Well, I think I've done some calculations for concrete.
I make the resistance for typical 20cm concrete floor (with some major
assumptions) with a wet foot sized imprint about 400 ohms, easily enough to
cause trouble for a human touching a 240V live. However, the introduction of
floor insulation or a damp proof course would seriously increase this,
possibly to safe levels. In any case, it would appear to me that an
uncovered ground floor concrete floor is not a safe surface for a bathroom.
If I was having to lay one, I would definitely include any internal
reinforcement mesh in the supplementary bonding. It should be noted that the
conduction path is through ionic transfer in the cement.
So you're saying it wasn't the bonding that caused the metalwork to
become live but two seperate faults elsewhere in the installation.
If you didn't have the bonding you would have found other faults or
problems from not having bonding anywhere.
If you would care to have go on google you will find plenty of
articles on bonding, why you should do it, how it works etc..
The floor isn't conductive so, no shock. You most certainly *can*
grap a live wire and not get a shock, in fact it's the commonest
situation in my experience as one nearly always wears rubber or
plastic soled shoes nowadays and floors are nearly always covered with
a layer of material which is afairly good insulator.
It's only of your other hand is leaning on something earthy that one
gets a shock usually. I can't remember when (if ever) I've got a shock
which was anything other than hand to hand (bad, across chest) or,
more often in my case, simply across my fingers.
It can only make things more dangerous if there's something conductive
at earth potential in the bathroom, if there is then the bonding is
faulty because that item should be connected too. If the bonding is
done correctly then even a damp floor will be at the same potential as
the bonded metalwork, where else can it get its voltage level from?
Thanks for making that clear, Bob. As Lurch has said, while it's possible
to create scenarios with multiple faults where the supp. bonding causes a
sustained potential difference between metalwork and surrounding surfaces
- 'the floor' in your example - the idea is that even in such a case, you've
got all of the possible *low resistance* paths for current flow at the same
potential, thus no current high enough to kill; leaving only a high
resistance path - viz., through the floor - to give a probably unpleasant,
but only pretty remotely possibly fatal, tingle. (Around 0.5mA is the lower
limit of a current you can feel, which neglecting the body's own impedance
means a resistance-to-earth below half a megaohm on a 240V supply gives you
a detectable tingle. Muscle spasm is induced at around 10mA, which needs a
resistance of some 20 times lower, i.e. about 25kohms; screwing up your
heartbeat starts being likely at about 40mA, so 4 times lower again at
about 6kohm.) I know of no commonly-used floor construction which will
give a resistance to earth in the 10kohm-and-under range... even the
'bath salts and metal drainage pipe' instance reported in this thread
was at the 'nasty tingle' rather than 'fatality' level.
Bonding only *some* of the stuff in the bathroom would give you the worst
case in the presence of the sort of earth faults we're discussing - leaving a
low-resistance path for the current from the live supply to continue on
after it's been through someone's body - and that's the reasoning behind the
obsessive 'every single bit of metal that might become live OR provide a
plausible path to earth' approach. Even the dear old IEE in trying to square
this circle gives some explicit guidance over limits to 'might' and 'plausible'
in the above: I seem to remember that they're the source of the example of
*not* bonding a metal windowframe in a bathroom, on the grounds that it's not
part of structural metalwork (hence doesn't afford a low-resistance path to
earth), isn't itself close to anything with a mains supply so isn't going to
go live directly, but if it's bonded to other metalwork which *might*
reasonably go live, would then become live and so present a risk to a
windowcleaner climbing up a metal ladder - for whom even mere muscle spasm
could result in falling, splat... :-(
Maybe I'm being dense, but I still don't see how the cross-bonding made
the fault "more" dangerous. I guess you're saying the following: (a) there
was a whole-house fault making all the "earth wires" in the whole installation
essentially useless - i.e., floating, capable of attaining a potential more
or less that of the incoming live feed and incapable of making any useful
fuse-blowing/MCB-tripping current flow; (b) there was then a second fault
in some part of the bathroom, which caused something, I'm assuming exposed
metalwork, to go live; and (c) that because of the supplementary bonding
in the bathroom, there was *more* exposed metalwork in that bathroom now
at more or less the potential of the incoming live feed than there would've
been without the bonding, hence more places from which you could get a
shock. Is that about it?
