My student fucked up and almost killed himself, what now?
April 25, 2010 8:19 PM Subscribe
Last night my student caused a catastrophic high speed centrifuge failure while alone in the lab. What do I do?
I am an undergraduate researcher at an institution which does not take graduate students in my field. This means that I am one of a few more experienced undergrads who functionally is treated like and expected to fill the traditional role of a graduate student would in most small to mid sized Molecular/Micro labs. I have been given two separate projects, one involving bench-work and one involving bio-informatics and I am supervising two advanced track students helping me for credit, one for each project.
The student:
One of my students, we'll call this person (S), is a phenomenal student. (S) regularly receives the top scores on (S)'s exams in the hardest year long series of courses at our college and is never outside the 90th percentile. (S) is a known and generous source of tutoring for other students, volunteered for a quarter in our lab before he was able to be here for credit, and now works as the lab dishwasher while working with me and training to replace the undergraduate leader of our sister lab for next year. (S) gets an amazing amount done in a day, learns something once and learns it right.
The cause:
(S) was working on the weekend with no one else in the lab when he needed to spin down 0.5 L of fluid at 12,500 RPM for an aspect of our project I had expected to do with him tomorrow and for one of three reasons the centrifuge failed.
1) Our centrifuge has been having chronic problems with the door sealing which may have in some way caused a problem. This is extremely unlikely and does not make much sense, but is possible.
2) The lid of the centrifuge rotor (PDF) was not tightened properly which should have been easily detected by lifting up on the rotor and seeing that it was not secure. I had recently switched the rotor to one he was less familiar with which may have contributed.
3) The four centrifuge tubes used to hold the fluid were not balanced properly.
The accident:
When (S) ran the centrifuge (S) noticed an unfamiliar whizzing sound and as it accelerated the rotor (S) became concerned and aborted the run at ~5,000 RPM. The brakes then caused the rotor to disengage spinning to rotor off of its axis and around the chamber causing superficial but possibly significant damage to the chamber, an unknown amount of damage to the column which is the axis the rotor spins around, and irreparable damage to the rotor. If (S) had not aborted the run it would have likely still failed but with more force and the rotor would have either punched through the chamber, likely killing (S), or caused the centrifuge to leap at him, again likely killing (S). Instead the rotor thankfully banged around the chamber to a stop without injury.
(S) first opened the centrifuge door, exposing himself to potentially dangerous fumes and particles, to look at the damage and them called me immediately to ask what to do. I told him to not touch anything, e-mail our central lab technician and that I would be there as quickly as humanly possible.
My arrival at the scene:
As soon as I arrived I made sure (S) was alright and saw that the rotor was indeed irreparably damaged. (S) had also almost finished cleaning up the powdered graphite and aluminum shavings that coated the rotor chamber and just gotten off of the phone from a half hour conversation with one of our two Principal Investigators (who is as much a grandmother to all of us as a PI) who was upset but supportive of (S). He showed me his e-mail which he had sent to our instructional technician, CCing both PIs, which was coherent, though uncharacteristically poorly worded and punctuated. It explained what had happened to the best of his understanding and that he understood that he had just made a large and dangerous mistake.
I made sure both the centrifuge and the lab were secure, arranged all of the pieces I could find coherently, made sure (S) was not impaired in any way, and responded to a one word e-mail from our (for these purposes) main PI which asked when I would be in tomorrow by saying everything was safe and relaying my schedule. I then took (S) out to pizza at a local spot where some of our mutual friends had planned to meet with whom we then spent the night hanging out.
Now here is the mystery I am unsure of how to address:
Before (S) and I left we decided that the most likely cause was lid related somehow, particularly since that was something we could both accept blame for, mitigating any direction our PIs' disappointment might form. However since then another theory has come up that I'm not sure I can ignore.
(S) needed to be balance the centrifuge tubes to the hundredth of a gram and I trust that he did exactly that. However, in sorting through the aftermath I did weigh the tubes and found two of the tubes were still balanced to each other to the tenth of a gram, while the other two were also still balanced to the tenth of a gram with the exception of 10 full grams missing. On the scale we use for this purpose it would have been very easy to miss the ten gram difference, particularly if one were in a hurry.
I am almost completely convinced that there was a ten gram difference and that would have been way more than enough to do the damage. (S) however is convinced that this could not have happened and asserts that the current difference in mass results from the tubes leaking as evidenced by some mysterious fluid in the centrifuge that he found but had wiped up. If the fluid was from the tubes my instinct would tell me any leaky tube should have lost much more than the 10 missing milliliters. Additionally I looked in the trash and found the paper towels (S) used to be crusty as if they were wet but not be stained with the TSB colored media in the tubes suggesting to me that what (S) saw was condensation that had formed in the cooled chamber after (S) called me.
I have already shared this theory with our grandmother PI and so I couldn't just hide it even if I wanted to. But the last thing I want to do is back (S) into a corner for something that, even assuming this theory is right, (S) simply did not notice. Particularly on top of the deep water (S) is already in, and the last thing the lab needs right now is conflict.
I have a meeting with the more relevant PI tomorrow morning. Is there something I've missed socially or technically I need to pay attention to? How should I address this mystery?
Thank you in advance.
I am an undergraduate researcher at an institution which does not take graduate students in my field. This means that I am one of a few more experienced undergrads who functionally is treated like and expected to fill the traditional role of a graduate student would in most small to mid sized Molecular/Micro labs. I have been given two separate projects, one involving bench-work and one involving bio-informatics and I am supervising two advanced track students helping me for credit, one for each project.
The student:
One of my students, we'll call this person (S), is a phenomenal student. (S) regularly receives the top scores on (S)'s exams in the hardest year long series of courses at our college and is never outside the 90th percentile. (S) is a known and generous source of tutoring for other students, volunteered for a quarter in our lab before he was able to be here for credit, and now works as the lab dishwasher while working with me and training to replace the undergraduate leader of our sister lab for next year. (S) gets an amazing amount done in a day, learns something once and learns it right.
