Ecmweb 16395 Pressure Washer Electrocution Pr
Ecmweb 16395 Pressure Washer Electrocution Pr
Ecmweb 16395 Pressure Washer Electrocution Pr
Ecmweb 16395 Pressure Washer Electrocution Pr
Ecmweb 16395 Pressure Washer Electrocution Pr

The Case of the Pressure Washer Electrocution

Sept. 25, 2017
Investigation reveals what caused one member of a cleaning contractor’s crew to receive a fatal electric shock

When a small independent cleaning contractor team arrived at a national fast-food chain restaurant to perform regularly scheduled procedures, no one could foresee the day would end in tragedy. Unfortunately, the routine job took an unexpected and deadly turn.

Shortly after starting work, one worker entered the exterior rear courtyard to find his partner (who had been pressure washing) immobilized on the wet pavement — still holding the pressure washer wand in hand. That worker received an electric shock when he touched his fallen colleague, at which point a third worker used a wooden broom handle to disconnect an extension cord, which was plugged into an exterior rear wall receptacle at the restaurant. This allowed the fallen worker to be touched without further shock effects. However, first responders were not able to revive the victim.

Setting the stage

The contractor, which reportedly had approximately 30 years of experience working at restaurants of this type, was contracted to perform specific cleaning procedures at the subject restaurant as well as other facilities for that same chain in the metropolitan area. One of the team’s assigned duties was regular cleaning of the kitchen vent hood to remove grease build-up, as is typically required by local ordinances, mechanical codes, and fire prevention/protection codes. The electrical service setup at this restaurant included a utility pad-mount transformer that delivered 208Y/120V through a 600A-rated, single-section switchboard.

At the time of the incident, the victim was within an enclosed roofless service courtyard (Fig. 1) outside the rear of the restaurant, where he was using a pressure washer to clean stainless steel grease baffles that had been removed from the kitchen exhaust hood system. The pressure washer wand was hose supplied from an engine-powered pressure washer mounted on the contractor’s utility trailer located outside the service courtyard. It was reported that the pressure washer engine alternator had failed; therefore, an automotive-type battery charger was being used to maintain engine battery voltage during operation. This charger was powered through an extension cord connected to the receptacle on the restaurant rear wall exterior, which was located in the service courtyard.

The aftermath

Following the incident, remaining contractor personnel coiled the extension cord and pressure washer hose with wand, placing them back on the utility trailer. Reportedly, the full trailer load of equipment was then stored in a shed until it could be initially inspected visually by the author. According to restaurant management, they requested inspection of the rear exterior wall receptacle by a licensed electrician soon after the accident. It was also reported that an electrician made some repairs and/or modifications to that receptacle, and retained components that were removed for possible later inspection.

Contractor’s equipment examination

The author, who was retained by an attorney representing the family of the deceased, was initially asked to perform a visual examination of the contractor’s trailer-mounted equipment. Based on the provided statements, special attention was paid to the pressure washer, battery charger connected to the pressure washer gasoline-fueled engine battery, and extension cord.

The trailer was also equipped with a diesel-fired water heater used with the pressure washer. Most of this equipment was in a well-used condition, but no visual evidence of an electrical anomaly or condition that might contribute to a shock hazard was found with one notable exception. The approximate 79-ft-long, 3-wire, hard-duty-type extension cord included a non-original replacement-type 3-prong plug. In addition, the outer jacket had been damaged and/or breached in at least 13 separate locations (with electrical tape repairs found in four of those locations).

After this inspection, the author advised the client attorney it was possible that the damaged cord (Photo 1) was a factor in the electrocution, based on its visible deterioration as well as the fact that the responding workers had received a shock sensation that ended when they disconnected the cord. This was important because the deceased’s employer is typically protected from a lawsuit due to workers compensation laws; therefore, a defect on their part (such as failure to maintain the cord) could prevent any litigation from being successful. The attorney decided to authorize further evaluation based on his knowledge that the restaurant had contracted with an electrician to make repairs after the incident, indicating that a defect at the facility may also have been a factor in the electrocution.

