0.005 21 21 0.069 2.00 0.372 2.4 Benzene 0.7 3 8 <0.7 1.9 0.8 0.58 HCN 0.01 16 20 <0.01 2.01 0.1 5.8 Particulate (mg/m3) •0.1 16 20 <0.1 2.3 0.2 3.7 C0 2 1 16 16 401 2456 •925 *558 CO 1 16 16 5 137 *36 *34 so 2 0.1 11 16 <0.1 8.4 *1.5 *2.1 *These distributions were normal and are expressed as the arithmetic mean and arithmetic stdev. The mean levels of formaldehyde, CO, and benzene, following venting, were higher than their respective exposure limits, while all other mean contaminant levels were below their respective exposure limits. The mean particulate level following venting was substantially reduced compared to the non-vented levels, though the maximum particulate level measured following venting was greater than the non-vented level. 3.4.4 Comparison of environmental levels immediately following venting and unvented levels (worst case environmental levels after fires) in both the ship mock-up and burn building To illustrate the effectiveness of venting, the maximum non-vented (worst case environmental levels) were compared to the mean environmental contaminant levels immediately following venting. The percentage decrease in levels was calculated and included to show the effectiveness of venting. These data are presented as table 11. Table 11: Comparison of unvented levels and mean levels immediately following venting, along with the % decrease in contaminant levels, in the burn building and ship mock-up Contaminant Maximum unvented level Mean level immediately following venting % decrease in levels following venting HCN (ppm) 3.5 0.095 3700 S0 2 (ppm) 52.1 1.5 3500 CO (ppm) 859 36 2400 Formaldehyde (ppm) 2.99 0.372 800 Particulate (mg/m3) 1.4 0.20 700 C02(ppm) 3456 925 500 Benzene (ppm) 0.4 <LOD not determined 36 3.4.5 Assessment of time for environmental levels (after venting) to dilute to insignificant levels Sampling in the ship mock-up and burn building following venting revealed that all contaminants (formaldehyde, particulate, HCN, CO, C0 2, and S02) were detected on >50% of occasions sampled. Therefore long-term sampling was conducted to determine the time for these contaminants to dilute to non-detectable levels. Data sets for formaldehyde, particulate, and HCN were found to be lognormally distributed, whereas CO, C0 2, and S0 2 data sets were found to be normally distributed. Since there were no differences in the trends between the two structures the data have been combined and are summarized in table 12. 37 Table 12: Summary of contaminants measured for 3 hours following venting in the ship-mock up and burn building Formaldehyde (ppm) Summary of Maximum Measured Values after venting at time ... . 0-30 min 30-60 min 60-120 min 120-180 min LOD 0.001 0.001 0.0005 0.0005 # above (LOD) 21 of21 16 of 16 16 of 16 14 of 14 % of time detected > LOD 100 100 100 100 Min 0.069 0.04 0.04 0.03 Max 2.00 2.43 0.898 0.51 Median 0.37 0.179 0.135 0.067 Geo. mean 0 3": 0.199 0 146 0 077 95% Confidence interval 0.001-0.9 0.001-1.4 0.001 -1.2 0.001-1.1 Geo. Stdev 2.4 2.7 2.4 2.2 Particulate (mg/m1) Summary of Maximum Measured Values after venting at time ... . 0-30 min 30-60 min 60-120 min 120-180 min LOD 0.09 0.09 0.0045 0.0045 # above (LOD) 14 of 19 9 of 16 7 of 16 4 of 12 % of time detected > LOD 74 56 44 33 Min <0.09 <0.09 <0.0045 <0.0045 Max 2.4 0.64 0.12 0.12 Median 0.16 0.08 0.06 0.06 Geo. mean 0.20 0 07 0 06 0.06 95% Confidence interval 0.09-1.9 0.09 - 1.3 0.09-1.5 0.09-1.7 Geo. Stdev 3.7 2.0 2.0 1.5 HCN (ppm) Summary of Maximum Measured Values after venting at time ... . 0-30 min 30-60 min 60-120 min 120-180 min LOD 0.01 0.01 0.005 0.005 # above (LOD) 15 of 19 11 of 16 9 of 16 5 of 15 % of time detected > LOD 79 69 56 33 Min <0.01 <0.01 <0.005 <0.005 Max 2.00 1.72 0.19 0.08 Median 0.11 0.08 0.01 0.01 Geo. mean 0.09 0.05 0.022 0.01 95% Confidence interval 0.01-3.5 0.01-2.6 0.01-2.2 0.01-2.3 Geo. Stdev 5.8 3.9 3.6 2.5 C O (ppm) Summary of Maximum Measured Values after venting at time ... . 0-30 min 30-60 min 60-120 min 120-180 min LOD 1 1 1 1 # above (LOD) 16 of 19 13 of 18 7 of 17 7 of 16 % of time detected > LOD 84 72 41 44 Min <1 <1 <1 <1 Max 137 35 26 6 Median 17 3 <1 <1 Mean 36 10 5 I 95% Confidence interval 22-50 5-15 0-10 0-6 Stdev 34 9.7 6.9 2.1 co2 (ppm) Summary of Maximum Measured Values after venting at time ... . 0-30 min 30-60 min 60-120 min 120-180 min LOD 1 1 1 1 # above (LOD) 14 of 14 14 of 14 12 of 12 8 of 8 % of time detected > LOD 100 100 100 100 Min 402 323 323 359 Max 2134 767 602 496 Median 721 480 396 374 Mean •Ml) 521 424 398 95% Confidence interval 653- 1167 450 - 592 381-467 366-430 Stdev 547 151 84 49 so2 (ppm) Summary of Maximum Measured Values after venting at time ... . 0-30 min 30-60 min 60-120 min 120-180 min LOD 0.1 0.1 0.1 0.1 # above (LOD) 11 of 16 8 of 15 3 of 14 2 of 13 % of time detected > LOD 69 53 21 15 Min <0.1 <0.1 <0.1 <0.1 Max 8.4 2.2 1.0 0.9 Median 0.9 0.1 <0.1 <0.1 Mean 1 5 0 5 0.1 0.1 95% Confidence interval 0.4-2.6 0-1 0-0.4 0-1.3 Stdev 2.1 0.76 0.23 0.56 38 Sampling for benzene was conducted only in the ship mock-up and not in the burn building and was conducted on 5 sessions. The data was found to be normally distributed. Summary statistics for these data are included as table 13. Table 13: Summary of benzene measured for 3 hours following venting in the ship mock-up Benzene Summary of Maximum Measurer. Values after venting at time .... 0-30 min 30-60 min 60-120 min 120-180 min LOD 0.7 0.7 0.7 0.7 # above (LOD) 3 Of 5 3 of 5 . 3 of 5 2 of 5 % of time detected > LOD 60 60 60 40 Min <0.7 <0.7 <0.7 <0.7 Max 1.9 1.2 0.8 0.8 Median 1.2 1.1 <0.7 <0.7 Mean 1.0 0.8 0.5 0.5 95% Confidence interval 0.1-1.9 0.2-1.4 0.2-0.8 0.1-0.9 Stdev 0.64 0.44 0.19 0.17 As can be seen in tables 12 and 13, all contaminants decreased in concentration over the sampling period. Plotting the mean concentration of each contaminant against the midpoint of the time interval during which it was collected made washout curves. Graphically represented these results have been included as figure 3. Note that CO is graphed on the secondary y-axis. CO2 has not been included, because of difficulties with the scale, however it follows the same trend as the other contaminants. 39 Figure 3 Mean contaminant level vs. time since active venting in the burn building and ship mock-up 1.6 T 1.4 t 1.2 E. S 3 O t a 1 T3 c (0 ,§I ' c c a> a> t> N c c at T3 at (0 £ z o O <0 0.8 0.6 + 0.4 4-0.2 — -© — S02 — h - HCN X — Formaldehyde - - o - - Particulate — -A- - Benzene — -n — CO t 40 35 30 25 20 + 15 A 10 E Q. a O u c o c u u c o o 150 Time since active venting (min) Note: C O is on a secondary axis 40 Deteirnination of the time until contaminants reached insignificant levels was done by plotting washout curves. Confidence intervals for each point on the washout curves were calculated, and the time when the upper confidence interval intersected the contaminant level corresponding to 54 the exposure limit was taken as the time period when contaminant levels were considered ^significant. These data are summarized in table 14. Table 14: Time period for contaminants following venting, in the burn building and ship mock-up. to diffuse below Vi the exposure limit (the action limit). Contaminant Time to diffuse below Vi the exposure limit (minutes) Vi Exposure limit 9 5 % Confidence interval C0 2 <15 2500 ppm 400- 1100 HCN <15 2.5 ppm 0.1-0.5 Particulate <15 5 mg/m3 0.1-0.8 S02 -40 lppm 0.2-1 CO -60 12.5 ppm 3 - 12 Formaldehyde -140 0.15 ppm 0.1-0.15 Benzene -380 0.25 ppm 0.1-0.25 *Benzene results are for the ship mock-up only For C0 2, HCN, and particulate the upper confidence interval for the first samples collected 30 minutes immediately following venting were below the action limit. The results are expressed as <15 minutes because this represents the midpoint of the time interval over which the first 30 minute sample was collected. Another trend, encountered on a few occasions, is a slight increase in the concentration of formaldehyde and HCN following venting. To illustrate this trend figure 4 is included. Figure 4 shows a representative curve that illustrates the increase in concentration followed by a decline. 41 Figure 4 Concentration of Formaldehyde and HCN in the burn building following active venting 42 3.4.6 Comparison of Contaminant levels in the Burn building and Ship mock-up For all contaminants measured in the ship mock-up and burn building, immediately following venting, comparison of the means of the values was conducted. These data are summarized in table 15. Table 15: Comparison of mean contaminant levels in the ship and burn building immediately following venting Contaminant Mean Mean Mann-Whitney result (Ship) (Burn building) Formaldehyde 0.159 0.485 p<0.05 Particulate 0.279 0.16 p<0.05 HCN 0.080 0.10 p>0.05 CO 44.6 32.5 p<0.05 so 2 3.7 0.957 p<0.05 co 2 1900 614 p<0.05 Benzene 1.37 no data not done As can be observed from table 15, only with HCN were the results not significant at p<0.05, thus the mean levels of HCN in the two structures were not statistically different, whereas the mean levels were statistically different for all other contaminants. For formaldehyde the mean contaminant level was greater in the burn building, and for particulate, CO, S02, and C0 2 the mean levels were greater in the ship mock-up. 3.4.7 Propane fires in the Burn Building Sampling was conducted on five separate occasions. On each occasion no detectable levels of particulate, HCN, or S02were recorded. Of all the contaminants measured only formaldehyde was detected consistently in the burn building and was log-normally distributed. The results of the formaldehyde levels are summarized below as table 16. Table 16: Formaldehyde levels in the burn building propane side Contaminant LOD # samples #of minimum maximum Geo. Geo. (ppm) >LOD samples value (ppm) value (ppm) Mean Stdev Formaldehyde 0.001 5 5 0.03 0.35 0.069 2.6 CO was detected on one occasion only, and was a brief peak of 5 ppm that dissipated after 1 minute. C0 2 levels were detected on each occasion but were always within 50 ppm of ambient levels that were recorded earlier on the same day (data not included). 43 Chapter 4: Discussions 4.1 T-pit and Round Tank Over the entire sampling regime the mean level of contaminants recorded at the outdoor sites were found to be much lower than the relevant 8 hour exposure limits, established by regulating bodies such as the WCB. Measured levels at these sites were: CO mean 2 ppm maximum 19, S02mean 0.1 ppm maximum 0.7, particulate mean 0.29 mg/m3 maximum 18.1, HCN mean 0.01 ppm maximum 1.03, formaldehyde mean 0.02 ppm maximum 0.13, benzene 0.24 ppm (one sample only). The rest of the summary statistics are summarized in table 6. The mean exposures of most contaminants were generally only a fraction of their respective WCB exposure limits. To contrast measured levels with the WCB regulated values, table 17 has been included. Table 17: Comparison of contaminants measured at the T-pit and Round tank with WCB exposure limits. CO so2 HCN benzene formaldehyde particulate Mean level 2 0.1 0.01 0.15 0.02 0.29 WCB PC 25 2 5 0.5 0.3 10 Peak level 19 0.7 1.03 0.24 0.13 18.1 WCB STEL 100 5 5 2.5 1 none The mean concentration of benzene was about 1/3 of its exposure limit, the closest any contaminant came to approaching its exposure limit. To address the possibility that the effects of a mixture may be additive, the AE was calculated (as discussed in section 1.2.2). The AE value for the mixture was calculated to be 0.53, which states that the effective mixture would be considered to be only 53% of an additive exposure limit. This supports the results of Leahey et. al. (1993) who found that exposure from outdoor burning did not exceed regulated exposure limits. Given that the levels measured were low compared to WCB limits, health risks should be rninimal at this site. 4.1.1 Micro-climate It is unquestionable these fires generate vast quantities of contaminants, as evidenced by the copious amounts of black smoke produced. However, the nature of the fire in the T-pit and round tank, generally 44 large fast burning fires, creates what is termed their own micro-climate i.e. the fire is of sufficient size and intensity that it creates its own burning conditions, or weather. These conditions are observable after the fire is lit and burning. Air rushes into the fire, noticeable by the loud sound, the combustion process intensifies, and the super heated gases are vented upwards in a massive plume of smoke. The rush of air into the fire, and upward venting of smoke, creates conditions (micro-climate) when smoke exposure may not occur. The data is consistent with the theory that most all contaminants are trapped wthin the smoke plume, as it rushes upwards. Consequently contaminants should be vented high above the facility to be deposited downwind at a considerable distance from the facility. At the facility, normal prevailing wind patterns are towards the east, and under circumstances when the micro-climate is occurring, the plume and contaminants are vented high into the air and should be dispersed distant from the facility. Study by Atlas et. al. (1985) who also looked at large outdoor hydrocarbon fuel fires, reported that atmospheric dilution and dispersion occurred to such an extent, that by the time the smoke plume settled to the ground, contaminant levels were not detectable! The FSTC may have differences in geography and wind conditions than the site studied by Atlas, however, one could presume a similar situation may occur at the FSTC, and atmospheric dilution should reduce or eliminate harmful exposure where the plume settles to the ground. Similar work with controlled petroleum hydrocarbon burning, by Booher and Janke (1997), reported that ground level measurements of poly aromatic hydrocarbons and volatile organic hydrocarbons were very low, if at all detectable. Thus there may be little, if any, risk to health when these fires are venting the plume high into the air. Previous study at the facility by van Netten (1997) has shown, under worst case scenarios, elevated CO levels, between 14-30 ppm, at various locations around the facility, van Netten also reported that gravitation or settling of contaminants might occur distant from the outdoor pits as benzene was measured at sites such as the office, parking lot, and between the maintenance shed and pump house. In all, these locations were further distant from the round tank and T-pit than the areas sampled in this study, and burning in van Netten's study was conducted at several sites at once to simulate worst case levels. It is 45 possible that these levels were a result of drifting smoke that may occur after the micro-climate from fires in the round tank and T-pit were extinguished, or likely from the greater fire activity used to generate a worst case scenario. van Netten's data has shown that worst case situations have the potential to create exposure, and the maximum level of CO, at 30 ppm, is above the WCB 8-hour exposure limit of 25 ppm. Since it is known that exposure can occur, it is useful to examine situations when contaminants were detected in this study. It was found that fire size, and stage of the fire suppression may affect contaminant dispersal. Fire size may affect exposure, and was attributed to smaller fires being unable to create as large a micro-climate as larger fires, if at all, due to less intense burning and a lesser volume of air drawn into the fire. Without the micro-climate, and rapid upwards venting of smoke, a situation can be created where the contaminants in the plume would be susceptible to wind dispersal. Observations of fire size were made, but there was no method to rate fires other than subjectively (staff at the site were asked to rate the fire size) as small, medium, or large. Given to the subjective nature of the observations no definitive conclusions were attempted. However, it was interesting that of the 41 samples that were above the limit of detection (out of a total of 109 samples), 39 occurred when fires were rated as medium or small. Another factor, that may effect exposure, is the stage of fire suppression. After the main body of the fire is suppressed the micro-climate is not present, due too insufficient burning intensity. At this time the smoke plume is susceptible to dispersal by wind or the spraying action of the hoses, as the firefighters continue spraying water into the fire pits to suppress fuel vapour and prevent re-ignition. Without the micro-climate wind may blow smoke and contaminants around or water sprayed from the hose can push contaminants in the direction of the stream travel. It was observed on occasion that steam, residual smoke, and possibly contaminants were pushed towards the sampling areas, by the stream of water from the hoses, after the micro-climate was eliminated. On these occasions CO, measured by real time detectors, was recorded, and it was during this situation when the maximum level of CO was measured (19 ppm). 