Send to a Friend View / Add Comments  
Freshwater Marinas; A Ground Fault Analysis

David Rifkin, Capt. USN, Ret.
Quality Marine Services, LLC
2418 Fallen Tree Dr., West
Jacksonville, FL 32246
James D. Shafer, AMS
Harbor Marine Consultants, Inc.
3062 SE Dominica Terrace
Stuart, FL 34997

Co-Authors - Proprietary Material
October 17, 2002


Research has been ongoing to determine why people are drowning due to electric shock in freshwater marinas across the nation. In addition to those occurrences where people are known to have been shocked, more drownings may indeed be caused by electric shock than are reported because autopsy results are not able to pinpoint this as the cause.

The freshwater environment is of most concern since salt water's high conductivity allows the majority of ground fault currents to pass harmlessly to the earth ground (this has been previously quantified through a significant amount of research in saltwater). In freshwater, where electrical resistance is high, a person's body represents the low resistance path that the current is seeking to return to ground. Hence, the occurrence of deadly electric shocks to immersed people in freshwater marinas, and none being reported in saltwater.

This research will substantiate the need to monitor the electrical condition of a marina using available marina monitoring tools to reduce the chances for personal injury and death from electrical shock induced by electrical faults in boats and marina electrical systems.

An opportunity recently became available to conduct electrical testing in a large, well-maintained marina facility located on a freshwater reservoir in the southeastern United States. The summary of this freshwater research is presented below. It will demonstrate the following:

  1. A low-level ground fault (less than shore breaker trip setting) can be induced on a boat connected to shore power.

  2. If the dock ground system is in good condition, the induced ground fault current will be safely carried off the boat to the earth ground ashore through the dock ground wiring system, therefore preventing lethal potentials from appearing on underwater gear.

  3. If the ground system is disabled (to simulate a faulty ground system) while the ground fault is being induced, a portion of the fault current will still find it's way to the earth ground directly through the boat's grounded underwater equipment (e.g. stern drives, propellers, shafts, struts, rudders, etc.) and the water.

  4. The fault current traveling through the water sets up an electrical field that is potentially lethal to any person encountering it.

Methodology, Test Results, and Analysis:

The boat used for the testing (30 ft express cruiser with twin stern drives) was being supplied with a single 30 amp, 120 volt shore cable. The dock was also capable of supplying 50 amp, 240 volt power. The test sequence involved the following steps:

1. Determine the integrity of the ground connection from the boat to the shore. The instrument used was a Suretest Model ST-1THD.

a. Measurements taken on the boat:

1) Line voltage: 123 volts

2) % Voltage drop when exposed to a momentary 1800 watt load: 5% (this is considered near the top limit by industry standards but has no bearing on the test results)

3) Ground to neutral voltage: 0.0 volts (this indicated that the system was very lightly loaded at the time of the testing)

4) Neutral leg impedance: 0.21 ohms

5) Hot leg impedance: 0.32 ohms

6) Ground leg impedance: 0.72 ohms.

b. Measurements taken at the pedestal (supply to boat with shore cord disconnected)

1) % Voltage drop when exposed to a momentary 1800 watt load: 1.1%

2) Ground leg impedance: 0.16 ohms


Ground integrity. Max current (assuming a worst case direct short to ground, and based on 0.72 ohms ground resistance), for 123 volts on this ground leg is 171 amps. Although this ground resistance is higher than normally observed, it should still allow enough fault current to flow to cause a 30 amp breaker to trip instantaneously. Based on the premise that the ground wire must be capable of carrying at least five times the breaker's rated current during a fault, the integrity of the ground wire from the boat ground to the earth ground ashore is acceptable.

Also notable is the difference in ground resistances and voltage drops between measurements taken on the pedestal and on the boat. These indicate that the dock pedestal supply and ground is in better condition than the boat, and that there may be a problem with relatively higher resistances at the shore connection (boat or pedestal side) or on the boat itself.

2. Determine the effects of inducing a known ground fault while simulating loss of ground continuity. To accomplish this, a jumper was installed between the shore cord and the pedestal. This jumper is equipped with a ground lead that can be opened to simulate a loss of ground continuity to the boat. Additionally, a device was installed on the boat to introduce a 5.2 amp ground fault between the boat's hot and ground legs. This simulates a ground fault that is of low enough magnitude so as to not trip any circuit breakers, a condition that can persist unnoticed by boat owners or marina personnel (unless continuous or frequent monitoring routines are in place).

