Official investigation reports are among the most valuable training material a rescue team has. The four cases below are summarised strictly from the published reports of the investigating bodies; the conclusions are theirs, and the full documents are linked at the end of each section.
Nitrogen in a reactor: Delaware City, USA, 2005
What happened. On 5 November 2005, during a maintenance turnaround at an oil refinery in Delaware City, Delaware, two contractor employees died of nitrogen asphyxiation in a hydrocracker reactor. The reactor had been loaded with new catalyst and was being purged with nitrogen to protect it; the gas vented slowly through the open top manway. A crew was preparing to install a pipe elbow on the manway when a roll of tape fell inside onto a tray about 1.5 m (five feet) below the opening. One worker tried to retrieve it with a wire hook and, by the investigators' reconstruction, ended up inside the reactor. His foreman pushed a ladder into the manway and climbed in after him; within a few breaths he collapsed as well. The plant's emergency responders reached the platform in less than two minutes, but an oxygen meter lowered through the manway showed a concentration near zero, and recovering the two men through the 24-inch (about 60 cm) opening took time. Both had been without adequate oxygen for nearly ten minutes and could not be revived.
Causes according to the report. The US Chemical Safety Board (CSB) found that the work permit marked the nitrogen purge as "N/A" although the reactor was still being purged, and that the required joint visit to the job site before the permit was approved had not taken place. Contrary to site procedures, no nitrogen purge warning sign or barricade was in place. Training had not covered the possibility of an oxygen-deficient atmosphere outside a confined space, near its opening — a risk the CSB found was not adequately addressed by industry guidance either.
What it shows for rescuers. Both victims were experienced and knew that nitrogen is an asphyxiant. The report notes that, had the foreman applied his emergency response training, he would have raised the alarm by radio and stayed on the platform until qualified responders arrived.
Lessons drawn in the report:
- an oxygen-deficient atmosphere can kill in a few breaths, without warning;
- such an atmosphere may also exist outside the opening of a vessel being purged;
- the instinct to help a colleague frequently produces multiple victims; only trained personnel with proper breathing equipment should attempt a rescue;
- warnings and barriers must stay in place for as long as an access opening is unsecured or a purge is venting.
Report: CSB case study, Valero Delaware City refinery (2006)
Toxic release in a process building: La Porte, USA, 2014
What happened. On 15 November 2014, about 24,000 pounds (roughly 11 tonnes) of methyl mercaptan, a highly toxic and flammable gas, were released inside an insecticide manufacturing building at a chemical plant in La Porte, Texas. Three operators and a shift supervisor died from a combination of asphyxia and acute inhalation exposure. Two of the victims had gone into the building to help. When an operator called for help on the radio, two colleagues ran from the control room and a third followed them; none of them knew of the release, and none wore respiratory protection. One of them died. Shortly afterwards, another operator took three five-minute escape respirators and entered the building, despite warnings from colleagues in the control room. He put an escape respirator on a colleague, who recovered and got out, but died himself while trying to help his brother, before he could connect his own mask to a self-contained breathing apparatus.
Causes according to the report. The CSB traced the release to a flawed engineering design and inadequate safeguards: liquid collected in waste gas vent piping and was routinely drained inside the building. Contributing factors included building ventilation that had not been designed for toxic hazards (and whose roof fans were not working), toxic gas detector alarms set well above safe exposure limits, and no automatic lights or sirens in the building to warn people in the field.
What it shows for rescuers. The emergency response itself ran into difficulties. A call for "rescue" was understood by the site emergency response team as a technical rope or confined-space rescue, so members first arrived with harnesses and ropes rather than breathing apparatus. The vehicle carrying the breathing apparatus and radios would not start, and air monitoring was delayed in the early hours, partly because the equipment was controlled by unit personnel and the shift supervisor who was to support the team was among the victims.
Lessons drawn in the report:
- workers must know the hazards of every toxic chemical on site;
- automatic alarms warn people faster than verbal relaying from a control room;
- emergency calls need clear protocols, including for situations where the nature of the emergency is not yet known;
- emergency vehicles and air monitoring equipment must be maintained, tested and accessible to the team;
- responders need building layouts, site familiarisation and drills inside the actual structures.