46 4.1.2 CO and SQ2 exposure (examination of real time recordings) At the outdoor fuel and depth pits these gases were detected a total of seven times. Unfortunately, the lack of data made detailed analysis difficult, however some interesting correlation could be made. Examination of the data revealed that on all five occasions when CO, and 1 of 2 times when S02, were detected, it was a cloudy overcast day. Intuitively one would consider wet rainy days, which are often windier, as ideal conditions for conterninant dispersal, especially if the micro-climate is not present. Looking deeper at weather conditions, it was useful to exarnine days that were warm, sunny, and had calm winds. Three days fit this description, they were Sept 12, Sept 26, and Oct 31. On these days, S0 2 was the only gas detected, and was detected once on Sept 26. The time that the recorded levels occurred it was noted to be at or near the time the fire was completely suppressed. This may be indicative that the recorded contaminants may have been dispersed by water spray after the micro-climate was extinguished. To further emphasize this point one could add up the total number of times CO and S0 2 was detected, which was 7, it was found that overcast weather accounted for 6 of the times that these gases were detected. The 7th time one of these gases was detected it occurred after the micro-climate was extinguished and may have resulted from hose spray pushing contaminants around. 4.1.3 HCN, formaldehyde, particulate and benzene exposure (examination of contaminants not measured as real time data) Formaldehyde, HCN, particulate and benzene samples represent the average exposure over the sampling period. The amount of contaminant collected represents the total collected during the sampling period (for example, on June 28 — 5.1 ug of HCN was collected which equaled an average air concentration of 0.05 ppm over the 24 minutes that the sample was collected). It is possible that the entire mass of contaminant was collected when a smoke plume, or plumes, drifted near the equipment and not over the entire sample duration. If so, the sample would represent a brief peak exposure where the HCN concentration was much greater than 0.05 ppm, but when averaged over the entire sampling period the value is 0.05 ppm. The sampling method can not differentiate between the possibilities that the recorded value may represent the 47 level of contaminant recorded over the entire sampling period, or an averaged value from a few brief peak exposures. Analyzing the HCN, formaldehyde, and benzene samples that were above the limit of detection, the data do not appear to follow a clear exposure pattern with respect to weather conditions. These contaminants were detected 13 times. Almost half of these times (5/13) it was during clear sunny weather, and did not show the same pattern as CO, which was only detected on rainy overcast days. In 11 of the 13 instances levels were recorded the fire size was observed to be small or moderate. Urrfortunately, it cannot be ascertained from these data if levels occurred after the fire was suppressed, or if levels may have been due to the hose spray pushing contaminants towards the recording equipment. Nov. 21 was the only day when contaminants were recorded, and fire sizes were large. On this day both formaldehyde and benzene were detected. Nov. 21 was the only day benzene was collected at any of the outdoor sampling sites. Unfortunately, no CO or S0 2 was collected by the gas detectors, which would identify the minute that the contaminant was collected. Thus it could not be determined if the benzene and formaldehyde were possibly collected after suppression of the fire, when the micro-climate is eliminated, or if these levels could be due to the action of hose spray. Of concern may be the peak level of particulate measured (18.1 mg/m3). It is known that emissions from hydrocarbon fires can release soot (Booher and Janke 1997), which is known to contain PAH's, many of which are carcinogens. This exposure may represent a health risk if repeated or prolonged exposure occurs. Mean particulate levels (0.29 mg/m3), tend to indicate that this exposure level occurs infrequently and may be an exception rather than the norm. Additionally, this level was recorded on the same occasion that the peak CO level (19 ppm) was recorded, which occurred as hose spray pushed contaminants toward the sampling area. It is possible that the particulate levels were a result of hose spray moving contaminants, which would allow peak levels to be anticipated, and allow exposure to be avoided: 48 Although the correlation of exposure related to weather, size of fire, stage of fire suppression, and micro-climate may not explain all instances of exposure, they do represent many factors that might be used to predict periods of potential exposure. It is arguable that when site staff recognize days when these factors occur, they can expect a much greater chance for exposure to occur, and accordingly, take precautions. One can easily observe situations when smoke drifts towards an area or when water spray may push contaminants around, and can simply move away from the smoke. Thus they can greatly reduce unnecessary exposure and alleviate health concerns. Ultimately it should be pointed out that care should be taken when interpreting these results. First, some are based on speculation about the relationship between times that contaminants were recorded and personal observations taken while sampling. Though all observations were made with the best of scientific intent they can be inaccurate, and very importantly speculation about the relationship does not mean causation. Additionally, it must be pointed out that there is more involved with exposure then was explained by the prior speculation about the relationship. Further in depth analysis and study would be required to fully characterize all situations when exposure may occur. 4.2 Extinguisher Pad Mean particulate and benzene levels were found to exceed their WCB exposure limit at this site, while the remaining mean levels of contaminants were much lower than the WCB 8-hour exposure limit. Measured levels were: CO mean 4 ppm maximum 81, S02 mean 0.1 ppm maximum 0.7, HCN mean 0.42 ppm maximum 1.03, formaldehyde mean 0.01 lmaximum 0.082, C0 2 mean 1873 ppm maximum 3190, particulate mean 417 mg/m3 maximum 1167.7, and benzene mean 0.9 peak ppm 1.29. The rest of the summary statistics are present in table 7. To contrast measured levels with their respective WCB exposure limits table 18 has been included. 49 Table 18: Comparison of contaminants measured at the extinguisher pad with WCB exposure limits. CO so 2 HCN benzene formaldehyde particulate c o 2 Mean level 4 0.1 0.42 *0.9 0.011 417 1873 WCB PC 25 2 5 0.5 0.3 **10 5000 Peak level 81 0.7 1.03 1.29 0.082 1167.7 3190 STEL 100 5 5 2.5 1 none 15,000 •Following censoring of the data the mean level was 0.9 ppm while the LOD was 1.0 ppm **The exposure limit for respirable particulate is 3 mg/m3 Particulate levels are exceptionally high and may constitute a serious exposure hazard, as levels were more than 40 times the exposure limit. Benzene, with a mean level of 0.9 ppm is almost 2 times the exposure limit of 0.5 ppm, and may also constitute an exposure hazard. Additionally, it was felt this level might underestimate the amount of benzene present, as the high limit of detection (1 ppm) may have precluded more detectable measurements. Previous work by Van Netten (1997) had recorded benzene at the facility during worst case levels of outdoor burning. It is believed that benzene may be present in most fire emissions, but at levels below the detection limit. It is recommended that if further study is to be done, a method with a lower limit of detection be used to measure benzene. The mean levels of the remaining contaminants were present at levels only a fraction of their exposure limits. To address the possibility that the effects of a mixture may be additive, the AE was calculated (section 1.2.2). The AE value for the mixture of conlaminants, excluding particulate and benzene (which already exceeded their respective exposure limits), was calculated to be 0.71, which states that the effective mixture, of the remaining conlaminants, is considered to be only 71% of an additive exposure limit. 4.2.1 Dry chemical clouds Fire suppression at this site creates large white clouds of dry chemical powder that slowly float around the facility, before eventually dissipating. These clouds are made up of sodium bicarbonate (NaHC03), the ingredient in the extinguisher, and may also have contaminants entrapped within. This may occur as the extinguishers use pressurized C0 2 as a propellant, to expel the NaHC03, and the force of the C0 2 can push smoke as well as contaminants in the direction the extinguisher was sprayed. This may result in capturing contaminants within the floating cloud of dry chemical. HCN, formaldehyde, CO, S02, and C0 2 (table 7) 50 were recorded, within these clouds, on the majority of occasions. Formaldehyde was detected on 5 of 10 occasions, HCN on 8 of 10 occasions, CO on 8 of 10 occasions, S0 2 on 4 of 10 occasions, and C0 2 on 6 of 6 occasions. That these contaminants were detected, within the cloud, on most occasions supports the possibility that contaminants may become entrapped within the cloud, creating an increased exposure hazard. Mean levels of contaminants, except particulate and benzene, associated with the cloud were low compared to their exposure limits (table 18). A few peak levels recorded may be noteworthy to highlight. CO had a peak level of 81, which is nearing the WCB short term exposure limit (100 ppm). The peak level for C0 2 was 3190 ppm, but may have exceeded this level, which is not unexpected as C0 2 is used as a propellant in the extinguishers. On the occasions that levels were 3190 ppm it is believed that the maximum detectable limit of the sensor was overloaded, because the values displayed on the screen did not drop, even after the detector was moved from the extinguisher pad. Thus the level recorded (3190 ppm) was suspect, but believed to be lower than the true levels: Unfortunately, it can not be proven. This phenomenon occurred only on two occasions and the highest level detected otherwise was 2160 ppm. Nonetheless, C0 2 appears to constitute a major component of the cloud. From observation of the burning there is no evidence to suggest the micro-climate will be present at these fires. There is no noticeable inward rush of wind, as the fires are too small, venting is not upwards in a massive plume, and often appears to be susceptible to dispersion by wind. There were too few samples to suggest how weather may affect contaminant dispersal, but it is felt the lack of a micro-climate may make contaminant dispersal at this site highly susceptible to wind, or environmental factors. Additionally, and importantly, contaminant dispersal may occur from the pressurized C0 2 in the extinguishers, which can entrap contaminants and push them in the direction the extinguisher was discharged. It is possible the levels of CO, reported by van Netten (1997), may have been created by fires in these pits, and occurred from smoke drifting from these smaller fires. 51 Readings collected at this site were taken on the extinguisher pad, generally within 7 meters of the pits, and were all sampled on the down side of the sprayed dry chemical. Thus they may represent worst case exposure from the cloud. It is expected that as the particulate settles out of the visible cloud, as evidenced by the diminishing density as they drift from the pit, that the exposure hazard will decrease. Unfortunately, no data was collected at a distant from the pits as one could not accurately predict the direction of drift, and it would be difficult to try to safely place the sampling equipment in the path of the drifting cloud. Nonetheless, the data at hand shows that the high levels of particulate, and benzene measured in these clouds may pose health risks. The sodium bicarbonate, used as the dry chemical extinguishing agent, is listed as a nuisance dust, and is not listed to cause health effects (Fisher Scientific MSDS). However, mean levels of 417 mg/m3 may cause irritation of the upper respiratory tract, reflex cough, tightness in the chest, and may coat the lining of the respiratory system (Biological exposure index). Benzene, at 0.9 ppm, should have few if any acute effects, but consequences such as bone marrow toxicity and cancer following long term or repeated exposure may pose a hazard (Biological exposure index). Thus, it would be prudent not to contact the cloud as it drifts around the facility and settles out, and fortunately, the clouds of dry chemical powder, with entrapped chemicals, are readily seen and easily avoided. 4.3 Propane Pad The potential for exposure appears minimal at this site. Sampling was conducted on three occasions and no detectable levels of SO2, C0 2, or CO were measured. The nature of the fire is exceptionally intense, and the propane burns very quickly and cleanly as evidenced by the lack of readily visible smoke or emissions. Venting of the fumes occurs upwards due to the intense fast burning, and appears unaffected by weather or water sprayed at the fire. Other factors observed to influence contaminant dispersal at the T-pit and round tank, such as size of the fire and stage of suppression do not appear to affect dispersal at this site. Fire size and stage of suppression are not relevant, as closing the propane fuel supply completes suppression, and 52 the fire is instantaneously suppressed, leaving no opportunity for the fire to become small enough to be influenced by weather or sprayed water. It would appear that little exposure potential exists at this site. 4.4 Indoor sites - Burn building and Ship mock-up This section is divided into sections to help present the data in a logical manner. The sections are background levels, unvented levels, levels immediately following venting, and detennining the time for residual contaminants to dilute to insignificant levels. The sections pertaining to levels immediately following venting and dilution time periods are again subdivided into sections that describe the burn building and ship mock-up combined, specific observations in the burn building, and specific observations in the ship. 4.4.1 Background levels (environmental levels in the absence of burning) From the data (table 8) it can be seen that CO, CO2, SO2, benzene, and HCN were not present in detectable background levels, in the burn building or ship mock-up, under the conditions described within the methods. Particulate and formaldehyde are present in detectable background levels. Particulate was detected in 10 of 15 samples from the burn building and ship mock-up, with a mean level of 0.10 mg/m3 and maximum level of 0.33 mg/m3. Formaldehyde was detected in the burn building only, with a mean level of 0.011 ppm and maximum of 0.12 ppm. The source of the particulate and formaldehyde levels is unknown. Particulate levels may be elevated due to natural dust, pollen, bacteria, and fungi in the air, or from soot and dust that is disturbed while walking through the structures. Formaldehyde levels may be present due to off-gassing from the soot encrusted concrete walls, or may be residual ambient levels from previous burning. Intuitively one would consider that the time since any previous burning in either structure, which was at least 120 hours (5 days), should be sufficient time for dilution of residual contaminants, into the atmosphere, to be complete. Therefore, it is more likely that these levels are a result of off-gassing in the burn building. Earlier study by Guidotti and Clough (1992), the WCB (report 1992), and Dyer (1976) has described the concrete sink effect, which is a description of how concrete can absorbed and later release contaminants. It 53 is postulated that formaldehyde, generated by the wood fires, is absorbed into the hot concrete of the burn building, and is later released, or off-gassed. Release may occur as a slow steady process or under certain circumstances. Previous study by Van Netten et. al. (1989) reported formaldehyde release from building materials (tile, board, brick), that release may occur for years from Swedish floor finishes Van Netten (1983), and that release is related to the ambient temperature and humidity of the surroundings Van Netten (1989). Thus, formaldehyde may be released from the walls of the burn building, following firefighting, as the rooms are hot and humid from the fire and sprayed water. Release may also occur later, without fire activity, during times when the temperature in the building rises, such as warming by sunshine. If so the cement walls and refractory material in the burn building may release formaldehyde for long periods of time following burning, or at other times when no burning has occurred, and thus be responsible for the background levels described above. If these background levels do represent off-gassing, apparently there exists the potential of low level formaldehyde exposure, at any time someone enters the building. The measured levels were low, about 1/25 of the exposure limit, which are below levels documented to cause health effects. However, the data cannot ascertain when levels may be present, under what conditions they may occur, or for how long they may last, but one should be aware that they do exist. 