“Insights from a Laboratory Fire”, Mitchell P. Jones, Kathrin Weiland, Claudia Mitterer, Philip Verdross, Robert T. Woodward, Alexander Bismarck2023-07-05 (Internet archiving, tech)⁠:

Fires are relatively common yet under-reported occurrences in chemical laboratories, but their consequences can be devastating. Here we describe our first-hand experience of a savage laboratory fire, highlighting the detrimental effects that it had on the research group and the lessons learned.

…We were unfortunate enough to experience the complete destruction of one of our chemistry laboratories by a fire ignited upon the failure of a laptop’s lithium-ion battery. Thankfully, no one was hurt, but it led to ~€1.4-million [$1.57m] worth of equipment and infrastructure losses, hundreds of hours of additional labor, at least 10 months of research delays, the loss of entire projects, and immeasurable stress and pressure on many of our students and staff…A forensic fire and arson investigator identified the ignition source as a battery inside a laptop, which was stored switched off and unplugged in a cabinet underneath a wooden laboratory bench (Box 1). Only two of the original 6 18650 cells in the laptop battery pack were found in the ash and the laptop itself was completely thermally converted (Figure 1c). The laptop was stored in the laboratory because it contained software necessary to operate an analysis instrument but had not been used for at least 3 years… Even lithium-ion batteries in a 0% state of charge can release considerable heat when ignited. Two charged Lenovo laptop battery packs each containing 6 cells exhibit an overall energy release of 3,470 kJ¹⁰, equivalent to the complete combustion of ~100 ml of petrol¹¹.

…The fire had started in the early hours in one of our basement laboratories (K05, see Figure 1a). The laboratory was fitted with smoke detectors and fire alarms, which went off at 01:42 but were neither noted by the central security service nor configured to be automatically forwarded to the Viennese fire brigade. The fire brigade was alerted at 02:00 by security personnel who smelled smoke during their routine patrol, and although they could not locate the fire, reported it nonetheless. The fire brigade arrived at 02:07. After locating the fire, firefighters broke ceiling-level basement windows from outside the building and extinguished the fire, flooding the laboratory with foam (Figure 1b) to a height of ~1.2 m.

…The damage and losses associated with the event were substantial, but that was just the start: we also had to face the complexity of the investigation, documentation and recovery. 3 independent fire investigations by the police, federal government and the insurance company were conducted in the fortnight following the fire. During this time, the laboratories had to remain untouched. In fact, equipment in the smoke-damaged laboratories started to corrode. Delays occurred as the university was understandably unwilling to allocate funding until the cause of the fire was determined and coverage by insurance guaranteed…The stress was aggravated by the thoughts of interviews and investigations that would follow with the fire brigade, police, insurance agents and university representatives, but much of this stress was in fact unwarranted. The insurers did pay out and the investigations were not confrontational or accusative, but were instead professional, compassionate and supportive.

…Time that should have been invested in science was now spent rebuilding. We took an inventory of what we thought was salvageable and recovered equipment from the burnt-out laboratory, packaging and shipping some of it to manufacturers for assessment: for 5⁄8 pieces, this was ultimately a waste of time and money because they were scrapped due to severe fire, heat, smoke and water damage. We got 45 quotes for replacement equipment and 38 for servicing and recalibration. For the major items that we bought, we had to negotiate prices, go through all the admin of ordering, and find space and resources for storage prior to the reopening of the laboratories. We were also faced with an often overlooked, exacerbating factor in the replacement of scientific instruments: what to do when attempting to service or replace equipment produced by companies no longer in business or that had changed ownership multiple times.

…Access to the smoke-damaged laboratories was restricted for 2 months while everything, including the ventilation system, was cleaned and restored by a disaster recovery and restoration company. When K01/02 reopened, we could still not access our instruments for an additional 2 months while they were all professionally cleaned, serviced and recalibrated. These closures considerably hindered the research and practical teaching of our entire team, totaling ~30 people. Fortunately, water damaged K04 only required drying and was functional again within a few weeks… It was made more challenging by the social distancing measures enforced during COVID-19.

…Whilst a triumph for the group spirit, the reopening of the laboratories only really represented a return to business as usual. Damage to student and staff progress could not be reversed. 4 group members lost samples relevant to ongoing investigations in the fire. Two master students lost their thesis work (one in its entirety and the other almost completely save a few samples). Such a blow severely demoralized the students, heavily delaying the completion of their degrees and the start of their subsequent careers. Affected students described feelings of helplessness and exhibited slower progress in their research compared to prior to the fire.

Less obvious costs incurred during the restoration of the laboratories included the human resource costs associated with putting the laboratory back together—as a conservative estimate this represented around 1,000 working hours distributed across 11 staff, amounting to an additional €50,000 [$56,240] in labor not covered by insurance.

[A good example of swiss cheese model of disaster and the burdens of complexity & regulation & the danger of safety measures: keeping the laptop undoubtedly seemed ‘prudent’ as a countermeasure to the danger of losing software, but backfired; the complexity of multiple integrated fire alarm systems inevitably resulted in failures and typically fail-deadly errors; the (doubtless well-intentioned) regulations leading to 3 redundant investigations wasted massive resources & damaged a lot of equipment itself; other safety measures like COVID-19 quarantine/controls impeded recovery, etc.]