IET logo
 
IET
Decrease font size
Increase font size
Topic Title: 240mm SWA
Topic Summary: I'm simply not man enough to handle it
Created On: 07 July 2014 08:04 PM
Status: Post and Reply
Linear : Threading : Single : Branch
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 07 July 2014 08:04 PM
User is offline View Users Profile Print this message



timothyarnold

Posts: 39
Joined: 12 January 2012

For me 240mm SWA is simply too big to handle and terminate.

I need to supply approx 275kVA so would need a fairly sizable cable and previous installs we have used 240mm

I'm considering single core AWA however one end is an existing switch panel with a steel gland plate and can't arrange the downtime to replace it with an aluminium or brass plate. I've seen others cut the plate to reduce eddy currents. Is this still good practice?

Other option is two multicore cables in parallel?

Or should I just man up?

Thanks
Tim
 07 July 2014 08:21 PM
User is offline View Users Profile Print this message


Avatar for slittle.
slittle

Posts: 3484
Joined: 22 November 2007

I'd say man up, but failing that get a cable installation company to do the humping.

We would slot the plate as I don't believe the laws of physics have changed


Stu
 07 July 2014 08:55 PM
User is offline View Users Profile Print this message


Avatar for mapj1.
mapj1

Posts: 2630
Joined: 22 July 2004

Two multi-core cables is certainly easier to route, if you have tight corners and so on, and so long as they are far enough apart not to be considered 'grouped', then, as for the same mass of copper you get more surface area so the cooling is better, the maximum total current rating is higher (or more usefully for the same rating you need less copper..).
I presume you mean 275 KVA at 230/400 volts so a touch under 400A per core.

Two 95mm cables in parallel manage a similar current rating to your 240mm (generally quite a bit higher actually), and will be a little bit cheaper and much easier to handle. However, you then run into voltage drop limits at nearer distances - so its a case of swings and roundabouts as to what is most reasonable.
If as an alternative you go for singles, yes where they enter the box, if tightly glanded, the magnetic circuit must be cut to effectively make one big dog-bone shaped hole (with a very thin neck - just the hacksaw thickness is enough) with all cores inside so it all cancels- always possible to fill the hackslot with mastic or epoxy to discourage rubbish blowing in.

-------------------------
regards Mike
 08 July 2014 09:09 AM
User is offline View Users Profile Print this message


Avatar for OMS.
OMS

Posts: 19593
Joined: 23 March 2004

If you can't easily get at the gland plate, put a spreader box on the panelboard (top or rear) with a non ferrous gland plate in that - as long as all current carrying cores go through the same magnetic "ring" then no problems.

I'd have to say however that 240mm2 isn't a particularly big cable to handle or terminate

If you do end up slotting, you can get the slot brazed if required - basically all you are trying to achieve is a break in the magnetic circuit.

If you can get enough access to slot it - why can't you change it (if you decide on single cores)

Regards

OMS

-------------------------
Failure is always an option
 08 July 2014 09:19 AM
User is offline View Users Profile Print this message



Zoomup

Posts: 264
Joined: 20 February 2014

Originally posted by: mapj1

Two multi-core cables is certainly easier to route, if you have tight corners and so on, and so long as they are far enough apart not to be considered 'grouped', then, as for the same mass of copper you get more surface area so the cooling is better, the maximum total current rating is higher (or more usefully for the same rating you need less copper..).

I presume you mean 275 KVA at 230/400 volts so a touch under 400A per core.



Two 95mm cables in parallel manage a similar current rating to your 240mm (generally quite a bit higher actually), and will be a little bit cheaper and much easier to handle. However, you then run into voltage drop limits at nearer distances - so its a case of swings and roundabouts as to what is most reasonable.

If as an alternative you go for singles, yes where they enter the box, if tightly glanded, the magnetic circuit must be cut to effectively make one big dog-bone shaped hole (with a very thin neck - just the hacksaw thickness is enough) with all cores inside so it all cancels- always possible to fill the hackslot with mastic or epoxy to discourage rubbish blowing in.


