Exposure assessment for automotive repair tasks in an attached garage

Transcription

1 University of Iowa Iowa Research Online Theses and Dissertations Spring 2011 Exposure assessment for automotive repair tasks in an attached garage Jacob Alexzander Krzystowczyk University of Iowa Copyright 2011 Jacob Alexzander Krzystowczyk This dissertation is available at Iowa Research Online: Recommended Citation Krzystowczyk, Jacob Alexzander. "Exposure assessment for automotive repair tasks in an attached garage." MS (Master of Science) thesis, University of Iowa, Follow this and additional works at: Part of the Occupational Health and Industrial Hygiene Commons

2 EXPOSURE ASSESSMENT FOR AUTOMOTIVE REPAIR TASKS IN AN ATTACHED GARAGE by Jacob Alexzander Krzystowczyk A thesis submitted in partial fulfillment of the requirements for the Master of Science degree in Occupational and Environmental Health (Industrial Hygiene) in the Graduate College of The University of Iowa May 2011 Thesis Supervisor: Assistant Professor T. Renée Anthony

3 Copyright by JACOB ALEXZANDER KRZYSTOWCZYK 2011 All Rights Reserved

4 Graduate College The University of Iowa Iowa City, Iowa CERTIFICATE OF APPROVAL MASTER S THESIS This is to certify that the Master s thesis of Jacob Alexzander Krzystowczyk has been approved by the Examining Committee for the thesis requirement for the Master of Science degree in Occupational and Environmental Health (Industrial Hygiene) at the May 2011 graduation. Thesis Committee: T. Renée Anthony, Thesis Supervisor Patrick O Shaughnessy Thomas Peters

5 To: My Parents ii

6 ACKNOWLEDGMENTS This study would not have been possible without the research support and environment created by the Heartland Environmental Research Center, funded by the National Institute for Occupational Safety and Health. The am grateful for the opportunity provided by The University of Iowa; this has allowed me to continue my education. I would specifically like to thank my thesis supervisor and advisor Dr. Renée Anthony for ensuring that this study could occur. My committee members Dr. Patrick O Shaughnessy and Dr. Thomas Peters have been instrumental in shaping the focus of this study. The Executive Council of Graduate and Professional Students at The University of Iowa provided additional funding for this study. I would also like to thank all of the Faculty and Staff within the department of Occupational and Environmental Health who have provided me with assistance and guidance through the research process. My fellow students have also been a key to shaping my research. Finally, I would like to thank my parents Ken and Kathy and my extended family that have provided me with the knowledge, ability, and drive to advance myself in addition to keeping automobiles functioning. iii

7 ABSTRACT The repair of automobiles is a critical aspect in vehicle ownership and is potential source of volatile toxic compounds being brought into a home when repairs are conducted in an attached garage. The goal of this study was to assess the impact of the repair of automobiles in an attached garage on the exposure of the home mechanic and degradation of indoor air. Five common automotive tasks were performed in two garages with the garage door either opened 30.5 centimeters (n=5) or closed (n=4). The exposure to the home mechanic, the behavior of contaminants within the garage, and infiltration of contaminants in the home were the determinants of interest. Integrative sampling incorporating charcoal sorbent sampling tubes analyzed by gas chromatography and directs reading photo ionization detectors were used to assess exposure. The tasks with the greatest contributions to the home mechanic s exposure were found to be brake pad replacement and oil change; these generated 95 th percentile concentrations of 51.2 ppm and 12.8 ppm, respectively, with the garage door closed. In contrast, the tasks of refueling and shock replacement had 95 th percentile contributions of 0.85 ppm and 2.99 ppm, respectively, in the closed garage. Equations were fitted to the aggregated concentrations during decay to estimate general ventilation (Q/V) in a closed garage. The contaminants within the garage were not found to infiltrate into the home as the average concentrations within the home never exceeded 1 ppm. It was found that automotive repair work in a closed garage may constitute up to 18% of threshold limit value of toluene over a 105 minute exposure at home. Automotive repair inside an attached garage has the potential to make a significant contribution to a mechanic s daily exposure and iv

8 should be incorporated into occupational exposure assessments of volatile organic compounds. v

9 TABLE OF CONTENTS LIST OF TABLES..... vii LIST OF FIGURES.. viii CHAPTER I. INTRODUCTION AND LITERATURE REVIEW...1 Regulations Sampling Safety and Health Issues..6 Organic Solvents in the Home Objectives 8 II. AUTOMOTIVE REPAIR IMPACT ASSESSMENT...10 Introduction Materials and Methods...11 Data Analysis. 15 Results Discussion.. 22 Conclusion III. CONCLUSIONS...43 Implications for Public Health..43 Limitations.43 APPENDIX A: RAW DATA 44 APPENDIX B: WEATHER CONDITIONS. 50 APPENDIX C: MATERIAL SAFETY DATA SHEETS.51 REFERENCES..71 vi

