Injury by Electric Fault: The Domino Effect


Project Info

Two employees of an electrical contractor were injured while renovating the top floor of a multi- floor office building in the center of a business district in an urban Virginia location. The contractor was in the process of updating the electrical power distribution system on the top floor of the building. In the process, the contractors were injured by an electric fault and arcing event that resulted in a fire that triggered the sprinkler system. RTI was retained to determine if the alleged defects of the equipment used in the electric power distribution system would have caused the injuries and property damage.

During the final phase of the electrical upgrade, an employee of the electrical contractor was tasked to install a switch on the building’s main bus duct that extended from the basement to the top floor, and connect the outlet to a transformer and a distribution panel. In order to access the bus duct, the bus duct manufacturer had constructed a small door-like panel that required a specific-sized screwdriver to unlock. Once the panel was open, the installer could mount and connect the switch to the bus bars, then run the electric feed to a transformer to service the floor. However, the contractor did not have the correct size screwdriver. Instead of getting the right tool, the contractor opted to try to force the door open by using the screwdriver that he was carrying, which was too large to fit the panel’s lock. Knowing the bus bars were “hot”, the contractor, who was not wearing any personal protective equipment, used the screwdriver to pry the panel door open. When he inserted the screwdriver into the panel side and began to pry the door open, the edge of the screwdriver came into contact with one of the bus bars, creating a phase-to-ground fault. The arc developed into a phase-to-phase fault in a 480 volt, three phase system. The arc was so powerful that within a few milliseconds, the arc had burned a three inch diameter hole through all four bus bars and the duct, explosively spewing molten aluminum out the sides and front of the duct. The molten aluminum and other flaming by-products ignited the hair and clothing of an employee who was at the side of the bus duct working on the distribution panel, the hair and clothing of the employee standing in front of the panel who had initiated the fault, and ignited other combustible materials located in the utility room in which they were working.

The fire eventually triggered the sprinkler system and the fire alarm. The two individuals were rescued by co-workers and survived the ordeal. However, since the fire was initiated by an electrical fault, the commander of the first responders correctly ordered the sprinkler system untouched until the electrical power to the building was terminated. After several hours, the electricity and water were finally turned off. The delay in turning off the top floor sprinkler resulted in water damage to every floor in the building. The businesses affected on the lower floors were out of business for nearly two weeks as crews worked to clean up the mess. Several of the tenants were law firms, whose file materials were damaged such that a great deal of effort and expense had to go into recreating the files. Fortunately, only two people were hurt and were able to recover from their injuries. However, the property damage and business interruption costs were significant.

 

The electrical contractor employee who had opted to use the wrong tool for the job and started the whole incident sued for personal injury claims. He had third degree burns to about five percent, mostly to his upper body, and had been out of work since the incident. The plaintiff’s basic complaint was that the building manager knew that work was taking place on the floor, but refused to turn off the electrical power, forcing the plaintiff to work on the system “hot”. Plaintiff also alleged that the bus duct system was defective in design and that a particular relay, used to sense over-current and shut off the main power in the bus ducts, failed to function as required and allowed the arcing event to transpire.

The relay in question is a power relay designed to terminate power in the bus duct within two-tenths of a second in the event that the amperage in the bus bars exceeded a certain level. This over-current protection device is designed to protect the mechanical equipment on the roof from damage and is not a Ground Fault Interrupt designed to protect people from shock or electrocution. During the incident investigation, experts for the plaintiff had tested the subject relay and found that it would function, but after six-tenths of a second has passed instead of within the two-tenths as specified. With this finding, plaintiff’s experts provided the opinion that had the subject relay functioned as intended, the plaintiff would not have been injured by the arc flash, and a fire would not have started. Additionally, the plaintiff alleged a defect in the design of the access panel on the bus duct, stating that had the panel utilized a different design in the locking mechanism, the plaintiff would not have been forced to pry the panel open, and the panel ‘s design should have somehow prevented the contact that triggered the phase-to-ground fault.

To address these allegations, RTI was retained to determine the time scale of the event, and provide an opinion regarding the effect of the relay’s response. Specifically, RTI was asked to find that if the relay did respond as specified, to what extent the injuries to the plaintiff from arc flash would have been mitigated, and to what extent would the building fire have been prevented. The plaintiff, through his experts, alleged that had the relay functioned, the plaintiff would have been exposed to arc flash, but his injuries would have been minimal. In addition, plaintiff’s experts alleged that had the relay cut power within two-tenths of a second, the combustibles within the utility room would not have been ignited and no fire would have occurred. RTI was also asked to review the design of the access panel and its locking mechanism and provide an opinion as to the adequacy of its design.

To address the question of time scale, RTI carefully measured the dimensions of the existing bus bar material and the bust duct, and determined the mass of missing material. By use of a custom heat transfer computer simulation for the bus bar material, and performing calculations regarding the heat transfer of molten droplets traveling through air, RTI was able to determine the time needed to generate the loss of bus bar material from an arcing event and determine the capacity for a fire to occur within the utility room. The key finding was that the available energy at the arc was sufficient to have caused the entire arc event to have transpired within one percent of the time that a properly functioning power relay would have operated. Had the power relay functioned as it was supposed to, there would have been no difference in the outcome. The personal injury to the plaintiff from arc flash would have been the same; the combustible materials would have been ignited and a fire started.

One interesting aspect of the investigation, particularly with the arc flash exposure, was that the expert for the plaintiff had utilized a computer program to calculate the effects to a human from exposure to an arc. The software and the basis were never disclosed by the plaintiff’s expert. RTI performed an analysis using calculations based on the IEEE standard as well as the NFPA standard for arc flash exposure and found very different results for exposure to the arc flash than was found by plaintiff’s expert. After some research, RTI found the software tool used by plaintiff’s expert.  Investigating the plaintiff’s expert’s software tool, RTI also discovered that the expert not only used calculation software that was inappropriate for the analysis, but had used it incorrectly. The software had been developed for internal use by a power utility, but had never been subject to peer review, and had been rejected for use by the American Society of Test and Measurement (ASTM). Additionally, the software itself had, in its own instructions, a provision for modifying the calculated results for exposure to an arc occurring either within an enclosure, or in the open. Plaintiff’s experts had neglected to modify their results appropriately. As a result of RTI’s findings, the arc flash calculations performed by the plaintiff’s experts, and their opinions based on those calculations, were excluded.

The effort used by RTI to verify the computer simulation of the arc-affected bus bar material was a significant factor in ensuring the opinions could survive a Motion in Limine. Simply creating a computer program to simulate the action of the bus bar material was not thought to be sufficient. Proof that the computer simulation was accurate was found by using the program to simulate a known condition. To validate the accuracy of the simulation, the code was tasked to solve several textbook transient heat transfer problems. The output results were compared against the known analytical solutions, finding the computer simulation results to be highly accurate. Re-tasking the computer simulation to address the specific problem resulted in reliable information about the transient heat energy distribution and movement in the bus bars and the associated time scales. From this basis, RTI was able to conclude the power relay response time would have been immaterial in the subject incident. Because of the rigorous testing of the simulation against accepted and published events, the technique and subsequent information was able to pass the test of reliability and relevance and its use upheld.

 

Categories

  • Electrical Engineering
  • Utilities