Cables in Trefoil

[Jereb]

Hello All,
I have a situation that i'm seeking some advice on. Basically we have a ~1500kVA transformer supplying power to our Motor Control Centre. The run is only short, around 20-30 meteres. However total current is rather high as each phase is made up of 3 400mm2 single cores unarmoured run in non trefoil. (i.e each phase is bunched like so Red-Red-Red, Blue-Blue-Blu, White-White-White). Having looked around the build i could not find anything ferrous enclosing the cables like tray etc. but the guys are still concerned about heat issues.
I tend to think that because these cables are bunched and run next to each other, could i not model them as single conductors of some arbitraty large size run flat next to each other with no enclosed ferrous materials around? Will it still cause all kinds of issues?

 

[zaza123]

Hi Jereb,
How are the cables in the short run laid, are they inside the duct bank? In my opinion heat dissipation is also affected by laying of cables & what is the Insulation type is it XLPE cables.

 

[jghrist]

Reactance is higher when phase spacings are higher. Therefore, with a flat configuration, there will be a higher impedance for outside cables and there will be more current in the inside cables. There will also be more mutual heating in the inside cables.

 

[Jereb]

Yes the cables are all XLPE with probably a 3-5mm gap between phase groups. The individual phase groups are bunched into a triangula config /_\ <--- like that i guess.

 

[WhiteyWhitey]

Hi,

The three phases bunched together will react with one another like heat sources.

One phases emitted heat will be transfered to the other phases. You should rate each cable assuming the other two are fully loaded and add them as a heat source to your thermal circuit.

Combining the phases into one big conductor would give you a higher result for ampacity than considering each cable independantly. You may be able to derive an overall rating for all three cables combined but individual ratings using this method would be questionable in my opinion.

Thanks,
Andrew

 

[Jereb]

Righto, so i should be more weary of the heating of the cables from them interfering with eachother's fields.
Would i be right to assume that induced currents would be negligible in objects that do not contain a fully enclosed path (i.e cable tray without a lid)

 

[FreddyNurk]

I assume from the units in use that you're in IEC land. Surely theres an IEC standard that deals with this sort of thing? Certainly in Australia theres a relevant standard that deals with loading of cables and installation factors, in order to deal with the sort of situation that you're referring to.

That said, the statements regarding mutual heating and so on are correct, but I'd be surprised if there isn't already a standard relevant for the installation location that already deals with this.

 

[sibeen]

The relevant IEC standard is IEC 287 Electric cables—Calculation of the current rating.

Most national codes in IEC land will reference this document, for instance AS-NZS 3008 Electrical installations—
Selection of cables is based on this document with some changes made to suit local conditions.

 

[cgrodzinski]

Jareb, I do not know rationale behind the decision of using 400mm2 cables. I would rather use 1000mm2 single conductor per phase. If, however, you have to use 3 conductors per phase there is something called "cable duct". If your voltage is 600V (I assume based on cables you mentioned) the cable load would be 1450A. Then you need two 400mm2 cables per phase and the duct size would be approximate 350 x 270mm.
If you opt for 3 cables per phase you have to remember that each conductor has its own insulation that thermal conductivity is somewhat less than perfect. As such you cannot consider a triple bunch of cables as a single conductor. IEC60287 must be employed with all consequences.
Chris

 

[ScottyUK]

1000mm² cable makes an installation difficult - the cable begins to resemble scaffold tube in terms of its flexibility - and you are paying for a lot of unproductive copper in terms of the current rating per mm². Two good reasons not to use it.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[cgrodzinski]

I think that using 1000mm2 per phase gets better current density than 3x400mm2 = 1200mm2 for the same load. In my opinion using multi-cables per phase circuit produce more copper than bigger conductors. It is easy to check by using IEC standard. I have seen 2000mm2 cables being installed.

 

[dpc]

For larger conductor sizes, the current-carrying capacity does not increase proportionally with the cross-sectional area, due to the difficulty in getting the heat out and the skin effect. So at a certain point, it is not cost effective to simply increase conductor sizes - it is better to go with multiple smaller conductors.

