An electrician suffers serious injuries after an arc-flash ignites the panelboard he is servicing.
When an electrician responded to a call at a downtown Washington, D.C., office building located just a few blocks northwest of the White House, he thought it would be just another routine maintenance check. Asked by the building engineer to troubleshoot a problem in the facility’s large 4,000A, 480/277V, multisection electrical panelboard, the electrician had no idea the assignment would take such a tragic turn.
After reaching the upper floor of the then-new facility, he decided to have a look inside the panel by removing the four slotted screws securing the top half of one of the panel sections, which held two circuit breakers. But as he began to remove the first screw, an arc-flash ignited from inside the panel, turning the seemingly typical job into a tragedy. Although the arc-flash lasted only one-tenth of a second, it burned through all three 2-in.- wide bus bars feeding a single breaker in the bottom of this same section (Photo 1 above) and torched a large hole through the face of the lower section’s cover (Photos 2 and 3 below).
Without warning, the heat from this sudden arc created an explosion that penetrated the electrician’s shirt and severely burned his stomach. Fortunately, the electric utility’s 8,000A fuses in the street vault opened after only 0.1 seconds, shutting off all power to the building. After hearing the huge explosion, the building engineer raced to the electrical room and found the injured electrician, who spent the next several months recovering from severe burns in the hospital.
Shortly after the accident, the facility’s insurance company asked my firm to conduct a forensic inspection. They wanted to know what had caused the arc and why the upstream switchgear had not removed power earlier, which would have lessened the electrician’s injuries. A number of people had already looked for the remains of any foreign object and found nothing. There was nothing about the melting of the various metals to suggest an easy answer. Perhaps something had been installed incorrectly. Perhaps not. We proceeded with the investigation.
By photographing the scene, we bought some time to plan the disassembly of one of the circuit breakers at the manufacturer’s facility. Then, drawing from a previous fire investigation, a possible correlation came to us: In a dwelling fire, when a flammable liquid is present at the origin, some of the liquid vaporizes without burning and condenses on cold objects at a distance from the fire. Depending on the liquid’s composition, detectable quantities remain for some time.
Though the temperatures in this incident were much higher and involved metals rather than liquids, the principles seemed to be similar. After presenting our theory to the insurance adjuster, we cut a large bolt from a location about a foot from the arc-flash area. This allowed us to examine the object for condensed atoms—a process similar to inspecting growth rings on a tree.
We didn’t believe using an ordinary microscope was the best way to search for deposited atoms. Instead, we placed the section of bolt head in a vacuum and used the beam of electrons in a scanning electron microscope (SEM) as a probe. This way, we could map even a thin layer of atoms. Plus, we could analyze the energy scattered during this process to identify the types of atoms present, such as aluminum, copper, iron, and oxygen. It turned out we could “see” the plating on the bolt head; it contained cadmium. Then, we worked our way outward. We were thrilled to find that atoms had condensed on the bolt head.
When we looked at the energy scattered from the first atoms, we found peaks for copper, iron, and oxygen as expected, but we also found peaks of molybdenum. Naturally, we were curious about the molybdenum because it is used to increase life and durability in products ranging from electrical contacts to steel tools. (Tungsten is also common in electrical contacts, but our first atoms did not include that element.) In addition, the electrical contacts involved in this incident were not overly heated. That made the common steel used in hand tools the more logical choice—molybdenum (without tungsten) is used there to facilitate the steel’s maximum hardness. Since molybdenum and iron were the first atoms deposited on the bolt head, this meant the initial short circuit across the conductors was caused by the conductive steel in a steel tool. Now, the missing piece to our puzzle fell into place. Based on this evidence, we presumed someone had inadvertently left a tool inside the panel. When the electrician began to remove the panel, he created small vibrations, which made the tool fall on one or two of the phase conductors and caused the short circuit. We now had the remains (or at least a few atoms) of the culprit that started this tragic arc-flash as well as a valuable forensic procedure for determining the sequence of events in arc-flash incidents.
Upon further investigation, we also determined the circuit breaker immediately up stream had a manufacturing defect involving tolerances of its moving parts. Although the breaker was not burned, this problem had effectively jammed the device, preventing it from opening to interrupt the flow of current feeding the arc-flash. Because of this defect, the injured electrician received an out-of-court settlement as compensation for his serious injuries.
Buske is a Principal with Buske Engineering, Benicia, Calif.
