Power sharing between multiple conductors 2

[Scrivs]

I have a UK site that is supplied by two 415v transformers, each feed to separate sections of a switchboard. A common standby generator can feed both sides through the relevant changeover switchgear upon a power failure.
This board feeds a distribution board via 22 off 240mm2 single cored aluminium armoured cables, 3 per phase and 2 per neutral on each of teh two sections. The cables are about 30ft long.
The power balance on the blue phase has been measured at 34%, 20% and 46%, the other two are balanced and we have been looking for an answer why.
The site is a computer centre so we know that there is a high harmonic content, especially the 3rd, on the supply.
It would seem that the harmonic distortion is primarily upon the lowest loaded cable.
Has anyone experience of this please?

 

[dpc]

For multiple conductors per phase, it's important that the conductors be the same length and routed in a similar manner. The shorter the run, the more likely that difference in current will occur. For a such a short run, a very slight difference in contact resistance at the terminations could make a substantial difference in current. Inductance also can vary depending on the geometry of the installation.

If you can visually inspect the terminations while energized, you might do an infrared check of the terminations. Otherwise, you may have to check the terminations.

 

[Scrivs]

I have already used our IR camera and this showed a very slight heating effect on about 4 of the 44 bolts - it was about 6'C higher than ambient.
The cables have been arranged on a galvanised tray structure and are tie wrapped together as individual phases about 2" apart.
One other consideration is that the cables are glanded and earthed at one end only, the other end has plastic glands. Both boards are earthed well and we have checked for a continuity / a potential between them.

 

[DanDel]

It has been my experience that parallel cables which are loaded significantly below their rating can definitely have a current mismatch as you have observed. As the loading approaches capacity, the mismatch minimizes.

It is important to make sure that the cable lengths are the same and there are no loose connections.

The IR scan is an important tool to determine a problem in this type of situation. Was the 6°C rise on just the bolts, or on the connection lug? I wouldn't be too concerned about this if it was just the bolts because the emissivity of the metal those bolts were made of or plated with may be just different enough to cause what you see.

As long as there are no obvious connection problems (at least 10-15°C rise at the lugs), I would not be too concerned. Keep checking with the IR camera and also check the loading once in a while. If the load increases, you will probably notice that the mismatch will decrease.

 

[Shortstub]

See thread 238-48487: "Mutual Inductance to Blame?"

 

[electricpete]

Good thread Shortstub. In linked format:

thread238-48487

 

[jbartos]

Suggestion: The subconductor sharing 20% of power flow may be experiencing some interaction with the subconductor sharing 46% of power flow. Namely, some arcing may have damaged one and become welded in couple places to the 46% power sharing subconductor.

 

[Shortstub]

Jbartos,
Without involving the armor covering?

 

[zohar]

Scrivs-if the distortion are the same at the three phases and the current not (like in your setuation)acorrding to the definition of the t.h.d the lower the current is the t.h.d will be higher, this is the reasone while doing harmonics anlyzing you have to check the amplitod of every suspced harmonic.

Dandel-if the 6 degree are on the bolts of the unlload cable i think it will be better to check the fastening moment of the bolts and the termination.

All the best.

zohar.

 

[Scrivs]

Chaps, thamks for your input.

Zohar - you are right, the lightest loaded of the three conductors carrying each phase has the highest THD ie they are in inverse proportions.

I have reviewed the half hour billing data from the main meter for the whole site now. They have a 1MVA availability from the supplier and their present peak use is only about 550kVA.

I have submitted my report to the client recommending that we have a brief shutdown to clean/re-torque all connections, prove all earth bonding/the ground connection, prove that all conductors are connected in teh same order at both ends etc. Once that those conditions are sound then review the situation again....

It's a major data centre so we will be lucky to get a 6 hour shutdown window, but it is achievable.

I will keep you posted