Document zzv06Ojrjw8gRoDbgp1qD9je6

FILE NAME Brakes BRK DATE 1982 DOC BRK104 DOCUMENT DESCRIPTION Journal Article - Asbestos Brake Emissions ENVIRONMENTal research 29 70 812982 Asbestos Brake Emissions RONALD L. WILLIAMS AND JEAN L. MUHLBAIER Environmental Science Department General Motors Research Laboratories Warren Michigan 48090 Received May 22 1981 An enclosed testing machine was built to determine the importance of asbestos from emissions from automobile brakes Higher emissions resulted from disc brakes than drum brakes due to the smaller surface area contacting the wheel in the former case Median emissions from one disc brake were 3.3 mg particulate 55 airborne 12 sedimentary 347 entrained 0.6 mg stop as hydrocarbons 0.7 mg stop as CO and 0.1 mg stop as CO Higher emissions were observed with increases in velocity and deceleration rate The mass median aerodynamic diameter of the emitted particles was 3.1 m Although the original brake material was about 50 asbestos the emitted particles averaged only 0.029 asbestos Therefore it appears that over 99.9 of the mass of original asbestos fibers was broken down into nonfibrous magnesium silicates In addition asbestos fibers which are on the order of millimeters in length in the brake had a median length of 0.5 min the emitted debris Based on the mass emission results and a stopping rate of 1.2 stops it can be estimated that 2.6 gof asbestos vehicle is emitted into the air 0.76 km settles to the roadway and 2.2 km is entrained in the wheel Although brake emissions are responsible for a minor fraction of ambient asbestos levels there may be a noticeable increase in asbestos levels near braking areas such as tolibooths INTRODUCTION The major component of the lining used in brakes is chrysotile asbestos com- prising 40-60 of the final product During a lifetime of 30,000 to 40,000 miles the material from the lining wears away and settles to the roadway becomes airborne or is entrained in the wheel If the asbestos survives the braking process there is the possibility for the release of large quantities of asbestos to the environment Several researchers have investigated asbestos emissions from brakes with varying results Early studies involved collecting entrained material from brakes Hickish and Knight 1970 Hatch 1970 This material was found to have an asbestos content of less than % indicating that the asbestos was broken down into nonfibrous magnesium silicate However in another study the brake dust was removed from a brake assembly and was analyzed by ray diffraction which indicated an asbestos content of 2-15 a value considerably higher than found by any other study Rohl et al 1976 It can also be questioned whether the material entrained in the brake is representative of the material which becomes airborne More sophisticated studies involved the collection of airborne particles from a brake attached to a dynamometer A study by Rowson 1978 showed the chemical constituents of asbestos magnesium and silicon were preserved in the wear debris indicating that asbestos had decomposed during the braking process 70 0013-9351 050070-13502.00 Copyright '1982 by Academic Press Inc. All rights of reproduction in any form reserved 30 kW DC Motor ASBESTOS BRAKE EMISSIONS Flywheels Electric Clutch ' ' i i = aoa t P i Sliprings Sliprings 71 Torque Tube -Torquemeter Ti] Sliding Fixture Drive Hub Enclosure | Tube | | | H FIG 1. A descriptive sketch of the emission test dynamometer Lynch 1968 first used electron microscopy to identify the fibers and found levels of less than % asbestos in the airborne dust near a dynamometer Anderson et al 1973 went one step further and used the electron microscope to quantitatively determine the asbestos in the debris and found that only 0.005 of the wear dust was asbestos fibers Another approach was taken by Jacko and DuCharme 1973 who constructed an enclosure around an automobile brake thus allowing the emissions to be sampled Their extensive testing program and electron microscope analysis indicated that 0.2 of the wear debris was asbestos Due to the enormous range of values of asbestos in brake emissions 0.005-105.05-15 0.005-15 a study was undertaken to quantify asbestos emissions from brakes This was accomplished