Hello mapj1,
that mastic or epoxy idea is a B.G. one. I would never have thought of that myself.

Regards,

Z.
 08 July 2014 09:19 AM
User is offline View Users Profile Print this message



Zoomup

Posts: 264
Joined: 20 February 2014

Originally posted by: mapj1

Two multi-core cables is certainly easier to route, if you have tight corners and so on, and so long as they are far enough apart not to be considered 'grouped', then, as for the same mass of copper you get more surface area so the cooling is better, the maximum total current rating is higher (or more usefully for the same rating you need less copper..).

I presume you mean 275 KVA at 230/400 volts so a touch under 400A per core.



Two 95mm cables in parallel manage a similar current rating to your 240mm (generally quite a bit higher actually), and will be a little bit cheaper and much easier to handle. However, you then run into voltage drop limits at nearer distances - so its a case of swings and roundabouts as to what is most reasonable.

If as an alternative you go for singles, yes where they enter the box, if tightly glanded, the magnetic circuit must be cut to effectively make one big dog-bone shaped hole (with a very thin neck - just the hacksaw thickness is enough) with all cores inside so it all cancels- always possible to fill the hackslot with mastic or epoxy to discourage rubbish blowing in.


Hello mapj1,
that mastic or epoxy idea is a B.G. one. I would never have thought of that myself.

Regards,

Z.
 08 July 2014 09:32 AM
User is offline View Users Profile Print this message


Avatar for OMS.
OMS

Posts: 19593
Joined: 23 March 2004

Just keep in mind IP ratings and forms of seperation depending on the orientation of the gland plate

Mastic or epoxy might well keep the "rubbish" from blowing in - but it's not quite the same as a substantial metal gland plate offering reistance to penetration or breakout (including fire) from within the panel board

Regards

OMS

-------------------------
Failure is always an option
 08 July 2014 09:44 AM
User is offline View Users Profile Print this message


Avatar for mapj1.
mapj1

Posts: 2630
Joined: 22 July 2004

True, and I was only thinking of cables getting in and out of equipment outside where you don't want leaves getting in with your terminals and no-one is too fussed.
It is more intended as an in-situ upgrade to an open slot and where the slot itself is not a regs problem - in the same vein pop rivited aluminium sheet might be another retrofit, that may be possible where the plate cannot be removed.
I would agree if it is practical to take to take it out and down to the workshops to be worked and the time spent with power off is no issue then there are indeed nicer options.

-------------------------
regards Mike
 08 July 2014 10:32 AM
User is offline View Users Profile Print this message



MrP

Posts: 836
Joined: 24 March 2006

Tim
Get a price of a specialist cable tugger that's what they do. The electrical industry is very diverse the man who knows everything knows jack sh#t everything is relative if you haven't done one before no prob they will put it in and terminate it while you're having you're second beer in the snug.

Good luck buddy don't be busting a gut when you can sub it out to someone who does this second nature still making a dollar

MrP on my way home I've heard I may need my sunglasses at home
 08 July 2014 10:55 AM
User is offline View Users Profile Print this message



Zoomup

Posts: 264
Joined: 20 February 2014

Originally posted by: OMS

Just keep in mind IP ratings and forms of seperation depending on the orientation of the gland plate



Mastic or epoxy might well keep the "rubbish" from blowing in - but it's not quite the same as a substantial metal gland plate offering reistance to penetration or breakout (including fire) from within the panel board



Regards



OMS


Hello OMS,
I think that the filler in the form of mastic or epoxy resin may have been intended just to fill the slot, so as to keep out say metal filings etc. which may bridge the gap and form an unintended accidental magnetic circuit for Eddy Currents that the slot is made to prevent.