4.4.2 Unvented Levels (worst case environmental levels after fires) Inspection of the data collected when the burn building was unvented (table 9), shows that maximum values of CO was 859 ppm, S0 2 52.1 ppm, C0 2 3456 ppm, HCN 3.5 ppm, formaldehyde 2.99 ppm, particulate 1.39 mg/m3, and benzene was 0.442 ppm. At these levels S02, CO, and formaldehyde are documented to cause health effects, the atmosphere may be classified as dangerous, and present a substantial risk to the health and well being of a worker unprotected by SCBA. One may conclude that SCBA is required to enter the structure before venting, a prudent practice currently employed at the facility. No data is available on levels of unvented gases inside the ship mock-up, but should be considered as hazardous, and precautions should be taken to protect staff health and safety. 54 Of all contaminants sampled, S0 2 and CO appear to represent the greatest threat to health, as both of these contaminants exceeded their respective short term exposure limits and approached the immediately dangerous to life and health levels. To the unprotected person the maximum level of S0 2 (52.1 ppm) can severely irritate the respiratory tract, eyes, and mucous membranes and adversely affect respiration (Biological exposure index for SO2 - ACGIH). At the maximum level CO was measured (859 ppm), and assuming a moderate work load, which would be an underestimate for the demands of firefighting, one could expect blood COHb levels to reach 3.5% in under 10 minutes. The eight hour exposure limit for CO was established at 25 ppm to keep the formation of carboxyhemoglobin (COHb) at or below 3.5% (ACGIH 1991). Though even 3.5% COHb may not be protective, as there are indications that COHb levels as low as 0.6% may aggravate symptoms of myocardial ischemia and angina, and that levels between 2-5% may affect psychomotor tasks (Biological exposure index). Levels of CO, at 859 ppm, may also cause unconsciousness and collapse in less than 1 hour (Biological exposure index, Committee on fire toxicology 1986). Therefore, there is the potential for adverse health effects to occur within minutes if someone, without SCBA, was to enter the structure. One other contaminant that may pose a health hazard is formaldehyde. The maximum level of 2.99 ppm exceeds the new exposure limit for this chemical (0.3 ppm) by 10 fold. At 2.99 ppm formaldehyde is documented to cause irritation, sore eyes, sore throat, and possible disturbed sleep (Biological exposure index). Additionally, formaldehyde is linked to chronic effects such as mutagenicity, teratogenicity, and sensitization. Formaldehyde has been classified as a possible human carcinogen by the International Agency for Research on Cancer (IARC), and this level may reflect a dangerous level of exposure. It should be mentioned that these levels might underestimate the true unvented contaminant levels in the burn building. Sampling was conducted in a hallway connected to the room where the fire had been burning. This was done as sampling in the room where the fire had burned could damage the equipment due to the residual heat. It was hoped levels in the hallway were similar to levels in the room with the fire, but 55 are not expected to be as high. Additionally, sampling was conducted after fires had been suppressed, and some venting of contaminants will generally occur as the firefighters perform their duties. Since the levels measured, underestimated or not, could create health effects within minutes, it would be a good practice to ensure that SCBA use remain mandatory, and compliance be monitored. 4.4.3 Environmental levels immediately following venting 4.4.3.1 Combined results (Burn building and Ship) Examination of levels immediately following venting (table 10) shows substantially reduced levels of contaminants: mean levels were; CO 36 ppm, C0 2 925 ppm, S0 2 1.5 ppm, formaldehyde 0.372 ppm, particulate 0.196 mg/m3, HCN 0.1 ppm, and benzene 0.8 ppm (benzene results from the ship mock-up only). When compared to the unvented level, all contaminant levels decreased significantly in concentration (table 11), and the percentage decrease in concentrations ranged between 500% to 3864%. Following venting, CO with a mean level 36 ppm exceeded the current WCB exposure limit, and formaldehyde (0.372 ppm) and benzene (0.8 ppm) were both in excess of the exposure limits adopted in April of 1998. S0 2 exceeded the action limit, whereas, HCN, C0 2, and particulate had mean values below their respective 8-hour exposure limits. Peak levels of CO, S02, and formaldehyde also exceeded the WCB short term exposure limits. To contrast measured levels with the WCB regulated values, table 19 has been included. Table 19: Comparison of environmental levels immediately following venting at the Burn building and Ship mock-up with WCB exposure limits. CO so2 HCN benzene formaldehyde particulate co2 Mean level 36 1.5 0.1 *0.8 0.372 0.2 925 WCB PC 25 2 5 0.5 0.3 10 5000 Peak levels 137 8.4 2.0 1.29 2.43 2.4 2134 STEL 100 5 5 2.5 1 none 15,000 * Benzene samples are from the ship only Exposure to these chemicals while in excess of their exposure limit may lead to adverse health effects following prolonged or repeated exposure. Mean levels of CO, at 36 ppm, can result in 3.5% COHb in about 200 minutes (assuming a moderate workload) after which health effects may occur (Biological 56 exposure index, Ginsberg 1995). At 0.372 ppm formaldehyde has produced eye, nose, and throat irritation in a small percentage of people (Biological exposure index, McGuire et. al. 1992). Benzene, at 0.8 ppm may not produce immediate acute effects, however, the exposure limit was lowered to 0.5 ppm, in part, due to chronic health effects and toxic action on bone marrow and blood cells (Biological exposure index). To address the possibility of health effects from exposure to the remaining contaminants, that did not exceed their exposure limit, the AE was calculated (section 1.2.2). The AE value for the mixture of particulate, S02, HCN, and C0 2 was calculated to be 0.98, which states that the effective mixture, of the retraining contaminants, reached 98% of a new exposure limit, derived for the mixture. Due to the mean levels of CO, benzene, and formaldehyde exceeding their exposure limits, S02, formaldehyde, and CO exceeding their STEL, and the rernaining contaminant mixture very nearly reaching a new AE exposure, it may not be prudent to enter the structure, until these levels clear. Comparison of the mean levels of contarriinants in the ship with levels in the burn building (table 15) shows that mean levels of formaldehyde and HCN were statistically higher in the burn building, and mean levels of CO, C0 2, and S0 2 were statistically higher in the ship mock-up. Benzene was not detected above the limit of detection in the burn building (LOD 0.3 ppm), but was detected on 3 of 5 occasions in the ship (mean level 0.8 ppm). It is hypothesized that differences may be due to different fuels used in the structures. The burn building had wood fires, which may be more likely to produce formaldehyde and HCN due to its chemical makeup (Ware 1991). The ship burns some diesel and stove oil, which may contain benzene and sulfur, and thus accounts for higher levels of contaminants containing these compounds. The increased CO and C0 2 levels, in the ship mock-up, may indicate incomplete combustion or poor venting. 4.4.3.2 Specific observations - burn building In the burn building, on some days, vented contaminant levels exceeded the exposure limits of one or more contaminants. On July 23 measurements were collected in a room that was observed to be filled with residual smoke, possible due to poor venting (the windows were left closed). The data indicated that S02, CO, and formaldehyde exceeded their respective exposure limits (CO 46 ppm, S0 2 2.1 ppm, and 57 formaldehyde 2 ppm). Formaldehyde remained at this level for 1 hour until (hffusing below its exposure limits, whereas CO and S0 2 both diffused below their exposure limits within 25 minutes. On Sept 28 and Oct 9, CO again exceeded its exposure limit, and again quickly diffused below its respective exposure limit within 40 minutes on October 9th, and 10 minutes on September 28th. On all other sampling sessions no conlaminants exceeded their respective exposure limits, after venting, in the burn building. The situation on July 23 when formaldehyde exceeded the exposure limit, it is noteworthy to point out that the second sample, taken 30-60 minutes following venting was higher than the first sample (2.43 > 2.00 ppm initial sample). It is not known why levels increased, but off-gassing is suspected and that formaldehyde was released from the concrete walls. Off-gassing can not be proven with the current data, but other situations occurred that could also lead one to suspect it may occur. Figure 4, illustrates that for HCN and formaldehyde, the levels recorded immediately following venting increased for a period of time and then quickly began to dilute away: Only HCN and formaldehyde have shown this trend on other occasions. In all this phenomenon occurred on 6 of 20 occasions that HCN and formaldehyde were sampled (data not included). It should be pointed out that even though a brief increase is occurring, levels have not yet been observed to again approach the levels recorded before venting. If off-gassing is occurring the exact conditions over which it occurs is unknown, can not be determined from the current data, and remains to be investigated later. Another possible explanation, for the increase in contaminant concentration noted in figure 4, is that air pockets containing contaminants in higher concentration may exist in the room. Pockets could be formed by residual layers of contaminants remaining at the ceiling or from pockets of air that may drift in from adjacent rooms that were not well ventilated. If these pockets exist they may drift past the sensor, at a later time, and be recorded as increased contaminant levels. No matter how the phenomenon in figure 4 is created, either by air pockets or by the cement walls off-gassing, the period that any contarninant increase in concentration appeared to last a maximum of 45 to 90 minutes, and levels subsequently decreased thereafter. 58 Existence of stratified layers of contaminants was observed on Nov 16. On this day the weather was 1.8 °C and snowing, and after venting a visible smoke layer, about 30 centimeters thick, was observed at the ceiling level of the room. Levels of CO recorded immediately following venting were about 20-25 ppm and remained at this level for 30 minutes before decreasing rapidly. During the sampling period, it was noted that the smoke layer did not dissipate, in fact, it remained thick and visible, and drifted slowly towards the floor. After 30 minutes when sampling tubes were changed it was noted that residual CO levels were low, and that the smoke layer was about 25 em's above the sampling apparatus. Personal observation was conducted while the layer drifted past the sampling equipment and it was noted that this corresponded to an increase in CO levels (from 7 to 42 ppm), as evidenced on the real time dataloggers. C0 2 also increased, but not HCN, formaldehyde, or particulate. S0 2 was not detected on this day. This data allows one to argue that air pockets or layers of conlaminants exist in the structure, which may result in differing exposures from one area to another. Another occurrence in the burn building, that provided evidence of air pockets, occurred during long-term sampling. On one occasion, C0 2 levels were above 600 ppm, and declined to less than 400 ppm in about Vi hour. It was later found that numerous sharp peak increases in the C0 2 concentration occurred throughout the sampling period, and was found to predominately coincide with the times I was present in the building, times I changed sampling tubes and checked equipment. Follow up study indicated the peak levels were not CO2 exhaled in my breath (data not included). It is postulated that my presence disturbed some air pockets present in the room. These "pockets" may simply be stratified layers of gas residual at the ceiling, where temperatures are warmer or actual pockets of air with increased contaminant levels. It is possible the peaks occurred when gusts of wind disturbed a pocket of air causing it to drift past the sensor. Leaving the indoor fire training structures open, and completely suppressing the fires, is important to ensure levels of contaminants dilute. This became particularly apparent on one occasion when long term sampling was to be conducted. The site was intentionally left closed with a small fire smouldering as staff 59 wanted the building to remain warm, to protect equipment over the cold winter night. Sampling was only conducted for Vi hour, as contaminant levels quickly became elevated. CO reached 137 ppm, C021674 ppm, S0 2 1.7 ppm, particulate 2.40 mg/m3, formaldehyde 0.65 ppm, and HCN 0.26 ppm. Afterwards the investigator deemed the levels were too high to be inside the structure, alone and without SCBA, and sampling for the night was promptly abandoned. Further examination of the times that HCN was detected in the burn building, showed that it was measured on more occasions, and in greater amounts, during times that the recruit training classes used the building than with non recruit classes. In total, HCN was detected on all 6 occasions sampling occurred after the recruits were in the building, and was detected on 3 of 6 occasions non recruit classes used the burn building. Mean level measured after the recruits used the structure was 0.516 ppm compared to 0.018 ppm when non-recruits used the structure, which was statistically significant (Mann-Whitney result p < 0.05). It is known that HCN is formed at higher burning temperatures. Since recruits are training as professional firefighters, technicians at the site stated that fires in the building are often made larger and hotter by increasing the burning duration and fuel load added to the fire. This is done to give the recruits more experience in the heat; afterwards, during sampling, the building is also noticeably warmer. Owing that HCN is formed more readily at higher temperatures (Bayer 1974), it is scientifically plausible that it could be present in