Over at incoherently scattered ponderings, there’s a post on safety at academic labs, which links to an article at Slate about an explosion at a lab which killed a worker, and discusses the difference in safety standards for students vs workers, and academia vs industry.
Why the difference between industry and academe? For one thing, the occupational safety and health laws that protect workers in hazardous jobs apply only to employees, not to undergraduates, graduate students, or research fellows who receive stipends from outside funders. (As a technician, Sheri Sangji was getting wages and a W-2. If she’d been paying tuition instead, Cal/OSHA could not even have investigated her death.)
I had not realized that students aren’t covered, but the disparity between the described situations is not surprising. I’ve spent time in academia (grad school) and worked in national labs (the NanoFabrication facility at Cornell, TRIUMF in Canada), and my current government job is a confluence of being industry/government and a quasi-national-lab (though not formally recognized as such). And I have to concur: lab safety in a university setting is not formally the priority is is in those other places. Academic safety leans far too much on the involvement of the PI, and leaves way too much to chance. A key difference of academia is that students are … students — they are still learning, and one cannot assume that they have the requisite experience to know much about the finer points of safety.
My first research at grad school happened the summer before classes — I got a grant to do six weeks of research, and did some work in a group doing some nonlinear optics. I was given a small project to build a device for making waveguides (one would couple light in and do four-wave mixing) and at some point I was using some chemicals, including some nasty solvent, 1,1,1 trichloroethane. It turns out the fume hood was not in the best shape, and I got a a little loopy, so I closed up shop and went home, and mentioned it to the professor the next day. He told me I made the right call, since dizziness was a symptom of inhalation, which could be followed by unconsciousness and death. Egads! I had no idea. But here’s what the MSDS says:
Inhalation of vapors will irritate the respiratory tract. Affects the central nervous system. Symptoms include headache, dizziness, weakness, nausea. Higher levels of exposure (> 5000 ppm) can cause irregular heart beat, kidney and liver damage, fall in blood pressure, unconsciousness and even death.
But there had been no formal training, no discussion of procedure or anything like that. There’s a lot more of “let’s do this and see what happens” going on, from what I could see. When I was working with Rubidium, we had a few bangs and booms that you get when alkalis mix with water. There was another group doing some solid-state investigations of materials structure which involved diffusing radioactive material into the sample. One day one of the students broke a sample holder and contaminated the lab, and another student got exposed. The whole thing was played down — the quote to the local paper was that it was the radiation equivalent of a few bananas, so it was no more dangerous than the produce department at the grocery store. (We did the calculation. The contamination on the lab bench was the equivalent of 30,000 bananas, and the fume hood radiation was the dose rate you’d get from 30,000,000 bananas.) So there were definitely some accidents stemming from inexperience.
During my time at national labs, I saw things were much more formalized. I had several hours of training at both the NNF and at TRIUMF, where there were formalized procedures for many common lab situations, and some level of enforcement. Accidents/incidents still happened, but there were people checking up on you to see that you were following procedure, which made for a safer environment. And though at school there was no program to run drills, e.g. practicing decontamination steps for those poor souls exposed to a few bananas, that isn’t the case elsewhere. At TRIUMF I recall a colleague practicing a repair (swapping out a target that wasn’t performing well) he would be doing in a radiation area, including people giving instructions along the lines of “at this point, if the part sticks and you can’t get it free in 30 seconds, leave it and abort,” because he was only allowed a certain amount of time in that area owing to the dose rate. At work I’ve had fire training — real extinguishers on real flames (along with the admonishment that it’s almost always a better idea to just get out and leave it to the professionals).
The one common theme in all of the places, though, was that when the bureaucracy got too involved, they focused on the wrong things. Everywhere I’ve been there’s a story of a coffee pot exploding, and some over-the-top reaction to it, while other situations are ignored. When I was in grad school the response was to ban coffee pots from offices, and only allow a large machine in the mail room, and it had to be unplugged whenever it was unattended by staff (i.e. the office staff, who worked 9-5 M-F. or put another way, there was at best a 50% overlap with the student hours. Thank goodness I get my caffeine from soda pop). All the while, equipment I had built — and certainly not inspected by the UL — was running 24/7 in the lab.
The danger of more supervisory involvement are safety mandates that basically mean shutting down the experiment if you complied. The laser safety folks at TRIUMF wanted an interlock which would turn the laser off when the door opened, which would have slowed experimental progress to a crawl because of all the things that would have needed to be reset each time. The last time I tried to buy Rubidium I was thwarted, because I wanted a different sized ampoule, only available from one vendor, and not the one for whom I had approval. Other than that, it’s the same product — still Rubidium, and Rubidium is Rubidium — but the approval was taking so long we decided to buy what we were allowed to get and modify the holder. More work for us, but it took less time than waiting for the red tape to get untangled. NIST had a plutonium incident where a sample was improperly handled, and the rumor is that there was widespread safety backlash. (One comment — how do you handle the safety of a laser frequency comb? The output is white light! Do you wear a welder’s mask? That reminds me of a complaint from other laser jocks about high-powered single-frequency lasers: if you wear goggles to block the light, you can’t see the beam and might burn yourself as a result.)
So I don’t know what the answer is. All things have risk, and it’s an illusion that everything can be made completely safe, so it’s important to understand that the goal is risk mitigation, not risk elimination. Certainly some more formal training and supervision would be helpful, if it’s properly done. But too often it isn’t properly done, and safety supervisors have a lot of power and aren’t motivated by your schedule, so a bad set of rules does a lot of harm of another sort.
I’ve never been content with the safety situation in any lab where I’ve worked, be it academic or government. (I’ve never worked for industry.) The safety rules were either too lax, too strict, or both.
For example, in the undergrad general chemistry labs at both of the universities I attended, students were required to wear full (hot, uncomfortable, fog-prone, vision-limiting) splash goggles at all times in the lab, even when no one was working with anything stronger than water. Meanwhile, directly over their heads in the research labs, grad students were transferring concentrated acids wearing nothing but T-shirts, shorts, and sandals. I generally knew (and know) enough to wear PPE appropriate to the job I’m doing, and I appreciated the flexibility offered by the absence of enforced rules, but some of my “cowboy” colleagues had their safety tolerances mis-set somewhere along the line and were (and are) courting disaster.
I’ve heard the “welding goggles” complaint (and made it myself) in the presence of visible OPO/OPA systems: “So, we need to block the Ti:Sapph output from 750 to 1000, the doubled Ti:Sapph from 375 to 500, the signal beam from 500 to 700…”