No one seems to agree. Should the electrical system be bonded to the hull or not?

Earthing in a marine installation

How should earthing arrangements be made on a boat?

This is one of the most common questions asked and one of the most misunderstood subjects in the field. Even so called "experts" very often get it completely wrong.

The subject is also, unfortunately, one of the most complicated to answer. Consequently this document is rather long.

The complexity problems arise because there are so many separate, unrelated, aspects to consider. In order of priority they are........

1. The matter of safety for those on board and for those not on board. The matter of safety for those not onboard is often overlooked or completely disregarded by the uninitiated.

2. The matter of galvanic corrosion.

3. The reliability of the equipment on board.

Now there are two ways we can answer this question. Some people will not believe a technical argument, perhaps because they don't understand it, perhaps because their interpretation of technical matters is incorrect. And for those people we will answer the question like this:-

The AC electrical system earth should be bonded to the hull because:-

1. The European Recreational Craft Directive says so.

2. The British Marine Electronics Association "Code of Practice" says so.

3. The book "The Boatowners Electrical and Mechanical Manual" by Nigel Calder (a world renowned expert) says so.

4. The ABYC (American Boat and Yacht Council) recommends so.

How many more references do you need?

The above people and organisation didn't come to the conclusion that the ground should be bonded to the hull on a whim. They came to this conclusion because they spent a lot of time and effort studying every possible fault and condition and drawing the conclusion that to bond the ground to the hull is, on balance, far safer than to leave the system floating.

The other way to answer the question is like this:-

Let's tackle the matter of safety for those on board and those not on board first.

The first safety matter is obviously that of electric shock. Clearly we have to reduce the risk of this as much as possible.

Electrical equipment is manufactured to 2 distinct standards. "Single insulated" (Class I) and "double insulated" (Class II).

"Single insulated" equipment is manufactured in such a way that the AC mains electricity is insulated from the casing so that the two do not touch. In the event of an internal fault, this may no longer be the case. For instance a live cable, inside the equipment, may become loose and contact the case of the equipment. This obviously presents a shock hazard if the equipment casing is metal. So the casing is earthed to the green/yellow conductor in the power cord.

In the event of the fault described above, the live cable connects to the case of the equipment, which is connected to the earth conductor, which causes an enormous current to flow, which blows the incoming fuse thus interrupting the supply.

That is the safety mechanism of single insulated equipment.

You will instantly see that the integrity of the earth conductor is vital to the safety of the equipment.

"Double insulated" equipment is manufactured in such a way that even if the same internal fault arises, the cable simply cannot contact anything metal that is accessible from the outside of the equipment. So even if the live conductor becomes detatched inside the equipment, it does not present a shock hazard to the user. "Double insulated" equipment, by definition, has two insulation barriers between the electrical parts and the user. This usually consists of the insulation on the cables and connections inside the equipment as the first barrier, then an insulating case as the second barrier. This equipment usually consists of an entirely non conductive outer casing (i.e. plastic).

"Double insulated" equipment does not require an earth conductor in the power cord.

These are the two accepted standards throughout Europe (and most of the rest of the world) for safety in AC powered equipment.

"Double insulated" still suffers from the problem that the equipment can still cause an electric shock if it gets wet and the water gets to the insides of the equipment. This danger cannot arise with "single insulated" equipment as the casing is "held" at the same voltage as the ground upon which we stand by the earth conductor in the power cord. Therefore no voltage difference can appear between the case of the equipment and the ground. Therefore no current can flow through someone holding the equipment and stood on the ground.

A vessel with AC and shorepower facilities

If we now take a look at a boat (let's take a metal hulled boat for this purpose), plugged into shorepower, we can instantly see that, from the point of view of someone not on the boat it is a piece of equipment on the end of a power cord. We can further see that it is a "single insulated" piece of equipment and therefore must have it's casing (the hull) connected to the ground conductor in order to ensure the safety of those outside the equipment (the vessel).