At this point, the restaurant was invited to retain an expert to participate in further examination of the contractor’s equipment. After that expert had a chance to visually examine the trailer and its contents, he and the author agreed on a protocol for more intrusive examination/testing, including internal examination, resistance testing, insulation integrity testing, and functional testing of the battery charger.

All resistance testing was conducted with a calibrated true-rms digital multimeter, and all insulation integrity testing was conducted with a calibrated analog meg-ohmmeter at 500VDC. The charger tests indicated proper expected conditions, including insulation resistance values exceeding 100 kΩ. The battery charger was then connected to a 120VAC energized GFCI receptacle and operated without tripping.

The visibly damaged and altered extension cord was also subjected to similar electrical testing, which indicated proper end-to-end continuity and insulation resistance values exceeding 70 kΩ. The cord plug was then connected to the calibrated GFCI receptacle, and its load end was connected to the charger power cord. The charger again operated without tripping. These cord tests were then repeated after the entire cord length was immersed and soaked in a low-resistance water and electrolyte solution with similar results, including when the cord was agitated within that solution during those tests.

After this testing, the electrical tape was removed from the cord, and the entire length was subjected to detailed examination, especially in the areas where the outer jacket had been compromised. In the areas where the inner wires were exposed, no evidence of exposure of the 16 AWG stranded copper conductors was found. The non-original plug attachment to the cord supply end was also dissected and found to be in proper condition, including all screw terminal connections.

The author and other expert also examined the pressure washer hose, reel, and wand. The hose actually consisted of two separate hose lengths connected with an intermediate coupling. Portions of the hose closest to the reel and the pressure washer revealed abrasion damage to the outer rubber jacket, which had exposed some of the imbedded steel mesh braid. Electrical continuity/resistance testing of the fully connected hose, reel, and wand indicated that the reel end of the hose was electrically insulated from the wand, and insulation resistance testing resulted in values exceeding 500 kΩ.

Restaurant examination

Inspection of the incident location by the author also was conducted cooperatively with the expert representing the restaurant. The single-story building included a rear open-topped enclosed service courtyard that included a walk-in type refrigeration unit, refuse containers, a cooking oil supply system for the kitchen, and a mop sink/drain-pan. Along the restaurant rear wall inside this courtyard were the electrical service entry, including the utility meter, two closets containing the electrical main switchboard, and a gas-fired, electronically controlled water heater.

Examination of the switchboard exterior evidenced heavy corrosion to the steel enclosure bottom front. The inspection conditions did not permit de-energization or internal examination of the switchboard, and the grounding features for that equipment were not visible due to the constraints of the surrounding closet. The water heater exterior visually appeared to be in good condition. Both the switchboard and the water heater enclosures were tested for continuity/resistance and differential AC voltage with respect to each other and ground embedded metal features within the courtyard. Measured values were less than 1 Ω and less than 1VAC.

Located on the rear wall near the entry door was a GFCI-type duplex receptacle with weatherproof cover that appeared to be new, all of which was horizontally mounted on a well weathered cast metal fixture box surface mounted on the brick masonry wall. During the inspection, personnel representing the restaurant confirmed that an electrician added this GFCI receptacle and cover after the incident date.

The author noted a rectangular hole was cut in the masonry behind the fixture box with unsealed gaps between the box edges and the hole. The fixture box had two conduits connected to the threaded ports at its top and right sides, which were both routed along the masonry wall exterior (Photo 2). The right side connection was to a rigid metal conduit (RMC) that was routed to a disconnect switch mounted on the restaurant rear wall, with no further connections to any appliance or load. The top fixture box connection was to electrical metallic tubing (EMT) through a compression-type threaded fitting. This conduit was routed approximately 48 lineal feet along the restaurant rear and courtyard side walls to a weatherproof box located switch, which was then EMT connected to a cooking oil supply system for the kitchen.

