Document N2y2ErVQmgy6kYewqzDYMw40g

COMPARISON BETWEEN AIR POLLUTION LEVELS CAUSED BY CARS OPERATING ON TML AND TEL GASOLINES *>y RICHARD S. BRIEF Medical Research Division Esso Research and Engineering Company Linden, New Jersey K 0003160 N9628 1. Summary As part of the over-all evaluation of tetramethyl lead as a gasoline antiknock additive (TML), the lead content of the air in and around foreign-make cars operating on fuels containing identical quantities of lead as TML or TEL was measured. The data were obtained simultaneously on the same makes and models of cars during the same driving cycle thus permitting a direct comparison to he made between TML- and TEL-induced lead contamination in the air. Results represent lead in air values inside the car, under the hood, general area behind the car and the exhaust gases. Statistical treatment of the data showed no significant difference between the atmospheric lead levels obtained from the combustion of TML- or TEL-containing gasoline. 2. Introduction The health aspects of exposures to tetraethyl lead (TEL) have been studied in considerable detail over the years in which this compound has been used extensively as an antiknock additive in gasoline. These studies have shown that no hazards to health need exist where recommended control procedures are conscientiously observed. However, all alkyl lead compounds are generally accorded a high degree of toxicity, so the proposed use of the more highly volatile tetramethyl lead (TML) as a substitute for or in addition to TEL in gasolines necessitated a study of potential health implications that might be associated with it. One important factor that needed to be determined was the quantitative differ ence, if any, between the lead content found in the atmosphere in and around automobiles burning fuel containing TML and that released into the air by automobiles operating on fuel containing TEL. The following atmospheres were sampled: (l) exhaust gases, (2) air directly above the engine with the hood closed, (3 ) in-car air at the driver's seat at breathing zone height, and (k) air in the general area, four feet behind the car at breathing zone height. 3- Methods Four pairs of foreign cars were used in the test which was conducted during the period February through March i960. The vehicles are listed in Table 1 Make English Ford Volkswagen Renault Fiat Model Anglia Sedan Dauphine 1200 Table 1 Engine Location front rear rear front Manufactured In England Germany France Italy 2. The cars were operated on the Mileage Accumulation Dynamometers (MAD) at the Esso Research Center in accordance with a schedule which simulated as closely as possible actual driving conditions in Europe. Some compromises were made because, due to the manual transmission in each car, low engine speed operation had to be manually controlled (idle and 1000 rpm). For the greater part of the test, cars were operated automatically in high gear by electronic programming on the MAD unit. This operation permits frequent changes in throttle position in high gear to simulate changes of speed on the road. For the manually controlled part of the test, mileage was accumulated at constant engine speed. Table 2 outlines the car operation schedule. Table 2 Car Operation Schedule Schedule Length - 8 Hours (5 Cycles) Cycle Number: Engine RPM Gear 500 10 0 0 2000 3000 3500 Neutral Top less 1 Top Top Top Total Time Per Cycle - Minutes 1 , 2 , and 3 h 5 10 20 20 (1 ) i+0 (l) 30 (1 ) 5 Shut 10 Down 10 One 20 Hour 15 Car Operation Manual Manual Tape Tape Tape (l) Car operation in top gear controlled by electronic tape. RPM will cycle between indicated speeds during the test with time at each rpm totaling indicated time. A. Sampling Procedure The sampler was designed to separate organic from inorganic lead, thus permitting the determination of both forms of lead in the air sampled. This equipment is shown in Figure 1. The sampling scheme is simply to meter the total flow with some suitable device, like the rotameter shown. The particulate matter in the air con taining inorganic lead is filtered on molecular sieve filter paper in a filter holder assembly. The air then bubbles through a collecting solution in an impinger designed for a maximum flow of lh liters per minute. This impinger contains a solution of KI-I2 which is a highly efficient medium for capturing organic lead vapor. The air stream is then passed through 0 u <-' i) Figure 1 SAMPLING EQUIPMENT THE VACUUM PUMP DRAWS A SAMPLE THROUGH A ROTAMETER, A FILTER ASSEMBLY TO COLLECT IN ORGANIC LEAD, AN ABSORPTION BUBBLER TO COLLECT ORGANIC LEAD, AND A DISENGAGING TRAP AND CHARCOAL ABSORBER TO PROTECT THE PUMP FROM CORROSIVE VAPOR. #- k glasswool and activated charcoal in separate traps to protect the pump from the corrosive effects of iondine vapor carryover. In practice, when sampling the exhaust gases, condensation in the rotameter made accurate flow measurement difficult. In its place, a 10 liter per minute critical nozzle was