with a specially designed apparatus which allowed a mass balance of all emissions to be made The particulate samples were analyzed for asbestos using a transmission electron microscope MATERIALS AND METHODS Brake dynamometer A sketch of the brake assembly is shown in Fig 1. A kW General Electric DC motor was used which was capable of accelerating the brake and wheel assemblies to speeds up to 100 -0.1 km A variable mass flywheel was mounted on the drive train which provided a variable inertial load equivalent to a vehicle mass from 1250 to 2500 kg The drive shaft was fitted with a hub to engage the wheel lugs which allowed the testing of complete brake assemblies of both disc and drum type The wheel assemblies were mounted to a torque tube which mated with a torque meter Torque signals to 1700 m were read out on a digital display The brake system used an actuated hydraulic system with a maximum pressure of 10,000 kPa Brake air pressure ( pedal pressure up to 700 kPa was controlled by a regulator Both brake air pressure and hydraulic pressure were read out on digital panel meters The brake was activated by a solenoid valve in the air supply line 72 - one ER WILLIAMS AND MUHLBAIER - wast aoe ~ The brake could be operated in a constant pressure mode or a constant deceleration mode Enclosure tube The brake assemblies were enclosed in a sectional 0.45 diameter x long gauge steel tube coated with polyurethane paint which could be disassembled for brake cleanup The enclosure tube and brake are shown in Fig 2. The inlet end of the tube was fitted with a 0.6 x 0.6 Cambridge absolute filter Air was pulled through the tube with an American Standard air blower A filter was also attached to the blower outlet to prevent contaminating the laboratory with brake dust Selection of test brake materials The disc pads and drum linings for this test program were selected to represent volume production components The brake set and the drum linings used for most of the tests were new original equipment components for standard size vehicles i.e. 4000-5000 pounds gross vehicle weight Two used brake sets one with 8000 km of use and one with 60,000 km of use were also tested All the brake results will be presented as a single set since there were no significant differences in emissions among the three sets The friction material used for testing consisted of about 60 asbestos 22 inorganic modifiers and % organic binder by weight An inertial load of 95.5 m for the front disc brake and 53.4 m for the rear drum brake was chosen as appropriate for a standard automobile ASBESTOS BRAKE EMISSIONS 73 Test conditions Three sampling modes were used for each test a dynamic mode a static mode and a combined mode The dynamic mode consisted of a brake application every 2 min and a total of 100 applications Air was drawn through the enclosure tube at a linear velocity of 1 sec past the brake This air flow was sufficient to transport particles up to 60 mdiameter to the collection site The static mode consisted of 100 brake applications without the blower operating The only air drawn through the enclosure tube came through the collectors This mode resulted in an improved background ratio The com- bined mode test consisted of 60 stops in a dynamic mode 30 stops in static mode and 10 stops with no air flow The final 10 stops were used to build up the concentration of emitted gases to measurable levels By combining all three modes each complete test consisted of 300 brake applications After a test was completed the brake system was disassembled The retained dust and the sedimentary dust were collected and weighed The weight of the brake pads before and after each test was also recorded The tests were made at one of the three simulated car velocities 40 64 and 88 km and one of four deceleration rates 1.2 1.8 2.5 and 4.9 secThe a temperature was measured with a thermocouple embedded in the drum or rotor During brake application the temperature rose rapidly to a maximum value which depended on the type of brake and the velocity of the wheel During the brake tests the average temperatures reached were 113 200 and 294 for stops from 40 64 and 88 km respectively For comparison the average temperature reached for a drum brake at 64 