View Single Post
  #11   Report Post  
Andy Wade
 
Posts: n/a
Default Generator Grounding

"Dale Shuttleworth" wrote in message ...

The neutral connection [to earth, ed.] must be made before any
earth leakage detection device.


Correct, obviously.

[Generator winding floating]
No, it depends entirely on how good your insulation is. Assuming
good insulation all round, it means that you become the neutral
reference when you touch one of the conductors and you don't feel
a thing. This is fine unless there is anything else drawing a
current to earth at the same time...


This is an IT system - /isole terre/ - floating source and earthed
metalwork. It's used in some industrial situations where supply
disconnnection in the event of an earth fault would be more dangerous then
leaving the power on. It's fine in the sense that a single earth fault
won't cause a problem - in fact the fault will go unnoticed until a second
one occurs. For that reason IT systems have to employ some form of earth
monitoring to draw attention to the fault.

Running your house as an IT system is not permitted, since the indirect
contact protection is designed on the assumption that it's a TN system -
either TN-S or TN-C-S (PME) - and no earth monitoring is employed.

[BW]
Then the phase you've chosen to stick in the earth will need less
than one ohm impedance to be of any in supplying a neutral at the
consumer unit and this would need to be calculated correctly in all
weather and ground effect situations. It may also trip any safety
devices on the generator itself. So that needs further
investigation.


Why is such a low earth impedance required?


It isn't - although the statement "will need less than one ohm impedance" is
pretty meaningless; did he mean impedance to earth, or something else?

Even the earthing resistance at public supply network susbtations only has
to be less than 20 ohms for PME networks (although it is typically far
less).

The rod only provides a reference for the whole
system. Think of the arrangement as having the very low earth fault
loop impedance of a TN-C-S system for faults to bonded equipment
(generator permitting!) but the relatively high earth fault loop
impedance of a TT system for true faults to ground. This does mean
that an RCD on the live and neutral from the generator is essential
to catch true earth faults though.


Without the RCD it's no different to a non-RCD protected circuit on the TN
public supply. This won't catch what you call "true earth faults" either,
but such faults shouldn't be possible since all exposed-conductive-parts
should be connected to the main earth terminal via their respective CPC's.
Anything supplied outside the main equipotential zone should be RCD-fed of
course.

I would also be wary of earth fault loop impedance figures when
looking at the supply from a small generator - I suspect there will
be quite a lot of resistance in the generator...


I agree. The higher impedance of the generator supply means that the value
of Ze (the part of the earth loop impedance exterenal to the house
installation) is higher than the value assumed for design, so you can't rely
on disconnection time requirements being met, etc. A 100mA time delayed RCD
is the best solution here. (If the house installation is TT-earthed then
this RCD will (should) already exist and there's no need for another one on
the generator.)

big snip

[BW]
But if the supply earth is receiving back fed half phase supply
from the generator, which has a phase stuck in the ground at less
than one ohm impedance, then the earth, no matter where it is
coming from, will have picked up this charge and will supply it
back to the connections that already exist.


???


Quite. Whatever he was getting at, it's so badly expressed that I haven't a
clue what he meant, except that he seems obsessed with generator outputs
being centre-tapped.

For a TT supply, there is no problem whatsoever. TT is convenient
in that there is only one possible earth reference, the earth rod.
This would provide the earth reference and the installation
protective conductors. One side of the generator, all metalwork
bonding, etc would be connected to this. The external supply only
gives you live and neutral and there is no risk of earth loops.


Yes, I agree with all that.

For a TN-C-S supply, you will have two earths, the supply earth
(derived from the supply neutral) and your own earth rod. Your own
earth rod is required to provide an earth in the event that the
supply cable is cut.


And that. The requirement for an independent means of earthing is in BS
7671 (Reg. 551-04-03).

I would be tempted to connect the PME earth and the earth rod
together but there are two caveats: 1) There is a risk that you
might experience more rapid than expected corrosion of your earth
rod due to any voltage present on the supply neutral derived earth
during normal operating conditions (this shouldn't be an issue since
otherwise any cross bonded buried pipework would also suffer)


I don't think that's a problem at all. There are other circumstances in
which local earth elecrodes bonded to the PME terminal are recommended.

and 2)
under fault conditions during generator operation the voltage on
your earth rod will rise taking the voltage on the supply neutral
conductor with it. This shouldn't be a problem assuming that
everything locally is RCD protected, cross bonded and the earth
through the rod is reasonable but it may cause some current to flow
into the supply neutral and I don't know if the local
electricity supplier would be expecting this. I don't know the
official answer in this circumstance - it will either be specified
in the relevant IEE regs or you will need to consult your
electricity supplier.


Again, no problem here that I'm aware of. If your generator to house TN
system is properly installed then earth fault currents are contained within
your own copper conductors. Anything flowing to earth should either be
negligible, or, as you say, quickly interrupted by the RCD. But in any case
you do need to consult the supplier - this requirement is in BS 7671, Reg.
551-01-01: "Requirements of the electricity supplier shall be ascertained
before a generating set is installed in an installation which is connected
to the public supply".

As I've said before, BS 7671 has three pages of requirements relating to
generators (which includes UPS's by the way) in section 551. Anyone
contemplating a permanent genny installation should purchase and study, and
consult their supplier. If in doubt consult a professional electrical
engineer.


Here's my attempt at an ASCII wiring diagram:

L o----------------o
\ SW1a
Mains o---------- L to CU
from
meter + -------o
| SW1: DPCO changeover switch
| 100A break before make!
N o----------------o
| \ SW1b
| o---------- N to CU
|
| +---o
| |
| |
Supplier's E o-----------------------------o Main earth terminal
earth | | /
| | |
| | 16mm^2 |
| | earth |
------- |
| | |
| G.P. | | G.P. = Generator
| | | protection,
------- | fuse or CB
| | | plus 100mA RCD
| | |
-- | - | -- |
| G( | | |
| G( | | |
| G( | | | G = generator winding
| G( | | |
| G( | | |
| | | | |
Generator | +---.B | |
frame - | | | | B = bonding connection
-------.-- |
B|\ |
| \-------------+
|
|
|
-----
--- Your earth electrode(s)
- Re 200 ohm


N.B. Reg. 551-02-03 requires automatic load shedding arrangements if the
generator is not rated to supply the whole installation.

--
Andy