Located just inside the building rear wall (in close proximity to the exterior mounted receptacle) were two 225A main-lug-only panelboards supplied from the main service switchboard. The legend within panelboard “B” indicated that the single-pole 20A circuit breaker in branch position #3 supplied that exterior receptacle and the cooking oil supply system. Subsequent voltage and continuity testing, including tests with the circuit breaker handle switched off, indicated that this legend entry was accurate. Voltage testing indicated that approximately 121VAC was supplied to that branch circuit with the circuit breaker ON, and there was no measurable voltage applied with the circuit breaker switched OFF. Testing also indicated there was less than 0.5 Ω resistance between the panelboard “B” bonding bus and the metal enclosure for the main service switchboard.

During the site inspection (with power switched off to the subject branch circuit), the current receptacle installation was inspected during systematic disassembly. During this process, it was noted that even though the new GFCI receptacle grounding yoke had been screw attached to the fixture box, multimeter testing indicated that there was an open-circuit condition (infinite resistance) between the yoke and the box. Further testing indicated that there was also an open-circuit condition (infinite resistance) between the fixture box (including the conduits connected to that box) and the bonding bus in panelboard “B” — and also to the enclosure of the main service switchboard. Multimeter resistance testing indicated a value of less than 3 Ω over the length of the EMT and switch box between the receptacle fixture box and the metal enclosure for the cooking oil supply system.

The author observed three conductor wiring (including a green insulated ground) between the receptacle and the panelboard, two conductor wiring (no ground conductor) through the EMT between the receptacle and the cooking oil supply systems switch, and no conductors routed through the RMC to the disconnect switch. Also noted was that some of the wiring insulation and insulation on crimp-type connectors within the receptacle fixture box evidenced exterior burn or scorch damage. There was also a heavy buildup of contaminants and scattered oxidation within the fixture box.

The site inspection concluded with removal of the GFCI receptacle, the wiring from that receptacle to the panelboard, and the surface-mounted fixture box for further laboratory examination. After removal of the surface-mounted box, it was observed that it had been screw connected to a yoke typical for a communications fixture that was attached to a flush-mounted metallic fixture box within the masonry wall (Photo 3). This flush-mounted box demonstrated severe deterioration, exposed sharp metal edges, and the end of an electrician’s “fish tape” protruding from the conduit routed back to Panelboard “B”. In addition, no conduit had been thread connected between the surface mounted box rear and the conduit connected to the rear of the old flush mounted box, which extended from Panelboard “B.”

During the site inspection, the courtyard area pavement was drenched with a hose to simulate water behavior during pressure washing. The courtyard pavement had a grated storm drain in the approximate center, but, due to unseen clogging of that drain, the water backed-up and pooled in the courtyard area. It was noted that this rising water level did make contact with the metal casing for the cooking oil supply system.

Laboratory examination

In addition to the items removed from the site during the cooperative inspection, the subject receptacle was also presented, along with its faceplate and some bundled insulated wiring. The 15A “Spec Grade” duplex receptacle exhibited heavy contamination and deterioration (Photo 4), including especially heavy corrosion to the hot leg side metallic hardware. The upper hot leg side screw terminal, which included a connected open lug with crimped connection to a cut stranded copper wire stub, exhibited burn and contamination damage to portions of the crimp insulation. The weatherproof faceplate was fully missing its hinged cover. The bundled wiring included three white insulated segments, with two joined with a twist-on type connector (totaling about 60 in. in length). There was also an approximate 75-in.-long red insulated segment. Both the red and white segments had “3” labeled near one end, and at the opposite end of the red segment a localized severe burn had destroyed about 1½ in. of insulation and a portion of the stranded copper conductors leaving melted ends (Photo 5).

Examination of the surface-mounted fixture box was performed both before and after extended ultrasonic cleaning. The empty threaded rear center conduit connection port exhibited severe metal loss over about one-quarter of its circumference with indications of metal melting (Fig. 2). The EMT transition fitting (threaded into the port in what had been the installed top of the box) showed localized melting and corrosion to the portion of the threads exposed within the box that were facing out toward the box front opening.