The cause:
(S) was working on the weekend with no one else in the lab when he needed to spin down 0.5 L of fluid at 12,500 RPM for an aspect of our project I had expected to do with him tomorrow and for one of three reasons the centrifuge failed.
1) Our centrifuge has been having chronic problems with the door sealing which may have in some way caused a problem. This is extremely unlikely and does not make much sense, but is possible.
2) The lid of the centrifuge rotor (PDF) was not tightened properly which should have been easily detected by lifting up on the rotor and seeing that it was not secure. I had recently switched the rotor to one he was less familiar with which may have contributed.
3) The four centrifuge tubes used to hold the fluid were not balanced properly.
The accident:
When (S) ran the centrifuge (S) noticed an unfamiliar whizzing sound and as it accelerated the rotor (S) became concerned and aborted the run at ~5,000 RPM. The brakes then caused the rotor to disengage spinning to rotor off of its axis and around the chamber causing superficial but possibly significant damage to the chamber, an unknown amount of damage to the column which is the axis the rotor spins around, and irreparable damage to the rotor. If (S) had not aborted the run it would have likely still failed but with more force and the rotor would have either punched through the chamber, likely killing (S), or caused the centrifuge to leap at him, again likely killing (S). Instead the rotor thankfully banged around the chamber to a stop without injury.
(S) first opened the centrifuge door, exposing himself to potentially dangerous fumes and particles, to look at the damage and them called me immediately to ask what to do. I told him to not touch anything, e-mail our central lab technician and that I would be there as quickly as humanly possible.
My arrival at the scene:
As soon as I arrived I made sure (S) was alright and saw that the rotor was indeed irreparably damaged. (S) had also almost finished cleaning up the powdered graphite and aluminum shavings that coated the rotor chamber and just gotten off of the phone from a half hour conversation with one of our two Principal Investigators (who is as much a grandmother to all of us as a PI) who was upset but supportive of (S). He showed me his e-mail which he had sent to our instructional technician, CCing both PIs, which was coherent, though uncharacteristically poorly worded and punctuated. It explained what had happened to the best of his understanding and that he understood that he had just made a large and dangerous mistake.
I made sure both the centrifuge and the lab were secure, arranged all of the pieces I could find coherently, made sure (S) was not impaired in any way, and responded to a one word e-mail from our (for these purposes) main PI which asked when I would be in tomorrow by saying everything was safe and relaying my schedule. I then took (S) out to pizza at a local spot where some of our mutual friends had planned to meet with whom we then spent the night hanging out.
Now here is the mystery I am unsure of how to address:
Before (S) and I left we decided that the most likely cause was lid related somehow, particularly since that was something we could both accept blame for, mitigating any direction our PIs' disappointment might form. However since then another theory has come up that I'm not sure I can ignore.
(S) needed to be balance the centrifuge tubes to the hundredth of a gram and I trust that he did exactly that. However, in sorting through the aftermath I did weigh the tubes and found two of the tubes were still balanced to each other to the tenth of a gram, while the other two were also still balanced to the tenth of a gram with the exception of 10 full grams missing. On the scale we use for this purpose it would have been very easy to miss the ten gram difference, particularly if one were in a hurry.
I am almost completely convinced that there was a ten gram difference and that would have been way more than enough to do the damage. (S) however is convinced that this could not have happened and asserts that the current difference in mass results from the tubes leaking as evidenced by some mysterious fluid in the centrifuge that he found but had wiped up. If the fluid was from the tubes my instinct would tell me any leaky tube should have lost much more than the 10 missing milliliters. Additionally I looked in the trash and found the paper towels (S) used to be crusty as if they were wet but not be stained with the TSB colored media in the tubes suggesting to me that what (S) saw was condensation that had formed in the cooled chamber after (S) called me.
I have already shared this theory with our grandmother PI and so I couldn't just hide it even if I wanted to. But the last thing I want to do is back (S) into a corner for something that, even assuming this theory is right, (S) simply did not notice. Particularly on top of the deep water (S) is already in, and the last thing the lab needs right now is conflict.
I have a meeting with the more relevant PI tomorrow morning. Is there something I've missed socially or technically I need to pay attention to? How should I address this mystery?
Thank you in advance.
Accidents happen. I don't see what the big deal is.
posted by mr_roboto at 8:29 PM on April 25, 2010 [3 favorites]
posted by mr_roboto at 8:29 PM on April 25, 2010 [3 favorites]
Best answer: 2nd-ing liketitanic. Accidents happen, your only responsibility is to deliver the facts and possibilities truthfully and in as neutral a fashion as possible. It's ok to stand up for (S) to prevent (S) from getting thrown under the bus, but don't look for blame elsewhere either.
(S) is probably scared shitless still, so I'd try to keep (S) at ease as much as you can.
posted by weaponsgradecarp at 8:30 PM on April 25, 2010 [1 favorite]
(S) is probably scared shitless still, so I'd try to keep (S) at ease as much as you can.
posted by weaponsgradecarp at 8:30 PM on April 25, 2010 [1 favorite]
Best answer: I worked in a lab as an undergrad, and a grad student and I were alone in the lab when a gasket catastrophically failed on a $400K machine. It wasn't immediately clear what happened, whether it was the machine's fault or ours, and we were both shaken by the fact that we could have been seriously injured. We called the lab manager, who came in at 2 a.m. to help us. We also called the PI in the middle of the night, and the guy in Germany who'd built the thing for us in the first place. In the end, it was pretty clearly an error on our end, and no one cared. I mean, they cared that the machine we used to do all of our experiments took months to get working properly again, sure. But no one made us feel bad about it — these things happen — and the PI wrote me a glowing recommendation for grad school the next year.
Regardless of whether your lab manager and PI take the stance mine did, it's not your responsibility to decide anyway. Stay out of it.
posted by adiabat at 8:38 PM on April 25, 2010 [8 favorites]
Regardless of whether your lab manager and PI take the stance mine did, it's not your responsibility to decide anyway. Stay out of it.
posted by adiabat at 8:38 PM on April 25, 2010 [8 favorites]
Best answer: It sounds like S made every effort to be honest and take responsibility for the incident. Everything you said about this person put them in a good light and makes them out to be a valuable asset to your lab. Whether the centrifuge malfunctioned, or it was a balancing error on S's part, these things happen. I would stop assigning blame to S and continue to encourage their personal development as a researcher. Leave any prying questions and decisions up to the P.I.
