Ship switchboard earth fault

This has happened to me too. That’s why I now switch out the Bender detector with one I know is working after the most probable causes have been investigated.

Last year during the coronavirus mayhem I took a temp job on a vesse that had a bad Bender detector (10 yr old) on a 480v switchboard. I reported the issue in the companies maintenance program. I didn’t have to switch it out with another one to test because it showed a ground with nothing connected to it. It was decided to send it ashore to attempt to be repaired. The problem I had was a lengthy ABS inspection was to take place & I didn’t want to have an empty 4" square hole in a switchboard during the inspection so placed one of those glow in the dark, high voltage warning IMO stickers on top of it. It looked almost completely normal. Even during all the standard reverse power, under current etc. tests the ABS inspector never said a word. If he had, I would have told him about it being written up waiting for parts. It couldn’t be repaired, a new one ordered & when I left the sticker was still covering the hole.


I agree. While ground fault detection can be a life saver spurious ground faults can be a matter of chasing ghosts. Moisture [ on ships?] poor connections and other annoying things can make them almost impossible to track down unless one has absolutely nothing else to do. I used explain to the crew what ground faults really meant while we were investigating. I asked them to let the electricians know if they felt a little tingle when touching something electrical so we could narrow the search area down. You’d be surprised how much more electrically aware and helpful they became.


Most cruise ships do actually have separate grounded systems for the pax cabins.

There really isn’t a better way other than shutting things off one at a time. Unless you have one of the bender systems that actually has monitoring on each feeder being protected on the MSB which they do offer.

Ships are ungrounded mostly out of convention and also because having separate systems would require transformers which cost money and take up space. The only things that need to be left ungrounded would be absolutely critical loads such as steering, M/E LOP, stuff like that.

The rest of heavy industry has moved on from ungrounded systems a long time ago, HRG(high resistance grounding) is pretty much standard in industry nowadays. Very easy to locate faults reliably with HRGs.


I’m going to assume based on the ground detection lights that your voltages referenced to ground are roughly even.

Have any new VFDs been installed recently? For various reason VFDs can trigger the ground fault alarm on the detector.

@prefixmiki No experience with Bender devices but with no fault being indicated on the ground fault indicating lamps you’d have to consider the device it self. Everything has a service life.

There is this on an IS-2 model of the Tokyo Keiki unit which seems similar to the IS-1 unit in your photo. They seem to be up to IS-3 at least.

Page 6 discusses testing the unit. Instead of using a test resistance less than the set point you could try one greater and see if it detects correctly. Since meter and detection LED are both showing low IR sure seems something internal to the device.

Probably unrelated and I know this is a translation to English so may be misread but I would have thought you select a single alarm/actuation point (page 4 )but your photo shows SW 1 and SW 2 both set to a value. Not sure is the circuit is expecting one switch to off and one used or if it would form an additive set point or would it create a fault inside the unit?


Maybe SW1 is warning, SW2 is alarm?

The linked IS-2 manual seems to suggest one on and other off results in various alarm levels. They may be additive but not sure why set at 0.91 Mohm vs 0.9? Also manual shows only a single output, alarm only on the diagram though it seems you have choice of NO or NC. .

1 Like

I get a dangerous file warning when selecting that link.

Ah understand. It does go into the principle a bit by explaining they apply a 50 volt DC to the AC line and monitor the ground current coming back to the device (purposefully grounded) and then somehow screen out the AC effects and using the I monitored and the 50 volts calculate R (insulation resistance). Compare to set point for triggering alarm and providing a 0 to 1 ma meter signal. Or something like that.

Got the manual via another source.
It looks like they saved 3 Yen by using a cheaper alarm relay. I see that circuit as the normally closed contact
maybe feeds a green light someplace that goes out when the SW1 level is reached. At that point the normally open contact closes and maybe turns on a red light or sounds an audible alarm. Who knows?

It does seem strange to have a system with an alarm that has (or can have) such a wide deadband between two levels and no indication that the system has exceeded a low setpoint but not reached the high setpoint. I don’t count a dark indicator light in this application as indicating much other than it might be burned out.

IIRC, floating ground AC outlets for passenger cabins can cause some power strips/surge suppressors to catch on fire.

  • be careful what you wish for re grounded systems, plenty of smaller vessels are wired with grounded AC systems that can cause their own set of problems.

Yes an ungrounded 1ø circuit will require an OCPD on both legs.

As far as “being careful what I wish for” whatever that means, every type of groundings scheme has pros and cons. I pointed out some of the cons of “ungrounded” systems. Yes there can be some special considerations with using grounded systems on vessels, that doesn’t mean ungrounded is the way to go for the entire LV distribution, it’s not. It’s just cheap.

Need help on ship switchboard 380 v earth fault problem. We try switching off all 380 v brakers still have problem. But all machinery are working properly. No trip. Any suggestion?

We rectify the problem. The fault was bridge AC compressor earth fault. Why we miss in first place was even though we disconnected all brakers the neutral line was directly connected from three generators to MSB and directly with out braker to distribution board. After we isolated the Ac compressor All is well now. Thank you everybody for your help.

1 Like

Interesting system. We have a common fault similar to this. It occurs with the laundry clothes dryers. Most of the time on the single phase models, they are wired with the neutral and frame ground tied together from the factory. This can show a 1 to 2 amp ground on the distribution system. Disconnecting the frame ground wire from the lead cures the problem.

1 Like

If you are talking about USA 220/240 volt domestic home dryers, dryer outlets used to only have 2 hots and a neutral, newer ones have 4 connections with a separate ground. Depending on how your house or boat is wired, as you discovered you can shift the chassis ground from its own connection or to the neutral wire.

It may cure the symptom but the unit still has a ground fault and since it is apparently electrically isolated from the hull (illustrated by removing the ground connection) it presents a very real electrocution hazard to anyone who touches the compressor frame and any other grounded item.

The problem isn’t the ground fault display, it is the ground fault itself.


The was what I was wondering about.

I had one hand on the machine and I leaned back to ponder the question and I put my other hand on a steel stanchion. POW - I got a huge electrical shock .

It is actually the display in this instance. With the neutral of the machine connected to the hull, the path of least resistance in the circuit was through the hull. The ground detection system ammeter in the branch breaker panel would indicate about an amp. Keeping the neutral isolated from the hull ground eliminated the ground indication.

Which begs several questions. Is the compressor motor 380V 3 phase?
Why and where is the neutral connected to the motor?
Does the compressor use a phase to neutral to power some small single phase function?
Where and why would the neutral be connected to the motor frame?
Has this ground fault been displayed from the moment the compressor was installed onboard or did it just show up one day?

1 Like

A post was split to a new topic: Electrician Joke