A reconstruction of the probable installed conditions inside that box with the original (non GFCI) receptacle and the damaged red insulated wiring removed by the electrician following the incident was performed. The receptacle was re-installed within the box with its hot leg side facing toward the box installed top position, based on markings found on the hardware. This simulation indicated that the burn damaged portion of the red conductor was likely in contact or close proximity with the melt damaged areas identified at the box top and rear ports. At that time, both the author and the expert representing the restaurant agreed that this simulation was probably the most accurate representation possible of the installed conditions present at the time of the incident.

Engineering evaluation

Examination and electrical testing performed on the contractor-supplied equipment indicated no evidence that it contributed to the electrocution. This was despite the observation that the pressure washer engine was being operated off of a 120VAC battery charger due to alternator failure. Even more significant, it was found that even though the extension cord used to power that charger evidenced visible deterioration — and was running across the wet pavement in the area where pressure washing was being performed — there was no indication that current flow would have occurred between the cord and its surroundings at a level known to cause serious electrical injury.

The cord minimum resistance to current flow at or above the system voltage was measured as at least 70 kΩ, which means the maximum ground fault current flow at 120VAC would be 1.8mA. Well-established references (such as the Electrical Safety Handbook) indicate that this is about the lowest current level at which someone might feel a shock sensation. Actual significant physical effects occur at much higher levels, including muscle paralysis/contraction (10mA), interference with or stoppage of breathing (30mA), and heart defibrillation (75mA). That maximum current level calculated for the damaged cord is also below the typical 5mA trip setting for a GFCI receptacle or other device designed to prevent ground fault injury.

Examination of the fixed wiring features that were present at the time of the incident revealed several indications that current flow had occurred at least intermittently between a hot leg conductor within the restaurant rear exterior wall receptacle fixture box and that metallic box and connected metallic conduit. That conduit was ultimately connected to the metallic casing for the kitchen cooking oil supply system, which was resting on the courtyard pavement. It was demonstrated that due to a clogged storm drain in the courtyard during pressure washing, standing water likely came in contact with that oil system casing. In addition, the victim, who was also probably standing in water, was also handling stainless steel panels resting in water and using a metallic washing wand that was streaming water onto those panels.

The conditions found by inspection of the original receptacle, connected wiring, and the fixture box where it had been located indicated that significant deterioration had occurred, including corrosion of receptacle conducting features and loss of insulation from the wiring. In addition, this same wiring had been routed without proper protection through the masonry wall and an old fixture box located behind the current fixture box, which was a probable factor in the wiring insulation damage. The fixture box in use at the time of the incident also had inadequate weather protection due to the missing hinged front cover and the box rear opening, which was open to weather entering through the unsealed hole in the masonry wall.

Of greatest importance, the fixture box and connected conduit were not effectively bonded back to the grounding system for this building. This meant that there was no return path to enable circuit overcurrent protection actuation for the ground fault conditions that the evidence indicated were occurring at least intermittently within the fixture box. Initially, it was thought that the cord must be the shock source because the reported disconnection of the contractor’s extension cord ended the shock sensed by the responding workers. However, after examining the receptacle conditions, it is probable that the receptacle disturbance caused by removing the plug (especially using a broom handle) probably shifted the internal wiring enough to break the intermittent current path to the metal conduits. Information provided by the restaurant indicated that the fixture box involved in this incident was probably installed as part of a renovation, and the older fixture box behind it within the masonry wall was most likely installed as part of the original construction.

Codes and standards

Based on the provided information, the National Electrical Code, 1993 Edition (NEC/NFPA 70) was in effect at the time the electrical branch circuit involved in this incident was installed or significantly modified. Section 250-51 requires establishment of effective grounding paths from electrical equipment and enclosures to safely conduct fault currents, limit currents to ground on surfaces, and facilitate operation of protective devices. Sections 250-70, 250-74, and 250-75 also require proper bonding of receptacles and their metal enclosures, and effective bonding continuity for metal raceways including conduits. These requirements clearly were not met by the subject receptacle installation at the time of the incident, and were probable contributors to the fatal injury.