placed between the collection devices and the pump. Sample times were varied from ten minutes to over two hours, primarily to collect enough organic lead for accurate analysis. For high flow rates, or where condensation is not a problem, the rotameter is more desirable than the critical orifice, mainly because of its lower pressure loss. The exhaust samples were obtained from stacks located near the cars. The exhaust from each car was directed into an oversized flexible duct which was connected to the base of a stack. Lead contents found In the stack may not truly represent lead concentration in the exhaust gas because of deposition and dilution in the connecting system, but they serve as good approximations for the purpose of this study. B. Analytical Procedure Inorganic lead on the filter paper was analyzed by a procedure developed at Esso Research and Engineering Company.^ The principle of this method involves heating the sample with nitric and sulfuric acids to destroy all organics and changing the metallic elements to the corresponding sulfates. Lead is then extracted with a chloroform solution of dithizone forming the color complex, lead dithizonate. By comparing the color intensity of the sample with that of the series of standards, the lead content of the sample can be determined. The precision of the method is of the order of + 0.U micrograms of lead. Organic lead collected in the Kl-Ig bubblers was analyzed by a colorimetric procedure developed by Snyder.2 The bubbler solution pH is adjusted to 11.0 and the unreacted iodine is reduced with 30 ml of an ammoniacal solution containing 10 gms potassium cyanide, 100 gms sodium sulfite, 20 gms ammonium citrate, 550 ml distilled water, and 1,950 ml ammonium hydroxide. The lead is then extracted with a chloroform solution of dithizone forming the color complex, lead dithizonate. The color developed (varies from colorless to cherry red) is compared with previously developed color standards to indicate the total quantity of organic lead collected. The precision of the method is reported to be + 0.5 micrograms of lead in low concentrations and + 10$ for high concentrations. b. Observations Samples were taken at random intervals during the car operation cycle. In Tables 3? and 5> adjacent data in the TEL and TML columns, under each sample location heading, were obtained simultaneously. K E 0uuo 1b 4 5 Table 3 ORGANIC LEAD, microg/cu.m Exhaust Gas TML TEL Front Engine Location Rear *Estimated 8o* 80 920 1+30 670 530 1+80 80 60 151+0- 57 10 0 530 Air Under Hood TML TEL 27 80 31 21 9 15 27 1+0 31 120 13 27 Breathing Zone Drivers Seat 1+' Behind Car TML TEL TML TEL 11 6 1+ 7 2 1+ 9 8 6 93 8 6 Table k INORGANIC LEAD, microg/cu.m. Exhaust Gas TML TEL Air Under Hood TML TEL Breathing Zone Drivers Seat 1+1 Behind Car TML TEL TML TEL Engine Location Front 18 70 1+270 21+00 Rear 8170 3^ 0 1+300 1720 2670 1*130 - 6620 3980 1330 310 0 3 8 6 20 90 50 -7 96 13 22 36 17 18 57 6 7 3 35 12 10 20 ll+ 6 Table 5 TOTAL LEAD, microg/cu.m Exhaust 'Gas TML TEL Air Under Hood TML TEL Breathing Zone________ Drivers Seat 1+' Behind Car TML TEL TML TEL Engine Location Front 1950 1+350 3320 Rear 8620 1+110 I+8 3 2200 2750 1+190 8l60 1+037 11+30 3630 30 39 15 1+7 121 63 87 107 3l+ 26 18 8 57 138 81+ 12 ll+ 7 kk 20 16 113 22 12 6. Comparison of absolute levels obtained within each sub-grouping (e.g., exhaustfront engine) was not possible because of (l) the wide difference in variance between the TML and TEL data within each sub-grouping, and (2) the limited number of data points. It is believed that high levels obtained within each sub-group cannot be discounted in an attempt to normalize the data. Culling of these peaks might lead to erroneous conclusions. Many more test points would be necessary to define the model accurately, if in fact this is felt to be necessary. Grouping of all the data for testing as a population of TML or TEL values was then done. The frequency distribution of the aggregate data for both TML and TEL gasolines was highly skewed toward the higher concentrations. This is to be expected since no concentration can fall below 0 and much of the data is near this lower limit. The data from Tables 3 and ^ were plotted on logarithmic frequency distribution paper as shown in Figures 2 and 3* The similarity of the distributions obtained for both TEL- and TML-fueled cars is easily seen. To compare the curves shown in these figures, the lead concentrations at the deciles were plotted against statistical deviation expected from a normal distribution at the deciles. The comparison curves are shown in Figure U. The two curves have a maximum spread of only about 0.1 standard deviations. With such similarity in the distribution of the aggregate data, it is appropriate to use nonparametric evaluations for further comparison testing. The sampling design and technique employed tended to eliminate car operation as a variable because of the fact that samples were taken during identical . operating cycles