Hot Tub - small potential differance between water and the ground

[Andy Wade]

In the open, or in some kind of outbuilding? What is the 'flooring' adjacent to the bath? Form later comments it does sound as it it's in the open.

Presumably you mean 35 V open-circuit voltage (meter on volts) and 0.12 mA short-circuit current (meter on mA). The latter reading is pretty meaningless with the short meter probes just stuck into the water and ground. Repeat the test between the metalwork of the bath (presumably earthed to the exported house earth) and a more substantial electrode in the ground - a 2 ft. offcut of 15 mm copper water pipe is handy for this.

And the house earthing system is what? TN-S, TN-C-S (PME) or TT? 35 V is quite a high difference between a supplier's mains earth and the local ground these days, with their multiply-earthed networks - but it clearly can occur. What's more to the point is that it might only take one fault in the network to turn that 35 V into 230 V and that would make getting in and out of the bath rather dangerous...

In fact the whole installation sounds quite dangerous. An exported mains earth (probably TN) into an outdoor situation and no attempt to make an equipotential zone around the tub. And the two 30 mA RCDs in cascade, no discrimination between them, sends out a message about the lack of competence of the installer.

To use an exported TN earth in this situation the very least I would have suggested is that there must be a bonded metal grid under the tub and immediate surroundings, and the bonding should meet the requirements for a main bonding conductor (min. 6mm^2 for TN-S or 10 for PME). Bits of 2.5 T&E (with 1.5 mm^2 earths) just aren't up to the job here.

The practical fix is to convert this to a local TT system. Isolate and insulate the exported earth at the outdoor end of the cable and provide a local earth system to which the tub and all its ancillaries should be earthed, via a local main earth terminal. The local earth system should consist of an adequate rod electrode (probably 8 ft.) and the aforesaid mat or grid under the tub. If the latter is impracticable then I'd advise putting in four earth rods to form an exscribed square enclosing the tub site and its immediate surroundings. These should be connected together and to the local main earth terminal using 16 mm^2 copper conductors. I presume the water supply is via plastic pipe, but any local copper plumbing that is in contact with the ground should be main-bonded to the local main earth terminal in 10 or 16 mm^2.

HTH

[Sparks]

Hi Andy, I was hoping you would reply! Yes, this is correct, meter on volts, one probe in the water and one stuck in the wet surrounding ground, then meter on mA with the probles in the same place.

There is the problem, the frame is wooden and the tub is fiberglass - there arn't any metal bits on it!

This is what I meant by "His earthing is provided by the armour of the main feed to the house" - So I guess TN-S or TN-C-S - how do we tell, or don't we care?) There are actually two, single phase, economy 7 meters, side by side in the house (Snot sure if they are on the same phase). A consumer unit directly connected to meter 1, then a switch fuse from the other meter, then on to anoter CU about 8M away in a different room. The supply to the shed is via the first CU (this has a 30mA RCD and about 8 MCB's) - The earth for this CU is connected to the armour of both incoming supplies (There are actually two seperate cables, side by side)

This was exactly my thought - not actually causing a physical problem right now, but it could later.

There is no metal anywhere near the tub though - the tub is just sitting on a concrete base about 1ft wider than the tub, then there is mud/grass around this at the moment - the plan is to deck all around the tub.

Here is a simalar tub

And the two 30 mA RCDs in cascade, no discrimination between them, sends

I assume this is because the house does not have a split load CU....yet!

They tub has an earth terminal inside it's control box, but there arn't any metal parts on or near the tub (Asside from the screws holding the side panels on, but these are screwed into the wooden structure of the tub)

The plan is to deck all around the tub wit hwooden decking. Does this change anything, as it will them be impossible to have one foot on the actual ground and another part of the body in the tub.

No water supply, it is filled with the aid of a hose (Connected to the kitchen mixer, so it was filled with warm water, so it was ready to go as soon as it was full!) (It only holds about 1000l)

Thanks for your reply Andy, most appreciated :-)

[Andy Wade]

[Quoting order changed to make arguments flow better.]

OK, this clarifies things a bit. The tub is a Class 1 (earthed) appliance but the tonne of water in it is not in contact with any large earthed metal surface. Nevertheless there are presumably pump(s) and/or heater(s) which might bring the water into contact with some earthed metalwork internally. I now think your 35 V reading is a complete red herring. The water mass is probably loosely capacitively coupled to live mains voltage somewhere "in the works" and, being poorly earthed, is picking up a voltage by capacitive coupling (aka electrostatic induction). This will be pretty inconsequential from a safety point of view as the impedances involved are very high. Certainly the 120 uA touch current you measured is nowhere near high enough to give a dangerous shock, although it could result in perceptible tingles.

Oh we do care...

... if the earthing is to the lead sheath (rather than armour, usually) of a supply cable then it's almost certainly TN-S.

That point still stands, although the danger level is perhaps not quite so high as I first assumed, as there are no exposed-conductive-parts (i.e. no accessible metalwork connected to the appliance's earth terminal).

It's not split-load, but you said it did have an RCD - so there are two

30 mA RCDs in cascade to the tub. An earth fault at the tub could trip either or both RCDs and, by the sound of it, possibly black out half the house. This is a violation of Regulations 314-01-01 and 314-01-02 of BS 7671. If there is an intention to convert the relevant house CU to split-load, with the tub feed on the non-RCD side then there is some logic to what's been done.

In principle no, I still think it needs to be converted to a local TT system by isolating the exported earth. Outdoors things like the concrete base and wooden decking could well be conductive enough to allow dangerous shock current to flow. But maybe you don't need to be quite so anal about the earth system as I suggested at first, if there really are no exposed-conductive-parts. A good single earth rod somewhere near the tub is probably adequate.

[Fred]

This may sound a silly question but; if the pump is double insulated, the tub and pipework are plastic, what's the earth rod going to be attached to?

[Andy Wade]

So it does appear to be Class 1 equipment with parts that need earthing. It would be interesting toe have sight of the installation instructions; are they on the Web anywhere?

[Sparks]

I will ask him if his manual has them in!

Thanks for your replies Andy, very helpful :-)

And yes, there is a heater in this tub, it has a metal outer body (Not sure about the insides!), also, when it was installed, there was a problem with the blower (that blows bubbles in the tub, I noticed this has an earth connection too, although, this does not (Should not!) come into contact with the water, maybe the earth s here too, in case this fan gets flooded somhow.