km was 145 In general disc brakes reach higher temperatures than drum brakes due to the smaller surface area in contact with the rotor In both cases the brake cooled 20 to 40 before the next application Sampling equipment Three fractions of particles were collected representing airborne material sedimentary material which would settle to the roadway and material which would remain entrained in the wheel The sampling equipment for the airborne fraction consisted of collector devices a set of Gast pumps manometers and flow meters The sample probes attached to collector devices are shown in Fig 3. The largest probe which was used for mass measurements withdrew 313 liters for deposition on a mm glass fiber filter The smallest probe was connected to a mm Nuclepore or Millipore filter which provided an additional particulate sample for microscopic analysis The third probe connected to an stage Andersen impactor which was used to separate particles according to aerodynamic diameter The flow rate through each sample probe was adjusted to match the velocity at the probe inlet with the mainstream velocity in the enclosure tube This isokinetic condition ensured that the sampling was indiscriminate with respect to particle size Using this approach the composition of the particulate obtained in the three collector devices will be identical The total mass of airborne particulate emitted from the brake assembly can be calculated from the mass of particulate collected by a device and its fraction of the total air flow in the enclosure tube using the principle of proportional sampling Sedimentary material fell out of the brake assembly during tests and accumu- 74 WILLIAMS AND MUHLBAIER 47 mm Filter Holder FIG 3. Sampling probes for brake emission studies lated in the enclosure tube This dust was collected in a polyethylene bottle through a short Tygon inlet tube The bottle was packed with glass wool and was attached to the inlet end of the hose to an industrial vacuum cleaner This device efficiently collected the brake dust which tenaciously adheres to most surfaces The final fraction entrained material was gently brushed into a tared vial and weighed The vacuum device was used for final cleanup of the brake assembly and the mass collected was added to the dust in the vial to measure the total entrained dust Asbestos determination Optical microscope measurments were made using the ASBESTOS BRAKE EMISSIONS 75 standard OSHA procedure for counting fibers greater than 5 min length Using a contrast technique at magnification of 400 10,000th of the total filter area was scanned Because asbestos fibers occur in a very wide size range it was appropriate to consider mass concentration in addition to fiber counts In order to make such calculations from the optical measurements we assumed that regard- less of length each fiber had a diameter of 0.25 ...mand a density of 2.56 cm Electron microscopy was necessary for counting and sizing the smaller fibers Transmission electron microscopy was found to be preferable to scanning electron microscopy The method used for counting and sizing fibers consisted of the following steps 1 A section of the Nuclepore or Millipore temperature asher to remove organic material 2 A methanol mixture was added to the trasonic nebulizer to disperse the fibers filter was ashed in a low- residue and placed in an ul- 3 The solution was filtered onto a mm pore Nuclepore filter 4 The filter was carbon and placed upside down on 3 TEM grids in petri dish containing chloroform to dissolve the filter 5 Two of the three grids were observed with a TEM at a magnification of 20,000x Asbestos fibers were counted and sized in 10 grid holes of each grid If the fiber was noticeably deformed a selected electron diffraction pattern was taken to confirm its integrity - 6 From the volume of the fibers and the density of asbestos the mass of asbestos was calculated The repeatability between the two grids averaged 32 RESULTS The results will be divided into two sections The first will describe the gaseous and particulate mass measurements under various conditions The second portion will describe the results of the asbestos analysis Particle and Gaseous Emissions The average results of the disc and drum