In the spring of 2002, an electric utility was upgrading a substation in the southwestern United States. Although the job seemed typical to the electricians hired by the utility from an outside electrical contracting firm to do the work, the substation configuration proved fatal to one of the crew members.
The electricians were tasked with installing cable connections between the new 138/13.8kV transformer and the distribution system control room, which included two banks of metal-clad switchgear. First, they needed to ensure that the “B” bus was de-energized before any work took place. To accomplish this, all of the circuit breakers for one bank of the control room were “racked out” — or physically removed from their operational position. The crew also grounded Bus B at the associated transformer terminals to ensure the equipment remained de-energized.
Reportedly, during the installation, the foreman expressed concern that the clearance between the terminations of the cables to the bus bars and the outside cabinet access door was inadequate and posed a potential safety hazard. With the intent to mitigate that hazard, he applied a heat-shrink insulating tape to the terminations of the transformer cable (Photo 1). Note: Although this approach is not required for all such installations, it is common practice, particularly in installations where there is limited clearance to other metal parts.
When the foreman finished wrapping the terminations of the transformer cable, located in cubicle 204, he performed the same task on the terminations in cubicle 203 and then moved on to Cubicle 202 (Photo 2). While wrapping the terminations in Cubicle 202 he noticed the hair standing up on his arms. He immediately stopped work and again sought and received verbal verification from another crew member that the B side of the control room was still de-energized. As part of the wrapping procedure, the foreman began applying heat to the shrink tape using a propane torch — at which point witnesses observed an explosion. Unfortunately, the foreman died as a result of the injuries he sustained in this accident.
The family of the deceased (plaintiff) sued the manufacturer of the substation equipment (defendant) and hired me to investigate the accident. This investigation led me to analyze the following areas and, in turn, draw certain conclusions.
The distribution side of the subject substation was made up of two banks of metal-clad switchgear housed within a control room — consisting of a metal enclosure with climate control and a number of cubicles. Each cubicle (except those in which connections are made to the large power transformers external to the switchgear) houses a circuit breaker and associated instrumentation. The main bus on the east side of the control room was designated Bus A (Cubicles 101-106), and the main bus on the west side of the control room was designated Bus B (Cubicles 201-205).
The electrical configuration of the switchgear in normal operational condition is shown in Fig. 1. Described as a “breaker and a half” configuration, this design allows each distribution feeder to be served from either bus, depending on the arrangement of open and closed circuit breakers. At the time of the accident, all of the circuit breakers on the west side (Bus B) were “racked out” — or physically removed from their operating position — to ensure that they could not be accidentally closed while personnel worked in the equipment (Fig. 2). Specifically, the circuit breakers that were racked out were designated 13A (in Cubicle 201), 15C (in Cubicle 202), 16A (in Cubicle 203), and 19A (in Cubicle 205). Circuit breaker 14B was closed prior to the incident and tripped as a result. The status of the remaining circuit breakers is not known. Transformer No. 2, which was being installed at the time of the accident, was de-energized. Transformer No. 1 and Bus A were energized at the time of the accident.
Cubicle 202 contained breaker 15C, which was connected to breaker 16A on one side and to cables extending to breaker 14B on the other. While breaker 16A was racked out, leaving half of the Cubicle 202 terminals de-energized, breaker 14B was closed, energizing the terminals associated with the connected cables. Thus, the terminations that the foreman was wrapping at the time of the accident were energized to 13.8kV. De-energization of those terminals would have required the opening of circuit breaker 14B.
Based on the testimony of eyewitnesses to the accident, the electrical crew worked on the equipment with the understanding that portions of the system were de-energized. Obviously, the circumstances of the accident point to the ultimate misunderstanding. Had the crew members fully understood the configuration and layout of the system, as described above, they could have taken measures to de-energize the appropriate circuits before working on them or foregone the work to enhance the safety of the switchgear terminations.
