On Thursday, 29 June 2017 03:38:22 UTC+1, John Rumm wrote:
On 27/06/2017 19:51, tabbypurr wrote:
On Tuesday, 27 June 2017 14:05:49 UTC+1, John Rumm wrote:
On 27/06/2017 09:34, tabbypurr wrote:
On Monday, 26 June 2017 23:22:59 UTC+1, ARW wrote:
On 26/06/2017 22:28, tabbypurr wrote:
On Monday, 26 June 2017 21:23:59 UTC+1, Dave Liquorice
wrote:

kW for a kettle plus a 1 kW of other load but remember
*everything* is going through that single MCB in the house
CU, so you loose diversity. Overload the house CU MCB and
everything goes off in the shop. Overload an MCB in the
shop CU fed with a sub-main only that MCB goes off, perhaps
the lights stay on...

Not mention that having a seperate supply available when
the other is off for some reason is handy.

It depends. If you have say a 32A & 6A shop circuit fed from
a 38A MCB then you're unlikely to trip both circuits if one
trips.



Not if it's a fault current:-)

The graphs at http://wiki.diyfaq.org.uk/index.php/MCB show that
discrimination is normally obtained with fault currents.

not all fault currents have looked at the graph though ;-)

In most cases discrimination will be maintained. The fact that
ocasionally it isn't is rather trivial, and hardly worth a protracted
discussion. The op knows the options.

In many cases of *overload*, discrimination will be maintained - at
least with a single downstream MCB one rating step or more lower than
upstream.

With multiple MCBs downstream (say a CU on a submain) its probably still
likely - but somewhat less so.

However with a 500A fault current why would you expect the magnetic
response of the lower trip current device to be necessarily faster than
that of the higher rating device?

(its for this reason one often elects to use a HRC fuse for the head end
of a submain)

Of the cu trips & fusings I've seen, not many have been such high current. The op can complicate the rewire, but I'm not covinced there will be much gain.


NT

On 29/06/2017 03:45, wrote:
On Thursday, 29 June 2017 03:38:22 UTC+1, John Rumm wrote:
On 27/06/2017 19:51, tabbypurr wrote:
On Tuesday, 27 June 2017 14:05:49 UTC+1, John Rumm wrote:
On 27/06/2017 09:34, tabbypurr wrote:
On Monday, 26 June 2017 23:22:59 UTC+1, ARW wrote:
On 26/06/2017 22:28, tabbypurr wrote:
On Monday, 26 June 2017 21:23:59 UTC+1, Dave Liquorice
wrote:

kW for a kettle plus a 1 kW of other load but remember
*everything* is going through that single MCB in the house
CU, so you loose diversity. Overload the house CU MCB and
everything goes off in the shop. Overload an MCB in the
shop CU fed with a sub-main only that MCB goes off, perhaps
the lights stay on...

Not mention that having a seperate supply available when
the other is off for some reason is handy.

It depends. If you have say a 32A & 6A shop circuit fed from
a 38A MCB then you're unlikely to trip both circuits if one
trips.



Not if it's a fault current:-)

The graphs at http://wiki.diyfaq.org.uk/index.php/MCB show that
discrimination is normally obtained with fault currents.

not all fault currents have looked at the graph though ;-)

In most cases discrimination will be maintained. The fact that
ocasionally it isn't is rather trivial, and hardly worth a protracted
discussion. The op knows the options.

In many cases of *overload*, discrimination will be maintained - at
least with a single downstream MCB one rating step or more lower than
upstream.

With multiple MCBs downstream (say a CU on a submain) its probably still
likely - but somewhat less so.

However with a 500A fault current why would you expect the magnetic
response of the lower trip current device to be necessarily faster than
that of the higher rating device?

(its for this reason one often elects to use a HRC fuse for the head end
of a submain)

Of the cu trips & fusings I've seen, not many have been such high current.

Who you kidding? 500A of PSSC only requires a Ze of around half an ohm
or less, common on many installations.

However the issue still stands at lower currents - say 180A then...
still enough to trip a B32 MCB in a head end.

The op can complicate the rewire,

How would be using a HRC fuse in a carrier in the head end CU
"complicating" it? Its the same design, just a different choice of
protective device.

but I'm not covinced there will be much gain.

There won't - hopefully there won't be many cases where fault current
protection needs to operate on the downstream installation in the first
place. If it does, and you trip multiple MCBs, then that may be
acceptable in the circumstances for an infrequent situation.

However that does not mean you can get away with glib "...show that
discrimination is normally obtained with fault currents" statements! ;-)




--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

On Thursday, 29 June 2017 12:14:43 UTC+1, John Rumm wrote:
On 29/06/2017 03:45, tabbypurr wrote:
On Thursday, 29 June 2017 03:38:22 UTC+1, John Rumm wrote:
On 27/06/2017 19:51, tabbypurr wrote:

In most cases discrimination will be maintained. The fact that
ocasionally it isn't is rather trivial, and hardly worth a protracted
discussion. The op knows the options.

In many cases of *overload*, discrimination will be maintained - at
least with a single downstream MCB one rating step or more lower than
upstream.

With multiple MCBs downstream (say a CU on a submain) its probably still
likely - but somewhat less so.

However with a 500A fault current why would you expect the magnetic
response of the lower trip current device to be necessarily faster than
that of the higher rating device?

(its for this reason one often elects to use a HRC fuse for the head end
of a submain)

Of the cu trips & fusings I've seen, not many have been such high current.

Who you kidding? 500A of PSSC only requires a Ze of around half an ohm
or less, common on many installations.

the point is that most trips aren't due to a zero ohm short

However the issue still stands at lower currents - say 180A then...
still enough to trip a B32 MCB in a head end.

The op can complicate the rewire,

How would be using a HRC fuse in a carrier in the head end CU
"complicating" it?