greater quantities after the hotter fires. Strangely however, no other contaminants appeared to follow a similar trend, and it could not be determined if levels of any other contaminants were significantly different when recruits did or did not use the structure. Particulate levels recorded in the burn building (table 11) seem unusually low for what one would anticipate following recent fires (mean level 0.16 mg/m3 and maximum level 2.26 mg/m3), as observation of smoke indicates it is heavily laden with particulate matter. Even background levels on one day (0.33 mg/m3) exceeded the mean level measured following venting. The reason for low particulate levels is unknown, it may be that levels are truly this low. As mentioned previously, sampling could not be conducted until after the crews had vacated the structure, which was usually about 10 minutes following 60 venting. It is postulated that much of the large sized particulate matter settled out during this brief delay. It is also possible that much of the large particulate matter is vented out of the building, and only smaller particles remained inside the building, resulting in lower measured levels. It is recommended that if further study is conducted, that one determines if particulate levels are indeed higher than recorded herein, and particle size in the structure be determined. In summary, it was found that venting reduced the dangerous non-vented contaminant levels to safer levels, except for circumstances where venting appeared to be poorly performed. On the days when venting was poorly done, the rooms were noted to be partially filled with residual smoke, or on one occasion, the fire was noted to have been insufficiently suppressed and smoldered for a short time afterward. From the data collected it appears that venting performs a necessary function to reduce the levels of contaminants, and if done well, can reduce the contaminants to below the exposure limit, on most occasions, for most all contaminants. The situations when poor venting occurred or when fires are left smoldering can easily be identified by staff, at which point they can ensure the buUding remains vacant to air out. 4.4.3.3 Specific observations - ship mock-up Venting in the ship mock-up was not found to be as effective as the burn building. On 60% (3/5) of occasions sampling was conducted, CO and S0 2 were detected at levels above their exposure limits, at levels reported to cause health effects. The recorded levels were: CO = 36,52, 95 ppm and S0 2 = 3.1, 3.3, 8.4 ppm. CO at 95 ppm very nearly reached its short term exposure limit of 100 ppm, and S0 2 exceeded its short term exposure limit of 5 ppm. Benzene was detected on 3 occasions, at a mean level of 0.8 ppm, a maximum level of 1.85 ppm, both of which are in excess of the permissible limit. Formaldehyde was detected on all occasions and once at levels above its exposure limit. The remaining contaminants were found on all occasions to be present at levels less than their exposure limits. Combustion gasses appear to clear exceptionally quickly, and on all occasions CO and S0 2 levels had diffused below their exposure limits in less than 10 minutes, and were below the limits of detection after 5 lminutes in the longest case. Both formaldehyde and benzene were detectable for three hours following 61 fire activity. Though formaldehyde had diluted to levels generally only a fraction of its exposure limit. On the other hand, benzene was found to exceed its exposure limit (0.5 ppm), after three hours, on all the occasions but one, it was detected. At these levels there is the potential for adverse health effects if any person was to enter the structure and remain inside. The peak SO2 level (8.4 ppm) can act as an irritant to the upper respiratory tract, eyes, and mucous membranes, and CO (95 ppm) can produce 3.5% COHb in about 40 minutes (assuming a moderate workload) (Biological exposure index). However, it was pointed out that the levels of combustion gases decreased rapidly, and were below their exposure limit after 10 minutes on all occasions. Benzene exposure may be of concern, as it was found that levels remained elevated, often in excess of its exposure limit for three hours following venting, which was the longest time sampling was conducted. Thus there is the potential for benzene exposure long after the structure has been vented. Analysis of the particulate levels in the ship revealed elevated levels on all five occasions: mean levels were 0.596 mg/m3 and maximum 1.01 mg/m3. The level recorded may underestimate the true amount of particulate in the air. All samples collected immediately following venting had substantial staining of soot on the backing pads of the sampling cassettes. This indicates that soot filtered through the 5 um pores of the PVC filters and was subsequently trapped on the backing pad. The amount could not be quantified, though on all samples, substantial amounts were present on the filter backing. Previous study by Van Netten (1997), at the FSTC, reported collecting soot residue on the 6th stage of the Anderson samplers. To collect on the 6th stage, soot particles would have to be between 2.3 — 3.3 um, which supports the data that the soot was small enough to filter through a 5 um pore. Particles of this size can be deposited in the terminal bronchi of the airways, and knowing that soot may contain poly aromatic hydrocarbons, some of which are known carcinogens, this may represent a potentially hazardous exposure. It was noted that the second samples, taken 30-60 minutes following venting, rarely had little, if any, soot staining, which indicates that most of the soot has settled within 30 minutes. 62 Of interest are the levels collected on March 14 (tabulated in appendix). On this day the highest levels of HCN, formaldehyde, CO, and S0 2 were recorded: levels were HCN 0.46 ppm, CO 95 ppm, S0 2 8.4 ppm, formaldehyde 0.324 ppm. This is not unusual as one may speculate that the firefighters did a poor job venting the structure. However, the levels of S0 2 (8.4 ppm), as observed with the real time gas detector, were noted to be considerably higher than any levels previous recorded. Levels were high enough that they exceeded the short term exposure limit, and could cause irritation, discomfort, constrict the upper airways, and reduce expiatory volume in susceptible individuals (Biological exposure index). This prompted further investigation immediately at the site. Discussion with technicians found that the fuel used on this day was diesel and that the main fuel tank was refilled completely only a few days earlier. Before this day the fuel tank had been filled with Jet B (aviation) fuel and the prior four sampling sessions were believed to have used Jet B fuel. It is speculated that the change in fuel type may explain part of the large increases in contaminants measured on March 14. Unfortunately, no samples of the exact fuel type were taken to verify the type of fuel used for each burning, and the above conclusions can not be proven. Also the possibility that the firefighters did a poor job venting the structure may have contributed to the elevated levels. 4.4.4 Assessment of time period for environmental levels (after venting) to dilute to insignificant levels Examination of the data immediately following venting (discussed in section 4.4.3) has shown that on most, but not all occasions, the burn building would be safe to enter, without SCBA, if adequate venting by the fire crews has been performed. However, the ship mock-up may not be as safe to enter, due to S02, CO and benzene levels that remained elevated. Table 14 (results) illustrates the time period for contaminants to diffuse below the action limit in the indoor burn sites. The mean level of C0 2, HCN, and particulate were below their respective action limits immediately following venting. Whereas S0 2 took about 40 minutes to diffuse below its respective action limit, CO about 65 minutes, formaldehyde about 140 minutes, and benzene took about 380 minutes to diffuse below its action limit. In summary, it was found that 380 minutes represented the longest dilution period for the mean level of any contaminant to diffuse below their action limit. From the data it appears that one could recommend that the 63 burn building not be entered, without respiratory protection, for at least 140 minutes, after which it may be safe to enter, but that the ship-mock up be left for at least 380 minutes before entry occurs. 4.4.4.1 Specific observations Burn building The time for contaminants to dilute to insignificant levels varied according to the type of contaminant, however, after 90 minutes combustion gases such as CO and S0 2 had diluted to a fraction of their original levels, or on most all occasions (21 of 31) were non detectable. C0 2, particulate, and HCN, though often detected after 90 minutes, were present in concentrations at least 100 fold lower than levels documented to cause health effects. Formaldehyde however, had been detected 3 hours following venting on all occasions (14 of 14), and levels were about 1/5 the initial concentration (mean initial levels 0.372: after 3 hours 0.0772 ppm). It should be pointed out that formaldehyde has a limit of detection one order of magnitude lower than the other contaminants, and that it has been detected for such long periods is most likely a result of the low limit of detection. 4.4.4.2 Specific observations: Ship mock-up On all occasions S0 2 was measured it diffused to non detectable levels within 20 minutes, and on all but one occasion CO was measured, it diffused to non detectable levels within 25. On one occasion, levels of CO remained detectable for 51 minutes, even though S0 2 was not detected after 19 minutes. Dilution time periods for HCN varied, though on all five occasions it became non-detectable after 120 minutes. Formaldehyde on all occasions (5 of 5) remained detectable after 3 hours, and on 2 of 5 occasions benzene was also detectable after 3 hours. After 3 hours the mean level of formaldehyde was 0.0738 ppm, and the mean level of benzene was 0.48 ppm. As can be observed formaldehyde was well below its exposure limit of 0.3 ppm, while benzene was marginally below its exposure limit of 0.5 ppm. It was calculated that these levels would remain elevated for 380 minutes until diffusing below Vi the exposure limit (action limit), and the lack of detectable background levels is evidence they do eventually dilute to non-detectable levels. 4.5 Propane side of burn building Contaminant levels in the propane side of the burn building were found to be substantially lower when compared to levels recorded in the wood side. On no occasion were detectable amounts of particulate, 64 HCN, or S02 found after propane fires. This may be attributed to propane fires burning cleaner than wood fires, or that the local exhaust fans, which are located on this side of the building, are remarkably efficient at removing contaminants when combined with normal venting. Only on one instance was CO detected (5 ppm maximum level). Unfortunately, the lack of additional detectable levels does not allow further data analysis. Formaldehyde was detected on all occasions sampled, with a mean level of 0.069 ppm and the maximum level 0.35 ppm. Previously it had been mentioned that propane is cleaner burning than wood, and thus may be considered unlikely to produce detectable levels of formaldehyde. Also consider that contaminants such as CO, which is more likely to be detected, were not observed to be present, with one exception noted above. Thus, formaldehyde levels present in the room are thought to be due to contamination by smoke produced from wood fires on the other side of the structure. Periods of smoke contamination from wood fires may result in absorption of formaldehyde into the cement walls in the propane side. Subsequent elevations in temperatures and humidity, resulting from firefighting activity in the propane room, may produce ideal conditions for off-gassing to occur. Thus the recorded levels may be a result of formaldehyde, that had been previously absorbed into the walls, being released. This cannot be proven, but provides a plausible explanation for the formaldehyde levels measured. Also it must be considered that the propane fire produced formaldehyde and that the low limit of detection allowed it to be measured. Even though formaldehyde was detected on all occasions, the mean level (0.05 ppm) was well below levels documented to cause health effects and exposure at this site appears to pose a minimal threat to health, if any. 4.6 Off-gassing On occasion it had been mentioned that off-gassing was suspected of occurring in the burn building. The fact of background levels of formaldehyde (discussed in section 4.4.1), the anomalous curves (figure 4, discussed in section 4.4.3.2), and the formaldehyde levels measured in the propane side of the burn building (discussed in section 4.5) all support the assumption that off-gassing may be occurring. 65 If off-gassing is occurring it can be argued exposure may occur at times when no fire activity has occurred. Circumstances such as warm humid summer days may result in off-gassing and exposure to some chemicals. It must be pointed out that these observations are not proof that off-gassing is occurring, but point that there is definitely the potential. It is recommended that more study, specific to potential off-gassing, be conducted to determine times at which off-gassing may occur, for how long it may occur, and what maximum levels may be. 4.7 Estimate of the risk of health effects to instructors at the facility Evaluation of contaminants at the FSTC found, in some instances, that concentrations of individual chemicals were at levels known to have negative health impacts. It was shown that these contaminants dilute to levels below those impacting health, but we must consider that contaminants in smoke are not emitted in isolation. Synergism, or interactions, may occur amoung the chemicals, and the effects of one may potentiate the effects of others. In all, there may be a quantifiable risk, from smoke exposure at the facility. Unfortunately, this study will fall short in quantifying the exact risk, but it can be said that the risk is small, because this study can be used to suggest means of managing the risk. Careful observation and prudent use of SCBA may successfully manage the risks of harmful exposure at the outdoor pits (round tank and T-pit). It was identified that hose spray, weather conditions, and possibly fire size may have an impact on potential exposure. Prudent observation of the factors creating exposure should occur during fire activities, and when conditions exist where exposure may occur, they can be corrected. For example, hose spray can be directed away from areas where people must be located, and staff can be extra cautious on cloudy overcast days, especially when fire sizes are small and no micro-climate can form. Exposure to the floating clouds of dry chemical powder may pose a definite risk to staff health. Though the sodium bicarbonate is listed as a nuisance dust, levels of this magnitude could undoubtedly pose a health risk. Benzene exposure, though it would be low if one calculated an eight-hour time weighted average, 66 should be avoided. Benzene is recognized as a powerful myelotoxicant (Ware 1991), and the International Agency for Research on Cancer (IARG) has published risk assessments purporting that exposure at 1 ppm may lead to 1-14 excess cancer cases per 1000 exposed people (Biological exposure index). Exposure should be kept as low as reasonable achievable, even short-term exposure, to ensure untoward effects can not occur. For this reason it may be advisable to relocate the extinguisher pad to an area further from the office and classroom complexes, if future site expansion will allow. At areas where staff technicians are required to monitor emergency switches (round tank and propane pad), the mean level of contaminants were below insignificant levels, and the peak exposures while they were not below insignificant levels, are manageable. It may be safe to assume that the risk of adverse health effects at these locations is low. Unfortunately, the study fell short on the third objective, detennining a safe distance from the pits where exposure is insignificant. Albeit, the study did identify that the majority of contaminants are vented high above the facility and that exposure from the large fuel and depth fires may be of less concern than those from the extinguisher pad. It is believed that drifting smoke from these small fires may pose a health risk, especially if repeated on a continuous basis. This too would reinforce the need to relocate the extinguisher pad, if at all possible. The indoor burn sites (burn building and ship mock-up) may pose a risk of health effects to those who would enter the structure following venting, and before a sufficient period of time has elapsed. Peak levels of formaldehyde, S02, and CO were measured at levels that could lead to acute health effects reasonably quickly after entry. In addition to the previously listed contaminants, benzene was measured at slightly elevated levels for a substantial period of time following venting in the ship mock-up, as was formaldehyde in the burn building. There exists the potential for chronic health effects following long term or repeated exposure. As with the outdoor pits, this risk can easily be controlled. Long-term sampling has shown that eventually all conlaminants would dilute to insignificant levels. For this reason it has been suggested that the burn building be left open and vacant for at least 140 minutes and the ship mock-up be left open for at least 380 minutes. 