TN ring mains in hospital theatres

mapj1:

I must say, as one who worries about EMC and earth loops a lot, I find that advice a bit surprising. Only a defective ring final is going to generate or respond to magnetic fields more than a radial.

The thing that really throws the earth currents off kilter is providing parallel paths that do not follow the same route as the L and N currents. This occurs everywhere we bond to any already earthed object for example  at a water heater with a metal feed pipe, or a socket box that attaches also to structural steels.

If it really is in the operating theatre, supplementary local equipotential bonding is required by BS 7671, so such an arrangement would be in place anyway?


I have to add, though, that whilst on occasion a fortuitous alternate path might form a "loop", these cannot be avoided in most real installations, even where bonding doesn't come into it - for example, simple connections between electronic devices: even unscreened wired Ethernet has a controlled static discharge path to Earth (well should have according to the standard, I know an unplugged laptop breaks this rule, but some laptop PSUs that don't have earthed DC output also have a high resistance discharge path DC to mains Neutral). These discharge paths are also serve to discharge common-mode noise.


Interestingly, the solution to this problem is to provide more earth paths using multiple bonding networks (CBN and mesh-BN) - see BS EN 50310 and BS IEC 61000-5-2.


So, how does ring vs radial stack up now? Well, arguably sightly better because it's a lower inductance path overall.

IF there really is an EMC problem, the advice must be to avoid these cases  too, but in fact they are not only not prohibited, but in effect mandated.


A radial will be fine, but don't make EMC the reason not to use a ring.  I presume whoever offers the advice  has not actually measured the fields around a ring.

mapj1:

I must say, as one who worries about EMC and earth loops a lot, I find that advice a bit surprising. Only a defective ring final is going to generate or respond to magnetic fields more than a radial.

The thing that really throws the earth currents off kilter is providing parallel paths that do not follow the same route as the L and N currents. This occurs everywhere we bond to any already earthed object for example  at a water heater with a metal feed pipe, or a socket box that attaches also to structural steels.

If it really is in the operating theatre, supplementary local equipotential bonding is required by BS 7671, so such an arrangement would be in place anyway?


I have to add, though, that whilst on occasion a fortuitous alternate path might form a "loop", these cannot be avoided in most real installations, even where bonding doesn't come into it - for example, simple connections between electronic devices: even unscreened wired Ethernet has a controlled static discharge path to Earth (well should have according to the standard, I know an unplugged laptop breaks this rule, but some laptop PSUs that don't have earthed DC output also have a high resistance discharge path DC to mains Neutral). These discharge paths are also serve to discharge common-mode noise.


Interestingly, the solution to this problem is to provide more earth paths using multiple bonding networks (CBN and mesh-BN) - see BS EN 50310 and BS IEC 61000-5-2.


So, how does ring vs radial stack up now? Well, arguably sightly better because it's a lower inductance path overall.

IF there really is an EMC problem, the advice must be to avoid these cases  too, but in fact they are not only not prohibited, but in effect mandated.


A radial will be fine, but don't make EMC the reason not to use a ring.  I presume whoever offers the advice  has not actually measured the fields around a ring.