Bye,

Z.
 08 July 2014 11:27 AM
User is offline View Users Profile Print this message


Avatar for OMS.
OMS

Posts: 19593
Joined: 23 March 2004

Originally posted by: Zoomup

Originally posted by: OMS

Just keep in mind IP ratings and forms of seperation depending on the orientation of the gland plate

Mastic or epoxy might well keep the "rubbish" from blowing in - but it's not quite the same as a substantial metal gland plate offering reistance to penetration or breakout (including fire) from within the panel board

Regards

OMS




Hello OMS,
I think that the filler in the form of mastic or epoxy resin may have been intended just to fill the slot, so as to keep out say metal filings etc. which may bridge the gap and form an unintended accidental magnetic circuit for Eddy Currents that the slot is made to prevent.

Bye,

Z.


Indeed - I think Mike explained that pretty well

I was merely pointing out that in the desire to fill the gap, we should be mindful of the environment the switchboard is situated and what protection it's required to offer (both from internal insults and to protect from external insults)

If you've got metal filings getting into the switchboard, then you have much bigger problems than filling up the magnetic gap, trust me.

If you have to gap a ferrous gland plate, then it's usual to fill that gap with a non ferrous weld to maintain strength and integrity - under a good short circuit event, the gland plate suffers from some pretty extreme forces - a hacksaw cut full of putty may well be subject to catastrophic failure

Ideally, you'd have a non ferrous gland plate of significant size so as to distance the cores from the surrounding ferrous metal frame of the switchboard

Regards

OMS

-------------------------
Failure is always an option
 08 July 2014 06:52 PM
User is offline View Users Profile Print this message



Jaymack

Posts: 4625
Joined: 07 April 2004

Originally posted by: timothyarnold

For me 240mm SWA is simply too big to handle and terminate.

Agreed, and is the reason why a limit of 120 mm² of the Cu variety is imposed by many; unless one has a source of labour spaced at a close distance apart. All together now lads.

Regards
 09 July 2014 05:12 PM
User is offline View Users Profile Print this message



bowmandj

Posts: 139
Joined: 04 October 2011

Are we saying that only ferrous steels suffer eddy current induction? Does not a moving magnetic field also induce into non ferrous metals (in fact any conductor)? Or is the induction so much less for aluminium and brass metals that eddy currents are too small to require a break in the induction conduction path.
 09 July 2014 05:40 PM
User is offline View Users Profile Print this message


Avatar for mapj1.
mapj1

Posts: 2630
Joined: 22 July 2004

Don't confuse the two cases

1) Eddy current losses, where an induced electric current flows parallel to the original -which is why we slice transformer cores so a thin layer of surface oxide allows magnetic continuity, but not electrical, and the central leg of the transformer then no longer allows an electric current to flow round the iron core like an un-wanted secondary turn, which it would if it was a solid bar of metal with wire wrapped around it.

With
2) Magnetic hysteresis losses, where the periodic magnetising and reversal of the rings of magnetic field that surround a current cause heating due to the movement of the magnetic domain boundaries, in the manner of friction between lots of little molecular sized compasses swinging back and forth.

Its the latter we worry about when a wire passes through a thin sheet or iron or steel, and the former in cores and when we have interference currents induced between parallel wires.

Because air is many hundred times less magnetic than iron or steel, or even thousand times if its specially prepared transformer core alloy, a few mm of air gap (or a gap filled with almost any non-magnetic material) will reduce the magnetic flux to the same level as if you had folded in a few inches of extra steel path length for the lines of magnetic field to pass through - equivalent to the flux at a much greater distance from the core. Its not that there is no magnetic field there at all, but it is much lower, and the magnetic path round the whole circuit looks more attractive than jumping the gap..
If the extended hole now includes the return conductors, then the magnetic flux circulating round the whole lot is only driven by the difference in current, which is much less.
This may not be my best explanation let me know if you need more clarification

-------------------------
regards Mike

Edited: 09 July 2014 at 06:00 PM by mapj1
 09 July 2014 09:16 PM
User is offline View Users Profile Print this message


Avatar for slittle.
slittle

Posts: 3484
Joined: 22 November 2007

Originally posted by: Jaymack

Originally posted by: timothyarnold



For me 240mm SWA is simply too big to handle and terminate.