10 LIST OF TABLES Table 1. Automotive Task Times..29 Table 2. VOC Concentration by Task (Total Hydrocarbons as Isobutylene), ppm...32 Table 3. Contribution of Repair Work to Total Daily Exposure Table A1. Matched Tube and PID Concentrations 45 Table A2. VOC Concentrations, Entire Cycle, ppm..46 Table A3. VOC Concentrations over Work Cycle ppm 47 Table A4. Calculation of Garage Residence Time 49 Table B1. Weather Conditions...50 vii

11 LIST OF FIGURES Figure 1. Experimental Sampling Layout Garage A.30 Figure 2. Experimental Sampling Layout Garage B..31 Figure 3. Example plot of Log PID Concentration during the continuous sampling period for an Open cycle at the worksite Figure 4. Example plot of Log PID Concentration during the continuous sampling period for a Closed cycle at the work.34 Figure 5. Time-weighted averages of VOC concentrations during all Open Cycles...35 Figure 6. Time-weighted averages of VOC concentration during all Closed Garage Cycles 36 Figure 7. Comparison of Inside PID Time-weighted averages and Door PID Time-weighted average.. 37 Figure 8. Comparison of sampling tube total hydrocarbons quantitated as Toluene and PID time-weighted averages converted to toluene...38 Figure 9. Comparison of sampling tube acetone and matched PID Time-weighted averages converted to acetone..39 Figure 10. Comparison of sampling tube toluene and matched PID Time-weighted averages converted to toluene Figure 11: Plot of Decay Behavior of Solvents for all closed garage Tests in Garage A at the Worksite...41 Figure A1. Comparison of matched Sorbent Tube and PID concentrations..44 Figure A2. Significant Compounds present in GC/MS Analysis.. 48 Figure C1. Material Safety Data Sheet for lubricant used during Shock Replacement.52 Figure C2. Material Safety Data Sheet for Brake Pad Replacement cleaner.58 Figure C3. Material Safety Data Sheet for the paint used during Touch-up painting...65 viii

12 1 CHAPTER I INTRODUCTION AND LITERATURE REVIEW Automobiles are an integral part of modern societies and are a known source of environmental contaminants that are toxic to humans (Williams, et al., 2011). Due to the inevitable wear of machines, repair work is essential to automobile use and ownership. The repair of an automobile creates a unique set of environmental contaminants that are separate from the normal emissions created by the day to day use of automobiles for commuting and the transportation of cargo In 2009, 606,990 vehicle repair technicians were employed in the United States (BLS, 2009). Repair technicians, otherwise known as mechanics, are required to apply chemicals which can be hazardous in order perform repairs on vehicles. The application of chemicals typically involves the spraying of aerosolized agents onto a part or surface (Spencer et al., 2007). The spraying of chemicals is the most effective means of applying solvents, penetrating lubricants, and cleaners to often oddly shaped or difficult to reach parts of a vehicle (Stidham, 1999). The downside of the ease of application for these chemicals is that they become airborne and are easily inhaled into the respiratory system (Daniell et al., 2010). Regulations The chemicals used in automotive repair are organic solvents. These substances are derived from petroleum and present a unique set of health risks (Wilson, et al., 2007). Organic solvents, otherwise known as volatile organic compounds, acutely affect the central nervous system (CNS). The main symptom of exposure to organic solvents is the depression of the CNS resulting in decreased heart rate, diminished breathing, and loss of

13 2 consciousness (Dennison et al., 2005). Organic solvents that are commonly generated as a byproduct of automotive repair work include; acetone, toluene, benzene, and xylene. Guidelines for exposures to these airborne contaminants have been established by governmental agencies including the Occupational Safety and Health Administration (OSHA) and the National Institute of Occupational Safety and Health (NIOSH). Industry Consensus standards for airborne exposures are also established by the American Conference of Governmental Industrial Hygienists (ACGIH ). These governmental regulations and industry standards attempt to set guidelines that limit worker s exposures to prevent any negative health effects. While all three of these institutions publish guidelines for exposure, the exposure limits promulgated by OSHA are the only standards that are enforceable by law. The guidelines published by NIOSH are the result of government research and tend to be lower than the OSHA exposure limits for any given chemical. The guidelines for exposure produced by ACGIH are industry consensus standards that tend to allow a lower level of exposure than OSHA and match the standards produced by NIOSH for most exposures. The exposure limits of interest in this study were chosen to be the ACGIH consensus standards given that they allow the lowest level of exposures (ACGIH, 2009). The ACGIH exposure limits or Threshold Limit Values (TLVs) for acetone are 500 parts per million (ppm) over an 8-hour shift, and 750 ppm for a short term exposure limit (STEL). In comparison, the OSHA occupational exposure limit (OEL) for an 8- hour shift is 1000 ppm. The allowable limit for toluene is much lower than acetone. The ACGIH TLV for toluene for an 8-hour shift is 50 ppm with no recommended STEL. The OSHA OEL for toluene is higher the TLV at 200 ppm. Benzene has a TLV of 0.5 ppm