If so... I'm afraid I still don't see the cross-bonding has made things
worse. I think we'd all agree that (a) is the root cause and gives the
greatest cause for concern. Given the undetected presence of this first
fault, it means any fault-to-earth anywhere - not just in the bathroon -
makes every bit of metalwork on all 'earthed' appliances float up to
'live' potential. Ugly, nasty, potentially 'orrible: but even so, the
cross-bonding of services (water, gas, and electric 'earth') across the
whole house is still trying to make this less of a fatally dangerous
situation: *either* the incoming water/gas service provides enough of
a path to earth that it makes a fuse blow, alerting the householder to
fault (a), *or* the services are well enough isolated from local earth
that even though stuff connected to them through pipework ends up at or
near 'live' potential, just about all the readily conductive Things in
the house have all floated up to the same potential, so the occupants
aren't going to get a fatal belt from touching two such surfaces
simultaneously. Given that your bathroom was the source of such a
fault-to-earth, it's not clear that things were made worse by having
*supplementary* bonding there in the bathroom which made sure that other
immediately accessible metal surfaces were really at the *same* potential
- rather than, say, the CH-connected radiator or towel rail having
a low-resistance-enough path back to earth - oh, say, 1kohm - so that
touching that and the actual point of the fault-to-earth would've given
someone a fatal belt.
I do appreciate there's a tradeoff - the window-frame example shows the
downside of *gratuitous* bonding which merely increases the number of
places which could rise to a fault potential (either for ages-and-ages
in the presence of a further fault which makes the 'earth' connection
ineffective, or just for the 0.4s/5s which the installation design allows
between a fault-to-earth occuring and the relevant final-circuit fuse/MCB
blowing); and I've mentioned before that there are alternative approaches to
preventing fatal shock currents, of which avoiding any 'earth' potentials
at all is one (and is the approach often used in creating electrical-test-bench
environments, alongside using isolating transformers to remove any
reference to local earth from the 'live' feed). But those other approaches
aren't generally practical for the nation's bathrooms (though isolating
transformers *are* specified for shaving points!), while sensibly-applied
cross-bonding rarely increases the risk of fatal shock currents, and
in most circumstances usefully reduces such risk.
Sounds OK to me, since you could *only* touch things at 240 volts
(everything conductive is bonded together) you still can't get a
This is the whole point of supplementary bonding, it gives a 'second
line of defence' if the earth is faulty for some reason.
Indeed. If you can eliminate earths altogether the protection is better. The
IEE recommend that you use plastic pipework so that you can avoid the need
for supplementary bonding.
However, if you DO have metal pipework, then supplementary bonding is safer
because that earth fault that makes something 240V does one of two things:
1. It causes a major short to earth, tripping an MCB/RCD before you have a
chance to be shocked.
2. If (1) doesn't occur, it means that everything is at 240V, so you don't
get shocked. Have you seen those pictures of people working on live
multi-kilovolt systems safely simply by only touching the live and not the
Unfortunately, you do. Even if all the metalwork is only at 120V.
Everything was fine until someone used bath salts so that the water
going down the plastic drain to the old [broken?] earthenware drain in
the yard was conductive.
Stand in draining water, get nasty belt off mixer.
I didn't altogether believe this, went over to the friends house and
tried it. Once was enough.
Contractor had used armour of cable for earth continuity, terminating it
very professionally in a proper gland onto the Consumer Unit casing.
Which was plastic.
Filters on a big PC system, TV's etc then took "earth" to 120V.
Not quite sure they go that far; certainly with plastic plumbing there is
less that needs bonding, e.g. taps don't; but other electrical gear
such as electric showers, heaters, shaver points, etc. will still need
Indeed. However, they are keen to have you install plastic plumbing so that
the metal bath and radiators don't need bonding. These are large areas
likely to be touched.
Yes, but these are often not large areas. Also, I have none of these devices
in my bathroom. Nor does my mother's house and nor did my last house. I do
have metal pipework, though, so I've still got to have bonding.
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.