Wiring and installing an isolation transformer - Page 2.

Here is the diagram again with the chassis earthed to the shoreline earth.

Now let's consider water between each of the points A, B, C and D and the transformer chassis and enclosure.

Water between point A and chassis.

This may or may not blow the incoming fuse (almost certainly not in the case of fresh water). It may or may not trip the incoming RCD (if fitted). No immediate danger exists as the case is safely held at earth potential i.e. 0 volts.

Water between point B and chassis.

This will either do nothing or possibly trip the incoming RCD (if fitted).

Water between point C and chassis.

This is a difficult one! The water will attempt to ground the live side of the secondary to the shorepower earth (albeit through a resistor - the water in the transformer enclosure). This will therefore try to put a voltage on the hull of the vessel. The ratio of the voltages between the hull and real ground and that between the secondary live and real ground depends entirely upon the ratio of the resistance of the two water paths. It is a fair assumption that the resistance between the hull and real ground will be substantially lower than that between the secondary live and the shorepower earth. The voltage between the hull and real earth will be lower by the same ratio. No circuit fuses will blow and no RCDs will trip.

Water between point D and chassis.

Nothing, no symptoms, no danger.

Now let's run through the same faults with the chassis returned to the boat earth.

Here is the diagram again.

Water between point A and chassis.

This produces the same final effect as the the last fault described above but with one major difference. The water will attempt to make the hull of the boat live (albeit through a resistor - the water in the transformer enclosure). The hull will try to prevent this due to the conductivity of the water that the boat is floating in. The ratio of the voltages between the hull and real ground and that between the primary live and the transformer chassis depends entirely upon the ratio of the resistance of the two water paths. It is a fair assumption that the resistance between the hull and real ground will be substantially lower than that between the primary live and the chassis. The voltage between the hull and real earth will be lower by the same ratio. However in this case there is a chance that an RCD (if fitted) on the shorepower may trip. This is far from certain (particularly in fresh water).

Water between point B and chassis.

Nothing will happen. An RCD on the shorepower may trip but this is doubtful. No danger exists.

Water between point C and chassis.

No danger exists, No fuses will blow, no RCDs will trip. The water will probably be blown away by the current.

Water between point D and chassis.

No effect.

So, in the case of water ingress to the transformer enclosure, both methods of earthing each produce one scenario that attempts to put voltage on the hull. Due to the massive difference in ratios between the resistance of the water the boat floats in, and the resistance of the water inside the transformer (say 1 litre of water compared to a cubic kilometre!), the possible voltage presented across the hull and real earth is most likely very low. However in the case of the chassis being earthed to the boat hull there is a chance (it is far from certain) that an RCD on the shorepower (if fitted) may trip. This isn't the case when earthed to the shorepower earth.

Note that it is, of course, possible that water ingress contacts more than one point, perhaps 2 of them or even all of them. However this then becomes very complicated to analyse with any level of meaning.

So the argument of which way to earth the transformer, when considering water ingress comes down to the fact that both methods produce one scenario that attempts to make the hull live. The chances of any level of voltage that is dangerous are much more remote than the chances of a dangrous voltage when we consider faulty wiring in the transformer enclosure.

And when we consider faulty wiring, it is clear that to bond the tranformer chassis to the shorepower earth produces less danger.

I therefore believe, on balance, mainly due to the extreme danger of the boat hull becoming live due to faulty wiring that it is safer to bond the transformer safety screen and chassis to the incoming shorepower earth.

We are not alone in this conclusion. Of all the references I found, books, websites, manufacturers manuals etc, all but three recommends bonding the transformer chassis to the shorepower earth. Of the remaining three, one recommends bonding it to the boat hull, the other two recommend simply not connecting it. Not connecting it at all leaves the installation open to a whole collection of other possible problems including, in the main, greater risk of electric shock.

Note that bonding the chassis to the incoming earth means the transformer and casing have to be isolated and insulated from the hull. Otherwise the whole purpose of installing the transformer in the first place (i.e. for galvanic isolation) becomes negated.

I have not considered the possibilities of transformer faults here because, in comparison to wiring faults and water ingress, they are extremely rare.