An arc that occurred while a worker adjusted the set points on a recently installed over and under-current relay caused a fire that burned down part of a line-up of single-pole DC circuit breakers and most of the control-circuit wiring in a 1,500VDC electrified railroad catenary tie-breaker substation. Photo 1 above shows the breaker line-up in the substation control room. The relay was triggered from a current-sensing shunt located in the feeder between the feeder circuit breaker and the corresponding section of overhead conductor (Photo 2). A critical high-voltage fuse was omitted in the design of the protection system between the current sensing shunt and the current relay at the location. As a result, the circuit-breaker cubicles were damaged by arcing (Photo 3).
Understanding the overall design of the railroad power system is helpful in determining what went wrong. Power in this particular substation was supplied to each of the four tracks via 1,500VDC overhead conductors. The system used rectifier substations and tie-breaker substations. Each section of the catenary was fed at its center from a 1,500VDC rectifier substation. The ends of the four catenaries were electrically connected to each other by means of eight feeder circuit breakers located in the substation. A fault in a catenary section should cause the normally closed feeder circuit breaker to open and the normally open bus tie breaker to trip, thus isolating the fault. At the time of the accident, the railroad was in the midst of completing the installation of a transfer trip system that would ensure that the breakers tripped to isolate the faulted section.
The single-pole DC circuit breakers were designed to be tripped by an electrically actuated solenoid, which released a mechanical latch. The circuit breaker would open by the force of a pre-set spring. The relay served to confirm that its breaker tripped in response to a control command. If the relay sensed that the feeder and the breaker were still carrying current after the breaker should have tripped, the control system would trip all of the breakers to ensure that no current could be carried into the substation from the adjacent substations via the catenaries. Since the shunt was located in the 1,500V feeder circuit, the interior of the relay — namely the springs and the armature — was at a 1,500VDC potential, too.
The fault, which occurred while the worker was adjusting the relay and caused an arc to ground that engulfed the tie-station control-wiring system and the circuit breaker cubicles, wouldn't have been possible had the transfer trip been operating — it had been installed, but it wasn't placed in service. The arcing was interrupted when the circuit breakers and the tie-breakers in adjacent substations were manually tripped from a central control facility. The damage to the circuit breakers in the tie-breaker station indicated that the arcing persisted for several minutes. If the fuses were located in the leads from the shunt to the relay, they would have limited the arc and fault current, and the damages would have been confined to the relay and its wiring.
In the arbitration that followed between the railroad and the consulting engineering firm that designed the modifications for the transfer tripping and other features, the question arose whether the fuses were omitted intentionally in the design because the relay was operating in a backup function. The question of the fuses was that they would not have been in an operating circuit. They were located in the back-up circuits to check on the circuit breakers. It would have only been possible to know if the fuses had blown at some time by routine checking or by failing to back up the circuit breaker.
Although no personnel were injured in the accident, including the worker conducting the adjustments, the interior of the tie-breaker substation had to be completely rebuilt. The arbitrator ruled for the railroad that the fuses should have been included in the design. In keeping with that ruling, high-voltage fuses were placed in the circuit between the shunt and the relay when the substation was rebuilt. In addition, a warning was placed on each relay that instructed workers to not change the relay settings when they were energized — in this case, when the shunt was carrying current.
The railroad continued to function without the tie circuit breakers between the catenary sections. They were still supplied with power by the rectifier substation.
Kusko is vice president of Exponent Failure Analysis Associates, Natick, Mass.
Sidebar: What Would You Do?
For nearly six years EC&M has used Forensic Casebook to inform its readers of cases of electrical mishaps and offered analysis explaining what happened and why. Now it's your turn to take the facts we give you and tell us what you would do. Michael J. Foley, president of Denver-based Technical Consultants Group, Ltd., presents the following case from a forensic engineer's point of view and gives you the opportunity to walk in his shoes.
An attorney contacts you with the news that one of his clients, a large insurance company, has incurred a loss at one of its insured properties — a shopping mall located six hours from your office. He goes on to explain that an “electrical failure” occurred at the mall's main electrical distribution equipment four hours ago, when an electrician, contracted by one of the retail stores undergoing remodeling, was performing work on the energized main electrical switchboards. According to the lawyer, “something happened” and the equipment “blew up.” The outage has closed about 25% of the retail stores.
An electrical contractor has been hired to restore power and to perform emergency temporary repairs. You explain to the mall manager that it's very important to your investigation, on behalf of the insurance company, that the condition of the damaged equipment not be altered before you arrive. He says he will do his best, but that he must have electrical power restored to the stores as soon as possible, no later than the next morning.
What would you do if you were the contractor hired by the mall to perform the emergency repairs? If you disturb any of the damaged equipment, will you be later accused of “evidence spoliation”? Will you be sued?
You arrive at the mall to find that the electrical crew has started repairs on the 480V/277V, 3,000A rated switchboard. One of the cubicles has sustained severe damage as a result of an arc fault. Portions of the switchboard have been partially disassembled (Photo). The electrical repair crew is working quickly to isolate the damaged switchboard cubicle and make temporary repairs to restore power. It's now 9:30 p.m. and the mall manager tells you that the mall must be operational by 7 a.m. the next morning.
You learn from the electrical repair crew that the arc fault set the injured electrician's clothes on fire while he was working in front of the energized cubicle. He suffered burns to his hand, face, and shoulders. The mall manager tells you that the work performed by the injured electrician was related to terminating recently installed cables for a new feed to a recently remodeled retail store.
What do you do? Based on your past experience with the attorney representing the insurance carrier for the mall, coupled with the fact that there was a personal injury, you anticipate that there could be litigation. Now you consider certain “standards of care” that must be followed. Do you allow the repairs to proceed at the current pace? Do you stop all repairs until you have documented and investigated the incident? What else needs to be done? Do you interview the injured electrician? Were there witnesses? What background information should you be gathering?
Oftentimes the desires of the building owner to get his system operating again conflict with the needs of the forensic engineer to investigate the scene of the incident. Do you know what you can do to secure the scene of the accident to protect your investigation but enable the ongoing repair of the system? How would you handle the situation? Do you have the authority to stop the repairs until you've completed the investigation? EC&M would like to know what you would do in such a situation. Send in your responses to firstname.lastname@example.org, and the best submissions will be included along with the outcome of the case in a future installment of Forensic Casebook.