Document 1gQ2oYVkEO3degXvMXg14NpbX

wTHE~ijPmRESENT STATUS OB' iAmH wIN^VmESTFInGAwTI<OrNmOFimAcN ijALRATIrF^A^CTi!raOCxCwURrR^ING -^-M-a-S--T--f-a-^-I--C-L-E---o-T-ZsFmSlnTicHs^ATMeIrNiEcTSAmI'rE^VTjlrtaiuPtfli-A-N-C-ES- -00-.--T-'-iP~E~ Tlie 123A electrostatic precipitator is generally operated, at a sampling rate of 5 cfm. or 1 cfm. when used in the usual type of investigation of atmospheric constituents of the industrial environment. Ve have found an apparent decrease in efficiency at these rates when sampling the air In the human respiratory* chamber at this Laboratory and we, herefore , use a sampling rate of o .3 cfm., which in on Hie plateau of the sampling efficiency curve, as we determine it, for lead seoquloxide particles of .05 micron median size at the approximate concentration of .075 ng/Ms . The efficiency of precipitation is apparently a function of particle size and -perhaps of concentration, and in all probability it is principally the extremely small size of the particles of lead sesquioxide in the chamber which makes this lower campling rat e necessary. Recently we have changed tie lead concentration in the chamber from .075 rag/M3 to .15 mg/M3 (determined at .5 cfm. sampling rate), and while we have not as yet determined the distribution of particle size at the new concentration we have no reason to believe it to be different, from that observed at -.'Lie concentration of .U75 mg/5Ss * Near the new concentration (.11 mg Pb/t!3 ) we ran a series of. samples with the precipitator at. rates below 0.5 cfm. and obtained fantastically high results at the lower sampling rates. A series of determinations were made at other concentrations below .15 mg/M3 (determined at 0.5 cfm. sampling rate). The results are shown in Pig. (l). The individual and familial shape of the curves clearly indicates the presence of an artifact of consistent and uniform characteristics. At this point, of the investigation Mr. . Younkers suggested that a possible explanation might he found to he related to the electric wind, set up in the sampling tube, which might cause a circulation of air into and out of the end of the sampling tube, deposition of lead taking place from the air being so circulated. Such an explanation seemed tenable. Mathematical treatment of the data of Pig. (l) assuming such circulation and a true concentration of lead in air as the asymptotic value for each setting indicated an effective circulation corresponding to complete lead recovery from approximately 7 liters of air per minute. Tests were carried out using an extension on the sampling tube insulated from the charged sampling tube . The results to date do not show the artificial effect, and indeed the results are slightly lower than those obtained with the regular precipitator at 0.5 cfm. sampling rate. We are continuing tests with modified precipitators, with precipitators in series, and with variable precipitator voltages . We are investigating further the decomposition of the TEL in the air line to the burner, the' amount of TEL vapor given off at various flow rates and other Inethods of analysis for the room air. Previously to this investigation we tried chemical scrubbing methods for sampling, with poor success. K ? 0020830 5 The suggested explanation, or something very similar to it, is probably correct but has as yet not been proved linequivocally. It may be said', however, that the true concen tration can not be much different from the concentration obtained with the regular precipitator operated at 0.5 cfm. sampling rate. Lester B . Roberts Lawrence Schafer Roland Burkey Herbert Schlecht April 15, 1955 KE 0020831