In the interim, perhaps minimize the unsupervised weekend lab sessions.
posted by clearly at 8:39 PM on April 25, 2010 [17 favorites]
In the interim, perhaps minimize the unsupervised weekend lab sessions.
posted by clearly at 8:39 PM on April 25, 2010 [17 favorites]
Having spent quite a some time in (chemistry) labs, this
was working on the weekend with no one else in the lab
should be addressed. Accidents happen for the most various reasons, as I'm sure you're aware of. Take a deep breath, thank your deity of choice (if available) that noone got hurt or killed, see if anything can be done to review safety policies.
posted by _dario at 8:51 PM on April 25, 2010 [4 favorites]
was working on the weekend with no one else in the lab
should be addressed. Accidents happen for the most various reasons, as I'm sure you're aware of. Take a deep breath, thank your deity of choice (if available) that noone got hurt or killed, see if anything can be done to review safety policies.
posted by _dario at 8:51 PM on April 25, 2010 [4 favorites]
Best answer: You might want to make clear to S that the accident itself matters much less than how he handles the accident.
posted by amtho at 8:54 PM on April 25, 2010 [19 favorites]
posted by amtho at 8:54 PM on April 25, 2010 [19 favorites]
S did fine in the aftermath, regardless of of what he may or may not have done to precipitate the mess. You, too, have done well, in taking care of him and keeping communication open.
Neither S nor you will have to pay (either literally or metaphorically) for this, so there are no material consequences for anyone's actions. It's just those awful life lessons that will agonize you and S (and make you balance your tubes with precision) for the rest of your days. In short, calm down.
If he stopped the spin at 5K rpm that was all he could do. it sounds like the thing was in dire need of repair regardless and it just happened to crap out on him. Rotten luck, but maybe it will get repaired now. Same thing just happened on my floor a few months ago. No one was hurt but the centrifuge was down for a week. C'est la vie.
For both of you, continue to be as responsible and responsive as possible, knowing that you're all adults, keeping in mind that blame is not the same thing as responsibility.
posted by Cold Lurkey at 9:01 PM on April 25, 2010
Neither S nor you will have to pay (either literally or metaphorically) for this, so there are no material consequences for anyone's actions. It's just those awful life lessons that will agonize you and S (and make you balance your tubes with precision) for the rest of your days. In short, calm down.
If he stopped the spin at 5K rpm that was all he could do. it sounds like the thing was in dire need of repair regardless and it just happened to crap out on him. Rotten luck, but maybe it will get repaired now. Same thing just happened on my floor a few months ago. No one was hurt but the centrifuge was down for a week. C'est la vie.
For both of you, continue to be as responsible and responsive as possible, knowing that you're all adults, keeping in mind that blame is not the same thing as responsibility.
posted by Cold Lurkey at 9:01 PM on April 25, 2010
It doesn't matter whose fault it was, nobody's perfect and things get missed all the time. Unless there's been signs of a more severe and systematic problem with S's lab techniques (and nothing in your post indicates this) there's no reason to inform your PI of your sleuthing.
More importantly, since S is still in training and under your supervision, why is he allowed to work alone in the lab? And on a weekend, no less, when campus security/emergency support services operate on a more limited capacity?
posted by cosmic_shoals at 9:24 PM on April 25, 2010 [1 favorite]
More importantly, since S is still in training and under your supervision, why is he allowed to work alone in the lab? And on a weekend, no less, when campus security/emergency support services operate on a more limited capacity?
posted by cosmic_shoals at 9:24 PM on April 25, 2010 [1 favorite]
Just nthing what people above said, basically, but:
Mistakes happen in labs. Sometimes big, expensive mistakes, sometimes small, cheap mistakes. It's an inevitable facet of lab work. All one can do and all any lab member of high authority wants is someone to be honest and own up to their mistakes - exactly what you and S did here. Furthermore, S should be commended at having the presence of mind to stop the spin at 5K. Seriously, that's more than I would expect from many undergrads.
There may be some awkwardness in lab for a week or two, but this'll blow over, no problem.
And P.S.: PIs don't care about perfect grammar and spelling when it comes to email communication. Trust me.
posted by sickinthehead at 9:29 PM on April 25, 2010
Mistakes happen in labs. Sometimes big, expensive mistakes, sometimes small, cheap mistakes. It's an inevitable facet of lab work. All one can do and all any lab member of high authority wants is someone to be honest and own up to their mistakes - exactly what you and S did here. Furthermore, S should be commended at having the presence of mind to stop the spin at 5K. Seriously, that's more than I would expect from many undergrads.
There may be some awkwardness in lab for a week or two, but this'll blow over, no problem.
And P.S.: PIs don't care about perfect grammar and spelling when it comes to email communication. Trust me.
posted by sickinthehead at 9:29 PM on April 25, 2010
I've never worked with a high-speed centrifuge before, though a few things about this seem incredibly odd. Is it common for these machines not to have some sort of mechanism to detect when they're unbalanced, or a mode of failure that doesn't involve the destruction of the machine and possible death of its operator?
Although I understand that labwork will never be child-safe or idiot-proof, I'm surprised that the machine requires an incredibly delicate balancing procedure in order to work without failing dangerously.
Also, the 'alone in the lab' scenario is a big no-no, which both you and the PI should already know about. Humans do screw up, and you need to be prepared for this eventuality. (Doubly so if this took place at night, and 10x so if the student wasn't on an actual "night shift." Bhopal and Chernobyl took place on the night shift. People make more mistakes at night even if they're well-rested, and it sounds like your guy wasn't, especially if he's an undergraduate at this time the of the year).
What it comes down to is: Either your undergraduate was violating the rules by being alone in the lab at night, and operating a piece of equipment that he wasn't familiar with, or you and your PI were grossly negligent by allowing this scenario to play out.
posted by schmod at 9:37 PM on April 25, 2010 [2 favorites]
Although I understand that labwork will never be child-safe or idiot-proof, I'm surprised that the machine requires an incredibly delicate balancing procedure in order to work without failing dangerously.