Current flow into the water was measured by clamping the cable with a Hyoki Model 3283 clamp-on ammeter. The meter reading indicated the current that is not coming back on the neutral or ground wires and therefore must be leaking into the water through the boat's ground system via grounded, underwater equipment. The following summarizes the tests and measurements taken.

a. Current flow into water with 5.2 amp ground fault induced:

1) Ground connected: 0.069 amps leaking into water (most of 5.2 amps carried off on the ground wire)

2) Ground open: 1.6 amps leaking into water (5.2 amps not achieved due to resistance of the water)

Analysis: With the ground intact, 98% of the fault current is carried safely off the boat to ground. This provides protection to personnel in the event of a ground fault. With the ground open, 31% of the maximum possible fault current enters the water seeking earth ground. This creates an electric field that is discussed in subsequent tests.

b. AC voltages measured in reference to dock ground at various locations with the ground open:

1) Across open ground connection, no fault: 0.69 to 1.25 volts, at jumper

2) Across boat hot leg and boat ground, no fault: 122.6 volts

3) Across open ground connection, fault induced: 84 volts, at jumper

4) Across boat hot leg and boat ground, fault induced: 37.6 v

Analysis: Note that the sum of the two faulted voltage drops above (c + d) equals the total no-fault voltage measured in part a. This was as expected. The 84 v measured in part c above represents the voltage on any grounded underwater metal component on the boat. In other words, touching any of these metal parts directly would most likely result in a fatal shock if the person is grounded through the water. The 1.6 amp fault current measured indicates that the water resistance in this part of the freshwater reservoir is approx 50 ohms (using Ohm's Law: 84 volts/1.6 amps = 52.5 ohms).

A confirmation calculation was made to ensure that the 0.069 amp current leaking into the water with the ground connected correlated with the total fault current (5.2 amps) and the measured ground resistance (0.72 ohms). The total fault current, with the ground connected, should split in proportion to the ground resistance and water resistance.

Amps leaking into water = 5.2 amps x 0.72 ohm ___________________ = 0.073 amps
(fault with ground intact)                                           (52.5 ohms + 0.72 ohms)

This calculation validates our ground resistance and current measurements.

3. Determine the presence and extent of any electric fields in the water generated by the ground fault. Examine any affect this field might have on neighboring boats and nearby dock structures. To accomplish this, a ground fault was induced and the ground opened as above (simulating the two conditions necessary to force current flow into the water and establish an electric field). A copper probe was immersed in the water in various locations around the boat and the voltage between this probe and the dock ground was measured (with the fault induced and normal ground open). The voltage on a nearby dock adjusting cable (attached to the bottom) was also measured.

a. Voltages measured at various distances from boat:

1) Directly behind test boat, approx 1 ft from outdrive: 42-52 volts

2) 10 ft away from test boat: 11.0 volts

3) 20 ft away from test boat: 3.0 volts

4) 30 ft away from test boat: 1.3 volts

5) Near boat directly astern, across dock, approx 10 ft away: 72 volts

6) On dock adjusting cable 15 ft from boat (measured between the cable and the copper probe in the water near the cable): 4.5 volts

b. Voltage across the open ground with no fault: 0.8 volts for all measurements

Analysis: An electric field was measured at and in the vicinity of the ground-faulted boat. It was strongest near the test boat and tapered off with distance. The exception to this was the field measurement made on the boat across the dock, directly astern of the test vessel. This has to be further examined to explain the finding. Since the depth of the water was reported to be approx. 18 ft, these measurements showed that the electric field was strong enough directly behind the boat (2.3 - 2.8 volts per foot) to be potentially lethal to a person entering the water in this field (2 volts per foot generally considered to be lethal). In other areas further from the boat, the effects of the field would likely be felt as a shock or tingle while in the water. The resulting panic from this sensation could cause a person to sustain injuries or death from the inability to swim safely out of danger.

The voltage measured on the dock adjusting cable indicates that some of the ground fault current in the water is seeking this path back to earth ground. It is possible that a person touching this cable could feel a shock or tingle if they are also grounded to the dock.

c. Current measured when shore cables on adjacent boats were clamped:

Proximity to test boat Ambient leakage With fault induced
Next boat 0.044 amps 0.169 amps
15 ft 0.0016 amps 0.044 amps
25 ft 0.007 amps 0.056 amps
30 ft 0.001 amps 0.020 amps

Analysis: On all boats clamped during the faulted condition, current increased in the shore cables. This again represents current escaping the water and returning to ground via the dock ground system and indicates how ground fault currents can travel in the marina when the ground on the faulted boat is lost. Being in this current path between boats is also a hazard.


We have shown that when a ground fault is induced in a purely freshwater marina, and the ground on the boat is disabled, potentially lethal electric fields are established near and around the affected boat. The voltage potential of other dock structures are also elevated, presenting further risk of injury.

Actions can be taken as follows to minimize the possibility of the two necessary conditions (ground fault and reduction in ground integrity) existing simultaneously:

1. NFPA 303 requires an annual verification of the integrity of the grounds in marinas. This test may be performed more often by trained marina personnel to ensure the safety afforded by a proper ground system is maintained.

2. Equipment is available to monitor docks continuously for ground faults (both dock and boat related). With a monitoring system installed, the marina operator would be alerted immediately whenever a ground fault occurs. Operators can then take swift action to locate and isolate the fault using the fault monitoring system and logical isolation procedure.

By periodically ensuring the ground system is in good condition, and by proactively eliminating ground fault conditions, the marina operator will be presenting the safest possible environment for marina clients and staff.

  Send to a Friend View / Add Comments  

  [ Back to Top ]

Copyright © 2004 Mike Holt Enterprises,Inc.
1-888-NEC-CODE (1-888-632-2633)