I didn't say it was. Running multiple cables to the shop & removing the CU would be.

Its the same design, just a different choice of
protective device.

but I'm not covinced there will be much gain.

There won't

so despite arguing you agree anyway

- hopefully there won't be many cases where fault current
protection needs to operate on the downstream installation in the first
place. If it does, and you trip multiple MCBs, then that may be
acceptable in the circumstances for an infrequent situation.

However that does not mean you can get away with glib "...show that
discrimination is normally obtained with fault currents" statements! ;-)

40A MCB takes more i squared t than 32A & 5/6A MCBs to trip. With a 5/6A circuit problem you'll nearly always get discrimination. With the 32A circuit you often will.

As I said you can complicate the job by losing the shop CU & adding more cables, but the real life gain will be minimal.


NT

On 29/06/2017 12:26, wrote:
On Thursday, 29 June 2017 12:14:43 UTC+1, John Rumm wrote:
On 29/06/2017 03:45, tabbypurr wrote:
On Thursday, 29 June 2017 03:38:22 UTC+1, John Rumm wrote:
On 27/06/2017 19:51, tabbypurr wrote:

In most cases discrimination will be maintained. The fact
that ocasionally it isn't is rather trivial, and hardly worth
a protracted discussion. The op knows the options.

In many cases of *overload*, discrimination will be maintained
- at least with a single downstream MCB one rating step or more
lower than upstream.

With multiple MCBs downstream (say a CU on a submain) its
probably still likely - but somewhat less so.

However with a 500A fault current why would you expect the
magnetic response of the lower trip current device to be
necessarily faster than that of the higher rating device?

(its for this reason one often elects to use a HRC fuse for the
head end of a submain)

Of the cu trips & fusings I've seen, not many have been such high
current.

Who you kidding? 500A of PSSC only requires a Ze of around half an
ohm or less, common on many installations.

the point is that most trips aren't due to a zero ohm short

However the issue still stands at lower currents - say 180A
then... still enough to trip a B32 MCB in a head end.

The op can complicate the rewire,

How would be using a HRC fuse in a carrier in the head end CU
"complicating" it?

I didn't say it was. Running multiple cables to the shop & removing
the CU would be.

If it were my setup, I would run a submain from house CU to shop CU.
Leave the tails from that meter parked in a henley block and disconnect
the service. I would put a suitably sized HRC fuse in a fuse carrier in
the house CU for the submain. Minimal change and new materials. The OP
could then have the service to the shop terminated and let the supplier
do what they want with the meter and tails etc.

As a DIY job its the cost of the submain cable, and probably no more
than a couple of hours of work - would likely pay for itself in the
first year.

Its the same design, just a different choice of protective device.

but I'm not covinced there will be much gain.

There won't

so despite arguing you agree anyway

I am disagreeing with your statement about discrimination, since it was
basically wrong for the specific case of handling fault currents.

That is a separate issue from the solution that I propose to Graeme,
since in this circumstance I suspect he could cope with a lack of
discrimination given the circumstance he describes, and the likelihood
of needing to clear a fault in the first place is relatively low.
However since some mitigation can be designed in at no extra effort or
expense it would seem daft not to.

- hopefully there won't be many cases where fault current
protection needs to operate on the downstream installation in the
first place. If it does, and you trip multiple MCBs, then that may
be acceptable in the circumstances for an infrequent situation.

However that does not mean you can get away with glib "...show
that discrimination is normally obtained with fault currents"
statements! ;-)

40A MCB takes more i squared t than 32A & 5/6A MCBs to trip. With a
5/6A circuit problem you'll nearly always get discrimination. With
the 32A circuit you often will.

True for overloads, but not faults.

As I said you can complicate the job by losing the shop CU & adding
more cables, but the real life gain will be minimal.

No point since there are easier ways, and for that matter *in this
circumstance* Graeme may not care anyway.


--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

On Thursday, 29 June 2017 18:40:53 UTC+1, John Rumm wrote:
On 29/06/2017 12:26, tabbypurr wrote:
On Thursday, 29 June 2017 12:14:43 UTC+1, John Rumm wrote:
On 29/06/2017 03:45, tabbypurr wrote:
On Thursday, 29 June 2017 03:38:22 UTC+1, John Rumm wrote:
On 27/06/2017 19:51, tabbypurr wrote:

In most cases discrimination will be maintained. The fact
that ocasionally it isn't is rather trivial, and hardly worth
a protracted discussion. The op knows the options.

In many cases of *overload*, discrimination will be maintained
- at least with a single downstream MCB one rating step or more
lower than upstream.

With multiple MCBs downstream (say a CU on a submain) its
probably still likely - but somewhat less so.

However with a 500A fault current why would you expect the
magnetic response of the lower trip current device to be
necessarily faster than that of the higher rating device?

(its for this reason one often elects to use a HRC fuse for the
head end of a submain)

Of the cu trips & fusings I've seen, not many have been such high
current.

Who you kidding? 500A of PSSC only requires a Ze of around half an
ohm or less, common on many installations.

the point is that most trips aren't due to a zero ohm short

However the issue still stands at lower currents - say 180A
then... still enough to trip a B32 MCB in a head end.

The op can complicate the rewire,

How would be using a HRC fuse in a carrier in the head end CU
"complicating" it?

I didn't say it was. Running multiple cables to the shop & removing
the CU would be.

If it were my setup, I would run a submain from house CU to shop CU.
Leave the tails from that meter parked in a henley block and disconnect
the service. I would put a suitably sized HRC fuse in a fuse carrier in
the house CU for the submain. Minimal change and new materials. The OP
could then have the service to the shop terminated and let the supplier
do what they want with the meter and tails etc.

