Help me track a GFCI fault
March 14, 2015 12:58 PM Subscribe
For the last week I've been getting intermittent GFCI trips on my houseboat shore power. Is there a way I can track down the source?
The shore power trip switch is very sensitive (which is normal with most marina installs).
When the power goes out my main consumer unit switchboard is completely untripped, so no circuits are getting overloaded.
We were okay for two days, which I had put down to not using the washing machine. I used the washing machine again and the power tripped, so I switched off it's wall switch. I'm still getting trips.
I rewired all my wall sockets to check they were not shorting and haven't seen any problems there.
Is there a way that I can trace the exact place this is happening?
I've kept a log of trips and there seems to be no one appliance or socket causing it. I am very much not happy about the situation.
The shore power trip switch is very sensitive (which is normal with most marina installs).
When the power goes out my main consumer unit switchboard is completely untripped, so no circuits are getting overloaded.
We were okay for two days, which I had put down to not using the washing machine. I used the washing machine again and the power tripped, so I switched off it's wall switch. I'm still getting trips.
I rewired all my wall sockets to check they were not shorting and haven't seen any problems there.
Is there a way that I can trace the exact place this is happening?
I've kept a log of trips and there seems to be no one appliance or socket causing it. I am very much not happy about the situation.
Any chance someone else could try your cable?
With GFCIs it's not so much continuity as any tiny bit of current that leaks to ground (tiny abrasion in the insulation, whatever) will trip the GFCI.
posted by BillMcMurdo at 1:21 PM on March 14, 2015 [1 favorite]
With GFCIs it's not so much continuity as any tiny bit of current that leaks to ground (tiny abrasion in the insulation, whatever) will trip the GFCI.
posted by BillMcMurdo at 1:21 PM on March 14, 2015 [1 favorite]
The simplest solution is to turn off your circuits one by one, but that's probably going to be very timing consuming. I'm not familiar with the setup on houseboats, but in your situation I'd get a multimeter, unplug the shore power cable and see what kind of resistance I got to whatever passes for ground in your situation from the hot side of the unplugged cable (figuring out what the best ground is might be tricky here).
If you get any kind of resistance (my guess is you're looking for a resistance from 100K to 10 megs), then shut off circuits until the resistance goes up. This is more or less assuming that the problem is lingering at a low level and occasionally gets worse so that it causes a trip. If it's an all or nothing thing (like a bare wire that occasionally contacts a metal ground) then this won't really work and you're pretty much back to flipping breakers until the issue goes away.
I'm skeptical of the cable being the source of this, but you can test it in a similar way by unplugging it entirely. and making sure there's no resistance between any of the leads. You could try to have someone flex the cable while you're doing this.
I think it goes without saying that you should only test resistance when you are 100% certain that what you're testing is unpowered.
posted by All Out of Lulz at 2:16 PM on March 14, 2015 [1 favorite]
If you get any kind of resistance (my guess is you're looking for a resistance from 100K to 10 megs), then shut off circuits until the resistance goes up. This is more or less assuming that the problem is lingering at a low level and occasionally gets worse so that it causes a trip. If it's an all or nothing thing (like a bare wire that occasionally contacts a metal ground) then this won't really work and you're pretty much back to flipping breakers until the issue goes away.
I'm skeptical of the cable being the source of this, but you can test it in a similar way by unplugging it entirely. and making sure there's no resistance between any of the leads. You could try to have someone flex the cable while you're doing this.
I think it goes without saying that you should only test resistance when you are 100% certain that what you're testing is unpowered.
posted by All Out of Lulz at 2:16 PM on March 14, 2015 [1 favorite]
How old is the cord from the boat to the shore power plug in? Can you try a neighbor's power cord to rule it out? Any corrosion on the cord or either end of the plug ins?
posted by littlewater at 4:34 PM on March 14, 2015
posted by littlewater at 4:34 PM on March 14, 2015
The cord should be meggered; an ohm test is of very limited utility with these symptoms. Most electricians can do this for you in 15 minutes and the cable is easily portable.
Past that your boat electrical system can be tested circuit by circuit. However that is both time consuming and therefor expensive and may not find your problem. You can't megger anything with electronics in them so if the problem is equipment rather than wiring you are back to process of elimination.
Sadly intermittent problems like this are be very hard to track down. You need to unplug everything and then add in devices one by one.
posted by Mitheral at 8:13 AM on March 15, 2015 [1 favorite]
Past that your boat electrical system can be tested circuit by circuit. However that is both time consuming and therefor expensive and may not find your problem. You can't megger anything with electronics in them so if the problem is equipment rather than wiring you are back to process of elimination.
Sadly intermittent problems like this are be very hard to track down. You need to unplug everything and then add in devices one by one.
posted by Mitheral at 8:13 AM on March 15, 2015 [1 favorite]
Intermittent problems are tough.
Shy of taking the whole circuit apart and re-doing every tap and connection, your best is a long slow process of elimination. Remove one item from the circuit. Give it a few days, and see if it trips. If it does trip, then you know that item is NOT the source. Then remove something else, etc. etc.
posted by Flood at 4:52 AM on March 16, 2015
Shy of taking the whole circuit apart and re-doing every tap and connection, your best is a long slow process of elimination. Remove one item from the circuit. Give it a few days, and see if it trips. If it does trip, then you know that item is NOT the source. Then remove something else, etc. etc.
posted by Flood at 4:52 AM on March 16, 2015
Response by poster: My current theory is based around a super high humidity problem. Does this seem reasonable? Like 9x% humidity could exacerbate underlying wiring issues?
posted by Just this guy, y'know at 3:53 AM on March 18, 2015
posted by Just this guy, y'know at 3:53 AM on March 18, 2015
Response by poster: UPDATE:
Got an electrician in tested all the circuits. No problems, perfect ramp tests across the board, all appliances worked fine.
Turns out that of my two connections the 63 Amp was dodgy and I apparently, occasionally draw more than 32A.
So I was getting overload trips on the 32A and switch failures on the 63A.
The diagnoses of assuming that the trip was happening for the same reason on both connections was ultimately why it was so hard to trace the problem.
Got the 63A circuitry replaced, all is fine.
posted by Just this guy, y'know at 4:13 AM on October 12, 2015
Got an electrician in tested all the circuits. No problems, perfect ramp tests across the board, all appliances worked fine.
Turns out that of my two connections the 63 Amp was dodgy and I apparently, occasionally draw more than 32A.
So I was getting overload trips on the 32A and switch failures on the 63A.
The diagnoses of assuming that the trip was happening for the same reason on both connections was ultimately why it was so hard to trace the problem.
Got the 63A circuitry replaced, all is fine.
posted by Just this guy, y'know at 4:13 AM on October 12, 2015
This thread is closed to new comments.
I haven't tried another cable because an armoured 63A rated 15m cable costs about £300, so I don't have a spare.
I could test the cable, but it would be quite an endeavour to get both ends together for a good continuity test, so I hope I don't have to.
posted by Just this guy, y'know at 1:18 PM on March 14, 2015