This thread has been quite fascinating.  Let me ruin it with some uninteresting musings (ramblings) below….

 

As someone who has authored hundreds of MSDSs and now SDSs for the two very TINY companies he’s worked for, I have largely enjoyed* the migration to HAZCOM 2012.  I look back at the initial 1986 regs and wonder how nothing could have even been in place before then.  Like…”wait, what?” <insert mind-blown emoji> What a shock to the system that had to have been for the old timers among us (no offense!).  I came on the scene in the mid-90s and it’s completely unfathomable now to think now that we’d only had mandatory MSDSs a few short years at that point.  They’re just so ubiquitous these days; yet, 35 years ago, there was literally…nothing?

 

I took over MSDS authoring from my boss in the late ‘90s.  I was never formally trained (in a classroom sense), but content origin mostly made sense right out of the gate.  Supplier MSDSs, of course, varied considerably back then.  Some used the very inadequate OSHA form and were only a page long.  Some were 16 or even 20+ pages long (Oxy’s Metso 20 comes to mind) and were amazingly thorough.  I appreciated the latter as I always tried to be as conscientious as possible when putting our MSDSs together.

 

* Then 2015 rolled around.  I never bothered to put SDS software money in my lab budget because I’d never needed something for MSDS authoring.  I won’t say that was a mistake, but it certainly made for a fascinating and tedious spring as my solution was to write an SDS content generator using M$ Excel.  After 3+ months of 12-14 hours days, I had about 8000 lines of simple calculations, ridiculously long IF() formulas, composite output statements, and everything in between, and I input base data for close to 100 raw materials.  What a beating.  However, it “saved” the company $20-30K (depending on your point of view) plus annual maintenance fees, and it gave me a fairly unique perspective of where every classification and H or P statement came from, how derived LD₅₀s are calculated, why sometimes you see the exclamation point and the exploding chest pictograms together while other times one trumps the other, etc.

 

That said, the problem of which manufacturer’s SDS to “believe” has been a constant challenge.  Which LC₅₀ do I use?  Why are PM flash points for the same chemical different?  Why does one SDS for n-propyl bromide say it’s flammable and the next one says it isn’t?  Finding quality, accurate data for MSDSs was challenging enough. but with the classifications and so many of the H and P statements for SDSs being based purely on numerical values, i.e. math, deciding which data to plug in for a calculation for a mixture became maddening.  I won’t admit to how long I struggled with ATEs for blends with methanol in them, but it did ultimately culminate with a phone call to OSHA.

 

Then, of course, there is the fact that there’s little global harmony in GHS, so that’s another ridiculously complex wrinkle that must be accounted for. 

 

Anyway, thanks for reading.  Stay in school.  Don’t do drugs.

 

The GHS was a terrific step forward in three ways – comprehensibility, globalization (in that it provided a ready-made system for countries that could not develop their own), and trade. But it has a serious flaw compared to the original OSHA Hazcom standard. (Full disclosure – I was a member of the coordinating committee that designed the SDS and labeling systems.) For some endpoints, the GHS takes a “weight of the evidence” approach, whereas the old OSHA standard mandated that a single positive well-conducted study was enough to establish carcinogenicity, mutagenicity, etc.  Weight of the evidence determinations can vary widely, and sometimes for reasons unrelated to the scientific evidence.  We’ve seen wildly different SDSs for the same chemical. Sometimes the same manufacturer will classify the same chemical differently in different jurisdictions. Clearly, this is not a science-based decision.

 

This is less a problem in the USA than in some other countries, because the easiest way for a manufacturer to lose a big product liability case is to withhold information from users. Thank God for our contingent-fee tort system. But we’ve still seen plenty of discrepancies.

 

I’ll leave to the lawyers whether OSHA could go back to the “one positive study” approach without running afoul of trade law. But even if they can’t, they could make it harder for a classifier to ignore positive studies. One way would be to require a transparent, publically available analysis, disclosing the evidence relied on and the evidence rejected, for any “weight of the evidence”  determination. Some of us discussed that possibility with OSHA during the Obama Administration, and some preliminary efforts were underway, but of course they ended in 2017.

 

Fortunately, OSHA’s current Hazcom rulemaking provides a splendid opportunity to push this issue again. The initial comment period ended last week, but hearings are scheduled for September, so there will be plenty of chances for written and oral testimony and post-hearing comments.  I know OSHA would welcome participation by members of the DCHAS.

 

 

 

> Requiring references is the norm in published scientific literature, why not this standard of quality and traceability in SDSs as a key source of information?

 

I think I can suggest the reason for this. The SDS did not begin existence as scientific literature, but as industrial safety information provided by the manufacturer for the _industrial_ user, and that in a much less formal era of the 1960s.  There have been significant modifications since then but dragging it into the realm of scientific literature is clearly a herculean task still in progress and unlikely to be completed in the near future.  That doesn’t mean you should not continue to seek that level, but there is little point in being incensed about the SDS not being at that level now or soon.

Peter Zavon, CIH
Penfield, NY

PZAVON**At_Symbol_Here**Rochester.rr.com

 

 

 

--- For more information about the DCHAS-L e-mail list, contact the Divisional membership chair at membership**At_Symbol_Here**dchas.org Follow us on Twitter **At_Symbol_Here**acsdchas

--- For more information about the DCHAS-L e-mail list, contact the Divisional membership chair at membership**At_Symbol_Here**dchas.org Follow us on Twitter **At_Symbol_Here**acsdchas

--- For more information about the DCHAS-L e-mail list, contact the Divisional membership chair at membership**At_Symbol_Here**dchas.org Follow us on Twitter **At_Symbol_Here**acsdchas