Also, the 'alone in the lab' scenario is a big no-no, which both you and the PI should already know about. Humans do screw up, and you need to be prepared for this eventuality. (Doubly so if this took place at night, and 10x so if the student wasn't on an actual "night shift." Bhopal and Chernobyl took place on the night shift. People make more mistakes at night even if they're well-rested, and it sounds like your guy wasn't, especially if he's an undergraduate at this time the of the year).
What it comes down to is: Either your undergraduate was violating the rules by being alone in the lab at night, and operating a piece of equipment that he wasn't familiar with, or you and your PI were grossly negligent by allowing this scenario to play out.
posted by schmod at 9:37 PM on April 25, 2010 [2 favorites]
I agree with the earlier posters. It seems to me that among your worries is the possibility that S is covering up his error. This is certainly possible, but it assumes that he thinks he made the error to start with. By your account, he can't be really sure that he measured the fluid correctly. I presume he's generally careful and would have tried to avoid mismeasurement. So at worst he's not trying to conceal the truth: he's merely someone who prefers a particular finding - that is, he's not dispassionate and disinterested in the outcome. But we already knew that.
My feeling is that 10ml from a 125ml tube is a large-ish discrepancy and one that even an amateur ought to have noticed. From your description of S I understand him to be meticulous; I can't imagine someone like that making an error of this magnitude. It's certainly possible that he made this error, but I don't think it's necessarily more probable than the other theories. I wouldn't push this finding, particularly when the accident revealed several serious safety risks - it could be that you subconsciously prefer an outcome that doesn't blame your lab's equipment!
Since you're doing the right thing by analysing possible causes, please don't ignore the problem with your scales. It needs to be written up. If S did make an error, the scales were a likely contributor and this is just unacceptable. Lab scales are relatively cheap. Students, even undergrads, are not.
posted by Joe in Australia at 9:44 PM on April 25, 2010 [2 favorites]
My feeling is that 10ml from a 125ml tube is a large-ish discrepancy and one that even an amateur ought to have noticed. From your description of S I understand him to be meticulous; I can't imagine someone like that making an error of this magnitude. It's certainly possible that he made this error, but I don't think it's necessarily more probable than the other theories. I wouldn't push this finding, particularly when the accident revealed several serious safety risks - it could be that you subconsciously prefer an outcome that doesn't blame your lab's equipment!
Since you're doing the right thing by analysing possible causes, please don't ignore the problem with your scales. It needs to be written up. If S did make an error, the scales were a likely contributor and this is just unacceptable. Lab scales are relatively cheap. Students, even undergrads, are not.
posted by Joe in Australia at 9:44 PM on April 25, 2010 [2 favorites]
Best answer: schmod I do work with ultra high centrifuges fairly often and you're off base with your assessment of the machine. Incredibly delicate balancing is par for the course and, while they do have failsafes built in, once they're getting up to speed those failsafes aren't necessarily enough to stop them. That's why there's a manual override, it's needed. At the same time they're easy to use and I'm certified to use ours after hours on my own and our health and safety rules are very strict.
To the OP: when a centrifuge makes a weird noise the right thing to do is turn it off straight away. For ours we stay nearby and listen carefully until it's up to speed just as a matter of course. So your student did the exact right thing and that's why he not dead. Possibly something worth mentioning to him as a small ray of light in the whole situation.
Then the right thing to do afterwards is to escalate up the chain of command, which both of you did in a timely and responsible manner. So again you've done the right thing. Now it's up to the lab manager and the University to deal with it, this is big enough that people right up the chain of command should be involved. You don't need to worry about that though, your PI and lab manager will know the protocol to follow and they'll tell you what to do. Don't worry about apportioning blame or trying to cover your butts, just lay out the information that you have and let them figure it out.
Part if what happens next should be reviewing the procedures to make sure it doesn't happen again, and part of that includes talking about what might have happened. That's where you'd mention the balancing issue, not as an accusation but as one of many pieces of information. Keep in mind that the failure could have easily caused liquid to leak in some unexpected way (lids deforming under uneven pressure for example) so you'll never really know if it was done wrong or not, but focus instead on the part where uncertainty is possible due to your balance design and see if there's something you can do about that. For example we actually use old fashioned scales, the kind with two arms sticking out that physically move up and down, so the tubes are always perfectly balanced even if you don't know exactly how much is in each of them.
You've dealt with this appropriately. Now it's time to let the higher ups worry about it, that's what they get paid for after all.
posted by shelleycat at 10:41 PM on April 25, 2010 [5 favorites]
To the OP: when a centrifuge makes a weird noise the right thing to do is turn it off straight away. For ours we stay nearby and listen carefully until it's up to speed just as a matter of course. So your student did the exact right thing and that's why he not dead. Possibly something worth mentioning to him as a small ray of light in the whole situation.
Then the right thing to do afterwards is to escalate up the chain of command, which both of you did in a timely and responsible manner. So again you've done the right thing. Now it's up to the lab manager and the University to deal with it, this is big enough that people right up the chain of command should be involved. You don't need to worry about that though, your PI and lab manager will know the protocol to follow and they'll tell you what to do. Don't worry about apportioning blame or trying to cover your butts, just lay out the information that you have and let them figure it out.
Part if what happens next should be reviewing the procedures to make sure it doesn't happen again, and part of that includes talking about what might have happened. That's where you'd mention the balancing issue, not as an accusation but as one of many pieces of information. Keep in mind that the failure could have easily caused liquid to leak in some unexpected way (lids deforming under uneven pressure for example) so you'll never really know if it was done wrong or not, but focus instead on the part where uncertainty is possible due to your balance design and see if there's something you can do about that. For example we actually use old fashioned scales, the kind with two arms sticking out that physically move up and down, so the tubes are always perfectly balanced even if you don't know exactly how much is in each of them.