67 Formaldehyde was detected in background levels in the burn building. At these levels it should pose little risk to health, except that it is a known sensitizer. The possibility exists that health effects may occur, from previously sensitized people, even at levels of exposure this low. This may pose problems as formaldehyde was detected up to 1 week following the last fire activity, and levels were attributed to off-gassing. If sensitization occurs, sensitive individuals could possible have to avoid entry into the structure, to ensure no adverse reactions occur. 4.8 Generalizibilitv of the study results Though the focus of this study was to assess chemical exposure at the Maple Ridge FSTC, some findings may be applicable to other fire related activities. There is an increasing amount of evidence that large ground level fires vent much of their emission high into the air, and ground level exposure is minimal (Booher and Janke 1997, Leahey et. al. 1993). This may relieve anxiety concerning ground level exposure at other fire trairiing centers. Additionally, it may also be assumed that similar burn buildings, such as the one in the trairiing center at Nanimo BC, may also need a minimum period of time to completely vent residual contaminants before entry occurs. Though not addressed in this study, the results may be applicable to similar situations, such as following suspicious structural fires (i.e. house fire). This may have implications on when arson investigators may safely enter a structure to investigate possible fire causes, and that entry into these structures should not be attempted until an undetermined period of time has elapsed. This may present itself as a future area where study could be directed. Finally, the potential of off-gassing may pose problems, as it may occur from cement, or possible other materials unburned in a fire. This too could lead to chemical exposure to arson investigators for days following a fire. Future studies looking at off-gassing could add important understanding of the periods when it may occur, for how long it may occur, levels at which it may occur, and possibly determine what contaminants it may occur with. 68 4.9 Study limitations • Measurement of outdoor ground level concentration of smoke emissions from fire is difficult, and techniques used to estimate ground level concentrations can be subject to errors. Atmospheric mixing of contaminants is highly variable and dependent on the prevalent conditions each day, and has been estimated to affect ground level concentrations by as much as one order of magnitude (Leahey et. al. 1993). • It was initially hoped that it could be attempted to characterize possible determinants of exposure, but the study fell short on this task. Partly due to subjective measures, such as estimates of fire size, the variegated effect of weather that can not be controlled, and due to the lack of samples above the detection limit. • It was not possible to examine the differences in exposure at each of the four sites sampled at the outdoor fuel and depth pits (T-pit and Round tank) due to small sample sizes. Nor was it possible to investigate the effect of fuel used on exposure. It was found, that to save on operating cost, aviation fuel had been mixed in with the diesel fuel and it could not be determined what type of fiiel was used at the pits. • Benzene sampling was inconclusive due to the high limit of detection of the method used to assess benzene. In many instances the LOD for benzene was greater than the WCB exposure limit. Thus, the study fell short on characterizing possible exposure situation for benzene that may result in exposure. • The longest period that long-term sampling was conducted was 3 hours. This was partly due to the lack of sunlight during the winter months. However, in the ship mock-up, the mean benzene levels were found to remain above the action limit for more than 3 hours. To estimate the time till it would be safe to enter the ship mock-up (defined as levels below the action limit) an extrapolation of the curve (figure 3) was required. • The study fell short on the objective of characterizing if there exists a safe distance from the outdoor pits where exposure would be insignificant. This was due to difficulties with estimating where smoke exposure may occur distant from the pits, and then placing equipment there. As well as the fact the 69 levels would be brief peak levels and the lack of datalogging gas detectors made multiple site sampling impossible. It is hoped that this study has provided reasonable insight into the ground level concentrations of contaminants at the facility, however, it is only a cross sectional description of the exposures, and can not guarantee that all burning will result in similar levels of emissions at the sites investigated. 70 Chapter 5: Conclusions 5.1 Outdoor fuel and depth pits (T-pit and Round tank) Similar to study by Booher and Janke (1997) and Leahey et. al. (1993) it was found that the majority of smoke emissions from these large fires rise into the air and are dissipated distant from the facility. At the four sites measured in this study, contaminants were detected on some instances. Though not statistically proven, it was found that rainy overcast weather and the force of the water stream sprayed from the hoses may account for some of the instances when contaminants were measured. Nevertheless, levels measured at these sites were low compared to the WCB allowable exposure limits. 5.2 Extinguisher pad Levels of contaminants measured directly in the clouds of particulate indicated that contaminants may become entrapped within the clouds, and that particulate and benzene represent exposure hazards. Due to increased levels of conlaminants in these clouds, contact should be avoided. Since these clouds can float considerable distances before dissipating, it is recommended that if future site expansion allows, that the extinguisher pad be relocated further away from the office and training classrooms. 5.3 Propane pad Due to the lack of any detectable samples at this site there appears to be little risk to the health and safety of the staff at the FSTC. 5.4 Burn building and Ship mock-up Without any fires it appears that particulate is present inside the burn building and ship mock-up and formaldehyde is present inside the burn building. Particulate levels may be due too natural dusts, pollen, organic matter, etc. present in the air, however, the formaldehyde levels may be indicative off off-gassing occurring inside the burn building. Further study is recommended to fully characterize the potential for off-gassing. Levels off contaminants, except benzene, particulate, and CO2, following a burn exercise and preceding venting, were found to be more than levels reported to cause health effects. They surpassed Vz the 71 immediately dangerous to life and health level for SO2 and CO. Venting of the structures, following a burn exercise, decreased these levels of contaminants. However, the mean level of formaldehyde and CO in the burn building and ship mock-up were still elevated in excess of their exposure limits, and in the ship mock-up benzene levels were elevated. To help allay staff health concerns, exposure to these contaminants, should be minimized. Long-term sampling indicated that after a burn session entry into the burn building should be avoided for at least 140 minutes, at which time contaminant levels have diffused below Vz the exposure limit (the action limit). Entry into the ship mock-up should be avoided for at least 380 minutes as benzene levels were elevated until this time. Small size particulate levels of < 5 um in size, were detected in the ship mock-up following burn exercises. Particles of this size may penetrate deeply into the lung, and as they may contain soot, a known carcinogen, inhalation should be avoided. It is suspected, though not proven, that off-gassing may occur inside the burn building. Staff should be aware that low levels of formaldehyde may be present inside the burn building even without burning for up to one week. General conclusions: Overall, the potential exists for exposure to chemical contaminants, however, the risk of health effects can be minimized through the use of effective site specific control procedures mentioned previously. 72 Bibliography ACGIH (American Conference of Governmental Industrial Hygienists), Documentation of the Threshold limit values and Biological Exposure Indices ., 5th edition, American Conference of Governmental Industrial Hygienists Inc., Cincinnati Oh. Allara, D. L., and Edelson, D., A computational modeling study of the low temperature pyrolysis of n-alkanes; mechanism of propane, n-butane, and n-pentane pyrolysis. Int. J. Chem. Kinet. 1975;7:479-483. Anon. Plastic fires create new hazards for both firemen and the public. Medical News JAMA. 1975;234:1212. Anon. Carbon monoxide risks at home. National Fire Protection Association., Quincy, MA. Anon. What's in wood smoke and other emissions. Burning issues. - From Internet world wide web. Anon. Firemen's lungs. Lancet 1975; 1(439). ANSI Z88.2 (1992) ANSI/CGAG-7.1. Commodity specification for air. 1972. ANSI Z88.6 American national standard for respiratory protection - respirator use - physical qualifications for personnel. Lawrence Livermore National Laboratory 1984. Alpaugh E., 1988, Temperature extreme, in B. Plog (ed.) Fundamentals of Industrial Hygiene (National safety council, Chicago), 259-281. Atlas E., Donnelly K., Giam C, Mcfarland A., Chemical and biological characterization of emissions from a fireperson training facility. Am. Ind. Hyg. Assoc. J. 1985;46(9):532-540. Axford T., Mckerrow C, Jones A., Lequesne P., Accidental exposure to isocyanate fumes in a group of firemen. Brit. J. Ind. Med. 1976;33:65-71. Bayer A., Aspects and methodology for the evaluation of toxicological parameters during fire exposure. JFF/Combustion Toxicology 1974;1:4-51. Beckner C , The role of cyanide in fires. Vet. Hum. Toxicol. 1985;27(6):487-490. Bendix S., Firefighters exposure to environmental carcinogens. J. of Com. Toxicol.l979;6:127-135. Biological exposure index (BEI). Formaldehyde CAS: 50-00-0 * See ACGIH Biological exposure index (BEI). Acrolein CAS: 107-02-8 * See ACGIH Biological exposure index (BEI). Benzene. CAS: 71-43-2 * See ACGIH Biological exposure index (BEI). Carbon Dioxide CAS: 124-38-9 * See ACGIH Biological exposure index (BEI). Carbon Monoxide CAS: 630-08-0 * See ACGIH Biological exposure index (BEI). Hydrogen Cyanide CAS: 74-90-8 * See ACGIH Biological exposure index (BEI). Hydrogen Fluoride CAS: 7664-39-3 * See ACGIH Biological exposure index (BEI). Hydrogen Sulfide CAS: 7783-06-4 * See ACGIH Biological exposure index (BEI). Hydrogen Chloride CAS: 7647-01-1 * See ACGIH Biological exposure index (BEI). Nitrogen Dioxide CAS; 10102-44-0 * See ACGIH 73 Biological exposure index (BEI). Sulfur Dioxide CAS: 7446-09-5 * See ACGIH Biological exposure index (BEI). Particulate not otherwise classified (PNOC) * See ACGIH Biological exposure index (BEI). Propane CAS: 74-98-6 * Sec ACGIH Biological exposure index (BEI). Gasoline CAS: 8006-61-9 * See ACGIH Birkey M., Clark F., Inhalation of toxic products from fires. Bull. N.Y. Acad. Med. 1981;57(10):997-1013. Bishop P., Smith G., Ray P., Beaird J., Smith J., Empirical prediction of physiological response to prolonged work in encapsulating protective clothing. Ergonomics 1994;37(9): 1503-1512. Brandt-Rauf P., Fallon L., Tarantini T., Idema C , Andrews L., Health hazards of fire fighters: exposure assessment. 1988;45:606-612. Booher L., Janke B., Air emissions from petroleum hydrocarbon fires during controlled burning. American industrial hygiene association journal. 1997;58:359-365. Burgess L., Crutchfield C , Tucson fire fighter exposure to products of combustion: a risk assessment. Appl. Occup. Environ. Hyg. 1995;10(l):37-42. Carter J. Thermoregulatory responses of fire fighters to work and recovery in heat. Master of Science Thesis Simon Fraser University, BC. Canada. April 1996. Clark J., Campbell D., Reid W., Blood carboxyhemoglobin and cyanide levels in fire survivors. Lancet 1981. 12(1134):12-15 Committee on fire toxicology. Fire and smoke: understanding the hazards. Board of environmental studies and toxicology, Commission on life sciences National research council. National academy press, Washington, DC. 1986. CrapoR., Smoke inhalation injuries. JAMA. 1981;246(15): 1694-1696. Davis P., Dotson C , Physiological aspects of fire fighting. Fire technology 1987;23(3):280-291. Davis P., Santa-Maria D., Relationship between simulated fire fighting tasks and physical performance measures. Med. Sci. Sports Exerc, 1975;14(1):65-71. Davis P., Santa-Maria D., Relationship between simulated fire fighting tasks and physical performance measures. Med. Sci. Sports Exerc, 1975; 14(1):65-71. Davis P., Dotson C , Santa-Maria D., Relationship between simulated fire fighting tasks and physical performance measures. Med. Sci. Sports Exerc. 1982;14(1):65-71. Decker W., Garcia-Cantu A., Toxicology of fires: an emerging clinical concern. Vet Hum Toxicol 1986;25(8):431-433. Douglass D., Douglas R., Oakes D., Scott G., Pulmonary function of London firemen. Brit. J. Ind. Med. 1985;42:55-58. Duncan H., Gardner G , Barnard R., Physiological responses of men working in fire fighting equipment in the heat. Ergonomics 1979;22(5):521-527. 74 Dyer F., Esch V., Polyvinyl chloride toxicity in fires: hydrogen chloride toxicity in firefighters. JAMA 1976;235(4):393-397. Faff J., Tutak T., Physiological responses to working with fire fighting equipment in the heat in relation to subjective fatigue. Ergonomics 1989;32(6):629-638. Gledhill N., Jarnnik V., Characterization of the physical demands of firefighting. Can. J Sports Sci., 1992;17:207-213. Faff J., Tutak T., Physiological responses to working with fire fighting equipment in the heat in relation to subjective fatigue. Ergonomics 1989;32(6):629-638. Fornal D., Understanding turnout technology: a matter of health and safety. Fire engineering. 1992; June :105-113. Friones J., Hinds W., Duffy R., Lafuente E., Liu W., Exposure of firefighters to diesel emissions in fire stations. Am. Ind. Hyg. Assoc. J. 1987;48(3):202-207. Feunekes F, Jogeneelen F., Lann H , Schoonhof F., Uptake of polycyclic aromatic hydrocarbons amoung trainers in a fire-fighting training facility. American industrial hygiene association journal. 1997 58:23-28. Garner R., A common car fire poses uncommon hazards. (Letter), Fire fighter Mason fire dept. Michigan. . GenovesiM., Effects of smoke inhalation. Chest 1980;77(3):335-336. Ginsberg M.D., Carbon Monoxide intoxication: Clinical features, neuropathology and mechanism of injury. Clinical toxicology 1995; 23(4-6):281-288. Gold A., Burgess W., Clougherty E., Exposure of firefighters to toxic air contaminants. American Industrial Hygiene Association Journal 1978;39:534-539. Guiditti T., Clough V., Occupational health concerns of firefighting. Annu. Rev. Publ. Health. 