Connecting up an ASHP

[PeterW]

Quick one then....

 

ASHP outside unit requires a 25A breaker but a max 3.6kw supply. Indoor controls to external unit is 4 core 0.75mm² control cable which will be SY Multiflex direct to the ASHP

 

Current plan for power is RCBO > 2,5mm SWA > 32A TP&N Isolator > ???? > ASHP.

 

Question is, what is the best solution for the last metre from the isolator to the ASHP..? 2.5mm² SY-Multi seems the best approach for flexibility but does it need to be SWA..?

 

@JSHarris what did you use..??

 

 

[Steptoe]

Armour flex is what I fit, done 100s of them, (YY or SY depending)

Make sure you gland it properly with pig tails, its not that hard really with the proper glands.

[Steptoe]

Oh, and check if it needs a C type breaker, depending on length of the run, you might need 4mm if that's the case.

[Nickfromwales]

4mm if it were me. 

SY all the way for ease and uniformity. 

[PeterW]

7 hours ago, Steptoe said:

Oh, and check if it needs a C type breaker, depending on length of the run, you might need 4mm if that's the case.

 

Ta

 

max load is 3.7kw so on a 7.5m run a 1.5mm would be fine hence the step up to 2.5mm for safety. 

 

@Nickfromwales 4mm seems overkill ..? Do you have shares in Doncaster Cables ..??!

[Nickfromwales]

At the stated 25a breaker rating there may well be shunt resistance allowed in that. 1.5mm2 is good for 18a nominal so 2.5mm2 would be the minimum.  

C curve breaker is prob best too. 

The cable needs to exceed the current rating of the breaker so, if 25a, then 2.5mm2 is very close. That's shows an estimated 5750w peak threshold, but they've obviously allowed some headroom. Is the 3700w stated as running or peak ( shunt ) ?

Remember that you should allow for over current as well as a short circuit. That's why I'd use 4mm. ?

[Jeremy Harris]

I used 3 core 2.5mm² flex, inside armoured flex conduit.  Our heat pump had punch outs to accept an armoured conduit glands, and I ran the conduit through the wall (foamed in place and sealed) and then the cable ran directly into the back of a 45mm single back box with a single gang, DP switch on the wall of the utility room, right behind where the heat pump is outside.  I also ran another armoured duct though the wall alongside, to take the low voltage control cable, that also terminates via a gland at the heat pump.  2.5mm² T&E runs from the DP switch up to the RCBO in the CU on the floor above (our CU is an all-RCBO one, I can't be done with the daft idea of 17th Ed split boxes, it makes far more sense to me to have an RCBO on every circuit).

Edited January 23, 2017 by JSHarris

[Steptoe]

46 minutes ago, JSHarris said:

I used 3 core 2.5mm² flex, inside armoured flex conduit.  Our heat pump had punch outs to accept an armoured conduit glands, and I ran the conduit through the wall (foamed in place and sealed) and then the cable ran directly into the back of a 45mm single back box with a single gang, DP switch on the wall of the utility room, right behind where the heat pump is outside.  I also ran another armoured duct though the wall alongside, to take the low voltage control cable, that also terminates via a gland at the heat pump.  2.5mm² T&E runs from the DP switch up to the RCBO in the CU on the floor above (our CU is an all-RCBO one, I can't be done with the daft idea of 17th Ed split boxes, it makes far more sense to me to have an RCBO on every circuit).

What's a 17th es split box,?

If your RCBOs are SP then you have less compliance with 17th edition amd3 than a dual RCD CU, and they are not in total compliance either , but generally accepted as the norm. Folks do tend to fit RCBOs tho due to being miseducated about their benefits, and not understanding the drawbacks or dangers of the neutral not being disconnected,

Totally non compliant in TT, and just plain lethal in TN with reverse polarity or dropped neutral.

[ProDave]

What I am finding up here is most people totally over state the circuit breaker requirements for a heat pump. Most have their heat umps fitted at an inflated price by an mcs approved contractor so they can claim the RHI. It is the installer that usually gives me the spec and it's quite normal for them to specify a C40 breaker, so I install an isolator with appropriate cable for that, and they connect from it with a bit of 2.5mm flex, and the unit turns out to have a max load of a little over 3KW.  I think they are stuck in the dark ages up here when heat pumps didn't have variable speed drive and soft start, so had a huge start up surge.
 

[Jeremy Harris]

22 hours ago, Steptoe said:

What's a 17th es split box,?