You've dealt with this appropriately. Now it's time to let the higher ups worry about it, that's what they get paid for after all.
posted by shelleycat at 10:41 PM on April 25, 2010 [5 favorites]
Aside from what everyone else here's said (and seconding not blaming the student for an innocent fuckup), can I recommend you stop using electronic scales for weighing out centrifuge tubes.
Go and get one of the old fashioned balances, where you have to put one centrifuge tube on each side. They'll give you enough accuracy for the rcf's you're using, and this way, your student will always be dealing with balanced pairs of tubes, so this kind of misreading cannot happen.
Reading numbers on digital scales dozens of times when you're stressed or tired is incredibly repetitive and prone to this kind of mistake.
posted by 7-7 at 10:42 PM on April 25, 2010 [3 favorites]
Go and get one of the old fashioned balances, where you have to put one centrifuge tube on each side. They'll give you enough accuracy for the rcf's you're using, and this way, your student will always be dealing with balanced pairs of tubes, so this kind of misreading cannot happen.
Reading numbers on digital scales dozens of times when you're stressed or tired is incredibly repetitive and prone to this kind of mistake.
posted by 7-7 at 10:42 PM on April 25, 2010 [3 favorites]
OP, that isn't your real name, is it (or is that your real name in your profile?) If it is, do you want to get the moderators to make you anonymous?If there is an investigation at your university/college/research institution, I don't think pontifications on what happened with a name attached should be in here. A lot of times it is easy to figure out if someone works in a lab, which lab, who his or her PI is, etc, all with a name
If that isn't your real name (or it doesn't matter to you), please ignore my comment.
posted by Wolfster at 10:57 PM on April 25, 2010
If that isn't your real name (or it doesn't matter to you), please ignore my comment.
posted by Wolfster at 10:57 PM on April 25, 2010
Best answer: It really doesn't matter who made the mistake (or if indeed this wasn't just a mechanical failure). Students, with the best of intentions, will fuck up. Blaming them for fucking up is counter-productive. It doesn't make them better students and it won't improve their lab abilities. It can drive someone away from the lab for life though.
What's the important things here? S is safe and hopefully learned something from this. Hopefully you and your PIs have learned something about this. S's takeaway is how to handle himself in a dangerous situation and how to own up to a mistake. Good character building stuff.
The lessons for you and your PIs are rather more important. When you are supervising someone in a lab, someone must be physically present in the lab with the student. Students should never, ever be working alone. In the training we get for supervising students, it's made quite clear that direct supervision students who work with high hazards is very serious, potentially go-to jail serious, if the worst happens. If he'd lost a hand or an eye, he would have been entirely within his rights to sue you personally, your PI and the school, even if he was the direct cause of the accident. You and your PIs really need to get a handle on this.
Review your work procedures, in particular the after hours policies. What you've discovered is that really bright folks with poor training and little experience are not immune to accidents. The really dangerous are those who don't know that they don't know something. That sounds very much like what happened here. It's your job, but mostly your profs' job to keep an eye on newbies and make certain that stuff like this doesn't happen.
Casting blame on the operator is the least sensible thing to do in this situation. It won't stop it from happening with the next bright young student. For that you need to look at your work procedures.
posted by bonehead at 10:57 PM on April 25, 2010
What's the important things here? S is safe and hopefully learned something from this. Hopefully you and your PIs have learned something about this. S's takeaway is how to handle himself in a dangerous situation and how to own up to a mistake. Good character building stuff.
The lessons for you and your PIs are rather more important. When you are supervising someone in a lab, someone must be physically present in the lab with the student. Students should never, ever be working alone. In the training we get for supervising students, it's made quite clear that direct supervision students who work with high hazards is very serious, potentially go-to jail serious, if the worst happens. If he'd lost a hand or an eye, he would have been entirely within his rights to sue you personally, your PI and the school, even if he was the direct cause of the accident. You and your PIs really need to get a handle on this.
Review your work procedures, in particular the after hours policies. What you've discovered is that really bright folks with poor training and little experience are not immune to accidents. The really dangerous are those who don't know that they don't know something. That sounds very much like what happened here. It's your job, but mostly your profs' job to keep an eye on newbies and make certain that stuff like this doesn't happen.
Casting blame on the operator is the least sensible thing to do in this situation. It won't stop it from happening with the next bright young student. For that you need to look at your work procedures.
posted by bonehead at 10:57 PM on April 25, 2010
Best answer: When I was an undergrad, we had a really nice vacuum system that contained a diffusion pump. (For the unfamiliar, this is a type of high-vacuum pump that involves boiling oil. It only works once you already have a moderate vacuum.) All the valves on it were numbered, there was a very clear procedure. Everyone knew the drill.
I'm sure you know where I'm going with this. One day, a guy I knew left a valve open and didn't check the pressure before turning on the diffusion pump. The oil caught on fire. Thousands of dollars of sensors were ruined by the smoke, and the whole system had to be taken apart and cleaned.
What I remember most about all this, and why I bring the story up at all, is because the professor who ran the lab dealt with it like our own private Challenger incident. There wasn't any placing blame. There was an investigation, but it was just to determine what, exactly, had happened. And then we all got to participate in fixing the system so that it would never. happen. again.
I hope — and from the sound of it I think there's a good chance — that the people in charge of your lab are taking a similar tack. But they can only do that if you're honest, and you encourage everyone else to be honest. And the best way to do that is to keep the atmosphere nonjudgmental. What's done is done, nobody got hurt, and now it's just time to move on and (hopefully) figure out how to prevent a similar thing from happening in the future.
So if you're asked, I would mention your suspicions, but not in any way that suggests that you're trying to deflect or place blame. If you're not explicitly asked, I think your responsibility extends only to suggesting those changes that you think would prevent something like what happened from reoccurring (e.g. doing the measurements on a different balance, one that would make a 10g error more obvious?). If that implies or communicates your theory about what happens, so be it. But I wouldn't point fingers.
posted by Kadin2048 at 11:05 PM on April 25, 2010 [4 favorites]
I'm sure you know where I'm going with this. One day, a guy I knew left a valve open and didn't check the pressure before turning on the diffusion pump. The oil caught on fire. Thousands of dollars of sensors were ruined by the smoke, and the whole system had to be taken apart and cleaned.