As a DIY job its the cost of the submain cable, and probably no more
than a couple of hours of work - would likely pay for itself in the
first year.

Its the same design, just a different choice of protective device.

but I'm not covinced there will be much gain.

There won't

so despite arguing you agree anyway

I am disagreeing with your statement about discrimination, since it was
basically wrong for the specific case of handling fault currents.

That is a separate issue from the solution that I propose to Graeme,
since in this circumstance I suspect he could cope with a lack of
discrimination given the circumstance he describes, and the likelihood
of needing to clear a fault in the first place is relatively low.
However since some mitigation can be designed in at no extra effort or
expense it would seem daft not to.

- hopefully there won't be many cases where fault current
protection needs to operate on the downstream installation in the
first place. If it does, and you trip multiple MCBs, then that may
be acceptable in the circumstances for an infrequent situation.

However that does not mean you can get away with glib "...show
that discrimination is normally obtained with fault currents"
statements! ;-)

40A MCB takes more i squared t than 32A & 5/6A MCBs to trip. With a
5/6A circuit problem you'll nearly always get discrimination. With
the 32A circuit you often will.

True for overloads, but not faults.

As I said you can complicate the job by losing the shop CU & adding
more cables, but the real life gain will be minimal.

No point since there are easier ways, and for that matter *in this
circumstance* Graeme may not care anyway.

So in short you do agree that it will generally discriminate on overcurrents, but not on dead shorts of zero ohms. Glad we cleared that up.


NT

On 29/06/2017 18:40, John Rumm wrote:


If it were my setup, I would run a submain from house CU to shop CU.
Leave the tails from that meter parked in a henley block and disconnect
the service. I would put a suitably sized HRC fuse in a fuse carrier in
the house CU for the submain. Minimal change and new materials. The OP
could then have the service to the shop terminated and let the supplier
do what they want with the meter and tails etc.

As a DIY job its the cost of the submain cable, and probably no more
than a couple of hours of work - would likely pay for itself in the
first year.


Just to add to that suggested set up - and I believe that it may be
important.

The old shop CU may well be include some form of main equipotential
bonding that is specific to the existing shop CU which would be lost if
the supplier physically removes their supply. You would have to run a
10mm earth along with the submain cable to incorporate any bonding back
to the house CU.


As for the fault current.

I'll give an example of a 40A MCB feeding a CU that has a 32A and a 6A
MCB. In this case I am going for a non RCD protected 40A MCB supplying a
RCD CU. I assume that this could be a realistic option for the OP
although I know very little about his setup.

The max ELI of a 40A MCB is 1.15 ohms and a current of 200A is needed to
trip the 40A MCB in a fault condition.

So the maximum Zs at the shop CU is 1.15ohms

Now a 32A MCB only needs 160A to trip in a fault condition and so can
have a maximum ELI of 1.44ohms.

Now what happens[1] if the shop CU has a Zs of 1.00ohms and a fault
occurs somewhere on the 32A circuit at a point where the Zs is less than
1.15ohms?

[1]Extra marks for working out the maximum 2.5mm ring circuit size that
will keep the maximum Zs under 1.15 ohms.



--
Adam

On Thursday, 29 June 2017 20:56:36 UTC+1, ARW wrote:
On 29/06/2017 18:40, John Rumm wrote:


If it were my setup, I would run a submain from house CU to shop CU.
Leave the tails from that meter parked in a henley block and disconnect
the service. I would put a suitably sized HRC fuse in a fuse carrier in
the house CU for the submain. Minimal change and new materials. The OP
could then have the service to the shop terminated and let the supplier
do what they want with the meter and tails etc.

As a DIY job its the cost of the submain cable, and probably no more
than a couple of hours of work - would likely pay for itself in the
first year.


Just to add to that suggested set up - and I believe that it may be
important.

The old shop CU may well be include some form of main equipotential
bonding that is specific to the existing shop CU which would be lost if
the supplier physically removes their supply. You would have to run a
10mm earth along with the submain cable to incorporate any bonding back
to the house CU.


As for the fault current.

I'll give an example of a 40A MCB feeding a CU that has a 32A and a 6A
MCB. In this case I am going for a non RCD protected 40A MCB supplying a
RCD CU. I assume that this could be a realistic option for the OP
although I know very little about his setup.

The max ELI of a 40A MCB is 1.15 ohms and a current of 200A is needed to
trip the 40A MCB in a fault condition.

So the maximum Zs at the shop CU is 1.15ohms

Now a 32A MCB only needs 160A to trip in a fault condition and so can
have a maximum ELI of 1.44ohms.

Now what happens[1] if the shop CU has a Zs of 1.00ohms and a fault
occurs somewhere on the 32A circuit at a point where the Zs is less than
1.15ohms?

[1]Extra marks for working out the maximum 2.5mm ring circuit size that
will keep the maximum Zs under 1.15 ohms.

It does strike me as missing the point once again. Most MCB trips are caused by overcurrent rather than a dead zero ohm short.


NT

On 29/06/2017 22:11, wrote:
On Thursday, 29 June 2017 20:56:36 UTC+1, ARW wrote:
On 29/06/2017 18:40, John Rumm wrote:


If it were my setup, I would run a submain from house CU to shop CU.
Leave the tails from that meter parked in a henley block and disconnect
the service. I would put a suitably sized HRC fuse in a fuse carrier in
the house CU for the submain. Minimal change and new materials. The OP
could then have the service to the shop terminated and let the supplier
do what they want with the meter and tails etc.

As a DIY job its the cost of the submain cable, and probably no more
than a couple of hours of work - would likely pay for itself in the
first year.


Just to add to that suggested set up - and I believe that it may be
important.

The old shop CU may well be include some form of main equipotential
bonding that is specific to the existing shop CU which would be lost if
the supplier physically removes their supply. You would have to run a
10mm earth along with the submain cable to incorporate any bonding back
to the house CU.