Report: CSB final report, DuPont La Porte toxic chemical release (2019)
Rescuers caught by a second collapse: Crandall Canyon, USA, 2007
What happened. On 6 August 2007 a massive coal burst during pillar recovery at the Crandall Canyon underground coal mine in Emery County, Utah, trapped six miners. Pillars failed over about 800 m (half a mile), and the section was also flooded with lethally oxygen-deficient air from adjacent sealed areas. Rescue workers began clearing debris from an entry with a continuous miner to reach the trapped men. On 16 August, while they were installing ground support behind the machine, a further coal burst threw coal, roof supports and fencing at them. Dust filled the entry, ventilation controls were damaged, and air containing about 16 % oxygen moved over the injured. Two mine employees and a federal mine inspector were killed and six other rescue workers were injured. None of the seven boreholes drilled from the surface located the trapped miners; their working area was found to have oxygen levels that would not sustain life.
Causes according to the report. The Mine Safety and Health Administration (MSHA) concluded that the 6 August collapse was not a natural earthquake but resulted from a mine design that could not control coal bursts, from faulty pillar recovery methods and from a failure to revise mining plans after earlier bursts. For 16 August it added causes specific to the rescue: insufficient information about underground conditions, an excavation method that reduced confinement of the failed pillars, and ground support that could not withstand the possible burst intensity.
What it shows for rescuers. The report states that, unlike for the original accident, viable alternatives did not exist for most causes of the second one: the rescue imposed greater risks than would be accepted in normal mining, and the prospect of saving lives warranted the rescuers' efforts. It underlines that the main protection for rescuers lies in preventing the first catastrophe.
Lessons drawn in the report:
- all available resources, including surface drilling, should be used to obtain information on underground conditions as quickly as possible;
- some rescue risks cannot be engineered away, which makes prevention by the operator all the more important.
Report: MSHA report of investigation, Crandall Canyon Mine (2008)
No entry without reliable gas data: Pike River, New Zealand, 2010
What happened. On 19 November 2010 an explosion occurred in the Pike River underground coal mine on the west coast of New Zealand's South Island. Twenty-nine men died; two men in the access tunnel escaped. Mines rescue crews from New Zealand and Australia were mobilised, but conditions never allowed them to enter. A second explosion occurred five days later, and two more followed before the mine was sealed.
Causes according to the report. The Royal Commission found that the immediate cause of the first explosion was the ignition of a large volume of methane. It described the mine's health and safety, ventilation and methane drainage systems as inadequate and identified several possible sources of the methane and of its ignition.
What it shows for rescuers. The commission found that the mine had no system for sampling its atmosphere after an explosion, so for the first days the risks of entry could not be assessed. It rejected criticism that rescuers should have used a "window of opportunity" to go in, stating that there is no predictable period in which a gassy coal mine can be safely entered before a further explosion, and that international practice is to re-enter only on the basis of representative and reliable atmospheric information. It also found that many decisions were made far from the mine by non-experts.
Lessons drawn in the report:
- emergency management plans must be comprehensive, prepared with workers and the mines rescue service, and audited and tested regularly;
- the incident controller should be on site, with the experts, and have relevant mining expertise;
- emergency facilities such as refuges, changeover stations and exits must be specified in the plan and actually provide what they promise;
- re-entry decisions depend on atmospheric monitoring that keeps working after an explosion.
Report: Royal Commission on the Pike River Coal Mine Tragedy, final report (2012)
Common threads
- Would-be rescuers. The urge to help a colleague is strong and immediate. The answer is to raise the alarm, stay outside and let a properly equipped team enter. Escape sets are for escaping, not for entering.
- Isolation. Inert gas and toxic product reached people because sources were not isolated or vents were not controlled. A purge, a drain or an open manway is part of the hazard zone.
- Atmospheric testing. In each case the decisive information was the composition of the atmosphere: near zero oxygen at a manway, a toxic gas in a building, lethal air in a mine. Measurements must be available before entry, continuously and after an explosion. See Gases and Equipment.
- Entry control and emergency planning. Clear permits, barriers, alarm protocols, working equipment and rehearsed plans decide whether the first minutes of an emergency go well. See Safety and Confined space entry.
