Why scary lab accidents happen

lab accident

If a chemist has never been in a lab accident, they have been lucky. Of course, luck is more likely to come to those whose mentors have learned from bad experiences and to those who have taken preventive measures seriously. Chemical reactions create products with behaviours that differ from those of the ingredients. That’s what makes them intriguing, and it’s also what makes them potentially dangerous. No matter how simple and controllable a reaction seems on paper, when it’s carried out in real life the exact conditions determine its rate. And when gases or acids acquire too much kinetic energy, no one wants eyes, lungs and flesh in their way.

As an adolescent I played with my chemicals more than my instructors did. Rarely did they carry out demonstrations while lecturing. Seldom did they deviate from the tight parameters of cookbook labs. So I unconsciously associated accidents with amateurs or with large scale industrial processes. But following freshman year of chemistry, I got my first summer job in the lab after a metallurgical company did not rehire a chemistry student previously involved in a serious analytical lab accident.

The scene of a lab explosion that severely injured a graduate student at Texas Tech University in Lubbock, Texas on January 7, 2010, http://www.csb.gov/csb-to-investigate-laboratory-explosion-at-texas-tech-university-chemistry-department/

One of the most ineffective ways of teaching something is to present it as a rule without explaining why: “Add concentrated acid to water; never water to acid.” But using a small quantity to show how things can go wrong will get the message across. Just do it on the side of a two-way fume hood while making sure that the glass is lowered on the side of observing students. As the less dense water remains on top, the exothermic reaction on the surface brings the top of the mixture to a boil, and of course acidic solution splashes all over the place. The acid is quite strong because the reaction occurs so quickly that little dilution occurs.

A university student working as a replacement for vacationing technicians had been pressured into quickly preparing a standard solution. To save time, so he thought, he measured concentrated sulfuric acid and poured it into a glass jug. He added water to it, screwed the cap on and shook it vigorously from side to side, across his chest. The heat, steam and closed container combined to increase the pressure enough to rupture the glass, and most of the acid ended up on his body. He was in such pain that coworkers had a hard time dragging him to the emergency shower. The result was scarring from third degree burns. What happened was consistent with a keen observation by one of the authors (Jacques Vilain) of a book1 about reducing chemical mishaps:

Each accident is a unique, often bizarre, event, albeit with repetitive patterns but great ‘variability’ in terms of cause/consequences ratio.

About 15 years ago, I wanted to show my students how soft sodium metal is, and of course I planned to subsequently demonstrate its classic, fiery reaction with water. Luckily I was wise enough to use the two-way fume hood setup. I cut the block, exposing its lustrous interior — the surface slowly oxidizes with time even when it’s kept under oil. But I had made the mistake of not taking tongs with me. I had actually used my pocket knife to pull it out of its bottle, and after cutting it, I poked my knife into one of the halves and placed the knife and sodium into a beaker of water.

Any time an exothermic reaction is carried out, the vessel should be checked for cracks. I only noticed the dark line in the glass after I dropped in the knife and sodium. At the time I was also unaware of the precautionary trick of placing sand at the bottom of a beaker. It helps prevent overheated Pyrex from cracking. My last error was my worst one: sticking a knife into too large a piece of sodium is a bad idea because it causes the chunk to sink into water. Without the added weight, sodium, with a density of 0.97 g/cm3, would not have sunk and not all of its surface would have come into contact with water; the reaction would have been slower and more controlled. The hydrogen would have ignited, but the reaction would not have generated as much pressure as when sodium is submerged.

To complete the disaster, our technician had recently left methanol in an open beaker in the fume hood. There was an immense explosion heard across our small school. The observing students gaped in astonishment but were unharmed since the protective pane on their side could not be opened. The fume hood glass on my side did not break, but all the glassware within it did. I had never totally lowered my window, and sodium hydroxide produced by the reaction along with bits of glass projected towards my belly. A couple of pieces actually got stuck in my shirt. When I unbuttoned it, I saw no blood but found NaOH reacting with my belly.

Although I washed it off, by the time I got home that evening I had developed a second rash. Coincidentally, during that week I had experienced an allergic reaction. (It proved to be the new detergent my wife had bought.) The next evening I told the dermatologist to ignore the NaOH rash and try to figure what was up with the rest of the redness. He looked at me as if I was from a different planet. But the reason I was in one piece and that my students were home safe was that in spite of all the errors committed, I had had the good sense of conducting a risky experiment in the safer environment of a fume hood.

About 10 to 50 times more lab accidents occur in schools and universities than in industry.2 If there’s more than one way to do an experiment, and one of those ways will result in disaster, then somebody will come along and try it that way.

This variant of Murphy’s Law materializes more frequently in academia due to inexperience on students’ part and the fact that both PhD candidates and faculty often work alone in the lab. The long hours they put in create overfamiliarity with the surroundings and lead to a false sense of security. But in industry, lab workers are very rarely allowed to solo. Dupont Chemicals and other companies send even minor accident reports directly to the chairman of the board. In the less rigid atmosphere of academia there is a greater chance that a lab accident will go unreported. Many lab accidents from all sources, however, do show up in blogs and other online sites.

At the other extreme, regulations can get overdone and safety officers can over-police. The resulting atmosphere can get stifling and hinder creativity. I remember walking into a pharmaceutical research lab a few years ago. They were shutting it down and donating most of their equipment. Nothing was plugged in. All the chemicals had already been removed, but I still wasn’t allowed to walk in without a lab coat and safety glasses. You can imagine what that safety officer was like when the lab was up and running!

What probably works best is a physical environment designed to minimize accidents. I recently visited McGill University’s new main graduate chemistry lab and was impressed. Every experiment was taking place under a long lineup of fume hoods. A spacious walkway separated it from their work stations, which were all behind Plexiglass and facing the fume hoods, allowing them to keep a watchful eye on their experiments while they wrote or researched. Separated from the labs, the chemical stockroom was easily accessible and actually maintained by someone with a sense of order. It was as clean as a pharmacy! The common work and lab areas minimize the probability that somebody will work alone.

Adopt a "safety first" approach. It will increase the probability that you and your students will enjoy your retirement.


References and notes:

  1. “The Nature of Chemical Hazards, their Accident Potential and Consequences” by Jacque Vilain in Methods for Assessing and Reducing Injury from Chemical Accidents. Edited by Philippe Bourdeau and Gareth Green, 1989 SCOPE. Published by John Wiley & Sons Ltd
  2. How dangerous is chemistry?, Nature, June 2006, 560-561  http://wvhschemteachers.wvu.edu/r/download/6644  


Recommended reading