As for the fault current.

I'll give an example of a 40A MCB feeding a CU that has a 32A and a 6A
MCB. In this case I am going for a non RCD protected 40A MCB supplying a
RCD CU. I assume that this could be a realistic option for the OP
although I know very little about his setup.

The max ELI of a 40A MCB is 1.15 ohms and a current of 200A is needed to
trip the 40A MCB in a fault condition.

So the maximum Zs at the shop CU is 1.15ohms

Now a 32A MCB only needs 160A to trip in a fault condition and so can
have a maximum ELI of 1.44ohms.

Now what happens[1] if the shop CU has a Zs of 1.00ohms and a fault
occurs somewhere on the 32A circuit at a point where the Zs is less than
1.15ohms?

[1]Extra marks for working out the maximum 2.5mm ring circuit size that
will keep the maximum Zs under 1.15 ohms.

It does strike me as missing the point once again. Most MCB trips are caused by overcurrent rather than a dead zero ohm short.


Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."





--
Adam

On Thursday, 29 June 2017 23:33:21 UTC+1, ARW wrote:
On 29/06/2017 22:11, tabbypurr wrote:
On Thursday, 29 June 2017 20:56:36 UTC+1, ARW wrote:
On 29/06/2017 18:40, John Rumm wrote:


If it were my setup, I would run a submain from house CU to shop CU.
Leave the tails from that meter parked in a henley block and disconnect
the service. I would put a suitably sized HRC fuse in a fuse carrier in
the house CU for the submain. Minimal change and new materials. The OP
could then have the service to the shop terminated and let the supplier
do what they want with the meter and tails etc.

As a DIY job its the cost of the submain cable, and probably no more
than a couple of hours of work - would likely pay for itself in the
first year.


Just to add to that suggested set up - and I believe that it may be
important.

The old shop CU may well be include some form of main equipotential
bonding that is specific to the existing shop CU which would be lost if
the supplier physically removes their supply. You would have to run a
10mm earth along with the submain cable to incorporate any bonding back
to the house CU.


As for the fault current.

I'll give an example of a 40A MCB feeding a CU that has a 32A and a 6A
MCB. In this case I am going for a non RCD protected 40A MCB supplying a
RCD CU. I assume that this could be a realistic option for the OP
although I know very little about his setup.

The max ELI of a 40A MCB is 1.15 ohms and a current of 200A is needed to
trip the 40A MCB in a fault condition.

So the maximum Zs at the shop CU is 1.15ohms

Now a 32A MCB only needs 160A to trip in a fault condition and so can
have a maximum ELI of 1.44ohms.

Now what happens[1] if the shop CU has a Zs of 1.00ohms and a fault
occurs somewhere on the 32A circuit at a point where the Zs is less than
1.15ohms?

[1]Extra marks for working out the maximum 2.5mm ring circuit size that
will keep the maximum Zs under 1.15 ohms.

It does strike me as missing the point once again. Most MCB trips are caused by overcurrent rather than a dead zero ohm short.


Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.


NT

In message , John
Rumm writes

If it were my setup, I would run a submain from house CU to shop CU.
Leave the tails from that meter parked in a henley block and disconnect
the service. I would put a suitably sized HRC fuse in a fuse carrier in
the house CU for the submain. Minimal change and new materials. The OP
could then have the service to the shop terminated and let the supplier
do what they want with the meter and tails etc.

I'm inclined to follow John's advice, keeping in mind Adam's earthing
comment. Thanks all. I confess that at least some of the discussion is
above my head, and whilst I can do some of the job, I'll leave the
actual connections and disconnections to a pro.
--
Graeme

On 29/06/2017 22:11, wrote:
On Thursday, 29 June 2017 20:56:36 UTC+1, ARW wrote:
On 29/06/2017 18:40, John Rumm wrote:


If it were my setup, I would run a submain from house CU to shop CU.
Leave the tails from that meter parked in a henley block and disconnect
the service. I would put a suitably sized HRC fuse in a fuse carrier in
the house CU for the submain. Minimal change and new materials. The OP
could then have the service to the shop terminated and let the supplier
do what they want with the meter and tails etc.

As a DIY job its the cost of the submain cable, and probably no more
than a couple of hours of work - would likely pay for itself in the
first year.


Just to add to that suggested set up - and I believe that it may be
important.

The old shop CU may well be include some form of main equipotential
bonding that is specific to the existing shop CU which would be lost if
the supplier physically removes their supply. You would have to run a
10mm earth along with the submain cable to incorporate any bonding back
to the house CU.


As for the fault current.

I'll give an example of a 40A MCB feeding a CU that has a 32A and a 6A
MCB. In this case I am going for a non RCD protected 40A MCB supplying a
RCD CU. I assume that this could be a realistic option for the OP
although I know very little about his setup.

The max ELI of a 40A MCB is 1.15 ohms and a current of 200A is needed to
trip the 40A MCB in a fault condition.

So the maximum Zs at the shop CU is 1.15ohms

Now a 32A MCB only needs 160A to trip in a fault condition and so can
have a maximum ELI of 1.44ohms.

Now what happens[1] if the shop CU has a Zs of 1.00ohms and a fault
occurs somewhere on the 32A circuit at a point where the Zs is less than
1.15ohms?

[1]Extra marks for working out the maximum 2.5mm ring circuit size that
will keep the maximum Zs under 1.15 ohms.

It does strike me as missing the point once again. Most MCB trips are caused by overcurrent rather than a dead zero ohm short.

That also seems suspect to me.

One will normally select and design circuits to be adequate for their
expected use, so overloads should be the exception rather than the norm.
While a design can be fault tolerant, its much harder to design the
possibility one will occur at all out of it.