If your RCBOs are SP then you have less compliance with 17th edition amd3 than a dual RCD CU, and they are not in total compliance either , but generally accepted as the norm. Folks do tend to fit RCBOs tho due to being miseducated about their benefits, and not understanding the drawbacks or dangers of the neutral not being disconnected,

Totally non compliant in TT, and just plain lethal in TN with reverse polarity or dropped neutral.

 

The standard split box with two DP RCDs and MCBs for all the rest is what I'd generally refer to as a "standard 17th Ed CU". 

 

I have a DP RCD master, plus SP RCBOs on each circuit, so am compliant, as the DP RCD master will pick up any odd ball faults.  We're TN-C-S in the house, and TT at the workshop CU, just because I didn't like the idea of exporting the earth out there without a local earth rod in addition to lower the earth impedance a bit there, as there's machine tools in there (or will be).  The outdoor CU is all DP RCBOs (for the borehole pump, treatment plant, and car charge point feeds) and the workshop CU is all DP RCBOs too.

 

I fitted RCBOs on every circuit, because 99% of the time they will pick up any common fault and just trip that single circuit, rather than half the box.  I couldn't see the advantage in the split box, TBH, losing half the house because of a single fault seemed little better than the old 16th Ed (and earlier) system of losing the whole house.

 

WRG to heat pumps, the the requirements depend very much on the type of heat pump.  A non-inverter type has to be treated for what it is, a high power motor starting under load, and so needs both hefty cable, a fairly high rating on the breaker and a slow response time, like a band C, to cope with the starting surge current.  On the other hand, if you have an inverter controlled unit like ours, although it's a 7 kW peak output unit, the max current is only around 10 A, and that current is only reached about 30 seconds after turn on, as the unit slowly ramps up from a standstill to full power.  I have mine on a 16 A type B RCBO.

 

Edited January 24, 2017 by JSHarris
Error in terminology - my fault entirely The house was TT, is now TN-C-S, the workshop has always been TT

[PeterW]

Its an inverter ASHP with a max current of 15.7A according to the plate. 

 

A 10m run of SWA clipped to a wall is fine under 17th at 1.5mm. So going up to 2.5mm would work fine, but not quite sure it's a C-curve as that's 5-10i from my reading and an inverter starts slow anyway so I would expect the 3-5i of a B-curve will be adequate ..?

[Jeremy Harris]

B curve is fine for an inverter unit, as they are all very slow start, with no current surge at all.  If the cable is clipped to a wall it doesn't need to be SWA, NYY-J is OK, but the section going to the heat pump needs to be flexible, as they vibrate slightly, and that rules out NYY-J for the final section, as it's not flexible, it's solid conductor.  That's why I used flex in armoured conduit.  You can use flex in rigid conduit along the wall, switching to flexible conduit where it goes into the ASHP, as long as it complies with the de-rating conditions.  Makes for a neater looking install in my view, having a dead straight run of conduit, rather than cable clipped to the wall.

[Steptoe]

46 minutes ago, JSHarris said:

 

The standard split box with two DP RCBOs and MCBs for all the rest is what I'd generally refer to as a "standard 17th Ed CU". 

 

I have a DP RCD master, plus SP RCBOs on each circuit, so am compliant, as the DP RCD master will pick up any odd ball faults.  We're TN-C in the house, and TN-C-S at the workshop CU, just because I didn't like the idea of exporting the earth out there without a local earth rod in addition to lower the earth impedance a bit there, as there's machine tools in there (or will be).  The outdoor CU is all DP RCBOs (for the borehole pump, treatment plant, and car charge point feeds) and the workshop CU is all DP RCBOs too.

 

I fitted RCBOs on every circuit, because 99% of the time they will pick up any common fault and just trip that single circuit, rather than half the box.  I couldn't see the advantage in the split box, TBH, losing half the house because of a single fault seemed little better than the old 16th Ed (and earlier) system of losing the whole house.

 

WRG to heat pumps, the the requirements depend very much on the type of heat pump.  A non-inverter type has to be treated for what it is, a high power motor starting under load, and so needs both hefty cable, a fairly high rating on the breaker and a slow response time, like a band C, to cope with the starting surge current.  On the other hand, if you have an inverter controlled unit like ours, although it's a 7 kW peak output unit, the max current is only around 10 A, and that current is only reached about 30 seconds after turn on, as the unit slowly ramps up from a standstill to full power.  I have mine on a 16 A type B RCBO.

 

I'm nor sure you have the correct terminology,

TN-C is not permitted in the UK,

So its either TN-C-S or TN-S, and unless you have a separate supply to the workshop, it must be TN-S, by default, although as its a submain, then generally regarded as such and the source is the type supply.