14 3 over an 8-hour shift and a STEL of 2.5 ppm. The OEL for benzene is set at a higher level of 1 ppm for an 8-hour shift with and STEL of 5ppm. Xylene has a TLV of 100 ppm with an STEL of 150 ppm. The OEL for xylene is set at 100 ppm with no recommended STEL (ACGIH, 2009) (NIOSH, 1994). Often, these substances and other solvents will be components of a mixture that comprises a commercially available cleaner. When solvents are combined within a mixture, the potential toxicity can be synergistically increased. Sampling The type of direct reading equipment used in this study is a photoionization detector (PID). These devices use ultraviolet light to breakdown molecules into subsequent atoms. The energy imposed on the molecules ionizes the electrons and creates an electric current. The current produced is then detected by the device and reported as a concentration relative to that of a calibration gas. Different lamps will ionize compounds that have an ionization energy that is less than the energy of the ultraviolet photons emitted by the lamp. The amount of current produced is proportional to the amount of chemical present (Nyquist et al., 1990). Like most scientific equipment, it is imperative that PIDs be regularly calibrated. The calibration of these monitors is accomplished through the use of a reference gas. The gas Isobutylene is commonly used to calibrate these monitors (Smith et al., 2007). Because different substances have different ionization energies, the manufacturers of these devices publish correction factors for various chemicals (Smith et al., 1978). These correction factors adjust for the sensitivity of the monitor. For any substance which the PID is more sensitive to than isobutylene, the correction factor will be less than 1. Conversely, for any substance with the monitor

15 4 is less sensitive to than isobutylene, the correction factor will be greater than one (RAE, 1997). There are limitations associated with the use of PIDs. While PIDs can detect the presence of volatile organic compounds in the air, they cannot differentiate between compounds when a mixture is present. Rather, these monitors can only detect for the presence of total hydrocarbons present, measured as isobutylene or any other reference gas used. The ambient environmental conditions in which PIDs are used can present device response issues. In reduced oxygen environments, the response of a PID will be amplified. In conditions where high levels of water vapor are present within the air, the response of the PID will be diminished (Pearce and Coffey, 2011). These performance issues require the operator to be aware of the limitations of photoionization detectors. The contaminants present within the chemicals used during automotive repair contain many petroleum based hydrocarbons. Integrative sampling was performed in this study consistent with previous studies that have compared direct reading instruments with sorbent sampling tubes in a laboratory setting (Coffey et al., 2009). When sampling in an occupational setting, these chemicals would be sampled for through the use of charcoal sorbent sampling tubes that are analyzed by gas chromatography and a flame ionization detector (Fioretti et al., 2010). Charcoal sorbent sampling tubes are used in exposure monitoring to trap chemical contaminants present in the air. In sampling, sorbent tubes are connected to sampling trains which consist of the tube, a filter to trap excess contaminants, and an air pump. The air pump pulls air from the atmosphere through the sampling train. As the air passes through the charcoal tube, contaminants such as hydrocarbons are adsorbed onto the surface of the charcoal grains. Charcoal tubes serve

16 5 as a means to trap and preserve contaminants present in the air. At the conclusion of sampling, the tubes are removed from the sampling train and sealed. The sealed tubes are then sent to a laboratory for analysis. Once at the laboratory, the tubes are opened and the charcoal media is removed and dissolved by a desorption chemical which is typically carbon disulfide. Laboratory analysis of the solution is performed through the use of gas chromatography with mass spectrometry (GC/MS). Gas chromatography involves passing the solution through a metal tube or column containing a solid phase material. The solution is pushed through the column by an inert carrier gas such as nitrogen or helium. An inert gas is used so that there is no reaction with the solution to be analyzed. The purpose of the solid material in the column is to slow the movement of the solution within the column so that only one chemical exits the column at a time. A flame ionization detector (FID) is located at the end of the column. The FID combusts the chemicals as they exit the column and analyzes the characteristics of the flame to determine specific chemicals (NIOSH, 1984). They most widely accepted means of exposure monitoring for organic solvents continues to be charcoal sorbent tube sampling. However, the use of PIDs has increased in recent years. This increase in use by field personnel places a decreased emphasis on the quality control that a laboratory provides. This situation requires that field personnel employing PIDs in exposure monitoring be familiar with the limitations of these equipment (Mouradian and Flannery, 1994). There has been an effort to determine a relationship between the responses that a PID will produce and the concentration of contaminants that a charcoal sorbent tube will retain. In the past, studies have been able

17 6 to achieve a relationship by which the response of the PID can be used to predict the mass of contaminant that a sorbent tube analysis will report. Similar to the PID s inability to differentiate chemicals within a mixture, the relationships that have been developed only pertain to a single substance or total hydrocarbons present (Coy et al., 2000). Safety and Health Issues The organic solvents used for automotive repair are monitored in the occupational setting. OSHA requires that employers protect the health of the worker. Employers which create unsafe work environments are subject to punishment under federal law. While regulations and guidelines are applicable to industry, businesses with fewer than ten employees and individuals who perform work themselves fall outside of these safety guidelines (Tuskes et.al, 1988). Many of the smaller vehicle repair operations lack the appropriate resources to identity and addresses their safety and health issues (Whittaker and Reeb-Whittaker, 2009) Similar to workers who are employed in a commercial repair shop, individuals who perform their own automotive repair are exposed to organic solvents. Many individuals who do not have the resources to have their vehicles professionally repaired within a commercial shop will perform their own repair work. Because the outcome of the functional automobile is the same regardless of where the work is conducted, the work that home mechanics perform is often similar to that which is conducted within commercial auto repair shops. The difference between work conducted within a home garage and commercial shop is that home garages typically lack the safety measures present within commercial shops. While a commercial repair shop contains the proper tools and equipment to perform automotive repair in the manufacturer s recommended manner, the home garage rarely contains tools and