I can only think of one example where I have seen a MCB do a thermal
trip, but loads on the magnetic response. (for example, filament lamp
failures, faulty appliance, faulty lamp, good old fashioned cable damage
(watching someone chisel through a live cable!), drilling into one
myself and so on.



--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

On Friday, 30 June 2017 10:55:54 UTC+1, John Rumm wrote:
On 29/06/2017 22:11, tabbypurr wrote:
On Thursday, 29 June 2017 20:56:36 UTC+1, ARW wrote:
On 29/06/2017 18:40, John Rumm wrote:


If it were my setup, I would run a submain from house CU to shop CU.
Leave the tails from that meter parked in a henley block and disconnect
the service. I would put a suitably sized HRC fuse in a fuse carrier in
the house CU for the submain. Minimal change and new materials. The OP
could then have the service to the shop terminated and let the supplier
do what they want with the meter and tails etc.

As a DIY job its the cost of the submain cable, and probably no more
than a couple of hours of work - would likely pay for itself in the
first year.


Just to add to that suggested set up - and I believe that it may be
important.

The old shop CU may well be include some form of main equipotential
bonding that is specific to the existing shop CU which would be lost if
the supplier physically removes their supply. You would have to run a
10mm earth along with the submain cable to incorporate any bonding back
to the house CU.


As for the fault current.

I'll give an example of a 40A MCB feeding a CU that has a 32A and a 6A
MCB. In this case I am going for a non RCD protected 40A MCB supplying a
RCD CU. I assume that this could be a realistic option for the OP
although I know very little about his setup.

The max ELI of a 40A MCB is 1.15 ohms and a current of 200A is needed to
trip the 40A MCB in a fault condition.

So the maximum Zs at the shop CU is 1.15ohms

Now a 32A MCB only needs 160A to trip in a fault condition and so can
have a maximum ELI of 1.44ohms.

Now what happens[1] if the shop CU has a Zs of 1.00ohms and a fault
occurs somewhere on the 32A circuit at a point where the Zs is less than
1.15ohms?

[1]Extra marks for working out the maximum 2.5mm ring circuit size that
will keep the maximum Zs under 1.15 ohms.

It does strike me as missing the point once again. Most MCB trips are caused by overcurrent rather than a dead zero ohm short.

That also seems suspect to me.

One will normally select and design circuits to be adequate for their
expected use, so overloads should be the exception rather than the norm.

That can be true, depending how you define overload. Eg it's normal for rings to supply much more than 32A, OTOH circuits shouldn't regularly trip.

While a design can be fault tolerant, its much harder to design the
possibility one will occur at all out of it.

impossible

I can only think of one example where I have seen a MCB do a thermal
trip, but loads on the magnetic response. (for example, filament lamp
failures, faulty appliance, faulty lamp, good old fashioned cable damage
(watching someone chisel through a live cable!), drilling into one
myself and so on.

Excess current that is not a dead zero ohm short can trip them in either thermal or magnetic mode. A failament lamp failure or faulty motor are good examples of this. Drilling into a 32A cable is of course more a dead short.


NT

On 30/06/2017 08:05, Graeme wrote:
In message , John
Rumm writes

If it were my setup, I would run a submain from house CU to shop CU.
Leave the tails from that meter parked in a henley block and disconnect
the service. I would put a suitably sized HRC fuse in a fuse carrier in
the house CU for the submain. Minimal change and new materials. The OP
could then have the service to the shop terminated and let the supplier
do what they want with the meter and tails etc.

I'm inclined to follow John's advice, keeping in mind Adam's earthing
comment. Thanks all. I confess that at least some of the discussion is
above my head, and whilst I can do some of the job, I'll leave the
actual connections and disconnections to a pro.

If you can decide on the cable required, and lay that in, it will cut
down the amount of work required.

Regarding the main equipotential bonds, there are some things worth
checking. Firstly find out what earthing system[1] is used in the shop
and the house. For example if it's TN-C-S (PME) then you will need to
ensure that the house's main equipotential zone is extended[2] to the
new CU. Also does the shop have other independent incoming services?
(e.g. water, gas etc). If it does, chances are those were previously
included in the main bonding to the shop's earthing system, but you will
need to check.

[1] http://wiki.diyfaq.org.uk/index.php/Earthing_Types

[2] Although in this case discussing feeds to outbuildings, much of what
is described could apply he

http://wiki.diyfaq.org.uk/index.php/...tent ial_Zone

--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

In message , John
Rumm writes

Regarding the main equipotential bonds, there are some things worth
checking. Firstly find out what earthing system[1] is used in the shop
and the house. For example if it's TN-C-S (PME) then you will need to
ensure that the house's main equipotential zone is extended[2] to the
new CU.

This, I confess, is where I get lost. Several times, over a good few
years, I have read the Wiki article and looked at the incoming supply,
but am still none the wiser. I'll need to take good photos of both CUs
and post them here, in the hope that someone will be able to see exactly
which system is in use. Alternatively, if I find a proper electrician
to handle the actual connections, he will know.

--
Graeme

In article ,
wrote:
It does strike me as missing the point once again. Most MCB trips
are caused by overcurrent rather than a dead zero ohm short.


Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.

Means you must have experienced lots of bad design, then?

--
*Rehab is for quitters

Dave Plowman London SW
To e-mail, change noise into sound.

On 01/07/2017 07:37, Graeme wrote:
In message , John
Rumm writes

Regarding the main equipotential bonds, there are some things worth
checking. Firstly find out what earthing system[1] is used in the shop
and the house. For example if it's TN-C-S (PME) then you will need to
ensure that the house's main equipotential zone is extended[2] to the
new CU.

