Cable sizing Query for Harmonic Filter

In terms of cable ratings and protections, the exam question is always " is there a credible fault mechanism that would not be enough to operate the trip, at its low setting, and still be enough to overheat the cables.?"

The general assumption is that  most equipment either fails to a clear open or a dead short, so you are left looking at the cable impedances and the PSSC at the origin and comparing that to the tripping times.


As I understand the   data   the ADF is just connected in parallel with the supply to the rest of the building and it injects or withdraws current at the right parts of the mains cycle to make the load as seen by the incoming supply look entirely real - that is to say the current is in phase with the voltage, and a reasonably pure sine wave, so the supply thinks it is driving a resistor.


Obviously the currents between the real loads and the ADF unit may be carrying any old wild waveform, within the range that the unit can add or subtract enough to compensate - about 360A if I have read the data sheet correctly.


You do not mention where the actual loads are connected or how big they are - this is also important.


regards Mike

Hi Mike

Thanks for your response and apologies for my delay in responding (it was the last thing I looked at before the Christmas break and just getting back to it now!).

OK so, some more info; yes the ADF is connected in parallel to the load via the dedicated ACB on the main switchboard (I actually have 4 switchboards fed from 4 individual Tx supplies and there's an ADF going on each, so 4 separate systems).

I'd appreciate if you would review the logic behind my cable sizing proposal, which is...maximum current draw capable on each ADF is 360A (based on data sheet); I also applied a de-rating factor of x 1.3 to the full load rating of each ADF to guarantee that the cables are not overheated due to the skin effect caused by the harmonics. This is in accordance with the manufacturer’s recommendations, so I'm selecting my cable based on a max load of 360A x 1.3 = 468A.

I contacted the cable manufacturers and they advised that the single phase figures could be used from table 4F2A, but that a 0.7 de-rating factor should be applied for trefoil formation. Based on this info, if I apply the inverse of the 0.7 de-rating factor (1.4285), to the max load of 468A, I get a nominal current carrying capacity figure of 668.5A which I have used to ref Table 4F2A. This brings the size of single core cables required for this installation method to 240mm.

The 1 second short circuit withstand current rating for the 240mm H07ZZ-F cable is 34.3kA, and I am satisfied that the cables will safely withstand a fault current long enough for the trip to operate.

However, the minimum trip setting on the supply ACB as mentioned above is 500A, as the max load capacity is lower that this setting, the ACB is not intended to provide overload protection for the equipment (as each ADF module has its own incoming fuse), just fault protection for the cables, but with the de-rating factor of 0.7 applied to the 240mm cable ccc value from Table 4F2A (673A x 0.7 = 471A), does this mean that the 240mm cables would be undersized for the minimum ACB trip setting??

My thoughts are that as the connected load is not capable of reaching 471A under normal operating conditions and the capacity for withstanding a fault condition is satisfied and unlike a sub-mains to a switchboard, there is no capacity to add-on load to the ADF, this proposal would be sound. I just have this nagging in my head from my years of sparking that the downstream cable should be rated at or above the breaker setting that feeds it...

Any advice appreciated,


Regards, Seán 

Don't worry about the delay - even if you had come back sooner,  something funny with the software meant that I could not login to respond at all over Christmas anyway - just got a 'down for maintenance' message, and I assume by the lack of regular  posters until yesterday, a number of folk who do not login with google or whatever but actually type in a name and password were in the same boat .

I know what you mean about the gut feel that it is better if the downstream cable should be rated at or above the breaker setting that feeds it... however it is not always required, and for big stuff it often is impractical - the cable to the house is protected at the substation by a fuse of many times the steady state rating, but the overload protection comes from the fuse at  the service head  end,  as to bring full sized 300mm2 wavecon or whatever  to the meter cupboard in every house and flat would be impractical in the extreme.

This is a similar case.

I suspect your 240mm cables will run pretty cool actually, you may even be able to drop a size, but I have not seen the route..

(I think open air trefoil 240mmm reaches 90C with about 600A  based on this) and there is no need to worry about the 500A breaker setting, as there is overload protection at the load end.

Even if there was not a load end fuse,  the likely fault condition is likely to be more or less 50Hz, sinewave, if one of the internal switching devices gets shorted or stuck 'on'. As the skin effect thing will only apply to fast changing waveforms, so you do not need the additional de-rating  of 1.3 during faults (!).


Mike.

I think your cable choice is entirely safe, and not unreasonable. You can make use of the derating factor in the opposite direction, if the ambient temperature is less than 30C, but in any case, your potential overload is very small at 30A, perhaps 2 or 3 degrees above 70C worst case which is unimportant.

Regards David CEng.

this nagging in my head from my years of sparking that the downstream cable should be rated at or above the breaker setting that feeds it

It's actually quote common to have things otherwise - just consider a 20A cable feeding an unfused spur from a domestic ring circuit fed by a 32A MCB.


For fault conditions the only considerations are the c.s.a & material of the conductor and its allowable temperature increase (usually determined by its operating temperature and maximum temperature the insulation can withstand for short periods) - i.e. "k" and "S". Factors that affect the long term current carrying capacity of a cable (grouping, thermal insulation and so on) don't come into it as they're all related to how well the cable can loose heat - while the fault conditions are (usually) considered to be adiabatic.


   - Andy.

Thanks for the replies and advice guys - much appreciated.

Regards, Seán