Supplementary equipotential bonding in group 2 medical locations

There are two things to consider.

The first is robustness, both mechanical and electrical - the wire has to be chunky enough it is  not going to snap if pulled by accident during cleaning/maintenance and it needs to be large enough to carry any current that may flow through it during a credible fault condition. (noting that credible fault may be very short lived - the time it takes the breaker to trip may be very short, so the cable can take much more than it;s steady state current rating without failure - but not for long.)


The second is actually the main function of a bonding conductor., which is to be low enough resistance that the voltage drop along in any normal use case (not all faults) it is small enough not to be dangerous.

This has to be converted into a maximum resistance, which in this case has been decided by the spec writers for you , as 0.2 ohms. This can be seen as  really  a couple of volts if ten amps was flowing. If you had one part of the patient that sort of voltage offset from another ,they are unlikely to come to harm. Actually for non medical use up to about 50V between body parts is considered to be usually OK, if a bit tingly,  for glancing contact with bare skin, but for medical facilities where the patient may not be well, may be wet, may not be fully clothed, may not be able to  flinch and self disconnect etc, the limits are set tighter.


So how do we translate 0.2 ohms (or any other number) into a cable size. Well we need to look at the resistance of copper (or whatever we are using, usually copper.) There are big tables of this but a 'get you close' rule of thumb is to remember one number - sixteen .


A one meter length of copper wire, of 1mm2 cross section, is 16 milliohms (0.016 of an ohm) when at room temp.

A longer length has more resistance pro-rata, say 10m of 1mm sq would be 160 milliohms.

A larger cross section would be less, also pro-rata, so a 4m length of 4mm2 is the same resistance of 1m length of 1mms, or a 300m length of 300mm2. (that's a sort of street main size..)

If you look at the tables you will see it is not quite true, but is good enough for a quick sum in the head to see if you are safely in spec by a factor of 2 or more, or if you  need to look it up and do it properly, or indeed if there is no way it could ever meet the spec without even reaching for a meter. to check.

Note that is for cold copper, and to re-use that rule of thumb for estimating voltage drop in poer cables should normally be calculated assuming  hot conductors  (70C or 90C) and is closer to 18 or 19 milliohms per metre of 1mm2.


Hope this helps.

Mike

A book on the subject.

The IET Shop - Guide to Electrical Installations in Medical Locations



Z.

Plus there's all the usual rules for supplementary bonding conductors as set out in regs 544.2.1 to 544.2.5 - all the stuff about not being smaller than the smaller c.p.c. where interconnecting two exposed-conductive-parts, or half the size of the c.p.c. if interconnecting an exposed-conductive-part to an extraneous-conductive-part and so on, as well as the usual minima of 2.5mm² if mechanically protected or 4mm² otherwise.

  - Andy.

I have this book, and I can see the touch voltage calcs but i'm struggling to see how to turn this into a CSA?


If you have an calcs that could help me out that would be much appreciated. 


Sam.

Well, it's Ohm's law.  But you need to know, either by assumption or measurement the expected  current, and often you do not really. And then a minimum to do with mechanical robustness.


Mike


Slightly more long winded- if you can point out where in my previous post above you feel yourself coming unstuck, I'm happy to go into that in more depth . Or if you can show an example of what you do not understand, can talk you through it in slow time.

I'm looking to do this calculation for bonding conductors to ceiling grids and pipes, so where do I find the values for expected fault currents? 


Would I calculate these from the highest rated circuit on the DB or using the impedance and fault current values on the DB?

If i do the calc based off the Medical Design Guide with an Ia of 200 


0.2 x 200 = 40v which isn't permissible as it is greater than 25V.


so to find the impedance needed to maintain the 25V do Ra = 25V/200 = 0.125Ohms 


So then to find the CSA im assuming ohms law to get I = 230 x 0.125 = 28.75A.


Then doing the adiabatic equation for CSA i get 0.034?

Stop at the 0.125 ohms.  ?


How long is your bond cable that you need to be less than this resistance ?


If we had a reel  of 4mm cable, then that would be  4m /16 milliohms approx  or about 0.04 ohms per metre by my rule of 16  (its not quite true, but first cut...)


(6mm cable would be 2/3  of the resistance of this..)


(watch the units - dividing something in ohms by something in ohms/m,  cancels the ohms and gives  one over inverse metres = metres on top, where you need them .... )


So if your ceiling bond is less than 125milliohms / 4 milliohms per metre   ~ 30metres long  then some off that reel of 4mm would be fine...

So without thought we could use 4mm cable for bonds up to perhaps 25m, and then think about it a bit more at greater ranges.


Any smaller cable than that would need mechanical protection, so is easier just to use 4mm.

But you could put 2.5mm in conduit for a run of perhaps 15metres or less.

If you needed to go further, or your Ia was higher than 200Amps,

then you'd need to either have multiple links in parallel, or fatter cable.


The 0.2 ohm figure assumes a lower Ia than you have chosen. I'm not sure why the book is like that.



Note also that I am being quite careless with mental maths to make the demonstration clearer,  I know that really 4*30 is 120, not 128, but then the tables suggest 4mm may be more like 5 millioms per metre, rather than exactly 4. I justify this arm waving approach by saying that  if the quick arm-wave answer is  close enough that we are looking for a few 10% it is usually worth using the next size up, or taking 30 mins and doing the sums properly, but it saves a lot of time in the majority of cases that are clearly OK  - say if that bond was only 5 or 10m long... 

Things designed to be within a cigarette paper thickness of the spec limit have no margin for later updates or minor errors in the assumptions - is your mains 230V today or 240, that sort of thing, and that can mean trouble later if the cable route is a tad longer than expected or whatever.)

 Mike.

Thanks for that.


As a standard where I am training we use 6mm for group 2 bonding conductors, they're wanting to know a how far a 6mm can get to stay below the 0.2 requirement, so I am kind of trying to figure out the length at the same time so I am just trying different numbers.

Ok so using the rule of 16, a 6mmsq cable is 16 milliohms (0.016 ohms )  for 6m.



soo how many metres would be 200 milliohms ?

(well, if we pretended that the limit was a bit tighter at 160 milliohms, and that is an easier sum at ten times 16 so ten lots of 6m at 16 milliomhs each -  so 60m of wire. The limit is a bit more than 160 millioms,  so you may get a bit further than 60m. But it is a good first estimate you can do in your head without the proper tables to hand.

As above I always like a bit in hand, as on test things tend to come out a bit 'off' , even after allowing for the last digits of the meter being as trustworthy as a fruit machine... ) 


There will not be many rooms so large that need two or more bonds run in parallel and if they do someone will need to pop back to the truck for a second reel.... hopefully the job in a larger place will be costed accordingly.

Mike.