July 2010 issue
By John Cavallaro, P.E.,
> Forensic Engineering, Inc.

A boy is electrocuted after a severed ground wire and water buildup in a junction box make for a fatal electrical combination at a lake house’s dock.

What started out as a beautiful day at a lake house quickly turned to tragedy for a grandfather/grandson duo, who decided to go swimming off the dock and were soon joined by some neighbor children. Because he did not own a boat, the grandfather raised the metal rails of the lift (Photo ) about 2 ft out of the water, operating a 1-hp motor connected to a pulley system with metal cables.

Noticing a child swinging from the metal rails of the boat lift, the grandfather immediately told the child to get off. After they began to have fun, another child, standing in the water and not touching anything, said he felt a tingle in his legs. Perplexed, the grandfather cautiously touched the metal rails. Feeling an electrical shock go up his arm, he immediately said, “Get out of the water, now!”

Standing near the lift controls, he looked down and saw a child floating face down in the water. The grandfather immediately brought the child to the deck’s surface and administered CPR while his wife dialed 911. When the emergency medical staff arrived, they couldn’t revive the child, pronouncing him dead at the scene (cause of death was electrocution).

After the tragic accident, the family of the deceased boy, as well as the grandfather (homeowner), wanted answers. My firm was called in to investigate this accident on behalf of the grandfather, who was being sued by the victim’s family (who also sued the electrician). 

My initial investigation uncovered several clues. The dock was fed by an underground feeder (UF) 12-2 rated cable with ground, which was routed from a junction box located in the house’s crawl space. A portion of the UF cable ran along the edge of a deck, exposed to the environment and foot traffic (see relative location of the severed wire under the house to the dock in the Figure ). Fed from a 20A circuit breaker at the main panel, a bare ground wire in the junction box was awkwardly twisted onto a shorter ground wire. I later learned from the electrical inspector that he had found this wire hanging out of the junction box, and it was severed at that point. Upon seeing this, he’d reconnected it to prevent any other ground issues.

The inspector stated the dock was originally built without electricity, and whoever did the wiring never obtained an electrical permit. He also informed me that he had opened a sealed standard plastic junction box under the wood dock and above the water line. He identified that the internal wires served the boat lift motor and were spliced to the house wires with wire nuts. When he removed the cover screws, water poured out of the box. He identified that the ground (green) wire and the phase (black) wire of the 120V UF cable were on the bottom of the box. I noticed a line on the inside of the junction box, suspecting that this was identifying the water level inside the box, with the two wires clearly being below the line.

The house panel’s circuit breaker for this circuit did not have a ground-fault circuit interrupter (GFCI) as per the 1995-1998 NEC, Art. 555, Marina and Boatyards. I noted that the boat lift motor had two GFCI devices connected in series in between the boat lift and the plastic junction box.

If there were two GFCI devices in series, how did the electricity get from the black (hot) wire to the metal rails of the boat lift? The motor casing was metal and was directly attached to a metal plate, which was attached to metal cables supporting the metal rails. Thus, an internal short in the motor’s casing would energize everything back to the metal railings. Or maybe the two wires in the junction box below the water line were the source of the short. Either way, the severed ground wire under the house would not have allowed anything to flow back to the panel’s ground bus. Hence, there would not be a GFCI-sensed voltage imbalance at the lift motor to trip the unit. Both GFCI circuits were on the up side of where the short to ground was located. Effectively, this said the GFCI became unusable as soon as the ground wire became severed.

I had to prove the motor had an internal short to its casing, which would place 120V on the metal cabling and metal boat rails, or that the plastic junction box was filled with enough water to allow a conductive path between the black (hot) wire and the “floating” green (ground) wire. To examine the motor for an internal short, I disconnected the ground wire going from the house to the motor and then energized the motor via the power switch. I made a voltage measurement between the motor’s ground wire and the ground wire going back to the house. I measured 0V, which told me there was no voltage on the motor’s casing. Thus, the only other path would have to be through the water-filled plastic junction box, back to the motor’s ground casing, and down the metal rails. No ground connection back at the house meant the metal rails remained energized until a grounding connection could be made somewhere.

Because the boat dock was made of wood, anyone standing on it would be isolated and wouldn’t receive a shock by touching the metal cables during normal boat lifting operation. When the grandfather lifted the rails out of the water, he disconnected the leakage path, allowing the electricity to remain on the rails. Unfortunately, when the boy hung from the rails with his feet in the water, he provided a direct path for current to flow. However, how the water got into the gasketed, sealed junction box remained a mystery. The wiring feeding the junction box was encased in two plastic conduits, but they were not sealed from the outside humid environment, possibly accumulating condensation over time. The section routed toward the house stopped about halfway along its length. The other section was routed vertically partially up to the GFCI connector. Thus, both ends were open to the elements. I also noticed the plastic junction box was located directly under a seam in the deck boards. If the cover was not sealed completely, then its location would be ideal for water to drip into the box. 

This case settled out of court for an undisclosed amount. We can learn several lessons from this accident. First, the electrician should have obtained an electrical permit. The inspection would have uncovered the missing GFCI at the house, and the connection under the house would have been installed to NEC regulations, preventing the severed wire. Eventually, water buildup in the plastic junction box would have tripped the GFCI. Then, the subsequent troubleshooting would have uncovered the problem and led to a fix. The quirky piece of this puzzle is that individually the severed ground wire and water buildup would not have caused this fatality. Together, however, a dangerous electrical condition was lurking under the surface.


To read the full story and view additional graphics, visit  http://ecmweb.com/design/case-floating-dock.