If the hull is not connected to the ground conductor, and we get the fault of a loose live cable touching the hull, we get the situation where the hull is at 230 volts (in Europe) and the ground around it is at 0 volts. It obviously depends upon the conductivity of the water (whether fresh or salt water, pollution levels etc) but it is far from certain (particularly in fresh water) that this fault will cause sufficient current to blow the fuse on the shorepower point.

Anyone touching the boat and the ground (perhaps climbing aboard) will instantly receive an electric shock.

Anyone (or anything) swimming in the water will have an enormous voltage differential presented across their body which may be sufficient to electrocute them, and even if not, will almost certainly paralyse them causing them to drown.

This fact cannot be argued against. The hull must be bonded to the incoming earth conductor in order to ensure the safety of those not on board the vessel.

And I really think this is the crux of the matter. When the safety of those not on the vessel is considered there is no argument whatsoever for not bonding the ground to the hull. It is that simple.

Now it could be argued that the presence of a Residial Current Device will sense this fault and trip thereby ensuring the safety of everyone in the vicinity of the vessel however there are a couple of problems here that means this cannot be relied upon.

Firstly, RCDs are not manadatory in many parts of the world (including the UK), so there may not actually be one fitted.

Secondly, even if one is fitted, whilst it is almost certain that it will trip in salt water, this is not the case in fresh water (because freshwater is a much poorer conductor than salt water). Especially on a small boat, or one with most of the underwater steelwork painted, or a GRP or wooden boat with some underwater metalwork.

Many people argue that the AC system should not be bonded to the earth conductor (for the reason that it can cause galvanic corrosion problems which we will come to later), but when presented with the above scenario, hold their head in shame and admit that it is something they had never even considered. They had only considered the safety of those on board the vessel.

That one argument alone confirms that, in the case of a vessel that has the facility to use shorepower, the AC ground must be bonded to the hull. To not bond them is leaving oneself open to electrocution of people (perhaps oneself or one's own family), sleepless nights in the event it should happen, litigation and possibly even criminal proceedings for manslaughter for ignoring all the codes of practice, guidlines and actual laws which state, quite categorically, that they should be bonded.

A vessel with an AC system but no facilities to use shorepower

The power in this case would come from either a generator or an inverter. It is still 230 volt (in Europe) power, it is just as dangerous as shorepower.

However in this case the risk to those not on board does not exist.

There are many possible causes of electric shock, the most common being someone touching live and earth at the same time. Now in the case of a generator or inverter, if the system is totally isolated from the hull (which we can do in this case as the risk to those not on board no longer exists) this danger no longer exists.

However, if the system is totally isolated from the hull, and a fault arises that connects live to the hull somewhere in the installation, no resultant problem will ensue. The system will continue to operate perfectly...... until someone touches the hull (which is at 230 volts) and the case of a "single insulated" piece of equipment (which is at 0 volts) at the same time. They will receive an electric shock.

Had the AC system had it's earth conductor bonded to the hull this situation could not arise because as soon as the live cable touched the hull either the fuse would blow or the inverter or generator would cut-out having detected an overload.

Now in order to ensure full safety, we need to also bond neutral and earth at the output of the inverter and/or generator and install RCCDs on the outputs of the inverter and/or generator.

Either way, we still need to bond the AC system to the hull.

AC in general

So whether the system can use shorepower or not, if we have AC on board, the earth conductor must be bonded to the hull.

DC system with no AC on board

From the point of view of safety it makes no difference whether or not the DC system is bonded to the hull.

From the point of view of galvanic corrosion it makes no difference whether or not the DC system is bonded to the hull (but the hull must never be used as a return path in the manner of vehicle wiring).

However, if the DC positive side is bonded to the hull this can have huge implications for galvanic corrosoion.

Remember, the most positive (voltage wise) point will be the point that erodes. The negative point will receive a plating from the most positive point.

DC system negative bonded to hull

This diagram shows a single battery, with the negative side bonded to the hull. It also shows a load switch, a load and two areas of dampness represented by the lines with arrowheads at each end.