18 7 equipment beyond simple tool sets and basic jacks and jack stands. This situation creates the potential for exposure to dangerous levels of hazardous chemicals with little means for the home mechanic to assess his or her own individual exposure Organic Solvents in the Home Previous studies have analyzed air infiltration rates between a garage and home. They have also associated benzene and other volatile compounds within the home to a vehicle located within an attached garage (Hun et al., 2011). The generation of volatile organic compounds within the home and attached garage has also been investigated. Studies on this subject have concluded that locating vehicles within an attached garage can contribute to volatile organic compounds within the home and that volatile organic compounds generated within the garage will infiltrate into the home unless control measures such as dilution ventilation within the garage are taken (Dodson et al., 2007), (Batterman et al., 2007). These studies have assumed that few individuals will spend any significant amount of time in the garage. It has been established in the previously mentioned studies that the presence of an attached garage contributes to volatile organic compound exposure within the home. Specifically, vehicles that are kept within an attached garage are known to produce the volatile organic compounds that can create exposures within a home. These chemicals can range from low risk substances such as acetone to high risk substances such as benzene. The repair of vehicles can generate these chemicals as a byproduct of the lubricants and solvents used during the repair. No studies have conducted specific, common auto repair tasks within an attached garage and then evaluated whether contaminants generated as a result of the work degrade a home s indoor air quality and

19 8 constitute a significant exposure to the home mechanic. While studies have assessed the exposures to the commercial mechanic, no studies have analyzed the exposures of the home mechanic. The home mechanic can be defined as any individual who performs automobile repair work outside of an occupational setting. The training of the home mechanic can theoretically range from technical degrees and certifications to no formal training and practical experience. While every automotive repair job is different, the model that is developed from this study has the potential to inform and empower the home mechanic on an issue which is not well understood. Objectives The repair of automobiles at home results in an unknown exposure group. The exposure to hazardous chemicals at home while performing automotive repair work is most likely to occur to those members of society who are least likely to be prepared for them. The purpose of this study was to create a profile of exposure for a home mechanic s work by using direct-reading equipment to measure real-time concentration levels and sorbent sampling tubes to determine exactly what chemicals comprise the contaminants generated by the work and if the contaminants move into the home. 1. Identify which airborne chemical contaminants are generated by automotive repair work in an attached garage through the use of integrated sampling. 2. Evaluate the effectiveness and comparability of charcoal sorbent tube sampling and direct reading photoionization detectors used in this study. 3. Establish the degree to which specific repair tasks contribute to volatile organic compound exposure for the home mechanic.

20 9 4. Develop a model to assess total daily exposure that combines occupational and home exposure.

21 10 CHAPTER II AUTOMOTIVE REPAIR IMPACT ASSESSMENT Introduction Automobiles are modern societies and are a known source of environmental contaminants that are toxic to humans (Williams, et al., 2011). Due to the inevitable wear of machines, repair work is an essential to automobile use and ownership. The repair of an automobile creates a unique set of environmental contaminants that are separate from the normal emissions created by the day to day use of automobiles for commuting and the transportation of cargo (Wilson, et al., 2007). Repair work often requires the idling of engines in confining spaces to prepare vehicles for work, the use of organic solvents to loosen fasteners, and the application of chlorinated cleaners in order to trap dust particulates. The application of chemicals typically involves the spraying of aerosolized agents onto a part or surface (Stidham, 1999). The spraying of chemicals is the most effect means of applying solvents, penetrating lubricants, and cleaners to often oddly shaped or difficult to reach parts of a vehicle. The downside of the focus on ease of application for these chemicals is that they become airborne and are easily inhaled via the respiratory system (Daniell et al., 2010). Government regulations and industry standards attempt to set guidelines that limit worker s exposures to prevent any negative health effects (ACGIH, 2009). While regulations and guidelines are applicable to industry, businesses with fewer than ten employees and individuals who perform work themselves fall outside of these safety guidelines. One such exposure group includes individuals performing repair on vehicles at home within a garage. The work that these home mechanics perform is often identical to that which is conducted within commercial auto repair shops, but these tasks typically