Also, putting a rod at the remote location has absolutely nothing to so with lowering the impedance, rods are measures using resistance, but again, that's totally irelevany here as well, mostly, you should also have a rod at the incoming supply btw,

The reason for the rod is to ensure the potential of the earth is the same potential as the general mass of earth, its something a lot of 'electricians' don't understand either, so your misunderstanding of it is forgiveable, but you really need to rod the incomer , and, preferably, where the submain exits the EZ as well,

Is your upfront RCD a 100mA TD,? in order to get discrepancy,.? 

 

 

[Jeremy Harris]

You're right, my error, I have TN-C-S at the main incomer - what most still call PME - where the PE is imported with the incoming supply, as the neutral.  At the workshop, I reinforced the earth with a rod to get the impedance down and get the slab ground closer to the local incomer ground potential under fast fault conditions.  Although it's only around 30m of 6mm² SWA feeding the workshop, that has a  bit of inductance that ups the impedance slightly to fast fault events.  I've added the additional earth rod out there really for equipotential ground reasons more than anything else.  No need for a rod at the incomer, we're very close to the sub-station, with a 90mm² 3 phase feed directly under my meter cabinet (so less that 2m of coaxial company cable from that to the head).

 

Yes, the upfront master RCD  leakage is rated above the smaller RCBOs, so I get the right trip sequences, and I accept that the higher leakage current is slightly less safe as a consequence.  TBH, given the way the house is wired, it's hard to come up with a way of getting the sort of failure mode where just SP RCBOs wouldn't trip for any feasible leakage fault. 

 

I can understand the dropped neutral failure issue, especially on some older lighting circuits, but realistically, how likely is that as a fault condition now?   In my view it was all about balancing risk, probability of failure and convenience.  Most faults now seem to be earth leakage faults, rather than overloads (I guess that's because overloads get picked up at the domestic equipment end 99% of the time, now)  and, provided no one buggers about with the fixed wiring, then the SP RCBO offers practically the same protection as a DP one, and takes up less space (although I see there are single width DP RCBOs about now, at a price).

 

 

 

 

Edited January 23, 2017 by JSHarris

[Steptoe]

DP RCBOs in single gang can be had for less than the price of some SP ones .

If you are that close to the Tx your PFC may be a concern,,,,,

OK, slight misconception about PME and TN-C-S ,

PME must be TN-C-S ,

TN-C-S is not necessarily PME,

although it should be,

and if you do NOT have a rod at the incoming point, then you do NOT have PME,! FACT,!

 no matter what you or anyone else says, PME must be rodded at point of entry. 

[jack]

To help those who might be reading this in the future (and me, if I'm honest) and trying to figure out what's important, is this about the formal definition of particular terminology rather than whether there's an issue with Jeremy's installation?  Eg, in the absence of a rod but being really close to the substation, is the issue that it doesn't meet the formal definition of PME rather than there being a potential functional problem?

 

Apologies if I sound like I don't know what I'm talking about (possibly because I don't!), but I'm mindful that that the thread is drifting away from the original question and into an area that might not be of practical concern for future readers.

 

Pardon me if my ignorance isn't helping!

[Guest Alphonsox]

The following link may help clarify things http://electrical.theiet.org/wiring-matters/16/earthing-questions.cfm?type=pdf

I didn't think PME had a formal definition in the eyes of the regs and being rarely used nowadays as synonym for TN-C-S

[Steptoe]

Sorry @jack

Being so close to the substation brings with it the possibility of the prospective fault current being extremely high, although the DNO will usually state they have measures in place to stop it rising above 16kA, most domestic consumer units are rated at 16.5 , so that's OK, and the main cutout fuse will be rated much higher, the issue can arise with domestic MCBs only being rated at (usually) 6kA, you are now reliant on the main incoming fuse of probably 80a and rated at 88kA to clear the maximum fault current possible.

 

As for the PME and no rods in place,

Jeremy's neither meets the formal, or practical definition of PME, although he has sort of (for want of a better word), bodged a bit of a half headed attempt at it , even though he thought he was doing it for an entirely different reason.

this can be a serious issue under certain fault conditions, lethal actually, especially when combined with SP protective devices, remembering that 50mA can kill, and that's why the 30mA threshold. 

100mA devices are only there to protect against fire, (etc) , not to preserve life.

I done a drawing a few years back of simplified earthing basics, I'll look it out.

[Jeremy Harris]

12 minutes ago, jack said:

To help those who might be reading this in the future (and me, if I'm honest) and trying to figure out what's important, is this about the formal definition of particular terminology rather than whether there's an issue with Jeremy's installation?  Eg, in the absence of a rod but being really close to the substation, is the issue that it doesn't meet the formal definition of PME rather than there being a potential functional problem?

