March 20, 1990

Charles R. Clark, PhD Manager, Toxicology and Product Safety Unocal Corporation 1201 West 5th Street Post Office Box 7600 Los Angeles, California 90051

Dear Dr. Clark:

Thank you for your letter of November 15, 1989, to Ms. Jennifer Silk of the Occupational Safety and Health Administration's (OSHA) Directorate of Health Standards. Your letter was referred to the Directorate of Compliance Programs for response, and I apologize for the delay of this letter to you. You requested a clarification of OSHA's policy on hazard warnings required on product labels as per the Hazard Communication Standard (HCS), 29 CFR 1910.1200.

In your letter you specifically ask: ".. is the consideration of exposure potential acceptable in determining whether health hazard warnings should be included on the labels?" OSHA policy on compliance with the provisions of the HCS is that exposure calculations are not permitted in determining whether a hazard must appear on a label. If there is a potential for exposure, (other than in minute, trace or very small quantities), the hazard must be included when well-substantiated. Suppliers may not exclude hazards based on presumed levels of exposure downstream (i.e., omitting a carcinogenic hazard warning because, in the supplier's estimate, presumed exposure will not be high enough to cause the effect). The hazard is an intrinsic property of the chemical. Exposure determines degree of risk and should be addressed in training programs by the downstream employer. It should be noted that the General Carbon decision confirmed OSHA's longstanding interpretation of the rule's requirements -- it does not represent a new or different position on those issues.

As you point out in your letter, OSHA's compliance instruction for the HCS, CPL 2-2.38B, states that "it will not necessarily be appropriate to warn on the label about every hazard listed in the MSDS." The weight of the evidence for each hazard reported on the MSDS is to be considered when determining if the hazard needs to be included on the label, not the predicted downstream exposure amounts. Of course, the hazardous chemical must be present in such a manner that it is available for exposure, i.e., if the hazardous chemical is an ingredient in a mixture, and the chemical is bound in such a manner that it is incapable of resulting in exposure, then information on its potential health effects would not have to be included on either the MSDS or the label.

As you may be aware, OSHA staff met with ANSI representatives during the development of the ANSI labeling guidelines. At that time, we discussed the implications of the language regarding exposure assessments, and also discussed our concerns that this approach was not consistent with the HCS. Our understanding, based on those meetings, is that the ANSI committee members believe the differences between the ANSI standard and the HCS are merely semantic in nature, and not substantive. They believe that a label properly prepared in accordance with the ANSI guidance will meet OSHA's requirements as stated herein. We appreciate their efforts in developing the ANSI labeling standard, which we believe serves as a beneficial tool to employers trying to comply with the requirements of the Hazard Communication Standard. However, employers must continue to ensure that their hazard determinations are based on the requirements of the HCS itself, using the ANSI standard or other such documents as supplementary information and guidance.

I hope this discussion helps clarify the concerns you raised. Please feel free to contact us again if we can be of further assistance.

Sincerely,

Patricia K. Clark Director Designate Directorate of Compliance Programs