tests are shown in Table 1. The particle mass is divided into airborne sedimentary and entrained mass The sedimentary and entrained portions were only collected during the brake tests because considerable material was lost when the drum brake was disassembled On the average 90 of the mass lost by the brake could be accounted for in the total emissions which was considered to be a good recovery In general the emissions were higher from disc brakes than drum brakes for a given set of conditions Median airborne particulate emissions were 1.6 mg brake compared to 1.1 brake from drum brakes Since disc brakes have a smaller friction surface area they tended to wear faster producing higher emissions Median emissions from one disc brake were 3.3 mg particulate 0.6 mg stop as hydrocarbons 0.7 mg stop as CO and 0.1 mg stop as CO The particles were approximately fractionated into 55 airborne % sedimentary and 34 entrained material These values were considerably different than those of Jacko and DuCharme i.e. 7-15 airborne 63-79 sedimentary and 7-15 entrained The differences are probably due to the differences 76 WILLIAMS AND MUHLBAIER oo 70'000'00 - seo - _-2 20 t0uo 70 zc'0 95 0 20 sco SuOISWA '09 z 00 600'0 - vv'l ~ _ v9 87'0 wo v3"0 09'0v2 _ cl SUD OH 0 400 30 0850 cv'0 18'0 96'0az 10 $00 67'0 zs'0 sau 9r'0 Leo poulesjug ofof SUOIEWA brakes brWaOkeWs 2]91Je Velocity SNOBING 40 JOVEIAY 88 401)8J9[3g40 19 0 as 0s 0a ZI a rt ee vio i0 m sec_ * LV'0 8eoAirborne ig"0 18 0 020 1.20 al 0% rz oul a's os'z 06 F02'1 sul 0s'Z 06'sel a'N *)X9} Brake Brake Airborne pausiEuldxa sayusq Wnp 0.70 4 sL0 I's 0.12 0.12 Lz Jo} pausw4jap jou sul OF oz! $3 osz 06 ?8't iam sojaied ADO}IA,64 or Or 4.90 2.50 2.50999 88 or Op r9e)9 v988 paulwjua pus and Jaqun sysa} entrained z z z 1 sig fwoirg in { t , { r4 I r4 \ I r4 y ASBESTOS BRAKE EMISSIONS 77 19 F wh 7 stop Z brake < brake Emissions Z Particle 4 Particle Y Particle Y)] N AY tf Y) Y4 40 64 88 12 1.8 2.5 49 FIG 4. Velocity km Deceleration Rate = 1.85 s Deceleration Rate s Velocity = 64 km Effect of velocity and deceleration rate on particle emissions from a disc brake in sampling Our test schedule may have overestimated entrained material since the brake was cleaned out after each test After prolonged usage the available surfaces would be filled causing additional dust to settle to the roadway or become airborne On the other hand it is possible that the sedimentary portion found by Jacko and DuCharme was overestimated The shroud surrounding their stop brake 3.0 brake Y mg Emisions 2.0 Z Emisions Z Hydrocarbon Hydrocarbon GaseousJ 40 64 68 ; AD 1.2 1.8 2.5 49 FIG 5. brake Velocity km Deceleration Rate s Deceleration Rate = 1.85 s Velocity = 64 km Effect of velocity and deceleration rate on gaseous hydrocarbon emissions from a disc 78 " 20 Stage / Most Most Most % 10 WILLIAMS AND MUHLBAIER A Drum Disc km km 2.5 s 1.8 s Stage Effective Cutoff Diameter ...m 04 0.7 " 2.1 3 2 4.7 1 6.8 9.0 FIG 6. Size fractionation of airborne brake particulate emissions brake system created artifically high surface areas for impacting material which otherwise might become airborne Velocity and deceleration effects The effect of increasing velocity and deceleration rate on particle emissions from a disc brake is shown in Fig 4. As velocity is increased with the deceleration rate held constant the particle emissions increase rapidly Gaseous emissions increase even more rapidly as shown in Fig 5 for hydrocarbons For instance at a deceleration rate of 1.85 seca doubling of the velocity causes a fold increase in particulate mass a fold increase in hydro- carbons a fold increase in CO and a fold increase in CO The increased particle ratio at higher velocities suggests that a thermal mechanism becomes more important while abrasive wear dominates at low velocities Increased deceleration rate also caused increased emissions but the effect was much smaller than for velocity A quadrupling of the deceleration rate at 64 km only increased particle emissions by 1.4 times and hydrocarbon emissions by 1.7 times Size distribution Several impactor tests were made to determine the size distribution of the airborne particulate Particle size distributions for typical drum- and brake tests are shown in