Interestingly, use of a GFCI receptacle wasn’t required by the NEC at the probable time the involved receptacle was installed. This requirement didn’t  appear until the 2005 Edition, in Sec. 210.8(B)(4), including reference to Sec. 210.63, where it was required for servicing the refrigeration unit within the service courtyard. But since the evidence indicated that the ground fault involved the hot leg wire between the panelboard and the receptacle, a GFCI receptacle would not have detected the ground fault, nor would it have been able to shut off power to the fault location.

The American National Standard Recommended Practice for Electrical Equipment Maintenance, 2006 Edition (NFPA 70B) provides extensive guidance to commercial building owners regarding establishment of inspection and preventive maintenance programs to minimize electrical equipment degradation and improved personal safety. This guide recommended monthly inspection of electrical receptacles for damage and deterioration. The facility was clearly not following these practices, since the obvious loss of the weatherproof cover on the subject receptacle had not been corrected. This edition was in effect at the time of the subject incident.

The American National Standard for Electrical Safety in the Workplace, 2004 Edition (NFPA 70E) provides requirements to protect employees in workplaces. Section 250.4 specifies that electrical raceway and enclosures bonding and grounding shall be maintained to ensure electrical continuity. Section 205.11 requires that electrical conductors be maintained free of damage and shorts that would present a hazard to employees. The subject branch circuit clearly was in violation of these provisions due to combination of installation defects and uncorrected deterioration. This edition was in effect at the time of the subject incident.

Therefore, failure to install and maintain the subject exterior branch circuit in accordance with the applicable codes and standards was considered a root cause for the subject electrocution.

Lessons learned

This forensic casebook example provides some valuable electrical safety lessons. First, the cleaning contractor was clearly not maintaining its electrical equipment with due regard for the safety of its employees. In particular, the damaged extension cord was close to becoming a potential shock hazard, although testing and evaluation indicated that its damage was (as yet) insufficient to result in an actual shock occurrence. At an opportune time —  and with permission of the client attorney — the author informally advised the owner of the firm that all flexible cords should be frequently inspected and discarded when any appreciable jacket damage was visible, especially if the internal wires were exposed. It was further advised that they obtain and use a portable GFCI unit (in accordance with the manufacturer’s instructions) for connecting cords to site power sources — whether or not they believe the site source is GFCI protected. This is especially important when working in a wet environment, which is always present during pressure washing.

The evaluation illustrates the need to both install and maintain building electrical features in accordance with the applicable codes and standards. This is especially true for branch circuits that involve direct human interaction, such as convenience receptacles in commercial facilities likely to be used by employees on a regular basis. Outdoor receptacles require even more care, because of the deteriorating effects of weather and the higher likelihood of other potential work hazards, such as widespread wet conditions.

For an entity such as a national restaurant chain, an electrical safety program in accordance with NFPA standards that ensures quality installation work (especially for repairs or remodeling), regular condition inspections and testing, prompt correction of degraded or unsafe conditions, and safety training requirements for employees and contractors is crucial for reducing liability and losses. It also helps to minimize the potential for unfortunate events such as the subject fatality, which are tragic for all involved and almost always preventable.    

A registered professional engineer in 10 southeastern states, Shiver is the president of the consulting firm Chris Shiver, PE in Roswell, Ga. Visit www.chrisshiverpe.com for contact and other information.

About the Author

Christopher Shiver | President

Voice your opinion!

To join the conversation, and become an exclusive member of EC&M, create an account today!

Sponsored Recommendations

Electrical Conduit Comparison Chart

CHAMPION FIBERGLASS electrical conduit is a lightweight, durable option that provides lasting savings when compared to other materials. Compare electrical conduit types including...

Don't Let Burn-Through Threaten Another Data Center or Utility Project

Get the No Burn-Through Elbow eGuide to learn many reasons why Champion Fiberglass elbows will enhance your data center and utility projects today.

Considerations for Direct Burial Conduit

Installation type plays a key role in the type of conduit selected for electrical systems in industrial construction projects. Above ground, below ground, direct buried, encased...

How to Calculate Labor Costs

Most important to accurately estimating labor costs is knowing the approximate hours required for project completion. Learn how to calculate electrical labor cost.