On Sep 13, 2012, at 3:13 PM, ACTSNYC**At_Symbol_Here**cs.com wrote:

1.  Transient nature of the workers.  Professors generally stop working in the lab after a few years - they are consumed by the almighty funding quest, teaching etc.  He/she trains the first couple students, those students train the incoming ones, and a giant game of telephone ensues.   The result is that people learn procedures that are not entirely safe or downright dangerous because SOP's are either not established, updated, or consulted.    And, gotta love this, that original professor who started the chain probably learned all his/her skills the same way.   I mentioned a specific example on the list previously - every student in the lab I worked in (including myself) had been taught the incorrect way to syringe t-BuLi and every one of us thought it was correct at the time even though the Aldrich bulletin had another way.

2.  Immortality.   Most of the students, postdocs, and young faculty aren't old enough to seriously appreciate the consequences of a safety screwup in the lab.  Unless you have witnessed a horrific accident, had an underwear-changing near miss, gone to the E-room in the middle of the night sure you were having a heart attack, or had kids who would be orphans if things went wrong, you generally don't tend to tune in that it Really Can Happen To You.

Industry has the same problem, but they solve it with enforceable policies.  Do This and you're fired.  Don't Do This and you're fired etc.  In my academic career I've seen students pushed out of PhD programs for poor academic performance, poor research, plagiarism, personal issues etc., but I have never heard of anyone being forced out or sanctioned for safety issues.  Perhaps that's something that should never happen - safety should be about education/improvement/partnership/support rather than blame, but with the UCLA prosecution that whole paradigm is on the brink of seismic shift.

3. Complacency/familiarity.   The more you work with something, the less dangerous it seems.   You may take every precaution using n-BuLi the first time, but use it enough and you'll start becoming complacent.  After you get some on your hand and it doesn't do anything but feel soapy, or you quench a big reaction without incident, you start to cut corners or not use the required PPE.  Then comes the day when your scaleup runs away thermally and you don't have an ice bucket, or it's humid and it does catch fire in air.   The first couple times you use a dichloromethane wash bottle and get some on your hand, you worry about it, but after a while you're hosing organic gunk off your hand with it.

Industry has the same problem, but in different ways and to a lesser degree.  Changes in scale (Texas Tech, UCLA) occur less often and in many cases expertise is available for the engineering/safety challenges of scaleup.  Procedures are more repetitive, and therefore you can draw boundary lines - cut this corner and you're fired etc.

5  No formal reporting/corrective mechanisms.  When a safety problem comes up in industry, that carries all sorts of direct consequences (workman's comp, down time, lost profits, OSHA worries, whistleblower fears, lawsuits, bad publicity).   Corrective action is generally straightforward - fire an employee, make a new rule, post a sign, put a new guard on a machine, whatever.