This, I confess, is where I get lost. Several times, over a good few
years, I have read the Wiki article and looked at the incoming supply,
but am still none the wiser. I'll need to take good photos of both CUs
and post them here, in the hope that someone will be able to see exactly
which system is in use. Alternatively, if I find a proper electrician
to handle the actual connections, he will know.


Yup photos would be good. Especially if they are of installs that are
different in look from the ones we already have - then we can pinch em
to update the article with ;-)


--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

On Saturday, 1 July 2017 11:59:10 UTC+1, Dave Plowman (News) wrote:
In article ,
tabbypurr wrote:

It does strike me as missing the point once again. Most MCB trips
are caused by overcurrent rather than a dead zero ohm short.


Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.

Means you must have experienced lots of bad design, then?

no.


NT

In article ,
wrote:
On Saturday, 1 July 2017 11:59:10 UTC+1, Dave Plowman (News) wrote:
In article ,
tabbypurr wrote:

It does strike me as missing the point once again. Most MCB trips
are caused by overcurrent rather than a dead zero ohm short.


Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.

Means you must have experienced lots of bad design, then?

How else do you explain regular over current trips, then?

--
*Experience is something you don't get until just after you need it.*

Dave Plowman London SW
To e-mail, change noise into sound.

In article , Dave Plowman (News)
wrote:
In article ,
wrote:
On Saturday, 1 July 2017 11:59:10 UTC+1, Dave Plowman (News) wrote:
In article ,
tabbypurr wrote:

It does strike me as missing the point once again. Most MCB
trips are caused by overcurrent rather than a dead zero ohm
short.


Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.

Means you must have experienced lots of bad design, then?

How else do you explain regular over current trips, then?

Halogen lamps failing is very common cause. Mind you, they often take out
a domestic dimmer, too.

--
from KT24 in Surrey, England

On Sunday, 2 July 2017 11:17:02 UTC+1, Dave Plowman (News) wrote:
In article ,
tabbypurr wrote:
On Saturday, 1 July 2017 11:59:10 UTC+1, Dave Plowman (News) wrote:
In article ,
tabbypurr wrote:

It does strike me as missing the point once again. Most MCB trips
are caused by overcurrent rather than a dead zero ohm short.


Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.

Means you must have experienced lots of bad design, then?

No.

How else do you explain regular over current trips, then?

Where do you get regular overcurrent trips from? If you keep getting them, something is wrong.
Sometimes the things you say are a bit odd.


NT

In article ,
charles wrote:
In article , Dave Plowman (News)
wrote:
In article ,
wrote:
On Saturday, 1 July 2017 11:59:10 UTC+1, Dave Plowman (News) wrote:
In article ,
tabbypurr wrote:

It does strike me as missing the point once again. Most MCB
trips are caused by overcurrent rather than a dead zero ohm
short.


Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.

Means you must have experienced lots of bad design, then?

How else do you explain regular over current trips, then?

Halogen lamps failing is very common cause. Mind you, they often take
out a domestic dimmer, too.

I'd call that a short.

--
*The most wasted day of all is one in which we have not laughed.*

Dave Plowman London SW
To e-mail, change noise into sound.

On 30/06/2017 01:56, wrote:

Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.



"Limited experience" would be a better.


--
Adam

On 02/07/17 17:45, ARW wrote:
On 30/06/2017 01:56, wrote:

Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.



"Limited experience" would be a better.


well since a dead zero ohm short is a physical impossibility, one has to
say he is right.


--
Microsoft : the best reason to go to Linux that ever existed.

On 02/07/2017 18:29, The Natural Philosopher wrote:
On 02/07/17 17:45, ARW wrote:
On 30/06/2017 01:56, wrote:

Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.



"Limited experience" would be a better.


well since a dead zero ohm short is a physical impossibility, one has to
say he is right.



It's zero for BS7671 regs and calculations.

--
Adam

On 02/07/17 18:34, ARW wrote:
On 02/07/2017 18:29, The Natural Philosopher wrote:
On 02/07/17 17:45, ARW wrote:
On 30/06/2017 01:56, wrote:

Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.



"Limited experience" would be a better.


well since a dead zero ohm short is a physical impossibility, one has to
say he is right.



It's zero for BS7671 regs and calculations.

Then they are meaningless


--
A lie can travel halfway around the world while the truth is putting on
its shoes.

On Sunday, 2 July 2017 15:33:50 UTC+1, Dave Plowman (News) wrote:
In article ,
charles wrote:
In article , Dave Plowman (News)
wrote:
In article ,
tabbypurr wrote:
On Saturday, 1 July 2017 11:59:10 UTC+1, Dave Plowman (News) wrote:
In article ,
tabbypurr wrote:

It does strike me as missing the point once again. Most MCB
trips are caused by overcurrent rather than a dead zero ohm
short.


Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.

Means you must have experienced lots of bad design, then?

How else do you explain regular over current trips, then?

Halogen lamps failing is very common cause. Mind you, they often take
out a domestic dimmer, too.

I'd call that a short.

So you think they conduct 1000s of amps during failure. They don't.


NT

In article ,
wrote:
On Sunday, 2 July 2017 15:33:50 UTC+1, Dave Plowman (News) wrote:
In article ,
charles wrote:
In article , Dave Plowman (News)
wrote:
In article ,
tabbypurr wrote:
On Saturday, 1 July 2017 11:59:10 UTC+1, Dave Plowman (News)
wrote:
In article
,
tabbypurr wrote:

It does strike me as missing the point once again. Most MCB
trips are caused by overcurrent rather than a dead zero ohm
short.


Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.

Means you must have experienced lots of bad design, then?

How else do you explain regular over current trips, then?

Halogen lamps failing is very common cause. Mind you, they often take
out a domestic dimmer, too.

I'd call that a short.

So you think they conduct 1000s of amps during failure. They don't.

Ah. So you wish to define how much current flows during a short circuit?

