Working the cables at higher frequency reduces the usable cross section of the cables due to skin effect. Thus, I guess the cables need to be derated. Manufacturers may be able to guide.
I haven't come across such a situation myself.
There is very litter information about the NEC derating cables due to frequencies ≠ 60Hz. Article 430.122(A) mentions vaguely frequency issues for the VFD application.
On the other hand, the ICEA Std P-43-457 addresses cable derating factor for 400Hz and 800Hz.
Below is a curve is shown various derating factors for cable sizes in a nonmagnetic conduit including a rough estimate for 100 Hz in a nonmagnetic raceway.
It should be noted that for a frequency of 100 Hz the following summarizes the results shown on the graph below: 1) Ampacity derating factor is significant for cable size above 350 kcmil.
2) For cable size ≤ 4/0 AWG the derating factor is negligible
3) For a higher size above 4/0 up to 350 kcmil the ampacity derating factor ≈ 3%.
4) The derating factor will increase for magnetic raceways.
NEC 2017 Art. 310.15 C) Engineering Supervision. Under engineering supervision,
conductor ampacities shall be permitted to be calculated by means of the following general equation:
I=SQRT((Tc-Ta)/Rac.(1+Yc)/Rca)
Neither Rca-thermal resistance- nor maximum conductor temperature and Rac [d.c. resistance at working temperature] depends on frequency. Only Yc [skin and proximity effect] depends on frequency. For skin effect calculation you need to know only d.c. resistance at working temperature but for proximity effect you need to know conductor diameter and the center line-to-center line distance between phases.
If no clearance is between phases the distance is the overall diameter of the phase conductor insulated core.
Let's say we have three conductors of 500 MCM copper at 75oC [Tc=75]
The d.c. resistance at 75oC[RdcTc] and the conductor diameter [Dc] will be according NEC Table 8
RdcTc=0.0258 ohm/kft=25.8 μΩ/ft ; Dc=0.813 inches O/alldia=0.943 inches[for XHHW NEC Table 5].
xs=0.875.sqrt(f.ks/RdcTc) Neher&McGrath formula 22
ks=1 for concentric round non coated Neher&McGrath Table II.
If f=60 xs=0.875.√ (60.1./25.8)=1.33
From Table III F(xs)=1.61%
skin effect Ycs=0.0161
Proximity effect kp=0.8 for concentric round non coated Table II
xp=0.875.sqrt(f.kp/RdcTc) Neher&McGrath formula 25
xp=1.193
From Table III F(xp)=1.03%
Proximity effect Ycp=F(xp)(Dc/S)^2x[1.18/(F(xp)+0.27)+0.312(Dc/S)^2]
Ycp=0.0103(0.813/0.943)^2x[1.18/(0.0103+0.27)+0.312(0.813/0.943)^2]=0.034
Yc=Ycs+Ycp=0.050 at 60 Hz.
If f=100 Hz xs=1.7227 F(xs)=4.4% xp=1.54 F(xp)=2.86% Ycp=0.0889
Yc=Ycs+Ycp=0.133
The derating factor then =√(1+0.05)/(1+0.133)=0.925
From cuky2000 curves the derating factor for 500 MCM it is 0.92 approx.
When working in induction melting business (frequencies in the range of 70 Hz to 1000 Hz at MW power level) we had our own tables for derating. But not only for derating, but also for recommended cross-sections and arrangement of parallel cables. I'm not able to post these guidelines, but I can give some general recommndations
- Use smaller cross sections as for line frequency
- Interleave parallel cable
- minimize magnetic fields to avoid heating up stell structures close by
