Shore Power Installation

[Gumpy]

I have to say that I am not talking cheap imports or chinese copies.
We only use RCBO by big name manufacturers I have now standardised on one make
Klockner Moeller as these are the most robust I have found.
One of the joys of doing so much testing to decide on which make to use is that I got a lot of free RCBOs that now are in use on the system here on the barge

 

[goboatingnow]

I for one agree with the current ISO spec ( which is what European boats use to meet the RCD requirements). USA boats never until recently speced whole boat RCD's.

I have huge trouble with main earth connected to DC negative, firstly if the ground ever goes hot, then the DC negative can go hot and (a) destroy every bit of DC systems on board and (b) electrocute you as its byepasses the RCD. Also earths can be elecrically very noisy and this system puts a lot of AC noise on the DC system.

If you have a metal boat then you really need a isolating traffo.


The ISO spec is likely to be rewritten to include isolation traffos as is the marina wiring guides.

To some extent this is all extreme event planning, RCD's rarely fail, boat AC systems dont kill people on a wide scale and in fatc lots of things in an extreme failure mode can kill you anyway ( including almost anything on the boat).

 

[DaveS]

MCBs

I am sure that with a Klockner Moeller RCD (RCCB or RCBO) operation of the test button correctly tests the trip function, as should IMHO be the case with any other reputable manufacturer. As I said, some confirm this by displaying a circuit diagram of the internals on the case, like this:

http://www.tlc-direct.co.uk/Images/Products/size_3/MK5860.JPG

You said, however, that you have come across examples from other manufacturers where this is not the case. I would be very interested to learn who these are.

Leaving aside the fact that arranging for the test button to create an electrical imbalance is probably the easiest and cheapest way to test the mechanism, is it acceptable for the test button to "check the operation of the mechanical parts" but not the electrical operation of the imbalance sensor etc.? In a "Wiring Matters" article the IET gives the following description:

5.2 Integral test device
An integral test device is incorporated
in each RCD. This device enables the
mechanical parts of the RCD to be
verified by pressing the button marked
‘T’ or ‘Test’.

which implies that it just might be. The authoratative answer, however, (at least for the UK) lies within this (and its equivalent for RCBOs):

Standard BS EN 61008-1:1995
Title Identifier Specification for residual current operated circuit-breakers without integral overcurrent protection for household and similar uses (RCCBs). General rules
Status CU
Publication Date 15/03/95
Committee PEL/23/1
Approximate Price £148.00
Notes
ISBN 0 580 23562 9
Pages 108
International Equivalent EN 61008-1:1994 IEC 61008-1:1990
Descriptors Circuit-breakers, Switchgear, Earth-leakage circuit-breakers, Domestic safety, Electrical protection equipment, Electrical safety, Domestic electrical installations, Electrical equipment, Definitions, Classification systems, Rated current, Marking, Performance, Environment (working), Performance testing, Type testing, Verification, Electrical testing, Dielectric-strength tests, Electrical insulation, Short-circuit current tests, Gases, Electrical components, Electric terminals, Screws (bolts), Clearance distances, Leakage paths, Breaking capacity, Power factor, Temperature-rise limit, Mechanical testing, Impact testing, Thermal testing, Environmental testing, Fire tests, Endurance testing, Approval testing, Certification (approval), Test equipment, Circuits, Test fingers, Testing conditions, Test specimens
Cross references IEC 60884-1:1994, EN 60051, EN 60529:1991, EN 61009, EN 61543:1995, EN 55014:1993, IEC 60038:1983, IEC 60050-151:1978, IEC 60050-441:1984, IEC 60051, IEC 60060-2:1994, IEC 60068-2-28:1980, IEC 60068-2-30:1980, IEC 60364-4-443:1990, IEC 60364-5-53:1986, IEC 60417:1973, IEC 60529:1989, IEC 60695-2-1:1980, IEC 60755:1983, IEC 60884-1:1994, IEC 61009, IEC 61543:1995, CISPR 14:1993, ISO 7000:1989, CENELEC HD 243 S10, CENELEC HD 323.2.28 S1, CENELEC HD 323.2.30 S3, CENELEC HD 444.2.1 S1, CENELEC HD 472 S1
Replaces BS 4293:1983
Replaces Notes Replaces BS 4293:1983 which remains current.
Replaced By
Replaced by Notes

but I'm not about to pay £148 to find out! Perhaps someone with work access to a suitable technical database and a few minutes to spare could take a look?

 

[Gumpy]

I am sure that with a Klockner Moeller RCD (RCCB or RCBO) operation of the test button correctly tests the trip function, as should IMHO be the case with any other reputable manufacturer. As I said, some confirm this by displaying a circuit diagram of the internals on the case, like this:

http://www.tlc-direct.co.uk/Images/Products/size_3/MK5860.JPG

You said, however, that you have come across examples from other manufacturers where this is not the case. I would be very interested to learn who these are.

Leaving aside the fact that arranging for the test button to create an electrical imbalance is probably the easiest and cheapest way to test the mechanism, is it acceptable for the test button to "check the operation of the mechanical parts" but not the electrical operation of the imbalance sensor etc.? In a "Wiring Matters" article the IET gives the following description:

5.2 Integral test device
An integral test device is incorporated
in each RCD. This device enables the
mechanical parts of the RCD to be
verified by pressing the button marked
‘T’ or ‘Test’.

which implies that it just might be.

Off hand I cant remember which ones that failed in the worst state of test button working and the RCBO not, since the R&D was a few years ago I don't have the results any more.
However the imbalance current that is put on the system by a test button is not going to be 30mA exactly it will have to be more how much more will vary from make to make so although it tests the mechanism the test does not confirm that the unit is operating to manufacturer's specification.