22 11 lack the safety measures present within commercial shops. This situation creates the potential for exposure to dangerous levels of hazardous chemicals with little means for the home mechanic to assess his or her own individual exposure. No studies have conducted several specific, common auto repair tasks within an attached garage and then evaluated their impact on a home s indoor air and the exposure to the home mechanic. The goal of this study was to determine what exposures occur as a result of automotive repair work conducted at home, within an attached garage and whether they are significant. Materials and Methods Automotive repair work must be simulated accurately in order for a typical exposure scenario to accurately be assessed. Automotive repair was conducted within a home s attached garage. Five automotive repair tasks were included in this exposure assessment. The work cycle comprised an oil change, the addition of one gallon of 87 octane gasoline, shock replacement, brush-applied touch-up painting, and brake pad replacement (Table I). These tasks were chosen based on their commonality, low injury risk, and generation of volatile organic compounds (VOC s) as a byproduct of their performance. Rather than actually performing a repair task, the chemicals were applied in the same locations and for the same durations as would be required to actually perform the work. Activities within each task that did not generate any VOC s were not performed; however, the time taken for these subsections was included in the work cycles. The total length of time for each task followed standard billing rates. For example, during the oil change task, the vehicle was idled for 5 minutes to warm up the oil which generates VOCs and introduces them into the garage. An additional 15 minutes was allotted for a

23 12 theoretical oil drain, filter replacement, and oil refill. Because the actual change of the oil emitted no VOC s in preliminary tests, the draining was omitted from the work performed. This method was applied to the brake pad and shock replacement as well. These two tasks require similar access to the vehicle. For both tasks, the tire was removed and the respective automotive aerosols were sprayed in their appropriate locations necessary for removal and installation. Similar to the oil change, the actual components were not removed from the vehicle. This was not deemed critical because the functional airborne contaminant profile was identical to conditions where the complete work would have been performed. The addition of fuel and painting were performed exactly the same as actual work. The repair work cycle was performed with each of the five tasks conducted in random order. The full set of five tasks in a work cycle were replicated five times with the garage door open 30.5 centimeters and four times with the garage door completely shut. The order of garage door condition was also randomly selected. After the completion of each work cycle, the concentration of airborne contaminants was given two hours to decay and the worker remained in the garage during the decay period. After the conclusion of the two-hour decay period, the garage overhead door was fully opened to facilitate the purging of contaminants from within the garage. The next work cycle was not begun until after the concentrations within the garage had returned to background. Sampling was conducted in two residential attached garages. These garages were chosen based on their physical characteristics. They both consisted of two-car garages that were fully dry walled, with the walls connected to a dry-walled the ceiling. Both garages also contained entry doors into the home. The entry doors were located on the

24 13 center of the opposite wall from the garage overhead door. The first garage (A) was the primary site for work sampling (Figure 1). Garage A and its attached home were less than two years old, representative of a newer home. The seals of this home s door were of excellent condition and permitted little if any airflow into the home from the garage. The dimensions of Garage A measured 7.1 m in length by 7.1 m in width by 3 m in height. The second garage (B) contained the same type of work but was intended as a means for assessing the variability between structures. Garage B and its attached home was representative of an older residence (24 years old) (Figure 2). The layout of garage B was similar in its orientation to garage A, but the seals and weather stripping around the windows and door frames were of much poorer quality than garage A, as light could be seen around the edges of the closed entry door between the home and the garage. Garage B also contained an additional personal door between the garage and the outside. The dimensions of Garage B measured 6.5m in length by 6.5m in width by 2.7m in height. The work was performed in the Midwestern U.S. during December in order to assess cold working conditions (mean outdoor daily temperature of -9.7 to -4.3 C). Automotive repair during this season would be performed in a closed or mostly closed garage for the comfort of the home mechanic. The garages remained unheated during each sampling period. The entry door to the home remained closed during each test. During all of the experimental runs, the contaminants were allowed to naturally move within the room. No artificial means of mixing the air, such as a fan, was used for all of the work cycles. Sampling was conducted at four locations during the repair work (Figure 1). The inside location was 30.5 centimeters inside of the home s door to the garage. The Door sampling site was located 30.5 centimeters away from the home s door to the garage

25 14 within the garage. The work sampling site was located 30.5 centimeters away from the zone of work. Three of these were area locations while the fourth was a sampler on the mechanic. At the area locations, sampling trains consisting of SKC Universal sampling pumps (Model PCXR4) were attached to SKC ANASORB CSC coconut charcoal 20/40 mesh 50/100mg tubes (lot 2000) were used to collect air samples. Eleven of the charcoal sampling tubes were sent to a contracted environmental health lab for analysis (EHL ESIS, Cromwell, Connecticut). The eleven tubes consisted of two blanks plus nine sample tubes. The samples were analyzed using OSHA method 7 to detect 60 different organic vapors and total VOC s. The accuracy of this analytic method is a minimum of +/- 25% of the true concentration for each tube (OSHA, 1989). The scan quantified the organic vapors generated from the automotive repair work. The solvents scanned for included acetone, benzene, cyclo-hexane, ethyl-benzene, heptanes, pentanes, toluene, and xylene. At each area location, a RAE Systems PPB Rae 3000 (Model PGM 7340, with a 10.6ev lamp) PID was positioned adjacent to the charcoal tube samplers. Each PID was operated at a flow rate of 500 cc/min with a data logging period of ten seconds. All PID s were calibrated with 10 parts per million (ppm) isobutylene. The manufacturer s stated accuracy of these PID s is +/-0.3% of the displaying reading at calibration. The fourth sampling location was on the worker s belt. This personal exposure was measured with a RAE Systems MultiRAE multiple gas meter with a PID sensor (Model PGM50-5P, with a 10.6ev lamp). The MultiRAE was operated at a flow rate of 200 cc/min with a data logging period of ten seconds. The accuracy of the MultiRAE was +/-3% of calibration (10 ppm isobutylene).