1992;13:151-71. Guzzardi J., Inhalation of products of combustion. Annals of Emergency Medicine 1983;12:628-632. Hartzel G., Packham S., Switzer W., Toxic products from fires. Am. Ind. Hyg. Assoc. J. 1983;44(4):248-255. Heinman E., Shy C , Checkoway H , Injuries on the foreground: risk factors for traumatic injuries amoung professional fire fighters. Am. J. Ind. Med. 1989;15:267-282. Jankovic J., Jones W., Burkhart J., Noonan G , Environmental study of firefighters. Ann. Occup. Hyg. 1991;35:581-602. Koseki H , Combustion properties of large liquid pools fires. Fire technology. 1989;25(3):241-255. 75 Larson T., Koenig J., Wood smoke: emissions and non cancer respiratory effects. Annu. Rev. Public Health. 1994;15:133-56. Large A., Owens G., Hoffman L., The short term effects of smoke exposure on the pulmonary function of firefighters. Chest 1990;97(4):806-809. Leahey D., Hansen M., Schroeder M., An assessment of air quality impacts of fires associated with fire fighting operations. Air and Waste. 1993;43:341-347. Lemon P., Herminston T., Physiological profile of professional firefighters. J. of Occup. Med. 1977a; 19(5):337-340. Lemon P., Herminston T., The human energy costs of firefighting. J. of Occup. Med. 1977a;19(5):558-562. Levine M., Hohman B., Do face masks provide protection? Yes according to this study. Fire Journal. 1978;52-55. Levin B., The development of a new small-scale smoke toxicity test method and its comparison with real-scale fire tests. ©1992 Elsiver Science Publishers B.V. Lockhart Risk Management Limited, Open air burning of diesel fuel: health aspects. 1992. Loke R., Farmer W., Matthay A., Putman C , Smith G , Acute and chronic effects of fire fighting on pulmonary function. Chest 1980;77(3):369-373. Lowrey W., Jaurez L., Petty C , Roberts B., Studies of toxic gas production during actual structural fires in the Dallas area. J. Forensic Science 1985;30( l):59-72. Lusa S., Louhevaara V., Smolander J., Kimimaki., Korhonon O., Physiological responses of firefighting students during simulated smoke-diving in the heat. Am. Ind. Hyg. Assoc. J. 1993;54(5):228-231. Markowitz J., Gutterman E., Link B., Rivera M., Psychological response of firefighters to a chemical fire. J. Human Stress. 1987 Svommer:84-92. McGuire M., Casserly D., Greff R., Formaldehyde concentrations in fabric stores. Applied Occupational and Environmental Hygiene 1992; 7(12): 112-119. Minty B., Royston D., Jones J., Smith D., Searing C , Beeley M., Changes in permeability of the alveolar-capillary barrier in firefighters. British Journal of Industrial Medicine. 1985;42:631-634. Morse L., Pasternuak G., Fujimoto G , Toxic hazards of firefighters. In Hazardous materials toxicology: clinical principles of the environment. 1992 Williams and Wilkins, Baltimore, Ma. Pages 545-550. Murrell J., Toxic hazards caused by fires. Trans IMar E © Vol 98, Paper C1/7 NFPA 1981 Standard on open circuit self contained breathing apparatus for fire fighters. 1987. NFPA 1500. Standard on fire department occupational safety and health program. 1992. NFPA 1404 Standard for a fire department self contained breathing apparatus program. 1989. 76 Phanuf D., Desgranges J., Plante N., Rodrigue J., contamination of local wildlife following a fire at a polyvinyl chloride warehouse in St. Basile le grand, Quebec, Canada. Arch. Environ. Contam. Toxicol. 1995;28:145-153. PolakoffP., Attention should be given to limiting firefighting hazards. [Viewpoint] in Occupational health and safety. 1984;53(4):55-56. PrienT:, Toxic smoke compounds and inhalation injury - a review. Burns 1988; 14(6):451-460. Rinke C , Early respiratory abnormalities from smoke inhalation. JAMA 1984;251(6):771-773. Report to the workers' compensation board on cardiovascular disease and cancer amoung firefighters. Industrial disease standards panel report #13, Toronto, Ontario, 1994. Reischl U., Water fog stream heat radiation attenuation. Fire technology. Romet T., Frim J., Physiological responses to firefighting activities. Eur. J. Appl. Physiol. 1987;56:633-638. Sammons J., Coleman R., Firefighters' occupational exposure to carbon monoxide. J. Occup. Med. 1974;16(8):543-546. Selala M., Couche F., Daelemans A., Joremns P., Beaucourt L., Schepens J., Fire fighting: how safe are firefighters? Bull. Environ. Contam. Toxicol. 1993;51:325-332. Sheppard D., Distefano S., Morse L., Beckner C , Acute effects of routine firefighting on lung function. Am. J. Ind. Med. 1986;9:333-340. Sivathanu Y., Faeth G., Soot volume fractions in the overfire region of turbulent diffusion flames. Combustion and Flame 1990;81:133-149. Sothman M., Saupe K., Pawelczyk J., Davis P., Dotson C , Landy F., Siljunas M., Oxygen consumption during fire suppression: error of heart rate estimation. Ergonomics 1991;34(12):1469-1474. Steenland K., Lung cancer and diesel exhaust: a review. American Journal of Industrial Medicine. 1986;10:177-189. Stevenson C , Effects of radiant heat in fire fighting instructors. British journal of industrial medicine. 1985;42:67-68. Sumi K., Tsuchiya Y., Toxic gases produced at fires. Canadian building digest. Division of building research. National research council of Canada. Dec. 1971. Terill J., Montgomery R., Reinhardt C , Toxic gases from fires. Science 1978;200:1343-1347. Trertman R., Burgess W., Gold A., Air contaminants encountered by firefighters. Am. Ind. Hyg. Assoc. J. 1980;41:796-802. 77 US dept of Health, Education and Welfare. A symposium: Occupational safety research specifically related to personal protection., Center for disease control, National center for occupational safety and health, Cincinnati, Oh. 1974. Van Netten C , Shirtliffe C , Svec J., Temperature and humidity dependence of formaldehyde release from selected building materials. Bull. Environ. Contam. Toxicol. 1989;42:558-565. Van Netten C , Shirtliffe C , Svec J., Formaldehyde release characteristics from a Swedish floor finish. Bull. Environ. Contam. Toxicol. 1988;40:672-677. Van Netten C , Analysis of sources contributing to elevated formaldehyde concentrations in the air in a new elementary school. Canadian Journal of Public Health. 1983;74:55-59. Van Netten C , Potential exposures and health hazards at the Pacific Marine Training Institute, Maple Ridge, BC. Unpublished, 1994. Van Netten C , Investigation of air quality at the Fire and Safety Training Center in Maple Ridge during a training exercise. Unpublished, 1997. Vedel S., Health effects of wood smoke: a report to the Provincial health officer of British Columbia 1993. Venghte J., Fire fighters' protective clothing: design criteria. (Lion Uniform Technical Bulletin. Dayton, Oh.) 1989. Washburn A., Lablanc P., Fahy R., Report on firefighting fatalities. NFPA Journal. 1993; July/Aug: 55-70. Westbrook C , Pitz W., A comprehensive chemical kinetic reaction mechanism for oxidation and pyrolysis of propane and propene. Combustion and science technology. 1984;37:117-152. Ware G., Toxicology of vegetation smoke. Reviews of environmental contamination and toxicology. 1991;119:1-46. White M., Vercruyssen M., Hodous T., Work tolerance and subjective responses to wearing protective clothing and respirators during physical work. Ergonomics 1989;32(9): 1111-1123. Workers' Compensation Board of British Columbia., Occupational health and safety regulations., WCB Richmond, BC. April 15, 1998. 78 Appendix 1A Description of byproducts of combustion measured and some general health effects Subdivision of the chemical byproducts of combustion into categories can be done to better characterize them, however, there is no one recognized classification scheme and many researchers have developed then-own to suit their needs. Of the classification schemes it appears that consensus has been reached regarding classifying the gaseous emissions according to those that are pulmonary irritants and those that are asphyxiants (Morse et. al 1992). Asphyxiants Carbon monoxide This is an odorless, tasteless, undetectable gas, present at every fire measured, in concentrations of up to 20% [under special circumstances], but normally rarely reaches or exceeds 5% (Utech 1975). CO is a byproduct of all fires and is believed to be one of the most hazardous chemicals encountered by firefighters (WCB report 1992, Bayer 1974). It has been clearly been established as one of the gases contributing to deaths from smoke inhalation (Committee on fire toxicology). Lowrey et. al. claimed that it is the only toxic gas produced in sufficient quantity at fires to produce death. Controversy exists if it is the main component responsible for deaths, if it is a contributory factor working synergistically with another gas, or if there are others gases responsible for deaths at fires. Effects of exposure are headache, chest tightness, nausea, which if exposure continues can lead to unconsciousness and death (Reinke et. al. 1978). Exposures during increased temperatures, heavy labour, or at altitudes more than 1600 meters have been found to lead to increased body uptake, and was part of the rationale for lowering the TLV to 25 ppm (Biological exposure index). Where the new TLV is designed to keep the blood carboxyhemoglobin (COHb) levels at 3.5% or lower (Biological exposure index). Table A l summarizes the effects of CO exposure. Table Al : CO exposure and effects according to the level of COHb in the blood Concentration % COHb in blood Physiological effects or symptoms < 35 ppm 0-5 none in 8 hours 200 5-15 mild headache after 2-3 hours 400 15-25 headache, nausea, dizziness in 1-2 hours 800 25-40 headache, nausea, dizziness in 45 minutes and collapse after 1-2 hours 1600 40-50 collapse and unconsciousness in 30 minutes and headache, nausea dizziness in 20 minutes 3200 50-60 Collapse and unconsciousness in 20 minutes and headache, nausea dizziness in 5-10 minutes, coma and convulsions. Above this level respiration is decreased and death is possible. 10,000 70+ coma, convulsions, above this level respiration is decreased and death is possible. 12,800 70+ immediate effects, death in 1-3 minutes Continuos low level CO exposure appears to have a deleterious effect on the body, especially the myocardium (Sammons and Coleman 1974). This is of concern if one considers that 50.6% of firefighter's deaths were attributed to heart attacks (Washburn et. al. 1994) and additionally, according to the US Department of Health Education and Welfare 95%, of firefighter's deaths in 1993, that were from heart attacks, were attributed to stress or overexertion. One may hypothesize that possible deleterious effects 79 from CO may play a contributing factor by damaging the heart so that it can not cope with stress and demands of firefighting. Carbon dioxide C0 2 is another major combustion product of fires, and has been documented to reach concentrations greater than 15% at fires (Committee on fire toxicology 1986, Utech 1975). It and CO have been termed natural products of combustion that are necessarily present at every fire (WCB report 1992). One of the most important physiological effects of CO2 is that it stimulates the respiratory centre, which will increase the inhalation of other toxic compounds formed during combustion (Bayer 1974), and in high concentrations it will act as an asphyxiant. Respiratory stimulation is noticeable at 27,600 ppm, where it has increased 15%, whereas exposure at 39,500 ppm produced few effects other than increased ventilation (Biological exposure index). CO and C0 2 appear to have a synergistic method of action that increases the toxicity of the CO (Decker and Garica-Cantu 1986) as C0 2 may increase CO absorption and prolong the effects (WCB report 1992). Oxygen depletion 0 2 is consumed during combustion and is displaced by the formation of other products of combustion. This process results in decreased 0 2 levels which acts as an asphyxiant, however this is not believed to be a problem (Committee on fire toxicology 1986). Measurements of 0 2 have found that the levels often remain between 19-21% (Lowrey et. al 1983, Utech 1975). On the other hand 0 2 levels can become depleted is when combustion occurs in confined spaces that are relatively airtight, when levels may decrease to 10-15%(Guzzardi 1983). Irritants First it should be noted that many of the gases listed below is highly toxic and capable of causing death. The fact that most have been measured at low concentrations at fires is the reason that they are considered irritants, as they are highly irritating at low concentrations, which is the main effect produced. Bayer (1974) stated that during the early stage of combustion that little CO is present in smoke and that most smoke inhalation injuries are due to irritant chemicals. Sulfur dioxide This is produced from the combustion of sulfur containing compounds (wool, felt, leather, wood) as well as synthetic materials (polystyrene, polyethylene, polysulfone and rubber formulations), and has been found at many fires in low concentrations (Trietman et. al. 1980). S0 2 has been classified as a mild respiratory irritant of mucous membranes, probably by formation of sulfurous acid when it contacts moisture (Biological exposure index). Concentrations of 1 ppm have been found to constrict the upper airways and reduce forced respiratory volume, while at 5 ppm irritation becomes noticeable (Biological exposure index). It does not bioaccumulate, and appears to have no carcinogenic effects (Trietman et. al. 1980). Hydrogen cyanide This product is not very and is rarely produced in concentrations that can be life threatening (Lowrey et. al. 1985, Beckner 1985, Committee on fire toxicology 1986). One study reported that it has been detected at 12% of the fires studied (Lowrey et. al. 1985), and was produced from incomplete combustion of nitrogen containing organic products and natural fibers (wool, silk, paper, wood) and synthetic polymers (nylon, acrylonitrile, nitrocellulose) (WCB report 1992). Although it has been found at fires it is uncertain if it contributes to smoke toxicity as it will only be produced if the fuel contains appreciable concentrations of nitrogen and carbon (Committee on fire toxicology 1986). It has been found to be a highly endothermic compound and formed more readily as the temperature of combustion increases (Bayer 1974). Some studies found that levels measured at fire fires were not of sufficient quantity to be considered a hazard, in 90% of the fires it was non detectable (Treitman et. al. 1980), and it may not play a role in many fire deaths (Terrill et. al. 1978). 