What I remember most about all this, and why I bring the story up at all, is because the professor who ran the lab dealt with it like our own private Challenger incident. There wasn't any placing blame. There was an investigation, but it was just to determine what, exactly, had happened. And then we all got to participate in fixing the system so that it would never. happen. again.
I hope — and from the sound of it I think there's a good chance — that the people in charge of your lab are taking a similar tack. But they can only do that if you're honest, and you encourage everyone else to be honest. And the best way to do that is to keep the atmosphere nonjudgmental. What's done is done, nobody got hurt, and now it's just time to move on and (hopefully) figure out how to prevent a similar thing from happening in the future.
So if you're asked, I would mention your suspicions, but not in any way that suggests that you're trying to deflect or place blame. If you're not explicitly asked, I think your responsibility extends only to suggesting those changes that you think would prevent something like what happened from reoccurring (e.g. doing the measurements on a different balance, one that would make a 10g error more obvious?). If that implies or communicates your theory about what happens, so be it. But I wouldn't point fingers.
posted by Kadin2048 at 11:05 PM on April 25, 2010 [4 favorites]
I just want to expand on that last point. I have worked about a dozen different labs employing centrifuges. It is absolutely standard to use beam balances for balancing centrifuge tubes together. I'd guessing S was using a JA20 rotor for the J25 here, which I'm very familiar with and have used up to the rpm you mentioned.
There's only one balancing procedure I'd ever used for this: balance opposite tubes on a beam balance. Using a digital scale would be awkward, time consuming and prone to failure. Whoever told S to use this procedure is the negligent person here.
It should not be possible for an inexperienced student to put herself in serious danger by missing a single digit on a digital readout.
posted by 7-7 at 11:08 PM on April 25, 2010 [1 favorite]
There's only one balancing procedure I'd ever used for this: balance opposite tubes on a beam balance. Using a digital scale would be awkward, time consuming and prone to failure. Whoever told S to use this procedure is the negligent person here.
It should not be possible for an inexperienced student to put herself in serious danger by missing a single digit on a digital readout.
posted by 7-7 at 11:08 PM on April 25, 2010 [1 favorite]
Aren't serious centrifuges supposed to have all kinds of interlocks that should have kicked in? One of my coworkers knows a guy who works at TRYING to crash centrifuges.
Also, How much did the full and empty tubes weigh? A ten-fold difference might not be obvious on the balance, but why would a ten gram difference not be obvious?
posted by Kid Charlemagne at 12:05 AM on April 26, 2010
Also, How much did the full and empty tubes weigh? A ten-fold difference might not be obvious on the balance, but why would a ten gram difference not be obvious?
posted by Kid Charlemagne at 12:05 AM on April 26, 2010
One more thing. Considering 0.5L is much less than the JA14 is designed for, what was the pressing reason that you couldn't use the much less massive (and easier to move around, and much less dangerous to screw up) JA20 rotor? It'll do 0.4 liters, so you could either just spin your 0.5 liters down in 2 batches or slightly reduce the amount of sample.
posted by 7-7 at 1:15 AM on April 26, 2010
posted by 7-7 at 1:15 AM on April 26, 2010
Best answer: I've seen a similar centrifuge failure where the run was aborted at a similar speed. It wasn't due to balancing, but the lid came off and the singed and broken bottle caps and altered volumes attested to how much heat is produced by emergency braking (and the rotor dragging against the walls of the centrifuge.) Depending on how things were damaged in your case, evaporation certainly could have altered volumes after things started going haywire. (Plus, from your description, I suspect S divided the solution into 4 125mL volumes in 200mL tubes, and 10mL in that volume is reasonably visible. S would have had to have not only misread the mass but also not noticed different liquid heights in the bottles.)
In the end, labwork is not without risk. Things go wrong. Bottles break, chemicals spill, radioactive areas need to be decontaminated, centrifuges end up unbalanced, machines worth tens of thousands of dollars are damaged. Scientists are human, and disasters can happen due to bad luck, momentary lapses of attention, and other fuckups that aren't really the result of "gross negligence." In the centrifuge incident I mentioned above - and a similar incident I recently heard about involving an ultracentrifuge (!) - the labs did not treat the incidents as something for which grad students/techs/postdocs (all of whom were very experienced) needed to be punished. Rather, they were taken as opportunities to go over the procedures people used to prep samples and run the machines, the training requirements, and so on. That's part of what we have to do when we design protocols, after all: build in what safeguards we can to minimize the risk we're exposed to as well as potential sources of error, because betting that you'll do everything perfectly every time is dumb.
S already has the stand-by-the-centrifuge bit of risk minimization down, and you're obviously going to get the centrifuge fixed, so you might want to focus on minimizing weighing error. Actual balances are great, and it would not be a bad idea to get one, but there are ways to minimize risk with digital scales too. For example, I re-zero the scale to the mass of one bottle and and adjust volumes in the other bottles as necessary to make their masses identical. Writing down the masses of each bottle and doing calculations is a great way to introduce error, but simply making sure the readout says 0.01g (or whatever's within the error tolerance) is harder to screw up. Honestly, other than being in lab alone, it sounds like S did precisely what anyone in this situation should do. They reacted quickly, stopped the machine, contacted all their supervisors, and did their best to help clean up the mess. Most lab safety people desperately wish all researchers were that attentive and responsible.
Is it common for these machines not to have some sort of mechanism to detect when they're unbalanced, or a mode of failure that doesn't involve the destruction of the machine and possible death of its operator?
and
Aren't serious centrifuges supposed to have all kinds of interlocks that should have kicked in?