26 15 The collection for all samples occurred during the 68 minute work cycle and the two hour decay period. The sampling occurred over a period of four days with the first three in Garage A and the fourth in Garage B. Three work cycles were performed on the first two days, consisting of two open and one closed garage conditions. On the third day a single closed garage cycle was performed. The fourth day sampled at Garage B, where one open and one closed garage cycle were performed. The information on tasks performed and their associated repair times are contained within Table I. The tasks were performed for exactly the same amount of time for all experimental runs. Figures 1 and 2 provide the layout and sampling locations within Garages A and B respectively. The main difference between Garage A and B is that Garage A appeared more airtight than Garage B. Data Analysis Descriptive statistics were obtained from direct reading and integrated samples. The time-weighted average of each experimental run was calculated by averaging the PID concentrations at each location. The time-weighted average of each task was calculated by averaging the concentrations during each task, then subtracting the background or previous task s concentration to assess only the contribution of each successive task. Using this method, the concentration contribution for each task calculated for both open and closed garage tests. The contribution by task was calculated to determine which tasks created the largest exposure. The infiltration of volatile organic compounds into the home was assessed using a linear regression comparing the time-weighted averages of inside concentrations to the time-weighted averages of the concentrations within the garage at the door site. For this

27 16 comparison, the time-weighted averages were computed using the entire work and decay cycle. It was assumed that the diffusion of contaminants into the home could occur slowly. For this reason, the entire cycle was used rather than simply the work cycle. The key contaminants from sorbent tube samples were identified and compared to the material safety data sheets (MSDSs) of products used (Appendix C). The concentrations of contaminants collected on sorbent tubes were paired to the averaged concentrations of each matched PID run. The PID time-weighted averages for each sorbent tube match were converted to toluene and acetone equivalents based on the manufacturer s correction factor of 0.5 and 1.1, respectively. Linear regressions were then performed with the converted PID time-weighted averages and sorbent tube concentrations of toluene, acetone, and total hydrocarbons as toluene. The decay behavior of the post work cycle concentrations was used to generate a model to estimate for solvent concentrations as they decayed within a garage after work was completed. The decay patterns of all closed garage experimental runs were analyzed together. In order to allow for comparisons, the start concentrations for each of the three closed runs were used to normalize the decay trend. The concentration at each time interval during the decay was divided by the initial concentration at the start of the decay after the work cycle had been completed. A trend line was then regressed through the data to determine an equation for decay behavior. The equation was structured to follow the same form as the ACGIH equation for rate of purging within a room (ACGIH, 2007). (1)

28 17 Where V = volume of the space in m 3, C0 = Initial concentration at the start of purging, C = concentration at any time, Q is the rate at which air moves out of the space in m 3 /minute, and t = anytime in minutes after the purging begins. The residence time of contaminants within the garage can be calculated within by (V/Q), otherwise known as τ. The value of 1/τ = (Q/V) was fitted to the decay behavior of volatile organic compounds in Garage A during the closed condition. Exposure estimates by task and duration spent in the garage post-work cycle were used to generate a model to estimate home-auto repair exposures. The experimental work cycle as conducted was 68 minutes, however, the time to complete many repair tasks can vary based on the condition of the vehicle and the skill of the mechanic. The contribution for any order of tasks was calculated by summing the products of the concentration produced by each task and the time length of each task. The decay equation was then applied to the concentration at the end of the work cycle. The total ppm*minutes from the integrated time series curve of concentration was then determined to be the total exposure associated with the complete repair cycle for as long as the home mechanic is present. The ppm*minutes for both toluene and acetone were calculated by multiplying the total ppm*minutes by the average fraction of each respective chemical and then multiplied again by the PID correction factor for each chemical. Finally, the contribution of home auto repair work exposure from a single work cycle to a potential occupational exposure was evaluated. The home repair work exposure time of 68 minutes and a theoretical reassembly time of 37 minutes were added to an 8-hour work shift to estimate a total exposure time for a 24 hour period of 9.75 hours. The 37 minutes for reassembly was based on the speed at which the home