for both cars. With this approach in mind, the best measure of whether or not the TML-fueled cars were producing more air pollution would be a differential analysis of Tables 3> ^ and 5* That is, one should consider the differences between TML-produced values and the TEL levels, and not the absolute values obtained. The composite Table 6 shows these differential levels: Table 6 DIFFERENTIAL LEAD, microg/cu.m (Lead in Air from TML Fuel Minus that from TEL Fuel) Breathing Zone_________ Exhaust Gas Under Hood Drivers Seat U ' Behind Car Org. Inorg. Total Org. Inorg. Total Org. Inorg. Total Engine Front 0* - 800 - 800 -53 20 lUO 160 10 -620 -h220 -bQbo - 6 Location Rear 373 570 0 *Estimated U210 2110 1200 4583 -13 2680 -8 9 1200 -lU -b -5 3 3 72 -7 -57 5 5 -3 - 3 -2 -10 -17 -21 90 15 3 95 12 1 -8h 0 0 15 -2 k -69 -2 b 7. Figure 2 FREQUENCY DISTRIBUTION OF ORGANIC AND INORGANIC LEAD IN AIR FROM TM L-FU ELED CARS LEAD CONCENTRATION/ MICROGRAMS/CUBIC METER LEAD CONCENTRATION, MICROGRAMS/CUBIC METER Figure 3 FREQUENCY DISTRIBUTION OF ORGANIC AND INORGANIC 8. Figure 4 COMPARISON OF FREQUENCY DISTRIBUTIONS OF LEAD IN AIR FROM T E L - AND TM L-FUELED CARS ^ LEAD CONCENTRATION, MICROGRAMS/CUBIC METER NORMAL DEVIATE 10. ^1 Nonparametric rank"5 and sign tests4 of the differential lead generated for organic, inorganic and total lead concentration in the air were done. These tests showed no significant increase in the air pollution caused by TML-leaded gasoline over that from TEL-leaded gasoline. 5. Discussion The fact that no significant difference appears between the atmospheric con centrations from TML and TEL gasolines is not surprising in view of the experimental set-up at the Mileage Accumulation Dynamometers. The fuel is piped directly to the cars on the unit eliminating tank-filling loss and possible spillage. Blowby, discharged unburned fuel and carburetor loss are other possible means by which lead can be released into the air. These are not believed to be significant, however, for the reasons given below. As a result, it is to be expected that the lead in the fuel supplied will be exhausted on the basis of lead content in the fuel and not on the lead volatility. In blowby some of the lead in the gasoline dissolves in the crankcase oil. With the more volatile TML more of this lead could escape by subsequent evaporation. Tests done in conjunction with this study have borne this out. Crankcase oil from TML-fueled cars had lower lead content than oil from TEL-fueled cars. In terms of air pollution, however, the difference in lead discharged should be small. The unburned fuel in the blowby and in the exhaust should approximate the composition of the fuel fed. Carburetor evaporation loss should be small inthe foreign cars tested. These cars have small carburetor bowls and at the low ambient temperatures existing during the test program little evaporation should occur. Thus the quantitative lead emissions from cars studied in this program should be quite similar for both TML and TEL fuels. This has been confirmed by the data collected in the study. In areas where gasoline can evaporate and the vapors can concentrate in the air, like a fueling garage, the higher volatility of the TML gasoline will probably produce higher levels of organic lead in the air. This, in fact, has been proven to be true by Cholak5 in studies in a California garage where the taxi fleet using the garage was serviced alternately with the same leaded quantities of TEL and then TML gasolines. The particulate lead concentration remained about the same while the organic lead in the air increased about three times. In conclusion, the general atmosphere should show little or no significant difference in the lead levels created by TML and TEL gasolines. However, occupational exposure should be checked in locations where the higher volatility of TML could increase the organic lead concentration in air. This would include garages, gasoline storage tank cleaning operations, gasoline loading racks, service stations, and areas of congested traffic. 11. 6. References 1. Analytical Technical Committee, Esso. "Analytical Methods Useful in Medical Research." Analysis Standards Circular Procedure #920.0^2T. Esso Research and Engineering Co., Linden, N. J. (Aug. 1955)* p. 18. 2. Snyder, L. J., Barnes, W. R., Tokos, J. V.: Determination of Lead in Air. Anal. Chem. 20(8);772-776 (August 19^8). 3. Wilcoxon, F. : "Some Rapid Approximate Statistical Procedures." American Cyanamid Co., N. Y. (July 19^-9)* P* 5* Dixon, . J. and Massey* F. J., Jr.: "Introduction to Statistical Analysis." McGraw Hill, N. Y. (1951)* P* 32^+. 5- Cholak, J. and Kehoe, R. A . : A Report of an Investigation of the Potential Hazards of Exposures to Lead Associated with the Handling and Use of Gasoline Containing an Antiknock Fluid in which Tetramethyl Lead is Substituted for Tetraethyl Lead. The Kettering Laboratory, Cincinnati, Ohio "(November 10, 1959) Confidential Report. RSB:mvv 6/ 8/62 ^ 0 J U O s. < 1