When an experienced electrician began working on the switchgear of a medical facility's main control room, he had no idea this job would be his last. The purpose of the electrical crew's activities on this particular day was rather routine: Upgrade the 4,160V buswork and switchgear from the original 2,000A rating to 3,000A, including the replacement of the main circuit breakers, tie breaker, and existing
When an experienced electrician began working on the switchgear of a medical facility's main control room, he had no idea this job would be his last. The purpose of the electrical crew's activities on this particular day was rather routine: Upgrade the 4,160V buswork and switchgear from the original 2,000A rating to 3,000A, including the replacement of the main circuit breakers, tie breaker, and existing buswork.
The project manager, foreman, Electrician A (the victim), and the rest of the crew (three electricians with less experience than the victim) arrived between 9 p.m. and 10 p.m. to carry out the project with a planned outage scheduled around midnight. The prep work included placing a tarp over electrical panels that would be exposed to weather and removing screws from the back of the cabinets enclosing the buswork to be replaced. The unfortunate sequence of events that followed led to the arc flash that ended the electrician's career that night and ultimately caused his death some weeks later.
Seeing the electrical crew standing idle after performing the initial tasks, the project superintendent instructed the group to proceed with the job, indicating that the system had already been de-energized by the electric utility. As a result, the crew removed the back of the enclosure, looking for a place to ground the system. Upon doing so, however, they found the buswork was taped, leaving no place to clamp a ground to it. Because the bus bars were about to be removed, the team decided that it didn't make sense to ground them — so they went inside the vault (Photo 2 below) instead and accessed the bus tie breaker enclosure. There, they found that the tie breaker had already been racked out (see Figure below). (When installed, the tie breaker provides connectivity between the south bus and north bus.)
At the time of the incident, the north bus was de-energized, and the south bus was energized from a secondary power source. The south bus was connected to the lower three terminals of the tie breaker, and the north bus was connected to the upper three terminals. Acting as “lead man,” Electrician A tested the upper tie breaker terminals for the presence of voltage. Using a pocket tester (rated up to 600V), he sensed the presence of voltage. Because it was possible that voltages from adjacent enclosures could be detected inside a medium-voltage switchgear, the lead electrician discounted the detection indicated and obtained test equipment suitable for use with high-voltage equipment.
He used a voltage detector “squawk box” in proximity to the tie breaker sockets and detected no voltage. Although that particular squawk box is effective on some high-voltage equipment without making contact, it required actual physical contact rather than just proximity at voltages below 2,500V. To resolve the apparent conflict, Electrician A then used a new high-voltage meter (Photo 1) that the company had recently acquired, touching each of the six terminals exposed within the tie breaker enclosure. The upper three terminals indicated 00.0 while the lower three indicated 2.4 (with respect to ground).
One of the other electricians reportedly asked whether or not the indication was showing kilovolts or volts. Electrician A reportedly stated that he believed it to be volts, being non-zero simply due to capacitively induced voltages. According to witness testimony following the event, none of the other three electricians told him to stop — nor warned that he was looking at energized terminals. After clamping one end of the ground cable to the enclosure frame, Electrician A was in the process of clamping the other end to one of the lower terminals of the tie breaker sockets when he drew an arc, which resulted in an arc flash. Immediately engulfed in flames, the victim died several weeks later in the hospital from injuries he sustained during this event.
After the accident, the victim's family sued the meter manufacturer. I was retained by the plaintiff to investigate the cause of the incident and analyze the forensic evidence associated with the case.
First, I compared the design and functionality of the meters in my possession, including multimeters, a power meter, an air speed meter, a temperature sensor, and other diagnostic equipment. I also examined a variety of other high-voltage meters and test devices (from flyers, brochures, etc.) — all of which showed the units of measure on the tester's display. Based on this information, as well as analysis of witness testimony and site conditions, I drew several conclusions.
I concluded that a number of factors had contributed to the incident. First, there was a communication problem coupled with a lack of understanding of the system being worked on. Obviously, if the victim had understood that the lower terminals were connected to the south bus, he would have chosen to ground the upper terminals instead. Based on the sequence of events leading up to the accident, it appears he didn't even realize that there would be terminals in the tie breaker enclosure that would be energized, or he mistakenly believed that he was working in the main circuit breaker cabinet instead of the tie breaker. (An appropriate warning sign indicating that the cabinet was energized from more than one source may have clarified his understanding relative to the switchgear configuration.)
In my mind, the biggest problem was that the high-voltage meter he used did not have any labeling or other marking on the display indicating that the displayed units were in kilovolts, although a small sticker on top of the unit (not visible during use) did state that the display was in fact showing kilovolts. In my opinion, if the meter in question had included a clear indication of the units associated with the displayed value, the victim could have easily seen and understood that the terminal to which he was about to connect the ground cable was indeed energized.
Although lockout/tagout (LOTO) procedures were under way at the time of the incident, they would not have de-energized the terminal Electrician A contacted. In placing a safety ground to avoid the potential for inadvertent energization of the portions of the system being worked on, Electrician A was implementing steps associated with LOTO. Furthermore, by using the three different voltage-sensing devices, he was exercising reasonable prudence to ensure safety in his activities. Therefore, I asserted that Electrician A believed the terminal to be de-energized because he misinterpreted the reading of the high-voltage meter.
Upon further investigation, I also uncovered that the victim had followed training requirements as outlined in NFPA and OSHA standards, and was qualified to perform the work he had been assigned. He had presumed the parts to be energized until he had verified the system had been de-energized, as required by OSHA and NESC. Similarly, he was wearing gloves and goggles, but did not have the arc flash personal protective equipment (PPE) that would have been prescribed for work on energized equipment — because he believed he had already verified it to be de-energized.
In this case, it is unreasonable to argue that the nature of the equipment speaks for itself because someone taking measurements should automatically know what he is measuring. The fact is that there are a number of different measurements an electrician might be making. For example, a high-voltage power system can have measurements of voltage (or kV), current, wattage, reactive power, or electric field intensity. An electrician may also have an interest in knowing temperatures of specific components, rotational speed of electric motors, resistance, inductance, or capacitance. On any given day, an electrician may use several of a wide array of different meters. This is the very reason it's standard practice to indicate units on the display of the meter, thus avoiding any potential ambiguity in the interpretation of results.
Ultimately, I concluded that the victim's reliance on safety protocols and procedures was trumped by a defect in his high-voltage meter, which could have been fixed with a very simple modification to the equipment design. The resultant lawsuit settled for an undisclosed amount before the case went to trial.
Palmer, Ph.D., P.E., C.F.E.I., is manager of electrical engineering and fire investigations at Knott Laboratory, LLC, Centennial, Colo.