29 18 mechanic was able to reassemble the vehicle during the decay period. Additionally for this calculation, it was assumed that the home mechanic would leave the workspace after reassembling the vehicle. The occupational exposure limits for chemical exposures are based on an assumed 8-hour work shift. For any shift longer than 8-hours, the exposure limit must be reduced to ensure the same level of worker safety (Brief and Scala, 1986). A home mechanic who is exposed on the job should combine exposures during both periods to assess his true personal daily exposures when occupational exposure occurs within the same day as home exposure. The chemicals that were present in measureable quantities were used as the exposure metrics. The ACGIH 8-hour threshold limit values (TLV s) were used as the determinants for overexposure. These TLVs were adjusted using the Brief and Scala method as shown in Equation 2 (Brief and Scala, 1986). Reduction Factor = [8/daily hours worked] x [24-(minus) daily hours worked/16] (2) The additional 105 minutes of exposure resulted in a reduction factor for the TLVs of chemicals present of This factor was used to adjust TLVs for all chemicals that were present in significant amounts based on the laboratory GCMS analysis. The VOC generation from the work was calculated by multiplying the parts per million (PPM) concentrations produced by each task by the time length of each task. This calculation method returns a measure of PPM*Minutes. For example, the TLV for toluene is 50 ppm. For an 8-hour shift, the TLV for toluene would produce a value of 24,000 ppmminutes. The percentage of daily allowable exposure was calculated by first reducing the TLV for an exposure of an 8-hour shift plus the 1.75 hours dedicated to repair work. Equation 2 was used for this reduction. The percentage of daily exposure was calculated

30 19 by dividing the ppm*minutes for each chemical by the reduced daily allowable ppm*minutes. The final ppm*minutes for each chemical was adjusted by multiplying by the correction factor of the PID s sensitivity to each chemical (RAE, 1997). Results A summary of the VOC exposure contribution for each task is located within Table 2. The contributions were calculated individually for the open garage conditions and the closed garage conditions within Garage A. The brake pad replacement within a closed garage condition was the greatest contributor to concentrations within the garage with a 95 th percentile contribution of 51.2 ppm. The smallest calculated contributor was the refueling with a 0.85 ppm increase. The large contribution of the brake pad replacement masked the contribution of tasks that followed it. In calculating the average contributions, the tasks that follow the brake pad replacement were omitted. With random task ordering, this resulted in only one measure for the refueling task. The sum total of the concentration contribution of tasks performed at home in an attached garage (H G ) is shown in Equation 3: (3) where: j = task reference order (1 st, 2nd, 3rd etc.), n = total number of tasks, C i = concentration at start of task in ppm (see Table 2), and HG = home task based exposure. Figure 3 illustrates the behavior of the PID VOC concentrations within opened garage conditions. The VOC concentrations in open garage tests increased as the tasks were performed, and then began an immediate, rapid decline when the tasks were completed. In contrast, the closed garage tests resulted in higher concentrations with

31 20 gradual decreases over time, as shown in Figure 4. Within these figures, oil corresponds to the oil change, fuel comprises the refueling task, paint is the brush applied painting task, shocks represents the shock replacement task, and brakes highlights the brake pad replacement. The decay of concentrations over two hours begins after the 68 minute work cycle. Figures 5 and 6 detail the PID time-weighted averages for all open and closed garage experimental runs at all sample locations. The work sampling site contained the highest concentrations for all but two of the experimental cycles with the maximum TWA concentration of 18 ppm during the third closed cycle. The door sampling sites, 1.7 meters away from the work, typically remained lower than the worksite. The sampling site on the person of the home mechanic varied as a function of the movement within the workspace. The inside concentrations remained at less than 1 ppm for all experimental cycles. The work-cycle PID VOC infiltration of contaminants into the home is represented within Figure 7. There was moderate correlation between total hydrocarbons as toluene present on the sorbent sampling tubes and the matched PID VOC concentrations (R 2 =0.62) as shown in Figure 8. When the PID VOC concentrations were converted to reflect compound specific PID responses for a specific chemical and compared to the chemical concentrations on the sampling tubes, there was an improved correlation. The comparison of acetone converted PID VOC and toluene converted PID VOC concentrations to sampling tube concentrations yielded a better correlation (R 2 =0.79 and 0.78 respectively) as shown in Figure 9 and Figure 10.

32 21 The residence time of the garage (V/Q) was estimated from the decay equation. The decay equation was fitted for the Garage A in the closed condition only. The rapid decrease in VOC concentration in the open garage condition did not allow for accurate modeling of the decay trend. The coefficient of determination was 0.77, as shown in Figure 11. The residence time of Garage A in the closed condition was found to be 220 minutes. The space can be assumed to have returned background conditions after four residence times or 880 minutes. The decay behavior of VOC concentrations in Garage A at any time in minutes after work can be calculated using the regression equation. The method can only be used in a closed garage setting however. The exposure during the time spent in the decaying conditions at home (H D ) is shown in Equation (4) Where n = total number of tasks performed, C i = task based concentration in ppm (see Table 2), T end = the time in minutes spent in the garage after the repair is complete, and H D = home decay exposure. The constant ( ) represents the fitted value for the 1/τ (Figure 11) from concentration decay in the closed Garage A. The theoretical model to construct a daily exposure can be expressed by Equation 5. The total daily exposure for a given chemical is the sum of any exposure at work and any at-home repair exposure in ppm*minutes. The at-home repair exposure is the sum of the contribution of all tasks performed (Equation 3) and any time spent in the work space during decay (Equation 4). The decay exposure is the integral area of the concentrations