This should be intersting...

--
*My wife has a slight impediment in her speech. She stops to breathe.

Dave Plowman London SW
To e-mail, change noise into sound.

On 30/06/2017 12:38, John Rumm wrote:
On 30/06/2017 08:05, Graeme wrote:
In message , John
Rumm writes

If it were my setup, I would run a submain from house CU to shop CU.
Leave the tails from that meter parked in a henley block and disconnect
the service. I would put a suitably sized HRC fuse in a fuse carrier in
the house CU for the submain. Minimal change and new materials. The OP
could then have the service to the shop terminated and let the supplier
do what they want with the meter and tails etc.

I'm inclined to follow John's advice, keeping in mind Adam's earthing
comment. Thanks all. I confess that at least some of the discussion is
above my head, and whilst I can do some of the job, I'll leave the
actual connections and disconnections to a pro.

If you can decide on the cable required, and lay that in, it will cut
down the amount of work required.

Regarding the main equipotential bonds, there are some things worth
checking. Firstly find out what earthing system[1] is used in the shop
and the house. For example if it's TN-C-S (PME) then you will need to
ensure that the house's main equipotential zone is extended[2] to the
new CU. Also does the shop have other independent incoming services?
(e.g. water, gas etc). If it does, chances are those were previously
included in the main bonding to the shop's earthing system, but you will
need to check.

Even on a PME supply you only need to extend the equipotential zone to
the shop CU if there is something that needs bonding. I was suggesting
that that shop may have it's own extraneous conductive parts.


--
Adam

wrote:

On Sunday, 2 July 2017 15:33:50 UTC+1, Dave Plowman (News) wrote:
In article ,
charles wrote:
In article , Dave Plowman (News)
wrote:
In article ,
tabbypurr wrote:
On Saturday, 1 July 2017 11:59:10 UTC+1, Dave Plowman (News) wrote:
In article ,
tabbypurr wrote:

It does strike me as missing the point once again. Most MCB
trips are caused by overcurrent rather than a dead zero ohm
short.


Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.

Means you must have experienced lots of bad design, then?

How else do you explain regular over current trips, then?

Halogen lamps failing is very common cause. Mind you, they often take
out a domestic dimmer, too.

I'd call that a short.

So you think they conduct 1000s of amps during failure. They don't.


NT

I'm pretty sure they conduct 100s of amps, they certainly arc, and they
trip MCBs in a fraction of a second. At least they do in this house.

--

Roger Hayter

The Natural Philosopher wrote:

On 02/07/17 18:34, ARW wrote:
On 02/07/2017 18:29, The Natural Philosopher wrote:
On 02/07/17 17:45, ARW wrote:
On 30/06/2017 01:56, wrote:

Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.



"Limited experience" would be a better.


well since a dead zero ohm short is a physical impossibility, one has to
say he is right.



It's zero for BS7671 regs and calculations.

Then they are meaningless

No, you just use the resistance of the existing supply and wiring
depending where you introduce the theoretical zero ohm short.

--

Roger Hayter

On Monday, 3 July 2017 00:17:43 UTC+1, Dave Plowman (News) wrote:
In article ,
tabbypurr wrote:
On Sunday, 2 July 2017 15:33:50 UTC+1, Dave Plowman (News) wrote:
In article ,
charles wrote:
In article , Dave Plowman (News)
wrote:
In article ,
tabbypurr wrote:
On Saturday, 1 July 2017 11:59:10 UTC+1, Dave Plowman (News)
wrote:
In article
,
tabbypurr wrote:

It does strike me as missing the point once again. Most MCB
trips are caused by overcurrent rather than a dead zero ohm
short.


Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.

Means you must have experienced lots of bad design, then?

No.
- you keep snipping that. Weird.

How else do you explain regular over current trips, then?

Halogen lamps failing is very common cause. Mind you, they often take
out a domestic dimmer, too.

I'd call that a short.

So you think they conduct 1000s of amps during failure. They don't.

Ah. So you wish to define how much current flows during a short circuit?

This should be intersting...

A lamp arcover is not a short circuit.


NT

In article ,
Roger Hayter wrote:
Halogen lamps failing is very common cause. Mind you, they often
take out a domestic dimmer, too.

I'd call that a short.

So you think they conduct 1000s of amps during failure. They don't.


NT

I'm pretty sure they conduct 100s of amps, they certainly arc, and they
trip MCBs in a fraction of a second. At least they do in this house.

Same here.

--
*Always remember you're unique, just like everyone else.

Dave Plowman London SW
To e-mail, change noise into sound.

In article ,
wrote:
So you think they conduct 1000s of amps during failure. They don't.

Ah. So you wish to define how much current flows during a short
circuit?

This should be intersting...

A lamp arcover is not a short circuit.

Who said they arc over? What usually happens here is the MCB just trips.

But I asked for your definition of a short circuit against an overload.
Since you seem to know the difference.

--
*The most wasted day of all is one in which we have not laughed.*

Dave Plowman London SW
To e-mail, change noise into sound.

On 03/07/2017 09:14, Roger Hayter wrote:
wrote:

On Sunday, 2 July 2017 15:33:50 UTC+1, Dave Plowman (News) wrote:
In article ,
charles wrote:
In article , Dave Plowman (News)
wrote:
In article ,
tabbypurr wrote:
On Saturday, 1 July 2017 11:59:10 UTC+1, Dave Plowman (News) wrote:
In article ,
tabbypurr wrote:

It does strike me as missing the point once again. Most MCB
trips are caused by overcurrent rather than a dead zero ohm
short.


Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.

Means you must have experienced lots of bad design, then?

How else do you explain regular over current trips, then?

Halogen lamps failing is very common cause. Mind you, they often take
out a domestic dimmer, too.

I'd call that a short.