80 Controversy about its role in fire related deaths exists because it has been found that HCN poisoning in the absence of CO is has found that high HCN (as blood cyanide) levels are found with high CO levels (as carboxyhemoglobin), and that the high HCN levels alone would have been deadly. Unfortunately the issue is clouded because but no low CO levels were associated with high HCN levels (Birkey et. al. 1981, Clark et. al. 1981). Cyanide is an extremely toxic rapid acting chemical that inhibits tissue oxygen utilization by inhibiting the terminal step in the electron transport chain. It is normally present in small quantities in tissue in concentrations up to 50 pg/lOOg (Biological exposure index). Acute effects of poisoning include lightheadedness, breathlessness, headache, and nausea, which may be followed by unconsciousness (Biological exposure index). Formaldehyde Formaldehyde is commonly found at fires, in low concentrations, as a byproduct of combustion of naturally occurring materials (wood, wool, paper) and synthetic products (upholstery, carpets, furniture) (Lowrey et. al. 1985). Most all fires measured contain detectable levels of formaldehyde in the smoke (Ware 1991). It has been found to be ubiquitous in the atmosphere, found in parts per billion at most locations sampled (Ware 1991), and as an endogenous chemical in most life forms (Biological exposure index). Formaldehyde is highly irritating and is likely to be responsible for some of the irritation produced by smoke (Ware 1991). Eye, nose, and throat irritation occur at about 0.1 to 3.0 ppm with severe irritation occurring over 10 ppm. Human eyes are exceptionally sensitive to formaldehyde, detecting concentrations as low as 0.01 ppm (BIOLOGICAL EXPOSURE INDEX - ACGIH). It has not been found to pose a risk of carcinogenicity at concentrations found at fires, largely due to low and infrequent exposures (Ware 1991). Others Polvaromatic hydrocarbons (PAH's) These compounds are found at many fires and are widely dispersed in nature as products of combustion of organic materials (WCB report 1992). Some can be absorbed through the skin and some have carcinogenic properties, however, it is believed that their concentration at structural fires is too low to be significant (Committee on fire toxicology 1986). Benzene Vedal (1993) reported benzene is a likely byproduct of any petroleum related fire and the WCB report (1992) stated that many organic materials release benzene when combusting. It has been found at almost every fire (Trietman et .al. 1980) and in significant concentrations at some fires. After CO and CO2 it has been found to be the most commonly found organic constituent at fires (WCB report 1992). Absorption of benzene can occur by inhalation, dermal, or by ingestion, though inhalation is the main route with up to 46% of inhaled benzene being absorbed by humans (BIOLOGICAL EXPOSURE INDEX - ACGIH). The most noticeable effects of benzene are carcinogenic as it acts as a powerful myelotoxicant (Ware 1991). The International Agency for Research on Cancer (IARC) has published risk assessments purporting that exposure at 10 ppm will lead to 14-140 excess cancer case per 1000 exposed individuals, and 1 ppm may lead to 1-14 excess cases per 1000 exposed people (Biological exposure index). Higher exposures can produce narcotic effects, similar to toluene, though levels must upwards of400+ ppm (Biological exposure index). Particulate 81 Particulate matter in smoke is the most visible byproduct of combustion. Particulate formed by combustion can serve as an aggregation that can contain condensed volatile organics, such as solvents (benzene, toluene etc.), as well as most any other contairiinant (Ware 1991). Research has shown that the majority of particulate released by combustion are respirable, less than 2.5 um in diameter (aerodynamic), and are capable of penetrating as deep as the alveoli during inspiration. Up to 70% of the total particulate mass is composed of particles 2.5 um in diameter that poses a significant respirable threat, which combined with the increased breathing rates of labouring firefighters, allow them to penetrate deeper into the lungs (WCB report 1992). Concern has been drawn to the potential that some chemicals can adhere onto the surface and be carried deep into the lungs with the particles (WCB report 1992). However, the extent of this problem remains unknown as at most fires particulate is not characterized beyond that of total mass (Burgess et. al. 1995) and that their role as irritants in fire is completely unknown (Ware 1991). Soot Soot is created at all fires and it is likely that firefighters are significantly exposed. They contain PAH's of which many are found to be carcinogenic (WCB report 1992). Another factor to consider is that it may behave like activated carbon, as it is primarily composed of carbon, and adsorbs chemicals onto its surface. In this manner it is believed that it can act as a vector and transport various chemicals into the body (Committee on fire toxicology 1986, Guidotti and Clough 1992). 82 0 Appendix IB Occupational exposure limits for chemicals The occupational exposure limits are set by regulating bodies. In British Columbia the WCB sets the limits, and in April 1998 new regulations come into effect, and for some chemicals the exposure limit will change: any limits, used in this thesis use the new 1998 exposure limits for comparison. Standards defining acceptable levels of chemicals to which workers may be exposed and not constitute a threat to a non susceptible individual has been a difficult issue. Many regulatory bodies such as the National Institute of Safety and Health (NIOSH), the Occupational Safety and Health (OSHA), and the Workers' Compensation Board of British Columbia all publish acceptable levels, however there seems to be little agreement between one agency and another. The current exposure limits for British Columbia, and the US (NIOSH and OSHA) is included as table A2. Table A2: Current exposure limits of common combustion byproducts Contaminant Exposure Limit STEL (ppm) IDLH (ppm) NIOSH B.C. NIOSH OSHA B.C. NIOSH OSHA Carbon monoxide (CO) 25 25 35 100 200 200 1500 Carbon dioxide (C02) 5000 5000 10,000 15,000 30,000 30,000 50,000 Hydrogen cyanide (HCN) 5 **none **none 5 4.7 4.7 50 Sulfur dioxide (S02) 2. 2 2 5 5 5 100 *Benzene 0.5 10 1 2.5 32 5 3000 •Formaldehyde (aldehydes) 0.3 0.1 0.75 1 1 2 . 30 Total particulate (not otherwise classified) 10 mg/m3 of whic diameter of 13 mg/m3 may be respirable or <2.5 um. No STEL established lave a **none -* substances with an asterix beside them are classified as IARC carcinogens or potential human carcinogens -**none: refers to not listed or not found, it does not necessarily mean that this organization has not established values Abbreviations: • STEL - acronym for - short term exposure limit. This is defined as the maximum allowable concentration that a worker may be exposed for 15 minutes. • IDLH - acronym for - immediately dangerous to life and health • ALARA - acronym for - as low as reasonably achievable • Exposure limits: This is the maximum concentration that a worker may be exposed 8 hours a day for 40 hours per week where non susceptible individuals should experience no adverse health effects. In BC they are called permissible and by the American Conference of Governmental Industrial Hygienists [ACGIH] they are referred to as threshold limit values (TLV). They may also be expressed as the TLV- time weighted average or average, which is the concentration a chemical can expose a worker over a 40 hour work week: the level may fluctuate above the level for brief periods (15 minutes) however the average concentration can not exceed the TLV-TWA. 83 Appendix II Individual measurements from sampling sessions 1. Outdoor sites: T-pit and Round tank Weather Conditions Outdoor pits (T-pit and Round tank) June 27: cloudy, light NW wind, temperature 15.8 °C June 28: rain 30 minutes prior to sampling, calm wind, temperature 13.6 °C Sept. 12: sunny, calm winds, temp 20.1 °C Sept. 26: sunny with few clouds, calm winds, temp 12.7 °C Oct. 24: cloudy, overcast, light rain, calm to light NW winds, temp 8 °C Oct. 31: clear, sunny, light southerly wind, temp 10.1 °C Nov. 21: partial clear, calm winds, snow accumulation 10 cm, temp -1.2 °C Nov. 28: mainly clear, calm wind, temp 6.7 °C Dec. 5: cloudy, constant rain, calm winds, temp 4.5 °C Observations of burning: 1. June 27 - samples were collected at the round tank and T-pit (areas 1 and 2) for a 46 minute duration. During this period 4 consecutive fires were lit and suppressed in the round tank with a total fire burning duration of 20 minutes over the 46 sampling period. Fire sizes were generally small (as pointed out by experienced staff at the site) with the wind carrying smoke over site 2 2. June 28 - samples were collected for 30 minutes, at areas 1 and 2, while 2 fires were lit and suppressed in the T-pit, and one fire being lit in the round tank. The two fires in the T-pit were suppressed within 3 minutes each while the round tank burned continuously for 23 minutes and reached a moderate size. Smoke plumes generally went straight up but occasionally drifted near site 2. 3. Sept 12 - samples were collected at the flange and T-pit (areas 1 and 3) for a 20 minute duration. During this period 3 consecutive fires were lit and suppressed in the round tank with a total fire burning duration of 10 minutes over the 20 minute sampling period. Fire sizes were generally very small with the smoke plumes traveling upwards. 4. Sept 21 - samples (area 3) were recorded for 13 minutes duration while 2 fires were lit and suppressed in the T-pit and fire sizes were generally considered moderate size. Smoke plumes generally went upwards. 5. Sept 26 - two sampling periods were conducted. The first sampling was at the T-pit (area 3) and lasted 28 minutes while 4 fires were lit and suppressed with the fire sizes generally being moderate size. Total burning time was 12 minutes. Smoke plumes generally went generally upwards and drifted partially towards the east, within a few meters of sampling area 3. The second samples were conducted at area 3 also while 4 fires were lit in the flange tank. The sampling duration was 23 minutes while fire burning occurred for 11 minutes, with fire size being moderate. Smoke plumes generally drifted up and east, but not toward area 3. 6. Oct 24 - samples were collected at the T-pit during morning burning and afternoon burning sessions. In the morning area 3 was sampled and area 3 and 4 were sampled in the afternoon. In the morning period 7 fires were lit with a total fire duration of 25 minutes; fire sizes were generally small to moderate in size with plumes traveling upwards. In the afternoon 7 fires were lit for a burning duration of 22 minutes, again fire sizes were generally very moderate or small with the smoke plumes traveling upwards. In the morning and afternoon it was noted that after suppression of the fires was complete that smoke fumes tended to he low to the ground and leave a lingering smell afterwards. 84 7. Oct. 31 - samples were collected at the T-pit during an afternoon burning sessions. During this period 7 fires were lit with a total fire duration of 21 minutes; fire sizes were generally moderate too large with plumes traveling straight upwards. 8. Nov. 21 - samples were collected at the T-pit (areas 3 and 4) for a 35 minute duration. During this period 7 consecutive fires were lit and suppressed in the pit with a total fire burning duration of 20 minutes over the 35 minute sampling period. Fire sizes were generally moderate too large with the smoke plumes traveling upwards. 9. Nov. 28 - samples were collected at the T-pit (areas 3 and 4) for a 30 minute duration. During this period 8 consecutive fires were lit and suppressed in the pit with a total fire burning duration of 19 minutes over the 30 minute sampling period. Fire sizes were generally moderate to large with the smoke plumes traveling upwards 10. Dec. 5 - samples were collected at the round tank (areas 1 and 2) for a 25 minute duration. During this period 2 consecutive fires were lit and suppressed in the pit with a total fire burning duration of 12 minutes over the 25 minute sampling period. Fire sizes were generally small. During the first fire the action of the firefighters hoses' streams resulted in pushing the smoke plume towards area 2 with exposure potential occurring at the site, while concurrent winds blew this smoke back towards site 1 with exposure potential occurring here as well. During the second fire scenario smoke plumes were not pushed towards area 2 but tended to drift towards area 1. 85 Summary of the concentration of contaminants collected at the round tank and T-pit Contaminant to/urca I.OI) Concentration 1JCN 27-Jun / area 1 0.02 ppm below LOD 27-Jun / area 2 0.02 ppm below LOD 28-Jun / area 2 0.03 ppm below LOD 28-Jun / area 3 0.03 ppm 0 05 ppm 12-Sep/area 1 0.02 ppm below LOD 12-Sep / area 3 0.02 ppm below LOD 21-Sep / area 3 0.04 ppm below LOD 26-Sep / area 3 0.02 ppm 0 02 ppm 26-Sep / area 3 0.02 ppm below LOD 24-Oct / area 3 morn 0.01 ppm below LOD 24-Oct / area 3 aft 0.01 ppm below LOD 24-Oct / area 4 aft 0.01 ppm below LOD 31-Oct/area 3 0.02 ppm below LOD 21-Nov/area 3 0.01 ppm below LOD 21-Nov/ area 4 0.01 ppm below LOD 28-Nov / area 3 0.02 ppm below LOD 28-Nov / area 4 0.02 ppm below LOD 5-Dec / area 1 0.02 ppm below LOD 5-Dec / area 2 0.02 ppm 0 ] ] ppm H i H & d e 27-Jun / area 1 0.005 ppm 0 012 ppm 27-Jun / area 2 0.005 ppm 0 01 ppm 28-Jun / area 2 0.005 ppm 0 011 ppm 28-Jun / area 3 0.005 ppm . below LOD 12-Sep/area 1 0.006 ppm below LOD 12-Sep/area 3 0.006 ppm 0.045 ppm 21-Sep/ area 3 0.01 ppm below LOD 26-Sep / area 3 0.005 ppm below LOD 26-Sep / area 3 0.005 ppm 0 13 ppm 24-Oct / area 3 morn 0.006 ppm below LOD 24-Oct / area 3 aft 0.006 ppm below LOD 24-Oct / area 4 aft 0.006 ppm below LOD 31-Oct/area3 0.004 ppm below LOD 21 -Nov/ area 3 0.009 ppm below LOD 21-Nov/area4 0.009 ppm 0.01 ppm 28-Nov / area 3 0.01 ppm below 1 Ol) 28-Nov / area 4 0.01 ppm 0 01*; ppm 5-Dec / area 1 0.007 ppm 0 04 ppm 5-Dec / area 2 0.007 ppm 0 02 ppm 86 tank and T-pit Cont Particulate 12-Sep / area 1 0. 3 mg/m3 0.374 me/m' 12-Sep / area 3 0.4 mg/m3 1.43 mg/m"' 21-Sep / area 3 0.6 mg/m3 below LOD 26-Sep / area 3 0.3 mg/m3 below LOD 26-Sep / area 3 0.3-mg/m3 below LOD 24-Oct / area 3 mom 0.2 mg/m3 0.31 nig/nr' 24-Oct / area 3 aft 0.3 mg/m3 0 38 mg/ni1 24-Oct / area 4 aft 0.3 mg/m3 0.32 mg/m1 31-Oct/area3 0.2 mg/m3 below I.OI) 21-Nov/ area 3 0.2 mg/m3 0 27 rag/nr" 21-Nov/area4 0.2 mg/m3 0 48 mg/m' 28-Nov / area 3 0.3 mg/m3 0.57 mg/inJ 28-Nov / area 4 0.3 mg/m3 0 36 iiig/mJ 5-Dec / area 1 0.3 mg/m3 1.86 ing/mJ 5-Dec / area 2 0.3 mg/m3 18 1 nig/in' Benzene 12-Sep / area 1 0.3 ppm below LOD 12-Sep / area 3 0.3 ppm below LOD 21-Sep / area 3 0.7 ppm below LOD 26-Sep / area 3 0.3 ppm below LOD 26-Sep / area 3 0.3 ppm below LOD 24-Oct / area 3 morn 0.2 ppm below LOD 24-Oct / area 3 aft 0.2 ppm below LOD 24-Oct / area 4 aft 0.2 ppm below LOD 31-Oct/area3 0.2 ppm below LOD 21-Nov / area 3 0.2 ppm below LOD 21-Nov/area 4 0.2 ppm 0 24 mini 28-Nov / area 3 0.3 ppm below LOD 28-Nov / area 4 0.3 ppm below LOD 5-Dec / area 1 0.2 ppm below LOD 5-Dec / area 2 0.2 ppm below LOD *note that benzene and particulate samp es were not done in the June sampling. 87 Summary of CO and SO? collected at round tank and T-pit C <ml;imiri :int Diiti7iircii i o n M : i \ Conccntnition CO 27-Jun / area 1 lppm below LOD 27-Jun / area 2 lppm 8 28-Jun / area 2 lppm below LOD 28-Jun / area 3 1 ppm below LOD 12-Sep / area 1 lppm below LOD 12-Sep / area 3 1 ppm below LOD 21-Sep/area3 1 ppm below LOD 26-Sep / area 3 lppm 2 26-Sep / area 3 lppm below LOD 24-Oct / area 3 morn 1 ppm 3 24-Oct / area 3 aft 1 ppm below LOD 24-Oct / area 4 aft 1 ppm 6 31-Oct/area3 lppm below LOD 21-Nov/area 3 lppm below LOD 21-Nov/area4 lppm below LOD 28-Nov / area 3 1 ppm below LOD 28-Nov / area 4 lppm below LOD 5-Dec / area 1 lppm below LOD 5-Dec / area 2 lppm 19 S0 2 27-Jun / area 1 0.1 ppm below LOD 27-Jim / area 2 0.1 ppm 0.7 28-Jun / area 2 0.1 ppm below LOD 28-Jun / area 3 0.1 ppm below LOD 12-Sep / area 1 0.1 ppm below LOD 12-Sep / area 3 0.1 ppm below LOD 21-Sep / area 3 0.1 ppm below LOD 26-Sep / area 3 0.1 ppm 0.2 26-Sep / area 3 0.1 ppm below LOD 24-Oct / area 3 morn 0.1 ppm below LOD 24-Oct / area 3 aft 0.1 ppm below LOD 24-Oct / area 4 aft 0.1 ppm below LOD 31-Oct/area 3 0.1 ppm below LOD 21-Nov/area 3 0.1 ppm below LOD 21-Nov/ area 4 0.1 ppm below LOD 28-Nov / area 3 0.1 ppm below LOD 28-Nov / area 4 0.1 ppm below LOD 5-Dec / area 1 0.1 ppm below LOD 5-Dec / area 2 0.1 ppm below LOD 88 Extinguisher pad Summary of concentration of contaminants collected at the extinguisher pad. Contaminant Date S:impk' t ) | K - 1 Ol) (finrriilrulion (ppm) HCN Nov 23 small tray 0.04 ppm 0.M ppm Nov 23 e-tray 0.04 ppm 031ppm Nov 23 large tray 0.04 ppm 0.13 ppm Dec 5 small tray 0.04 ppm 011 ppm Dec 5 large tray 0.04 ppm 1.03 ppm Dec 5 large extinguisher 0.04 ppm below 1 Ol) Jan 7 small tray 0.04 ppm 0 51 ppm Jan 7 e-tray 0.04 ppm below 1 OI) Jan 21 small tray 0.04 ppm 0.1'J ppm Jan21 e-tray 0.04 ppm 0 -16 ppm Nov 23 small tray 0.01 ppm below LOD Nov 23 e-tray 0.01 ppm below LOD Nov 23 large tray 0.01 ppm 0 03 ppm Dec 5 small tray 0.01 ppm 0 01 ppm Dec 5 large tray 0.01 ppm below l.Ol) Dec 5 large extinguisher 0.01 ppm • 0.06 ppm Jan 7 small tray 0.01 ppm 0 082 ppm Jan 7 e-tray 0.01 ppm below LOD Jan 21 small tray 0.01 ppm below LOD Jan 21 e-tray 0.01 ppm 0 038 ppm lBcnzene Nov 23 small tray 1.0 ppm 1 26 ppm Nov 23 e-tray 1.0 ppm 1 29 ppm Nov 23 large tray 1.0 ppm below LOD Dec 5 small tray 1.0 ppm below LOD Dec 5 large tray 1.0 ppm below LOD Dec 5 large extinguisher 1.0 ppm below LOD Jan 7 small tray 1.0 ppm below LOD Jan 7 e-tray 1.0 ppm below LOD Jan 21 small tray 1.0 ppm below LOD Jan 21 e-tray 1.0 ppm below LOD I'arlirulalc Nov 23 small tray 0.5 mg/m3 XI 7 nm/m' Nov 23 e-tray 0.5 mg/m3 17 2 nm/m1 Nov 23 large tray 0.5 mg/m3 152 7 inp/m' Dec 5 small tray 0.5 mg/m3 331 3 mg'm' Dec 5 large tray 0.5 mg/m3 474 2 I I I « L ' H I * Dec 5 large extinguisher 0.5 mg/m3 573.7 irig/'m1 Jan 7 small tray 0.5 mg/m3 481 ni^ /m5 Jan 7 e-tray 0.5 mg/m3 11 (>9 7 m^ /i-n * Jan 21 small tray 0.5 mg/m3 388 9 mg/nf Jan 21 e-tray 0.5 mg/m3 196 <> mfi'm' 89 Summary of CO. CO?, and SO? collected at the extinguisher pad Date Site M;i\. nuic. recorded CO Nov 23 small tray below LOD Nov 23 e-tray below LOD Nov 23 large tray 81 ppm Dec 3 small tray 5 ppm Dec 3 large tray 3 ppm Dec 3 large extinguisher 10 ppm Jan 7 small tray 7 ppm Jan 7 e-tray 1 ppm Jan 21 small tray 1 ppm Jan 21 e-tray 2ppm s o 2 Nov 23 small tray below LOD Nov 23 e-tray below LOD Nov 23 large tray 0.5 ppm Dec 3 small tray below LOD Dec 3 large tray below LOD Dec 3 large extinguisher 0.7 ppm Jan 7 small tray 0.3 ppm Jan 7 e-tray 0.1 ppm Jan 21 small tray below LOD Jan 21 e-tray below LOD c o 2 Nov 23 small tray 3190+ppm Dec 3 small tray 3190+ppm Jan 7 small tray 1256 ppm Jan 7 e-tray 789 ppm Jan 21 small tray 658 ppm Jan 21 e-tray 2160 ppm Note that C 0 2 was not measured at the extinguisher pad on Dec 3 and Nov 23. 