33 22 of VOCs and the time spent in the decay situation as the concentrations within the garage decrease: where H G = home task based exposure (see eq. 3), H D = home decay exposure (see eq. 4), and any potential exposure at work is the time-weighted average (TWA) of chemical exposure in the occupational setting. Using Equation 5 and assuming one complete work cycle (68 minute) and 37 minutes in the work area post work cycle, the contributions of selected hypothetical scenarios in closed garage conditions to total daily exposure for acetone and toluene were estimated based on their respective TLVs (Table 3). If all tasks are performed and generate their respective average concentrations, then the total contribution to daily exposure is 0.5% for acetone and 1.8% for toluene. If all tasks are performed except for brake pad replacement and generate their average concentrations, the total percentage of allowable acetone exposure will be 0.2% and toluene will be 0.8%. Finally if all tasks are performed and generate their 95 th percentile worst case concentrations, the contribution to total allowable daily exposure will be 5.2% for acetone and 18.2% for toluene. Discussion The analysis showed that brake pad replacement was the greatest source of contamination. The oil change and painting were the second and third most contaminating. The refueling and shock replacement were the lowest generating tasks. Contributions for each task in each experimental run were only used if they preceded the brake parts cleaner spray. Preliminary testing indicated that the shock replacement task

34 23 would constitute a larger exposure. This was not supported in these results and may have been a result of improper procedural testing during the preliminary phase. While the brake pad replacement generated the largest contribution to total hydrocarbon concentrations within the garage, the main ingredient of the chemical used was acetone which is not a significant concern due to its high exposure limit and lower relative toxicity when compared to toluene and benzene. Benzene was of major concern prior to experimental testing, but was not found to be present in measurable quantities (see Appendix A). The large contribution of this task and spray masked any subsequent task contribution and the randomization of tasks resulted in only one usable Refueling contribution. The repair tasks as performed resulted in concentrations with a large degree of variability. The average VOC concentrations generated from each task are most likely more representative of the true contribution of each task than the 95 th percentile contributions. The VOC concentrations generated by closed garage tests were much higher the concentrations produced in open garage tests. The time-weighted average for the open garage tests never exceeded 7 ppm. In contrast the time-weighted averages for the closed garage tests consistently exceeded 10 ppm, especially at the site of the work with a maximum of 18 ppm. The concentrations at the work site tended to be higher during all experimental runs. The inside concentrations had little correlation with the garage concentrations, and it was determined that contaminants did not infiltrate into the home. While this was expected in Garage A, this was an unexpected outcome for Garage B. The seal between the garage and home in the latter was in poor condition. The weather stripping was

35 24 cracked and as stated previously, light could be seen through the cracks. The additional openings in Garage B such as the window and extra door may have counteracted the poorer seal to the home. Dodson et al., 2007 and Batterman et al., 2007 have found that volatile organic compounds generated within the garage migrate into the home. This was not supported in this study. The reason for this may be that previous studies have employed a tracer gas to model the flow of contaminants into the home. The mixture of volatile organic compounds may behave differently in the temperature ranges present within this study. These studies performed their work in warmer conditions. This study, like the previous two, performed experimentation with the interface door between the garage and home closed at all time. The measure of total hydrocarbons on the charcoal sampling tubes did not correlate well with the PID readings when comparing total hydrocarbons as isobutylene. Acetone and Toluene constituted the two main chemicals present based on the laboratory analysis to sample for in the future. When comparing specific chemicals present on the tubes to PID concentrations adjusted for the chemical, a stronger correlation was achieved. Previous studies have been able to achieve a correlation between the log Tube values and log PID values (Coy et al., 2000). The best correlation achieved in this study was with non-transformed concentration values. The studies that have developed a relationship between PIDs and sorbent sampling tubes obtained samples with concentrations of total hydrocarbons ranging from mg/m 3 (Wilson et al., 2007). In this study, the concentrations of total hydrocarbons never exceeded 22 mg/m 3 on the sorbent tubes. The inability to calculate a relationship between the tubes and PIDs may have been a result of the samples collected in this study containing concentrations below

36 25 a level which permits the discerning of a relationship from noise in the samples. The PIDs reported consistent readings with each other; however the tube error percentages rarely overlapped the error percentages of the PIDs. This method of comparison has also been used to determine agreement between sampling tubes and PIDs in other studies. The PIDs and tubes may not have correlated well due to the stated error in the laboratory analysis. The behavior of the decrease of concentrations during the closed garage tests yielded a model for the decay of the VOCs after work is concluded with ventilation conditions remaining constant. The decay of organic solvents within a garage has not been measured in previous studies. This is useful for calculating the concentration at any time after work in a closed garage is concluded. The open garage tests were not used in the calculation of this model. Because the concentrations decayed rapidly after the work concluded in open garage tests where natural ventilation dilutes and removes VOCs, even with the garage door only open 30.5 centimeters. This results in the intercept for any regression line not matching the actual start concentrations. This study was able to achieve a model predicting decay concentrations only in a closed garage. The model for exposure during the work was developed to allow estimation of exposures for home-mechanics. The total exposure is represented by the work exposure plus the time in the workspace during decay. This model can be useful in determining the total daily exposure. An industrial hygienist can use this model to determine whether workers exposed to volatile organic compounds in an occupational setting are overexposed. The incorporation of non-occupational exposure into total exposure already occurs during noise exposure modeling. Similarly, the home mechanic can use