"Theory is when you know all and nothing works. Practice is when all works and nobody knows why. In this case we have put together theory and practice: nothing works... and nobody knows why! (Albert Einstein)

 

[markcraig]

cgrodzinski have you ever tried installing this size cable yourself? I am an Engineer from a Trade background so have had the pleasure of trying to install cables that have been sized with no thought of the site particulars. This is a short run coming from a transformer to an MCC if 1000mm2 cables where chosen this would make it neigh on impossible to install cables, maintain bending radius and keep the job looking neat and tidy. A little though for the installation not just the technical aspects goes a long way to making for a timely and cost effective project in my experience.

 

[cgrodzinski]

I know that it is difficult to install such conductor but it is not impossible. The bending radius is usually 15 x cable diameter. If 1000mm2 the cable diameter is ~41mm so the radius is say 700mm.

Regarding ampacity - I would say in reverse: that the ampacity of a circuit does not double when we use two cables per phase.

 

[7anoter4]

Let's say the transformer secondary rated voltage is 400 V. So the current per phase will be 1500/sqrt(3)/0.400= 2165 A. The current per one cable will be 2165/3=722 A.
In your actual configuration buried in 1 K.m/w soil thermal resistance, depth of 0.7-1.2 m and soil temperature 20 oC no more than 540 A per one cable will be permitted.
As jghrist said if you change the configuration to flat the current will be unbalanced up to 20% so the cable close to the center will be more heated.
I would suggest first of all to change to Red-Blue-White configuration for each group and to keep at least 150 mm clearance between groups. In this case a 650 A will be permitted per cable. In order to increase the load up to 722 A per cable you have to keep at least 40 mm clearance between the cables in the same group.
In a flat configuration the cable order could be R-B-W W-B-R R-B-W so all the cable will get only 722 A [and also keep 150mm distance between groups].

 

[dpc]

cgrodzinski,

Maybe we are talking about two different things. Here's my point:

(In NEC land)

One 500 kcmil copper conductor has a 75 deg C ampacity of 380 A in conduit.

TWO 250 kcmil copper conductors have the same total cross-sectional conductor area but each has an ampacity of 255 A, for a total of 510 A.

So the ampacity more than doubles.

"Theory is when you know all and nothing works. Practice is when all works and nobody knows why. In this case we have put together theory and practice: nothing works... and nobody knows why! (Albert Einstein)

 

[electricpete]

I haven't read the whole thread but just wanted to chime in on one apsect if not already addressed:

Would i be right to assume that induced currents would be negligible in objects that do not contain a fully enclosed path (i.e cable tray without a lid)

Less but not necessarily negligible.
1 - Let's say you have a plate next to a cable so they are both in the same plane. Draw a loop around the outside of the plate. Flux is enclosed and eddy current can flow.
2 - Hysteresis losses can still occur in steel even without loop current flow.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[cgrodzinski]

dpc, it is something wrong with my math?

... One 500 kcmil copper conductor has a 75 deg C ampacity of 380 A in conduit.
TWO 250 kcmil copper conductors have the same total cross-sectional conductor area but each has an ampacity of 255 A, for a total of 510 A.
So the ampacity more than doubles.

Is this what you meant?:

510A > 2 x 380A

or I misread your post?

 

[jghrist]

One 500 kcmil copper conductor has a 75 deg C ampacity of 380 A in conduit.

TWO 250 kcmil copper conductors have the same total cross-sectional conductor area but each has an ampacity of 255 A, for a total of 510 A.

These are ratings for no more than 3 cables in a raceway. For 6 cables, you have to use an 80% derating factor so the 250 kcmil conductors have an ampacity of 408 A.

 

[Jereb]

Wow, sorry didn't mean to start the flame war haha, anyway. Just to clear a few things up for you guys, this is a design given to us from an overseas client in which we are attempting to make sure everything is above board here in australia. The cable is not buried in the ground but is supported by cable ladder rather high up. The cable ladder does not have a lid as it is above areas where debri could potentially fall onto it and the worst it should see is some, probably nasty, dust build up and other gunk. Looking at the area it is installed the cable was likely chosen to be 3x 400mm2 conductors per phase because of the benind radius as it is unlikely that 1000-1200mm2 cable would bend the way we have this.

Anyway, thanks for all the advice guys.

Cheers,
Jereb