Document Rp5vzryxN2xRb0gd5M543dxnE

-tf-mrnt rtfiy oc\ Kfii-PH Le ti -torJ t y___ Received JUfi 05 1974 V CMMICA1S AND PU9RCS * N- WHEELER. JR. P. a IOX SMI, SOUTH CHARLESTON, WEST VIRGINIA 25303 To (Mow) DivMJon Copy w Mr. R. N. Wheelv / South Charleston, Bldg. 189 Mr. R. L. Andersen, 511 Mr. T. W. Carmody, NYO-31 Mr. R. E. Graebert, 511 Mr. H. R. Guest, 511 Mr. C. D. Hendrix, 511 Mr. A. W. Lutz, NYO-28 Dr. A. B. Steele, NYO-28 Mr. R. W. Wemon, NYO-32 Mr. J. W. Whittlesey, NYO-45 Dr. N. L. Zutty, NYO-32 aw QriplRMtMp D^t. June 3, 1974 Research and Development Proposed Standard for Stpasure to Vinyl Chloride Dear Mr. Wheeler: In accordance with your telephone requetf of May 28, I hovo rovlowod tho wording of tho proposed ponnonont OSHA standard far vinyl cHoride with rospoet to tho stotomont of background Information quoted bolow. A. Laval of exposure. Tho proposod standaid for employee exposure is sot at no detectable Ioval, os datonninod by a sampling and analytical method capable of detecting viryl chloride at concentrations of I ppm with an accu racy of I ppm 50 percent. It is the conscious of the analytical chemists end statisticians I consulted, and my own feeling, that the statement ce written above leads to varied interpreta tions with little chance of resolving the differences in viewpoint. It could be mgued that an ana(yds of 1.5 ppm or less would show compli ance. It could also be argued that an analysis of 0.5 ppm or more would show lack of conpliance. There are two main reasons for diverging interpretations. One is that the description of the method to be used is net precise. The other Is that the statement of I ppm 50% refers to the method, not the permlssfcle limit. A permissible limit of "none detectable*. Is virtually undefinafale, even when coupled edth a definition of the method to be ined. ucc 028701 -2 In order to minimize varying interpretations,the I ppm sensitivity and the reproducibility of the method could be worded as follows: "The proposed standard for employee exposure is set at no detectable level/ as determined by a method capable of detecting I ppm with a reproducibility (2) of 0.5 ppm at the 95% confidence level (2)f as defined by ASTM Stan dard E 180-67." The above statement would provide a valid technical basis for a mutual understanding of the requirements of the method to be used. Even with the improved definition of the method, it would still be true that for all practical purposes a concentration level of 0.3 ppm or less would be required in order to assure an analysis showing compliance. (At 0.3 ppm the reproducibility of the method would be expected to be about 0.15 ppm). Similarly a concentration of 3.0 ppm or more would be required in order to give an analysis conclusively showing non-compliance. (At 3.0 ppm the re producibility of the method would be expected to be about 1.5 ppm.) These relationships can be illustrated by the attached diagranv which shows the three zones of reproducibility associated with the use of the specified method. N. H. Ketcham (1) Federal Register/ Vol. 39 No. 92, May 10, 1974, page 16896. (2) "Reproducibility" and "Confidence Level" are defined in ASTM Standard E 180-67, Reapproved tn 1972. These definitions are in paragraphs A 2.14 and A 2.12, re spectively. Part F of E 180 illustrates the use of these terms in precisian statements. NHK/ml Attachment ucc 028702 Reproducibility of the method when analyzing a known concentration of 3.0 ppm -3 * t lb % * 1 I ki Analysis (PPM by Vol.) {r 4.5 3.0 I 1.5 3 50% Reprotfcjcibilfty of the method at the "detectable concentretion " of 1 ppm Reproducibility of the method when analyzing a known concentration of 0.3 ppm wJ u. 0 m* 1^ W S" SN * u % * C \js * L0 0.5 f 0.45 1 0.3 L 0.15 1 50% 0.3 50% t 'r ucc 028703