First a little background for those unfamiliar with Dan's reference to the compliance inspection manual etc. Start with our SDS FAQ question "Who can write an SDS?" at http://www.ilpi.com/msds/faq/partc.html#whocan
which begins with:
In theory, anyone can write an SDS as long as they are capable of filling out the OSHA-required elements outlined in Appendix D of the Hazard Communication Standard. Whether you have the expertise it takes to produce an accurate and complete sheet is another matter-
Note the reference to OSHA Publication 3844-02 2016, "HAZARD COMMUNICATION: Hazard Classification Guidance for Manufacturers, Importers, and Employers" as well as other resources and tips.
While 29 CFR 1910.1200, The OSHA Hazard Communication is the primary regulation governing SDS's in the US, the underlying rule book for OSHA inspectors is CPL 02-02-079, Inspection Procedures for the Hazard Communication Standard (HCS 2012). Our hyper-text improved version of this resource is http://www.ilpi.com/msds/osha/cpl0202079.html
Safety data sheets are discussed in Section G: http://www.ilpi.com/msds/osha/cpl0202079.html#sds_g
Section 6 (Inspection Procedures) describes what an inspector looks for in the content of a sheet.
Getting to the answer that Dan is seeking, I think the statement "full range of available scientific literature and other evidence concerning the potential hazards" makes it clear that a simple Google search effort is clearly not sufficient. However, the Standard is a performance-based standard that recognizes that one-size fits all methods and procedures can not apply to all of industry in the US and therefore looks for employers to use the steps necessary to achieve compliance with the intent (as well as specifics) of the regulation. For example, there is no point in requiring a company selling a Windex knockoff to be poking around SciFinder, however, a company that is marketing a new product based on a novel application of a new material certainly would be wise to do so. For another example of how performance criteria work with HCS see http://www.ilpi.com/msds/faq/partd.html#paperless
and the FAQ question that follows it.
I encourage anyone writing an SDS to thoroughly document their research efforts to show that they acted in good faith on the available data at the time they wrote the sheet, including screen shots of literature/web searches. I am aware of at least one (ancient) lawsuit stemming from differences between two SDS's for the same common material.
HCS's Hazard Classification process can be mind-numbing, especially for mixtures: http://www.ilpi.com/msds/osha/1910_1200_APP_A.html
The conversion from the old system of hazard determination to the new, more rigorous system of hazard classification has created a higher bar to the authorship of SDS's: http://www.ilpi.com/msds/ref/hazardclassification.html
That's generally a Good Thing. However, if someone is armed with insufficient knowledge they could create a mess of a sheet. And given that there is no place to submit SDS's for Official review, and because folks are lazy and copy other people's sheets without reviewing the underlying data, bad data can propagate. Which brings us right back to the Covid aerosol discussion again-
SDS quality (and the quality of the data that goes into the SDS) is a major issue. Although the HazCom Standard says that SDS authors "shall identify and consider the full range of scientific data and other evidence," there's no real standards for the quality of the literature search a manufacturer performs, either in the reg, interpretation letters, nor in the compliance inspection manual. Should a SDS writer do a Google search, consult specific government databases (e.g., PubChem, ToxNet, etc.), do a search on SciFinder, or what? There doesn't appear to be an answer.
I asked OSHA a while ago whether they ever actually enforced this provision, but they dodged the question. I'm reformulating the query as a FOIA request: In which inspections since the revised HazCom Standard took effect did they cite based on that requirement? If I get an answer, I'll write a paper on the topic.
Since there appear to effectively be no rules, I expect that SDS writers do literature searches either to the limit of their conscientiousness or to the minimum necessary to alleviate fear of liability. I've got a small collection of "bad SDSs" that contradict one another, list drastically-different hazards for the same compound, or give stupid first aid advice. I particularly like the SDS for H2O that advises that in the event of a skin exposure, one should rinse with water for 15 minutes.
I for one would welcome data search and quality standards for SDSs.
Regards, Dan Kuespert
Previous post | Top of Page | Next post