Fig 6. The distributions are basically identical peaking between 1 and 2 mThe median aerodynamic diameter averages 3.1 ...mfor the 10 impactor tests with the submicrometer mass making up only 10 of the total mass Velocity km Disc brakes 40 40 " " " " " " " " " 64 65 88 Drum brakes 40 40 40 64 64 64 64 88 ASBESTOS BRAKE EMISSIONS 79 TABLE 2 AVERAGE ASBESTOS EMISSions from BRAKES Deceleration rate sec) Number of tests Percentage asbestos Range Mean 292233 1 0.004 4.9 2 0.0033-0.014 0.0087 292233 1 0.016 292233 4 0.00095-0.18 0.049 292233 4 0.0028-0.11 0.0028-0.11 0.038 4.9 4 0.0021-0.056 0.027 0.3 4 0.0032-0.052 0.025 1.8 M 0.0027-0.0085 0.051 22222322 2142YNTn 0.0046-0.025 0.015 22222322 2142YNTn 0.0075 22222322 2142YNTn 0.040-0.098 0.065 22222322 2142YNTn 0.0055-0.0081 0.0068 22222322 2142YNTn 0.0053-0.034 0.013 22222322 2142YNTn 0.0028-0.0093 0.0060 22222322 2142YNTn 0.0020-0.087 0.026 22222322 2142YNTn 0.0040-0.19 0.051 Asbestos Twenty filters from brake tests and 24 filters from brake tests were analyzed for asbestos fibers using both optical microscopy and transmission electron microscopy The ratio of small fibers which can be seen with an electron microscope less than 5 min length to large fibers as determined with an optical microscope more than 5 ...min length was about 460 from disc brakes and 230 from drum brakes Obviously most fibers will be missed if the only analytical technique used is optical microscopy Although the large fibers are few in number they contribute almost twice as much mass as all of the small fibers The small fibers dominate the number distribution while the large fibers dominate the mass distribution The average asbestos concentrations found for each set of velocity and deceleration conditions are shown in Table 2. The large range found within a set results from the enormous effect of large asbestos fibers Typically only 0 1 or 2 large fibers were seen in the counting area but they greatly affected the mass when extrapolated to the entire filter The brake test results ranged from 0.00095 to 0.18 asbestos with an average of 0.027 The brake test results ranged from 0.0020 to 0.19 with an average of 0.031 There is no significant difference between the asbestos fraction from drum or disc brakes In addition there are no significant trends in asbestos content with deceleration rate or velocity However since the total particle emissions increase at higher deceleration rates and velocities the asbestos emissions also increase 80 WILLIAMS AND MUHLBAIER Assuming an initial asbestos content of 50 in the brakes and a final asbestos content of 0.029 over 99.9 of the asbestos has been broken down to nonfi- brous material during the braking process This is in agreement with the work of Jacko and DuCharme 1973 and Anderson et al 1973 The particles entrained in the wheel and the sedimentary particles had a slightly higher asbestos content 0.04 than the airborne particulate The median length of the airborne asbestos fibers was 0.5 ...mThis can be compared to the average fiber length in brakes which is on the order of millimeters in length brake The average mass emission of particles from one disc brake is 3.3 of which 55 is airborne 12 sedimentary and 34 entrained Based on four vehicle and an asbestos content of 0.030 in the airborne material and 0.040 in the sedimentary and entrained material the total asbestos emission per vehicle per stop is 2.2 ...gairborne 0.63 ...gsedimentary and 1.8 ...gentrained asbestos Based on an average of several different driving cycles there are about 1.2 stops 2.0 stops Jacko and DuCharme 1973 Therefore the total asbestos emission is 5.5 ...gasbestos of which 2.6 km is airborne 0.76 ...g settles to the roadway and 2.2 km becomes entrained in the wheel DISCUSSION It is possible to use these asbestos emissions to calculate the influence of braking on ambient asbestos levels This will be done for three situations 1 total nationwide asbestos emissions 2 typical urban asbestos levels and 3 the situation near a tollbooth where a large amount of braking occurs Nationwide Asbestos Emissions There were 114,000,000 registered automobiles in the United States in 1977 averaging 16,000 km each for a total of 1.8 ^ 10 km traveled MVMA 1978 If airborne