Most big centrifuges (especially, of course, ultras) do have some sort of mechanism to detect imbalances and other problems. Depending on what causes the breakdown (and when it occurs in the spin), this mechanism - and the emergency braking mechanisms - may or may not be sufficient to brake the rotor safely and without any damage to anything. Bottles breaking, mechanical issues with the centrifuge, structural issues with the rotor, user error with regards to lid, speed, balance, etc. can all cause different stresses.
posted by ubersturm at 1:28 AM on April 26, 2010
In the end, labwork is not without risk. Things go wrong. Bottles break, chemicals spill, radioactive areas need to be decontaminated, centrifuges end up unbalanced, machines worth tens of thousands of dollars are damaged. Scientists are human, and disasters can happen due to bad luck, momentary lapses of attention, and other fuckups that aren't really the result of "gross negligence." In the centrifuge incident I mentioned above - and a similar incident I recently heard about involving an ultracentrifuge (!) - the labs did not treat the incidents as something for which grad students/techs/postdocs (all of whom were very experienced) needed to be punished. Rather, they were taken as opportunities to go over the procedures people used to prep samples and run the machines, the training requirements, and so on. That's part of what we have to do when we design protocols, after all: build in what safeguards we can to minimize the risk we're exposed to as well as potential sources of error, because betting that you'll do everything perfectly every time is dumb.
S already has the stand-by-the-centrifuge bit of risk minimization down, and you're obviously going to get the centrifuge fixed, so you might want to focus on minimizing weighing error. Actual balances are great, and it would not be a bad idea to get one, but there are ways to minimize risk with digital scales too. For example, I re-zero the scale to the mass of one bottle and and adjust volumes in the other bottles as necessary to make their masses identical. Writing down the masses of each bottle and doing calculations is a great way to introduce error, but simply making sure the readout says 0.01g (or whatever's within the error tolerance) is harder to screw up. Honestly, other than being in lab alone, it sounds like S did precisely what anyone in this situation should do. They reacted quickly, stopped the machine, contacted all their supervisors, and did their best to help clean up the mess. Most lab safety people desperately wish all researchers were that attentive and responsible.
Is it common for these machines not to have some sort of mechanism to detect when they're unbalanced, or a mode of failure that doesn't involve the destruction of the machine and possible death of its operator?
and
Aren't serious centrifuges supposed to have all kinds of interlocks that should have kicked in?
Most big centrifuges (especially, of course, ultras) do have some sort of mechanism to detect imbalances and other problems. Depending on what causes the breakdown (and when it occurs in the spin), this mechanism - and the emergency braking mechanisms - may or may not be sufficient to brake the rotor safely and without any damage to anything. Bottles breaking, mechanical issues with the centrifuge, structural issues with the rotor, user error with regards to lid, speed, balance, etc. can all cause different stresses.
posted by ubersturm at 1:28 AM on April 26, 2010
ubersturm:
Yeah, it's also possible the rotor just wasn't locked to the spindle but was just resting on top of the pins. That may actually be more likely: I wouldn't expect an unbalanced rotor to come right off of the spindle, rather just bend it and inflict some scars on the wall. In that case, the problem would be due to a lack of experience with the large and heavy rotor that was being used - it can be difficult to tell when they are locked in place.
posted by 7-7 at 1:43 AM on April 26, 2010
Yeah, it's also possible the rotor just wasn't locked to the spindle but was just resting on top of the pins. That may actually be more likely: I wouldn't expect an unbalanced rotor to come right off of the spindle, rather just bend it and inflict some scars on the wall. In that case, the problem would be due to a lack of experience with the large and heavy rotor that was being used - it can be difficult to tell when they are locked in place.
posted by 7-7 at 1:43 AM on April 26, 2010
how does the centrifuge turn on/off? is the switch ON it? maybe a big kill switch at the end of the bench (big red button, kills power to the whole bench) would be a prudent investment, esp. if you're gonna have undergrads working alone at night...an electrician could set you up with that pretty cheaply...
posted by sexyrobot at 2:07 AM on April 26, 2010
posted by sexyrobot at 2:07 AM on April 26, 2010
I run several undergrad labs. For me there are really only two important pieces of data here:
1) "working on the weekend with no one else in the lab"
2) "called me immediately"
Point (1) should never, ever happen -- nobody, anywhere in the chain of command, should be working in the lab by themselves at any time if there's something that dangerous present. On a purely functional level, you need a second person to call the ambulance when the first person is unconscious on the floor.
And in dealing with these situations, point (2) is all you can ask for. There's a certain amount of realpolitik involved here. From the standpoint of the lab, the project, and the equipment, you need to be able to get good information about (a) what happened, (b) how the failure came about, and (c) how to improve procedures so it doesn't happen again.
You're not going to get any of that if you punish students who tell you the truth. If it turns out the student really was at fault, you can deal with it by improving their training, or supervision, or something else. If you try to make students (for example) financially on the hook for expensive equipment all you're doing is incentivizing lying, and in the long run the lab will suffer for it.
posted by range at 5:39 AM on April 26, 2010 [6 favorites]
1) "working on the weekend with no one else in the lab"
2) "called me immediately"
Point (1) should never, ever happen -- nobody, anywhere in the chain of command, should be working in the lab by themselves at any time if there's something that dangerous present. On a purely functional level, you need a second person to call the ambulance when the first person is unconscious on the floor.
And in dealing with these situations, point (2) is all you can ask for. There's a certain amount of realpolitik involved here. From the standpoint of the lab, the project, and the equipment, you need to be able to get good information about (a) what happened, (b) how the failure came about, and (c) how to improve procedures so it doesn't happen again.
You're not going to get any of that if you punish students who tell you the truth. If it turns out the student really was at fault, you can deal with it by improving their training, or supervision, or something else. If you try to make students (for example) financially on the hook for expensive equipment all you're doing is incentivizing lying, and in the long run the lab will suffer for it.
posted by range at 5:39 AM on April 26, 2010 [6 favorites]
Considering that one of potential consequences of equipment malfunction is death; it seems should be a minimum of two operators. Also, the manufacturer might provide some good feedback and would probably be very interested in knowing about this accident.
posted by rmmcclay at 7:20 AM on April 26, 2010
posted by rmmcclay at 7:20 AM on April 26, 2010
I've worked in both university labs and for industrial startups and I'd like to reiterate the need for supervision in the lab, but not limited to just students.