So you think they conduct 1000s of amps during failure. They don't.


NT

I'm pretty sure they conduct 100s of amps, they certainly arc, and they
trip MCBs in a fraction of a second. At least they do in this house.

Indeed, and for a normal B6 MCB any current over 30A should do that.
(changing to a C6 (60A magnetic trip threshold) will eliminate some (but
not necessarily all) filament lamp trips IME.


--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

On 02/07/2017 18:58, The Natural Philosopher wrote:
On 02/07/17 18:34, ARW wrote:
On 02/07/2017 18:29, The Natural Philosopher wrote:
On 02/07/17 17:45, ARW wrote:
On 30/06/2017 01:56, wrote:

Care to state the reference for "Most MCB trips are caused by
overcurrent rather than a dead zero ohm short."

experience.



"Limited experience" would be a better.


well since a dead zero ohm short is a physical impossibility, one has to
say he is right.



It's zero for BS7671 regs and calculations.

Then they are meaningless

Not when you calculate (or measure) the Prospective Fault Current or
Prospective Short Circuit Current, values of which will be dictated by
the fixed wiring and the nature of the supply and earthing system.

You can't meaningfully quantify the resistance of the fault itself since
at design time its unknown.


--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

On Monday, 3 July 2017 09:51:38 UTC+1, Dave Plowman (News) wrote:
In article ,
tabbypurr wrote:

So you think they conduct 1000s of amps during failure. They don't.

Ah. So you wish to define how much current flows during a short
circuit?

This should be intersting...

A lamp arcover is not a short circuit.

Who said they arc over? What usually happens here is the MCB just trips.

Pretty much every lamp expert that has written about the issue.


But I asked for your definition of a short circuit against an overload.
Since you seem to know the difference.

Of course I do. You appear not to know what you're talking about again.

In article ,
wrote:
On Monday, 3 July 2017 09:51:38 UTC+1, Dave Plowman (News) wrote:
In article ,
tabbypurr wrote:

So you think they conduct 1000s of amps during failure. They don't.

Ah. So you wish to define how much current flows during a short
circuit?

This should be intersting...

A lamp arcover is not a short circuit.

Who said they arc over? What usually happens here is the MCB just
trips.

Pretty much every lamp expert that has written about the issue.

They may or may not. Depending on the lamp design.


But I asked for your definition of a short circuit against an overload.
Since you seem to know the difference.

Of course I do. You appear not to know what you're talking about again.

So please give your definitions. A simple enough thing for you to do
surely? Except you've opened mouth before engaging brain again.

Rather obviously any short circuit is an overload. But not every overload
is a short circuit.

--
*You sound reasonable......time to up my medication

Dave Plowman London SW
To e-mail, change noise into sound.

On Monday, 3 July 2017 16:03:04 UTC+1, Dave Plowman (News) wrote:
In article ,
tabbypurr wrote:
On Monday, 3 July 2017 09:51:38 UTC+1, Dave Plowman (News) wrote:
In article ,
tabbypurr wrote:

So you think they conduct 1000s of amps during failure. They don't.

Ah. So you wish to define how much current flows during a short
circuit?

This should be intersting...

A lamp arcover is not a short circuit.

Who said they arc over? What usually happens here is the MCB just
trips.

Pretty much every lamp expert that has written about the issue.

They may or may not. Depending on the lamp design.


But I asked for your definition of a short circuit against an overload.
Since you seem to know the difference.

Of course I do. You appear not to know what you're talking about again.

So please give your definitions. A simple enough thing for you to do
surely? Except you've opened mouth before engaging brain again.

Rather obviously any short circuit is an overload. But not every overload
is a short circuit.

you're hovering right on the edge of the plonk filter.

In message , John
Rumm writes
On 01/07/2017 07:37, Graeme wrote:

This, I confess, is where I get lost. Several times, over a good few
years, I have re

ad the Wiki article and looked at the incoming supply,
but am still none the wiser. I'll need to take good photos of both CUs
and post them here, in the hope that someone will be able to see exactly
which system is in use.

Yup photos would be good. Especially if they are of installs that are
different in look from the ones we already have - then we can pinch em
to update the article with ;-)

Right. Photos here :

http://www.binnsroad.co.uk/misc/leccy/index.html

Thanks!
--
Graeme

On 04/07/2017 11:17, Graeme wrote:
In message , John
Rumm writes
On 01/07/2017 07:37, Graeme wrote:

This, I confess, is where I get lost. Several times, over a good few
years, I have re

ad the Wiki article and looked at the incoming supply,
but am still none the wiser. I'll need to take good photos of both CUs
and post them here, in the hope that someone will be able to see exactly
which system is in use.

Yup photos would be good. Especially if they are of installs that are
different in look from the ones we already have - then we can pinch em
to update the article with ;-)

Right. Photos here :

http://www.binnsroad.co.uk/misc/leccy/index.html

Oh, nice job. ta.

(I think your description probably wants a "(far left)" where it
currently has a "(far right)" ;-)

That looks like its TN, now the next question is TN-S or TN-C-S - hard
to tell from just the picture. If you were to take out those two screws
on the rectangular faceplate just under the main fuse, and have a peek
in there, that would probably tell you.

For example:

http://wiki.diyfaq.org.uk/index.php/...PMECutout3.jpg

Shows what could be a TN-C-S cutout, but the link between the neutral
and earth blocks on the right is missing, and a discrete earth
connection from the armour of the split concentric cable is connected to
the bottom of the earthing block on the far right.

Compare with:

http://wiki.diyfaq.org.uk/index.php/...ePMECutout.jpg

That has the link in place to join the earth terminal to the incoming
neutral at the cutout.

I would guess yours will be like the former (i.e. TN-S)

Adam may be able to spot more from the photos.


--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/