I’ve got no major problem with working alone, so long as the person doing so uses good judgment in deciding what type of work is reasonable in these situations. When alone, it is prudent to limit yourself to experiments that don’t require especially hazardous reagents, dangerous conditions, or large scales. That said, I don’t think there are any black-and-white rules you can institute. Experience should also enter the analysis; you don’t want to try something dodgy for the first time when you are alone.
There are a bunch of other questions that can arise with respect to any outright ban of working alone. First off, what counts as “alone”? The institutional policies I’ve come across aren’t specific. Must the researchers working be located in the same bay? The same room? Same floor? Same building?
Another thing Gallagher highlights about his department’s safety culture is a prohibition on working alone—something that can be tricky to get right, he says. One approach is that no one works outside of 8 a.m. and 6 p.m. because no one else will be there. Another is to get people to collaborate to enable longer or later time in the lab. In Gallagher’s department, “We have a culture where students will work with one another to enable their experiments,” he says, noting that people must work within the line of sight of another person. Having someone in an office down the hall doesn’t cut it.
On one level, I agree with Paul that the work in question should dictate the circumstances–that is, after all, what risk assessment is all about. But are there not still some fundamental rules that should be in place? I always wear my seat belt in a car, for example, even when it’s daytime, the roads are dry, and the driver is sober, well-rested, and has been driving for many years without incident. (Full disclosure: I did work alone as a graduate student, principally to collect magnetic circular dichroism spectra. I’ve also been in a car accident.) It’s not always the hazards you know about that will cause a problem, as Gallagher also illustrated:
An incident at Bristol [in 2009] left a student’s face and hands badly cut when an experiment exploded and shattered the safety glass on the fume hood. With the benefit of hindsight, Gallagher says that the most likely cause was a side reaction that produced a small amount of an alkyl peroxide, which detonated when it came into contact with a ground-glass joint. But the peroxide formation was not something anyone had foreseen.
Are there basic safety policies that you think should be in place for all labs, all the time?