December 10, 1986

P. W. Hess, Ph.D.
Director, Environmental Affairs
Hershey Foods Corporation
Technical Center
Post Office Box 805
Hershey, Pennsylvania 17033-0805

Dear Dr. Hess:

This is in response to your letter of November 5, to Mr. John B. Miles, Director, Directorate of Field Operations, Occupational Safety and Health Administration (OSHA) asking: "How does OSHA calibrate its dosimeters to see if they respond accurately to impulsive signals or noise?" Your letter was forwarded to this office for response. Note that on October 10 this office (the Directorate of Technical Support) responded to a similar letter from you to Mr. Miles dated July 31.

The OSHA policy regarding field calibration procedures of noise measurement instruments was explained in our October 10 letter, namely that field calibrations on all OSHA noise measurement instruments are performed according to manufacturers' instructions and that all such instruments are returned yearly to the OSHA Measurement and Calibration Laboratory in Cincinnati, Ohio for checkup and calibration.

The restrictive clause in your inquiry - "to see if they respond accurately to impulsive signals or noise" - is a question that does not apply to calibration, especially field calibration, of noise dosimeters, but rather to the intrinsic measurement abilities of such instruments themselves. This is a question that can best be answered directly by instrument manufacturers.

Measurements made by any noise measurement device employing A-weighted slow response of a pulse train output (106 dB for .5 seconds and 86 dB for 9.5 seconds) form a GenRed Omnical generator can not result in a workshift time-weighted average of 90 dB. Please note that both Table G-16 and Table G-16a of the OSHA 29 CFR 1910.95 Occupational Noise Exposure regulations refer to measured dBA slow response sound levels.

This is clearly spelled out in both paragraph 1910.95(a) and the equation for reference duration, T, in section I.(3) of 1910.95 Appendix A. Your calculated 90 dB time-weighted average clearly results from the arbitrary choice of a response time effectively equal to zero.

If you need additional information, please contact Mr. Raymond G. Kunicki of my staff at (202) 523-7065.

Sincerely,



Edward J. Baier
Director
Directorate of Technical Support