Agreed, and is the reason why a limit of 120 mm² of the Cu variety is imposed by many; unless one has a source of labour spaced at a close distance apart. All together now lads.



Regards



The last one we done, had 6 electricians and two b.....y huge eastern european farm workers. We were lifting about 1 metre each at a time (so around 7 metres), the farm workers were doing that on their own !!

It's all in the planning when the cables are large and I find plenty of tea breaks help too.

That said, I've now found a cabling company so they are getting the next one

Stu
 09 July 2014 11:14 PM
User is offline View Users Profile Print this message



bowmandj

Posts: 139
Joined: 04 October 2011

To mapi1

Thanks for your explanation. Made me do some head scratching and some research of my own. Yes the changing magnetic field around the cable entry point to the steel box will induce eddy currents in the steel and the field changes in the steel will cause hysteresis loss due to the magnetic dipoles switching around against their will.

The eddy currents will produce heat in the metal and add to the inductance affects of the cable at that point of entry. A slot in the metal will prevent the eddy current flow being so strong (resistance increase) and reduce that heat and inductive losses.

The slot will also cause the magnetic field to spread out further in the sheet metal and reduce the density in any area of the metal (but occur in a greater area overall). So the hysteresis losses are not reduced but spread out over a larger area reducing the temperature over said larger area. This diffusion removes the risk of a 'hot ring' at the cable entry point. Again there is a reduction in inductance effects with the reduced flux density.

Non ferrous materials only suffer the eddy current loss and not nearly so much of the hysteresis losses.

This fits in with my experience of cables carrying powerful near microwave frequency currents passing through a steel bulkhead. The hole punch for the gland hole did not cut a round hole but a series of slots giving a spoke effect (the hole looked like a Torq screw). Maybe 10 or 12 spokes (or more importantly 10 or 12 slots).

With the electric cables then a slot between both cable entry gland holes will allow the inverse field to balance out somewhat provided the cables are not too far separated.

That was a lovely journey back 40 years to college days in my late teens.
 10 July 2014 09:10 AM
User is offline View Users Profile Print this message


Avatar for mapj1.
mapj1

Posts: 2630
Joined: 22 July 2004

Glad it made some sense - it is indeed a complex business, and for many folk, not helped by the fact that magnetics is often the poor relation and badly taught relative to pure electrics. Then the use of 3 different sets of magnetic units in textbooks written over the last 50 years or so is a further undesirable complication as factors of pi, 1E7 and "c" can seem to dissapear from one side of the formula and appear in the other in alternative descriptions of the same thing without appparent reason. For this reason my explanations tend to be a bit arm wavey.
As the frequency goes up it gets worse, so losses at 800Hz are much higher than at 50 (although above a certain frequency the little domain 'compasses' can't keep up dancing to the faster music, and when they decide to just sit still and look a bit breathless, so magnetic losses that were rising continuously with frequency suddenly plummet as the frequency rises further Thenfor transformers and EMC chokes we use ferrites which have smaller more nimble domains, yes they have much less magnetic response at lower frequencies than iron, but equally less of a catastrophic failure to do anything at all above a few tens to hundreds of kiloherz..)
Eventiually we find ourselves insisting that flow and return must always stay side by side for RF welding,or be twisted for high speed data, and eventually one inside theother so the fikleds are perfectly centred and cancel out, i.e. only coaxial or carefully designed transmission line strucutures are enough to keep the fields where we want 'em.
This sort of thing takes up quite a lot of my life, and gently slides into antenna design. Indeed wierd and wonderful shapes exist, that are both grounded and live at the same time.

-------------------------
regards Mike
Statistics

See Also:



FuseTalk Standard Edition v3.2 - © 1999-2014 FuseTalk Inc. All rights reserved.