90 Indoor sites: Background levels: Ship and burn building Summary of background concentration of contaminants in the burn building Contaminant Date & sample 1 iine since l.OI) Concentration n I I in her last burning (ppm) CO •July 11 120 hours 1 ppm below LOD October 15 72 hours 1 ppm below LOD SO: July 11 120 hours 0.1 ppm below LOD October 15 72 hours 0.1 ppm below LOD July 11 120 hours 1 ppm below LOD October 15 72 hours lppm below LOD *licn/cnc July 11 #1 120 hours 0.3 ppm below LOD #2 120 hours 0.3 ppm below LOD #3 120 hours 0.3 ppm below LOD #4 120 hours 0.3 ppm below LOD •IICN July 11 #1 120 hours 0.01 ppm below LOD #2 120 hours 0.01 ppm below LOD #3 120 hours 0.01 ppm below LOD #4 120 hours 0.01 ppm below LOD October 15 #1 72 hours 0.01 ppm below LOD #2 72 hours 0.01 ppm below LOD #3 72 hours 0.01 ppm below LOD #4 72 hours 0.01 ppm below LOD October 15 72 hours 0.02 ppm below LOD l-'onnaldel July 11 #1 120 hours 0.004 ppm 0 06=: ppm #2 120 hours 0.004 ppm below LOD #3 120 hours 0.004 ppm below LOD #4 120 hours 0.004 ppm below LOD October 15 #1 72 hours 0.007 ppm 0 045 ppm #2 72 hours 0.007 ppm below LOO #3 72 hours 0.007 ppm 0 12 ppm #4 72 hours 0.007 ppm 0 10 ppm Particulate July 11 #1 120 hours 0.09 mg/m3 below LOD #2 120 hours 0.09 mg/m3 below LOD #3 120 hours 0.09 mg/m3 below LOD October 15 #1 72 hours 0.08 mg/m3 0.33 mg/in' #2 72 hours 0.08 mg/m3 below l.OI) #3 72 hours 0.08 mg/m3 0.15 nig/in' #4 72 hours 0.08 mg/m3 0.1 mg/in' *Note: for Benzene no samples were taken in October 91 Summary of background concentration of contaminants in the ship mock-up Contaminant Date and Time since LOD Concentration sample number last burning (ppm) CO Jan 24 168 hours lppm below LOD Feb 5 168 hours 1 ppm below LOD so 2 Jan 24 168 hours 0.1 ppm below LOD Feb 5 168 hours 0.1 ppm below LOD Benzene . Jan 24 #1 168 hours 0.6 ppm below LOD #2 168 hours 0.6 ppm below LOD • #3 168 hours 0.6 ppm below LOD #4 168 hours 0.6 ppm below LOD Feb 5 #1 168 hours 0.6 ppm below LOD #2 168 hours 0.6 ppm below LOD #3 168 hours 0.6 ppm below LOD #4 168 hours 0.6 ppm below LOD Formaldehyde Jan 24 #1 168 hours 0.004 ppm below LOD #2 168 hours 0.004 ppm below LOD #3 168 hours 0.004 ppm below LOD #4 168 hours, 0.004 ppm below LOD Feb 5 #1 168 hours 0.004 ppm below LOD #2 168 hours 0.004 ppm below LOD #3 168 hours 0.004 ppm below LOD #4 168 hours 0.004 ppm below LOD HCN Jan 24 #1 168 hours 0.01 ppm below LOD #2 168 hours 0.01 ppm below LOD #3 168 hours 0.01 ppm below LOD #4 168 hours 0.01 ppm below LOD Feb 5 #1 168 hours 0.01 ppm below LOD #2 168 hours 0.01 ppm below LOD #3 168 hours 0.01 ppm below LOD #4 168 hours 0.01 ppm below I .Ol) Particulate Jan 24 #1 168 hours 0.09 mg/m3 0.123 ma/nr1 . #2 168 hours 0.09 mg/m3 0.134 mg/m3 #3 168 hours 0.09 mg/m3 0.122 iiic/m1 #4 168 hours 0.09 mg/m3 0.156 ms/m1 Feb 5 #1 168 hours 0.09 mg/m3 0.09 mg/iiv' #2 168 hours 0.09 mg/m3 0.11 mg/m' #3 168 hours 0.09 mg/m3 0.091 ine/in' #4 168 hours 0.09 mg/m3 below LOD Unvented contaminant levels: Burn building Summary of unvented contaminant levels in the burn buile ling. Contaminant Date LOD Mux. Cone, recorded Benzene July 0.3 ppm 0.4 12 ppm Benzene October 0.3 ppm below LOD HCN July 0.01 ppm 3.50 ppm HCN October 0.02 ppm 2.04 ppm *Formaldehyde October 0.0075 ppm 2:99 ppm Particulate October 0.08 mg/m3 1.39 mg/m3 CO July lppm 859 ppm CO October lppm 544 ppm C0 2 July 1 ppm 3190 +ppm CO2 October 1 ppm 3456 ppm S02 July 0.1 ppm 52.1 ppm S02 October 0.1 ppm 9.6 ppm *Note formaldehyde and particulate were lost in the July session. 92 Vented contaminant levels in the Ship and Burn building Summary of contaminant levels collected in the burn building collected 10-40 minutes following SiiinpK- typo Date I'lillrclt'd Locution • (>•> Concentration July 19 level 1 0.01 ppm 0.45 ppm *July 23 # 1 level 2 0.01 ppm 0.17 ppm *July 23 #2 level 3 0.02 ppm 1.79 ppm July 24 level 1 0.03 ppm 0.084 ppm July 25 level 3 propane 0.02 ppm 0 18 ppm Sept 12 level 1 0.02 ppm below LOD Sept 21 level 1 0.02 ppm below LOD Sept 22 level 1 0.02 ppm below LOD Sept 28 level 1 stairs 0.03 ppm below LOD Oct 8 level 1 0.02 ppm 0 42 ppm Oct 9 level 1 0.02 ppm 0.11 ppm Nov 16 level 1 0.01 ppm 0 04 ppm Nov 17 level 2 0.01 ppm 0.03 ppm Nov 23 level 1 0.02 ppm 0.26 ppm (glima Idol i> lie •July 19 #1 level 1 0.005 ppm 0 80 ppm *July 19 #2 level 1 0.005 ppm 0 52 ppm * July 23 #1 level 2 0.01 ppm 0 ln ppm *July 23 #2 level 3 0.01 ppm 2.0 ppm July 24 level 1 0.01 ppm 0 413 ppm July 25 level 3 0.006 ppm 0 318 ppm Sept 12 level 1 0.008 ppm 0 43 ppm Sept 21 level 1 0.008 ppm 0 605 ppm Sept 22 level 1 0.01 ppm 0 \7 ppm Sept 28 level 1 stairs 0.008 ppm 0 ppm Oct 8 level 1 0.02 ppm 0 7 ppm Oct 9 level 1 0.01 ppm I 6S ppm Nov 16 level 1 0.01 ppm 0 1X ppm Nov 17 level 2 0.01 ppm 0.1" ppm Nov 23 level 1 0.01 ppm 0 6> ppm Particulate *July 19 #1 level 1 0.08 mg/m3 0 167mg.ni' •July 23 #1 level 2 0.12 mg/m3 below I Ol) •July 23 #2 level 3 0.09 mg/m3 0 182 mg'm' July 24 level 1 0.09 mg/m3 2 26 mu m July 25 level 3 0.2 mg/m3 below 1 Ol) Sept 12 level 1 0.1 mg/m3 0 15(> mg in" Sept 21 level 1 0.09 mg/m3 below I Ol) Sept 22 level 1 0.09 mg/m3 0 842 mg m' Sept 28 level 1 stairs 0.09 mg/m3 0 145 iniym' Oct 8 level 1 0.2 mg/m3 below I.OD Oct 9 level 1 0.1 mg/m3 0 193 me in1 Nov 16 level 1 0.2 mg/m3 below 1 Ol) Nov 17 level 2 0.2 mg/m3 below I Ol) Nov 23 level 1 0.2 mg/m3 2 40mgm' **llonzenc July 19 level 1 0.2 ppm below LOD •July 23 #1 level 2 0.3 ppm below LOD •July 23 #2 level 3 0.5 ppm below LOD July 24 level 1 0.5 ppm below LOD July 25 level 3 0.3 ppm below LOD *Note that #1 refers to samples collected at noon after the morning burning session and #2 refers to samples collected in the evening after the afternoon burning session. **Also note that benzene was not collected after the July sampling schedule. 93 The peak concentration of combustion gases collected immediately following venting in the burn (Das Dutr collected Location Max. com-, recorded Ambient outdoor com CO July 19 level 1 11 ppm N/A "July 23 #1 level 2 5 ppm N/A •July 23 #2 level 3 46 ppm N/A July 25 level 3 17 ppm • N/A Sept 21 level 1 12 ppm N/A Sept 22 level 1 19 ppm N/A Sept 28 level 1 stairs 38 ppm N/A Oct 8 level 1 8 ppm N/A Oct 9 level 1 46 ppm N/A Nov 16 level 1 33 ppm N/A Nov 17 level 2 18 ppm N/A Nov 23 level 1 137 ppm N/A so2 July 19 level 1 0.1 ppm N/A •July 23 #1 level 2 0.0 ppm N/A •July 23 #2 level 3 2.2 ppm N/A July 25 level 3 0.0 ppm N/A Sept 21 level 1 0.4 ppm N/A Sept 22 level 1 0.5 ppm N/A Sept 28 level 1 stairs 0.3 ppm N/A Oct 8 level 1 0.0 ppm N/A Oct 9 level 1 1.5 ppm N/A Nov 16 level 1 0.0 ppm N/A Nov 17 level 2 0.0 ppm N/A Nov 23 level 1 1.7 ppm N/A co2 July 19 level 1 653 ppm 354 •July 23 #1 level 2 417 ppm 338 •July 23 #2 level 3 .984 ppm 322 July 25 level 3 590 ppm 350 Sept 12 level 1 401 ppm 322 Sept 21 level 1 952 ppm 349 Sept 22 level 1 410 ppm 347 Sept 28 level 1 stairs 561 ppm 347 Oct 8 level 1 405 ppm 342 Oct 9 level 1 720 ppm 342 Nov 16 level 1 527 ppm 338 Nov 17 level 2 748 ppm 354 Nov 23 level 1 1674 ppm 323 N/A: not applicable as the ambient outdoor concentration was below the limit of detection of the equipment (1 ppm for CO, CO2 and 0.1 ppm for S02). *Note that #1 refers to samples collected at noon after the morning burning session and #2 refers to samples collected in the evening after the afternoon burning session. Highest recorded vented contaminant samples in the burn building Siunpli- n p e D.IU mi l l H i d I.OI) ( ' i inu ' i i i ra t iun HCN July 23 0.02 ppm 1.79 ppm Formaldehyde July 23 0.01 ppm 2.43 ppm Benzene N/A 0.3 ppm no samples > I Ol) Particulate July 24 0.09 mg/m3 2.26 mg/m3 S02 July 23 0.1 ppm 2.2 ppm CO July 23 & Oct 9 lppm 46 ppm C0 2 July 23 lppm 984 ppm 94 Summary of concentration of contaminants in the burn building collected over 3 hours following venting Sample type Date 0-30 minuii's JO-MI minutes MM 20 minutes 120-ISO minutes folleilid after u-miii". after \enting after venting after venting HCN July 23 1.79 ppm 1.72 0.16 0.08 July 24 0.084 ppm 0.1 0.19 0.05 July 25 0.18 ppm 0.17 0.05 0.04 Sept 21 0.03 below LOD below LOD below LOD Sept 22 0.03 below LOD below LOD below LOD Sept 28 below LOD below LOD below LOD below LOD Oct 8 0.42 ppm 0.12 0.14 0.03 Oct 9 0.11 ppm 0.12 below LOD below LOD Nov 16 0.04 ppm 0.08 0.04 below LOD Nov 17 0.03 ppm 0.21 0.09 0.04 Formaldehyde July 23 0.77 ppm 0.413 0.389 0.137 July 24 2.0 ppm 2.43 0.898 0.51 July 25 0.413 ppm 0.228 0.182 0.102 Sept 21 0.43 ppm 0.215 0.188 0.121 Sept 22 0.605 ppm 0.28 0.15 0.045 Sept 28 0.37 ppm 0.06 0.04 0.03 Oct 8 0.39 ppm 0.16 0.08 0.05 Oct 9 0.7 ppm 0.48 0.41 lost Nov 16 1.68 ppm 0.54 0.32 0.20 Nov 17 0.18 ppm 0.15 0.05 0.03 Particulate July 23 0.167 mg/m3 0.182 0.055 below LOD July 24 0.182 mg/m3 0.167 below LOD below LOD July 25 2.26 mg/m3 0.089 0.059 0.069 Sept 21 0.156 mg/m3 0.082 below LOD below LOD Sept 22 below LOD below LOD 0.09 below LOD Sept 28 0.842 mg/m3 0.34 0.09 0.11 Oct 8 0.145 mg/m3 0.094 0.051 0.1 Oct 9 below LOD 0.15 below LOD 0.13 Nov 16 0.193 mg/m3 0.64 0.11 0.12 Benzene July 19 below LOD below LOD below LOD below LOD •July 23 #1 below LOD below LOD below LOD below LOD •July 23 #2 below LOD below LOD below LOD below LOD July 24 below LOD below LOD below LOD below LOD July 25 below LOD below LOD below LOD below LOD 95 Summary of peak concentration of combustion gases in the burn building collected over 3 hours following venting Sample type Date 0-311 in i null's 30-60 minutes 60-120 minutes 120-ISO miiiiiu-s collected after tenting .itiu \eiilni<! alter \ en linn siller \eiitiiii> CO July 23 46 12 10 8 July 24 10 11 4 1 July 25 14 8 7 6 Sept 21 12 10 5 2 Sept 22 19 15 10 7 Sept 28 37 5 below LOD below LOD Oct 8 8 3 below LOD below LOD Oct 9 46 37 10 2 Nov 16 33 38 8 4 Nov 17 19 14 14 9 S0 2 July 23 2.1 2.2 0.9 0.1 July 24 below LOD below LOD below LOD below LOD July 25 below LOD below LOD below LOD below LOD Sept 21 0.3 0.5 0.1 below LOD Sept 22 0.4 0.5 0.1 below LOD Sept 28 0.3 below LOD below LOD below LOD Oct 8 below LOD below LOD below LOD below LOD Oct 9 2.2 0.7 0.7 0.3 Nov 16 below LOD below LOD below LOD below LOD Nov 17 below LOD below LOD below LOD below LOD C 0 2 July 23 1063 701 496 370 July 24 606 527 401 386 July 25 590 480 469 433 Sept 21 402 323 323 323 Sept 22 995 475 396 365 Sept 28 556 414 387 363 Oct 8 406 390 374 359 Oct 9 721 626 548 390 Nov 16 559 496 464 386 96 Summary of concentration of contaminants in the ship mock-up collected 10 to 40 minutes following venting. . S:mi|)li- Dale uilleeteil LOD \mliii 111 iiiiidiinr levels Max cone, over the sampling perind HCN Jan 17 0.02 ppm N/A 0 24 ppm Jan 31 0.02 ppm N/A 0 024 ppm Feb 28 0.02 ppm N/A 0 17 ppm Mar7 0.02 ppm N/A 0 21 ppm Mar 14 0.02 ppm N/A 0 40 ppm Formaldehyde Jan 17 0.01 ppm N/A 0163 ppm Jan 31 0.01 ppm N/A 0 069 ppm Feb 28 0.01 ppm N/A 0152 ppm Mar 7 0.01 ppm N/A 0 182 ppm Mar 14 0.01 ppm N/A 0 '24 ppm Particulate Jan 17 0.09 mg/m3 N/A 0 25 m j W Jan 31 0.09 mg/m3 N/A 0 523 nm/in' Feb 28 0.09 mg/m3 N/A 0 57 ing/m1 Mar 7 0.09 mg/m3 N/A 1 01 mg/m' Mar 14 0.09 mg/m3 N/A 0 i>29 mg/mJ Benzene Jan 17 0.7 ppm N/A 1 85 ppm Jan 31 0.7 ppm N/A below 1 Ol) Feb 28 0.7 ppm N/A 1 17 ppm Mar 7 0.7 ppm N/A 1 2 ppm Mar 14 0.7 ppm N/A below LOD CO Jan 17 1 ppm N/A 2 1 ppm Jan 31 1 ppm N/A 16 ppm Feb 28 1 ppm N/A 52 ppm Mar7 1 ppm N/A .-**>: 36 ppm -Mar 14 1 ppm N/A 95 ppm S0 2 Jan 17 0.1 ppm N/A 1 9 ppm Jan 31 0.1 ppm N/A 1 8 ppm Feb 28 0.1 ppm N/A 3 1 ppm Mar 7 0.1 ppm N/A 3 3 ppm Mar 14 0.1 ppm N/A 8 ^ ppm *co2 Jan 17 1 ppm 354 2287 ppm Jan 31 1 ppm 338 1290 ppm Feb 28 1 ppm 322 1684 ppm Mar 7 1 ppm 350 MOO ppm Mar 14 1 ppm 322 24^ 0 ppm Note that for C 0 2 the ambient outdoor levels are included, but for all other contaminants N/A (not applicable) has been entered as ambient levels if they occur are below the limit of detection for these methods. 97 r Summary of concentration of contaminants in the ship mock-up collected over 3 hours following Sample t\ pe Date lolleiled 0-30 minutes nt'ii- r \eiiliii". 30-60 minutes after venting 60-1211 minutes after venting 120-ISO minutes .liter \enting HCN Jan 17 0.24 ppm 0.024 „ 0.02 below LOD Jan 31 0.024 ppm below LOD below LOD below LOD Feb 28 0.17 ppm 0.024 below LOD below LOD Mar 7 0.21 ppm 0.025 below LOD below LOD Mar 14 0.46 ppm 0.17 0.01 below LOD Formaldehyde Jan 17 0.163 ppm 0.154 0.1 0.067 Jan 31 0.069 ppm 0.04 0.047 0.043 Feb 28 0.152 ppm 0.179 0.135 0.087 Mar 7 0.182 ppm 0.104 0.124 0.081 Mar 14 0.324 ppm 0.253 0.178 0.091 Particulate Jan 17 0.25 mg/m3 0.125 below LOD below LOD Jan 31 0.523 mg/m3 below LOD below LOD 0.09 Feb 28 0.57 mg/m3 0.32 0.09 below LOD Mar 7 1.01 mg/m3 below LOD 0.095 0.12 Mar 14 0.629 mg/m3 0.17 0.01 below LOD Benzene Jan 17 1.85 ppm 1.17 0.8 0.75 Jan 31 below LOD below LOD below LOD below LOD Feb 28 1.17 ppm 1.09 0.76 0.58. Mar 7 1.2 ppm 1.14 0.63 0.48 Mar 14 below LOD below LOD below LOD below LOD CO Jan 17 24 ppm 2 below LOD below LOD Jan 31 16 ppm below LOD below LOD below LOD Feb 28 52 ppm below LOD below LOD below LOD Mar 7 36 ppm below LOD below LOD below LOD Mar 14 95 ppm. below LOD below LOD below LOD S0 2 Jan 17 1.9 ppm below LOD below LOD below LOD Jan 31 1.8 ppm below LOD below LOD below LOD Feb 28 3.1 ppm below LOD below LOD below LOD Mar 7 3.3 ppm below LOD below LOD below LOD Mar 14 8.4 ppm below LOD below LOD below LOD *co2 Jan 17 2287 ppm 738 376 ambient Jan 31 1290 ppm 478 390 ambient Feb 28 1684 ppm 767 456 ambient Mar 7 1400 ppm 678 421 ambient Mar 14 2456 ppm 876 467 ambient Sample type Date collected K M ) ( iiiu'eniratiiiii HCN Mar 14 0.02 ppm 0.46 ppm Formaldehyde Mar 14 0.01 ppm 0.324 ppm Benzene Jan 17 0.7 ppm 1.85 ppm Particulate Mar 7 0.09 mg/m3 1.01 mg/m3 S02 Mar 14 0.1 ppm 8.4 ppm CO Mar 14 1 ppm 95 ppm C0 2 Mar 14 1 ppm 2456 ppm 98 Propane side of the burn building Summary of concentration of contaminants in the propane side of the burn building collected 10 to 40 minutes fol owing venting. S.implc t\pe Date collected 1 OD Max cone, O U T the \ainplint; period HCN Sept 26 0.02 ppm below LOD Oct 24 0.02 ppm below LOD Oct 31 0.02 ppm below LOD Nov 21 0.02 ppm belowLOD Nov 28 0.02 ppm below LOD Formaldehyde Sept 26 0.007 ppm 0.35 ppm Oct 24 0.007 ppm 0.06 ppm Oct 31 0.007 ppm 0.05 ppm Nov 21 0.007 ppm 0.05 ppm Nov 28 0.007 ppm 0.03 ppm Particulate Sept 26 0.09 mg/m3 below LOD Oct 24 0.09 mg/m3 below LOD Oct 31 0.09 mg/m3 below LOD Nov 21 0.09 mg/m3 below LOD Nov 28 0.09 mg/m3 below LOD Benzene Sept 26 0.7 ppm below LOD Oct 24 0.7 ppm below LOD Oct 31 0.7 ppm below LOD Nov 21 0.7 ppm below LOD Nov 28 0.7 ppm below LOD CO Sept 26 1 ppm below LOD Oct 24 1 ppm below LOD Oct 31 1 ppm belowLOD Nov 21 1 ppm belowLOD Nov 28 1 ppm below LOD so2 Sept 26 0.1 ppm below LOD Oct 24 0.1 ppm below LOD Oct 31 0.1 ppm below LOD Nov 21 0.1 ppm below LOD Nov 28 0.1 ppm below LOD 99 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0088547/manifest

Comment

Related Items