 

Apologies if I sound like I don't know what I'm talking about (possibly because I don't!), but I'm mindful that that the thread is drifting away from the original question and into an area that might not be of practical concern for future readers.

 

Pardon me if my ignorance isn't helping!

 

To be frank, it's getting a bit pedantic about terminology that changed as a consequence of harmonisation.  In general, we had two earthing schemes in common use pre-harmonisation.  One was where the earth was "imported", i.e. came in with the incoming supply cable, and one where the earth was supplied only by a local earth rod, near the point of entry to the house.  The former was known as PME, Protective Multiple Earth (a term that shouldn't now been used, as it hasn't been in the regs for years) and was normal in urban and built up areas.  Local earthing was the norm in rural areas.

 

Come harmonisation and we adopted EU-wide standards, with letter codes, set out in the link in the post above.  The snag is 99% of electricians STILL use PME as a term, hence the confusion.  Strictly speaking most installations are either TT (seperate earth rod as the only protective earth) or TN-C-S as described in that link Alphonsox gave.

[Steptoe]

I don't see it being pedantic at all, there are 4 totally different earthing systems in general use,

 

Here's a basic drawing , you can clearly see the differences, so how can anyone possibly say its being pedantic, trying to say they are the same is just being ignorant of the facts.

 

 

 

[Guest Alphonsox]

I don't think that is correct. I've just had a TN-C-S system installed and it looks like what you have labeled as PME. i.e a supply earth followed by additional earth connections on the run to the house.

[Jeremy Harris]

18 minutes ago, Steptoe said:

Sorry @jack

Being so close to the substation brings with it the possibility of the prospective fault current being extremely high, although the DNO will usually state they have measures in place to stop it rising above 16kA, most domestic consumer units are rated at 16.5 , so that's OK, and the main cutout fuse will be rated much higher, the issue can arise with domestic MCBs only being rated at (usually) 6kA, you are now reliant on the main incoming fuse of probably 80a and rated at 88kA to clear the maximum fault current possible.

 

As for the PME and no rods in place,

Jeremy's neither meets the formal, or practical definition of PME, although he has sort of (for want of a better word), bodged a bit of a half headed attempt at it , even though he thought he was doing it for an entirely different reason.

this can be a serious issue under certain fault conditions, lethal actually, especially when combined with SP protective devices, remembering that 50mA can kill, and that's why the 30mA threshold. 

100mA devices are only there to protect against fire, (etc) , not to preserve life.

I done a drawing a few years back of simplified earthing basics, I'll look it out.

 

 

Nothing bodged about it at all - it was all done by a qualified electrician - not the one that wired our house, but the one that put in the external wiring on the site, before we started building the house.  Bodging implies that I'm dangerous, or my work is, and frankly I don't think that's on.

 

As for the earth system, then we initially had TT when the supply (exactly the same box wiring as now) was initially put in, as the DNO wouldn't accept TN-C-S on a "site supply".  It wasn't a site supply at all, it was the permanent supply, just fitted inside a box that wasn't built in to the house, but separate from it.

 

When the house wiring was complete, the electrician asked why we were TT and not TN-C-S.  I told him it was at the DNO's insistence, and pointed out that the DNO had measured and written Ze (0.26 ohms) in felt pen inside the box, with the words "no PME for site supply" (see, even the DNO use the wrong term!).  My electrician rang the DNO on completion and asked whether the PE could now connect to their head, as normal for TN-C-S.  They said yes, but please remove and disconnect the TT local earth rod.  He measured Ze again to be sure it was still under 0.35 ohms, and then swapped the earth over and we removed the rod.  The system is certificated on that basis - no bodging, no unqualified work done by me, everything tested and signed off by a qualified electrician.

 

 

Edited January 23, 2017 by JSHarris

[Steptoe]

If this isn't allowed please delete it, but here is a link to a post I wrote a few years ago highlighting the basic differences in earthing systems

 

https://talk.electricianforum.co.uk/topic/14481-simplified-earthing-arrangements/#comment-202531

[Jeremy Harris]

4 minutes ago, Alphonsox said:

I don't think that is correct. I've just had a TN-C-S system installed and it looks like what you have labeled as PME. i.e a supply earth followed by additional earth connections on the run to the house.

 

 

The correct diagrams are as you gave in your earlier link from the IET, and are definitive.

Edited January 23, 2017 by JSHarris

[Steptoe]

4 minutes ago, Alphonsox said:

I don't think that is correct. I've just had a TN-C-S system installed and it looks like what you have labeled as PME. i.e a supply earth followed by additional earth connections on the run to the house.

The DNO are obliged to make all new TN-C-S connections as PME