No current will flow through the damp area at A as both ends are held at the same voltage by the bonding to the hull.

Current will flow through the damp area at B and, due to electrolysis, the wire will erode (as it is more positive) and plate the section of the hull at B (which is more negative) with a small amount of copper.

DC system positive bonded to hull

This diagram shows exactly the same installation but this time the positive side has been bonded to the hull.

This time no current will flow through the damp area at B because both ends are held at the same voltage by the bonding to the hull.

Current flows through the damp area at A, but this time the hull will be eroded (as it is more positive) and plate the wire (as it is more negative) with a small amount of steel from area A.

This is stray current erosion. The hull is being eaten away. Purely and simply by bonding the positive to the hull instead of bonding the negative.

If the DC system is not bonded to the hull at all then obviously this cannot happen.

However, if the positive becomes bonded as a result of a fault (a frayed wire perhaps), the system will continue to operate, the fault will not show itself with any symptoms. So the vessel now has a positive bonded system without the owner knowing anything about it. And with it, the vessel also has all the problems of a positive bonded DC system - i.e. greatly accelerated stray current erosion.

There are two options to protect against this....

1. Bond the negative to the hull. If a fault occurs that attempts to bond the positive to the hull, the main fuse will blow, alerting the owner to the problem.

2. Keep the system isolated from the hull and fit a device that detects if the positive somehow becomes bonded to the hull. I am not aware of any such device and do not see a market big enough to warrant designing one.

So even with a DC system and no AC system, bonding the DC system to the hull is still required. Without doing so, you may find you have inadvertently created a positive grounded system.

AC system and DC system

If both electrical systems are installed then all of the above applies. i.e. it is imperative that both systems be bonded to the hull.

There is also another scenario in this case to further convince the doubters.

Assume the AC system is bonded to the hull but the DC system is not.

Some equipment is connected to both systems. This sounds rare - in actual fact it isn't, dual voltage fridges, battery chargers and inverters are all connected to both systems.

A fault in one of these items could cause AC mains to be presented to the DC side. If both systems are bonded to the hull, this will instantly cause the incoming fuses or circuit breakers to blow.

If one of the systems is isolated from the hull this will not happen. The result will be that the DC system (which we all assume is safe to touch, and which usually has components with insulation rated for about 50 volts) will be sat at 230 volts with respect to the hull or the other electrical system. Clearly this is highly dangerous.

In summary, whatever electrical system is fitted, it is imperative that the system is bonded to the hull.

Galvanic corrosion problems as a result of bonding the AC system to the hull

When plugged into shorepower, and the AC system ground is bonded to the hull, the quayside, other boats and your boat creates a battery which, 9 times out of ten, causes a current to flow in such a direction that your hull erodes. This is galvanic erosion.

There are 2 simple remedies to this problem. One is to fit an isolation transformer, the other is to fit a galvanic isolator. Both will cure the problem.

Bonding the DC system cannot affect this.

All bonding should be done at one central point. It is not acceptable to bond various parts of the system in various separate places. This can cause voltage differentials between various parts of the hull which can lead to stray current erosion.

Finally, a few people with steel hulled narrowboats have mentioned that the resistance between the hull (which is usually almost completely bare steel in a narrowboat) and the actual ground (of the world) is so low (even in fresh water) that an RCD will always trip in the event of a live-earth fault and in fact some have gone on to say that the resistance is so low that a main circuit breaker or fuse will blow.

We accept that this is true most of the time however there are narrowboats out there which have completely or almost completely painted hulls and in this case is it far from certain that an RCD or circuit breaker will blow.

Further, narrowboat owners seem to forget that not all boats have steel hulls. Many are fibreglass or wooden. Some are carbon composite etc. One has to think further afield than "my boat". These regulations and guidelines have to cover all boats not just one or two!

There is a further discusion of galvanic corrosion here.


Web site and all contents Copyright SmartGauge Electronics 2005, 2006, 2007, 2008. All rights reserved.
Page last updated 02/04/2008.
Website best viewed on a computer of some sort.