A few random things I think are worthy of mention:
1) The most important thing here is preventing this in the future. Punishing S (if S did something wrong) should be COMPLETELY out of the question. Put it in writing, get your superiors to sign it, and hand it to S as a promise that they won't suffer any negative consequences as a result of this. There's absolutely no way to get to the bottom of this if you're putting the person with the most knowledge of the events leading up to the near-accident in the hot seat. The best thing organizations can do to ensure safety of their employees/students is to make protecting each other a team effort. One person did not screw up. YOU ALL SCREWED UP.
2) Nobody should touch any dangerous equipment when there aren't at least two people in the room. Period. Accidents will happen, but things always turn out worse when there isn't someone there to call 911 or kill the machine. I had to go to a funeral for a coworker who was repairing a multi-ton rotary table alone, forgot to disconnect the power, tripped a micro switch and had the machine start slowly crushing him. Open casket. Even after all the work the mortician did, his head looked like humpty dumpty. A friend knew a janitor that caught his arm in an automated lathe while sweeping around the machine late at night. He was stuck flipping over and over the machine for a couple hours. He lost the leg that was slamming into the ground on each rotation. I almost lost a hand while absentmindedly working on 40kg chuck spinning at 6K rpm...something I wouldn't have done if someone else had been in the room. Having someone there to chastise you when you do dumb stuff is a Very Important Thing, as is fostering a culture where people call each other out on stupid shit.
3) Before a potentially dangerous machine powers up, someone else should take a look at it. Have a safety checklist and log. This is a huge hassle, but it saves lives.
4) Anything that rotates deserves a LOT more respect than people give it. Hammer this into students' heads. A centrifuge is every bit as dangerous as dynamite if something goes wrong. Anything that can be exposed to hands while turning can grab you and twist your arm off or worse.
posted by paanta at 7:43 AM on April 26, 2010 [8 favorites]
A few random things I think are worthy of mention:
1) The most important thing here is preventing this in the future. Punishing S (if S did something wrong) should be COMPLETELY out of the question. Put it in writing, get your superiors to sign it, and hand it to S as a promise that they won't suffer any negative consequences as a result of this. There's absolutely no way to get to the bottom of this if you're putting the person with the most knowledge of the events leading up to the near-accident in the hot seat. The best thing organizations can do to ensure safety of their employees/students is to make protecting each other a team effort. One person did not screw up. YOU ALL SCREWED UP.
2) Nobody should touch any dangerous equipment when there aren't at least two people in the room. Period. Accidents will happen, but things always turn out worse when there isn't someone there to call 911 or kill the machine. I had to go to a funeral for a coworker who was repairing a multi-ton rotary table alone, forgot to disconnect the power, tripped a micro switch and had the machine start slowly crushing him. Open casket. Even after all the work the mortician did, his head looked like humpty dumpty. A friend knew a janitor that caught his arm in an automated lathe while sweeping around the machine late at night. He was stuck flipping over and over the machine for a couple hours. He lost the leg that was slamming into the ground on each rotation. I almost lost a hand while absentmindedly working on 40kg chuck spinning at 6K rpm...something I wouldn't have done if someone else had been in the room. Having someone there to chastise you when you do dumb stuff is a Very Important Thing, as is fostering a culture where people call each other out on stupid shit.
3) Before a potentially dangerous machine powers up, someone else should take a look at it. Have a safety checklist and log. This is a huge hassle, but it saves lives.
4) Anything that rotates deserves a LOT more respect than people give it. Hammer this into students' heads. A centrifuge is every bit as dangerous as dynamite if something goes wrong. Anything that can be exposed to hands while turning can grab you and twist your arm off or worse.
posted by paanta at 7:43 AM on April 26, 2010 [8 favorites]
Why is it in-policy for a relatively inexperienced undergrad to work alone on equipment that could kill them if something is ten grams out of balance?
The larger issue everyone at your lab should be addressing is this- S accidently did you a favor; you have a chance to fix this without anyone having to get killed.
S made a simple error, took steps to mitigate his screw-up, notified the relevant people in a timely fashion and took responsibility. There's no need to cover for him or condem him. Lay out what happened, what may have caused it, and leave out your junior sluething with the paper towels. The school should be a lot more interested in how they narrowly avoided a multi-million dollar lawsuit than nailing some kid.
posted by spaltavian at 8:41 AM on April 26, 2010
The larger issue everyone at your lab should be addressing is this- S accidently did you a favor; you have a chance to fix this without anyone having to get killed.
S made a simple error, took steps to mitigate his screw-up, notified the relevant people in a timely fashion and took responsibility. There's no need to cover for him or condem him. Lay out what happened, what may have caused it, and leave out your junior sluething with the paper towels. The school should be a lot more interested in how they narrowly avoided a multi-million dollar lawsuit than nailing some kid.
posted by spaltavian at 8:41 AM on April 26, 2010
I agree with Spalt here. It looks like the accident has been handled as best as it could be in the time directly following the events. The real issue here is the safety policies your lab has in place. In all the labs I've worked in we never left techs alone with dangerous equipment. You should always have a buddy watching your back if there is an element of risk involved. You should all be OSHA, etc trained and should be wearing and have access to appropriate safety equipment. If machines require special skill sets to 'disarm' them in emergencies it should be paramount that their users be provided with that knowledge.
I'm not faulting your supervision. I've been a tech and a supervisor, so I know that the policies aren't yours to control. I will however warn you that people above you are now in a position where finding a scape goat might be easier than addressing the real issue if it means accepting blame and jeopardizing their own standing. I've seen it happen more than once in a lab setting. Document in writing as much as you can, including interactions with your superiors.
posted by Gainesvillain at 9:46 AM on April 26, 2010
I'm not faulting your supervision. I've been a tech and a supervisor, so I know that the policies aren't yours to control. I will however warn you that people above you are now in a position where finding a scape goat might be easier than addressing the real issue if it means accepting blame and jeopardizing their own standing. I've seen it happen more than once in a lab setting. Document in writing as much as you can, including interactions with your superiors.
posted by Gainesvillain at 9:46 AM on April 26, 2010
This thread is closed to new comments.
posted by i_am_a_fiesta at 8:26 PM on April 25, 2010 [2 favorites]