asbestos emissions from autos are 2.6 km annual nationwide emissions are 4.7 metric tons This can be compared to annual estimates of made disposals and emissions of asbestos to the air of 2093 metric tons Levine 1978 Therefore brake usage is responsible for approximately 0.2 of total asbestos emissions Urban Asbestos Emissions The input of asbestos from braking to urban ambient concentrations can be determined by comparing with lead emissions from cars and urban lead concen- trations This comparison assumes that the major source of lead in the atmosphere is leaded gasoline and that particulate lead and asbestos fibers will behave the same in the environment Between 1970 and 1974 when all cars used leaded gasoline the average lead content of gasoline was 0.571 liter and the average fuel economy was 5.13 liter U.S. EPA 1977 Huntzicker 1975 estimates that - 75 of the lead in the fuel is emitted as airborne particulate matter for a lead mass ww emission rate of 0.083 km The average lead concentration in the 1960s in three - urban areas was 2.4 g in Los Angeles 1.4 g in Cincinnati and 1.6 g in Philadelphia Landau 1969 We will use an average urban lead concentration of 2 m Based on our airborne asbestos emission rate of 2.6 km and by ASBESTOS BRAKE EMISSIONS 81 comparing to lead an expected urban concentration due to braking of 0.063 m can be predicted This is a small fraction 0.06 to 0.6 of the measured urban asbestos concentrations of 10-100 m Nicholson et al 1971 Tollbooth Asbestos Values In areas of high braking such as near tollbooths ambient asbestos concentra- tions may be affected as was shown in Connecticut Bruckman and Rubino 1974 Ambient asbestos concentrations of 25 mwere measured near a tollbooth through which 55,000 cars passed per day Based on our emissions experiments and dispersion estimates from the General Motors Sulfate Dispersion Experiment Cadle et al 1977 it is possible to estimate the fraction of asbestos resulting from braking During the sulfate dispersion experiment 5500 cars per hour passed the sampling point with average sulfate emissions of 23 km for an emission rate of 127 km This emission rate resulted in a maximum increase in ambient con- centration of 15 g At the tollbooth 55,000 cars passed through in a 24 period or 2300 cars per hour To relate our results to the tollbooth situation we will assume an average deceleration rate of 2.4 secfrom 88 km which corresponds to a deceleration time of 10 sec and stopping distance of about 120 m Braking from 88 km leads to airborne particle emissions of 29 stop and asbestos emissions of 8.4 stop As a result the asbestos emission rate in the braking zone is predicted to be 160 km Comparing this to the roadway experiment the maximum expected asbestos concentration would be 19 m Therefore a significant fraction of the 25 m concentration at the Con- necticut tollbooth could arise from braking ACKNOWLEDGMENTS The authors appreciate the assistance of A. E. Fincham in building and operating the brake assembly and the careful analytical work by J. L. Johnson and H. W. Sturer of the Analytical Chemistry Department REFERENCES Anderson A. E. Gealer R. L. McCune C. and Sprys J. W. 1973 Asbestos emissions from brake dynamometer tests Society of Automotive Engineers Meeting Detroit MI May 14-18 1973 Paper No. 730549 Bruckman L. and Rubino R. A. 1978 Monitored asbestos concentrations in Connecticut J. Air Pollut Contr Assoc 28 1221 1226 Cadle S. H. Chock D. P. Monson P. R. and Heuss J. M. 1977 General Motors sulfate dispersion experiment Experimental procedures and results J. Air Pollut Contr Assoc 27 33 38 Hatch D. 1970 Possible alternatives to asbestos as a friction material Ann Occup Hyg 13 25 29 Hickish D. E. and Knight K. L. 1970 Exposure to asbestos during brake maintenance Ann Occup Hyg 13 17-21 Huntzicker J. J. Freidlander S. K. and Davidson C. I. 1975 Material balance for automobile- emitted lead in the Los Angeles Basin Environ Sci Tech 9 448 Jacko M. G. and DuCharme R. T. 1974 Brake emissions emission measurements from brake and clutch linings from selected mobile sources EPA Report 68-04-0020 Landau E. Smith R. and Lynn D. A. 1969 Carbon monoxide and lead environmental apprai- sal J. Air Pollut Contr Assoc 19 684 .