The electric utility's Electrical Safety Manual sets forth specific guidelines for working around high-voltage circuits and equipment. For example, the guidelines require that all circuits and equipment be considered energized at full voltage until de-energized and grounded. Given the nature of the accident, it is evident that the electrical crew did not de-energize and ground the terminations on which they were working. Reportedly, the electrical foreman had sought and obtained verbal verification that the circuit was de-energized, but the portion of the circuit he was working on was not electrically contiguous with Bus B, which apparently was de-energized and grounded. The lack of de-engergization and grounding of the terminations prior to applying the heat-shrink tape to those terminations is apparently the result of a lack of understanding of the electrical configuration of the substation.
According to the utility's incident report, the victim was wearing “all personal protective equipment (PPE).” While the report does not specify, it is assumed his PPE included high-voltage gloves, goggles, and a hard hat at minimum. However, given the extent of his injuries, it is unlikely he was wearing arc flash protection equipment, which is typically worn while working on energized terminations. Based on the testimony of witnesses, there is little doubt the foreman believed he was working on de-energized terminations; therefore, the use of arc flash protection equipment would not have been expected.
The National Electrical Safety Code (NESC) includes a requirement that a safety sign shall be placed in each cubicle containing more than one high-voltage source (180 B 11). The purpose of this requirement is to warn workers, such as electricians, that more than one voltage source in a given cubicle should be de-energized to ensure a safe working environment.
Cubicle 202, in which the electrical foreman was working at the time of the accident, contained two high-voltage sources, namely a connection to Bus B through breaker 16A and a connection to Bus A through breaker 14B. Hence, Cubicle 202 should have had a safety sign indicating the presence of more than one high-voltage source.
Potential design improvements
The accident can ultimately be traced back to confusion about the electrical status of the system the electrician was working on. While posting a safety sign in each cubicle that had more than one voltage source would have enhanced the awareness of potential hazards, an active warning system with a direct indication of the presence of voltage may have proved more effective.
This could be implemented in a number of different ways with relative simplicity and without adding substantially to the cost of the system. Following are several conceptual implementations of active warning systems that could be used alone or in combination to indicate the presence of high voltage in a switchgear cubicle:
Use existing status contacts on the circuit breakers to operate a warning light. Existing codes require the indication of breaker status on the exterior fronts of breaker enclosures. The same signals that are used to indicate breaker status could be used either directly or through signal relays in appropriate series/parallel combinations to activate lamps in the back of the cubicle that would indicate when circuit breakers associated with any potential high-voltage source are closed. An additional indication could show when circuit breakers are open but not racked out.
Use voltage transformers or other voltage-sensing technologies for voltage detection to operate a warning light or audible alarm. With the addition of a voltage transformer on the transfer bus (direct connection between two circuit breakers), direct sensing of the voltage on each of the critical points could be implemented. While this approach is more expensive than the use of status contacts, it would provide direct indication of voltage. In the unlikely event that the circuit breaker should fail to operate properly, direct indication of voltage would still warn of the presence of voltage.
Interlock. As an alternative or a complement to the use of a warning light or audible alarm, the same signals could be used to operate electromechanical latching mechanisms that would prevent the opening of any cubicle with high voltage present.
While additional engineering would be required for the specific implementation of these alternate design concepts, each is technologically feasible.
Although this case was settled out of court, it doesn't change the fact that a man died on the job. Ultimately, the key to preventing an accident of this type is a correct understanding of the configuration of the electrical system. The complexity and connectivity of substations can vary widely from location to location, so an intuitive understanding of substation layout and design is crucial to personal safety. For this reason, it is incumbent upon manufacturers and owners of equipment to effectively communicate the configuration of a substation. It is also imperative that electrical workers fully verify and ground all high-voltage parts prior to the initiation of work — unless appropriate precautions are taken for energized work practices.
Palmer, Ph.D., P.E., C.F.E.I., is manager of electrical engineering and fire investigations at Knott Laboratory, LLC, Centennial, Colo.