 

[lw395]

I think the OP is talking of using mains while his boat is ashore?
What is the considered view regarding bonding the earth at the incoming cable to local ground/cradles/ladders etc?
If the cable is long, there can be a difference, was there not a thread on here regarding somebody getting a tingle off his guardwires when boarding?
I'm off to test my RCD with a few resistors to earth (or is that GND?)
I've always harboured doubts about breakers with moving parts in a marine environment, but from my dim and distant lab days, I will say that RCD's are generally a GOOD THING!
Cheers,

 

[bedouin]

Let's keep things in perspective here. I don't recall hearing on anyone being killed or even seriously injured because of a faulty shore power installation (although I am sure it must have happen). People are beginning to hypothesise about multiple failure situations "The Battery Charger fails and the RCD fails". Such multiple failures are very unlikely (assuming basic checks and safety precautions are carried out).

I would certainly regard using electrical power in the garden (lawnmowers, hedge cutters) far more dangerous than using shorepower on a boat. And if people are that worried how on earth do they get to their boat in the first place. I remain convinced that by far the most dangerous aspect of sailing is driving down to the boat in the first place.

 

[Gumpy]

Let's keep things in perspective here. I don't recall hearing on anyone being killed or even seriously injured because of a faulty shore power installation (although I am sure it must have happen). .

It happens many times its just not always reported as electrocution due to the fact that the person or animal is paralysed by the electricity and then drowns.

Here is one from a quick google
http://www.commanderbob.com/art53.pdf
There are many more

Aricles describing the dangers
http://wwyc.info/Files/2008March-AShockingStory.pdf
http://ecmweb.com/ops_maintenance/houseboat_electrocution_case/index.html

 

[bedouin]

Interesting articles that certainly gives cause for thought. It actually surprises me that someone swimming in water (without contact with anything else) can get shocked. However in both cases (a) the "vessels" concerned were houseboats rather than yachts on shorepower and (b) neither had an RCD fitted.

I am certainly not saying that it doesn't happen - but as with all risks (especially wrt boats) it needs to be kept in perspective and an informed risk assessment made.

FWIW my own opinion (and it is no more than that) is that I don't like permanent shorepower installations on a boat and I prefer to use a fly lead and portable appliances on the rare occasion I want mains on the boat - but that is a different thread....

 

[dnickj]

Just a quick sideline - on fitting an inverter and or genny do the panel think they should be grounded to DC negative - if so at what point ?

 

[Gumpy]

Interesting articles that certainly gives cause for thought. It actually surprises me that someone swimming in water (without contact with anything else) can get shocked. However in both cases (a) the "vessels" concerned were houseboats rather than yachts on shorepower and (b) neither had an RCD fitted.

Its to do with the fact that the body is a better conductor than water so the electricity will pass through the body rather then round it in the water.

I live on a steel boat and believe me that makes you very aware of electricity, water and galvanic action.

 

[Gumpy]

Just a quick sideline - on fitting an inverter and or genny do the panel think they should be grounded to DC negative - if so at what point ?

I will say how mine is done (on a steel boat) and I believe that it will pass all present standards and those that are about to come in.
Generator Neutral bonded to main AC earth bus
Inverter/charger Neutral bonded to main AC earth bus ( with a relay that lifts it when on genny or shore power .
Shore power connected by isolation transformer
DC -ve bonded to engine bed
AC earth bus bonded to engine bed
The AC and DC should be bonded to the hull as near as is practicable to each other, there is the danger that if you use the same bolt that it could become detached and leave any AC fault current with nowhere to go except the DC system hence the use of two bonding points.

 

[bedouin]

Its to do with the fact that the body is a better conductor than water so the electricity will pass through the body rather then round it in the water.
.

What I find surprising is that there can be sufficient potential difference in the water to generate enough volts across the human body (that is relatively small in comparison to the amount of water involved) to cause damage.

I am guessing that in these cases (a) the water was probably fresh rather than salt and (b) that they must have been pretty close both to the boat and to a reasonably good ground point

 

[VicS]

What I find surprising

My understanding is that the risk is greater in fresh water than in sewater, as you have deduced for yourself.

 

[Gumpy]

Just had a Moeller unit fail (test button ceased to work) decided to take it apart. The test button trip is done by imbalancing between seconary neutral and primary live using a 5k6 resistor, now unless my maths are really rubbish I make this 40mA so as said before the test button only tests that the unit is working mecanically not that it is working within spec and the extra 10mA could be the fatal bit.

 

[lw395]

Just had a Moeller unit fail (test button ceased to work) decided to take it apart. The test button trip is done by imbalancing between seconary neutral and primary live using a 5k6 resistor, now unless my maths are really rubbish I make this 40mA so as said before the test button only tests that the unit is working mecanically not that it is working within spec and the extra 10mA could be the fatal bit.

I agree with your maths, but I'd be relatively happy with 40mA. I don't imagine there are lots of units out there with trip points of 39mA, and the fault that hurts you is likely to have amps behind it. Most will eiither work or not.
I am slightly surprised, normally if the requirement is a max of 30mA, I would test at 25 or so to allow for tolerances.
Aparently 20mA is roughly the lower limit for possible death!
http://www.mpoweruk.com/shock.htm
I thought it was more than that!, note the webpage is about DC fwiw.

I'd also be surprised if the test button ever worked in a very different way, I can't see any advantage in doing it differently?

 

[Gumpy]

Aparently 20mA is roughly the lower limit for possible death!
http://www.mpoweruk.com/shock.htm
I thought it was more than that!, note the webpage is about DC fwiw.

In the USA RCD (GFI) are usually 10mA as opposed to 30mA in EU
Personally I think that 10mA although safe is very low and would lead to a lot of nuisance tripping, 30mA is to high as it can be fatal, 15-20mA would have been a better standard but then I am in no position to have influenced this.