300kva XF calculations

[bill@usps06492]

I will be installing a 300kva XF 3ph 480p/208-120s (800a panel on the secondary) . I've cal'd the primary at 480v@500a ocpd, on the seconday I've come up with 833a@208-120 and decided to keep the ocpd secondary at 800a because none of the loads are continous, and because of the length of the run on the primary, I didn't want to over stress the transformer. The length of the run is my cocern! It's over 550ft from the primary circuit breaker to the spot where the XF is to be installed. At 500a I could use 250kcmil THHN parralleled with a 2awg cu ground in a 2.5" emt, but because of the length of the run, and I am using 600ft as the max distance, I'm concerned about the voltage drop. On a 408v circuit 14.4v=3%vd, and when I did my calculations I need a 350kcmil conductor, and a 3" conduit to get my 3% max vd.
300 x 1000/480 x 1.73= 361.4a

361.4a x 1.25%= 451.7 (500a ocpd primary)

300 x 1000/208 x1.73 = 833.3a (800a secondary)

cm = 1.72 x12.9 x 361.4a x 600ft/14.4vd

cm = 4839218.28/14.4 = 336056.825cm per ch9 table 8 is a 350kcmil and because the ocpd is 500a on the primary I need to size my wire to the 500aocpd, and thats 2sets of 350kcmil.

On the secondary it will be 2 sets of 600kcmil with a 1/0 awg cu ground. From the xf to the 208-120/800a panel is less tha 20ft.

So after all this am I on the right track, or am I over sizing my feeders? I will be using this information to order the material to do this job, and the price just for the panels, breakers, and the transformer is almost $10k, and God only knows what the wire is going to cost.
The loads to be servered for now,will be 3 pieces mail processing equp, with a max ocpd for the circuits @208-120v/125a 3ph using 1/0 cu awg 5wire, a 225a panel to feed 4-30a/208 3ph central vacuum units, and any future needs.
Bill

 

[bob]

The length of the run is my cocern! It's over 550ft from the primary circuit breaker to the spot where the XF is to be installed. At 500a I could use 250kcmil THHN parralleled with a 2awg cu ground in a 2.5" emt, but because of the length of the run, and I am using 600ft as the max distance, I'm concerned about the voltage drop. On a 408v circuit 14.4v=3%vd, and when I did my calculations I need a 350kcmil conductor, and a 3" conduit to get my 3% max vd.
300 x 1000/480 x 1.73= 361.4a

361.4a x 1.25%= 451.7 (500a ocpd primary)
cm = 4839218.28/14.4 = 336056.825cm per ch9 table 8 is a 350kcmil and because the ocpd is 500a on the primary I need to size my wire to the 500aocpd, and thats 2sets of 350kcmil.

Bill

Bill
I used 361 amps and 600 ft an VD = 2.4% or 11.4 volts with 2# 250 kcm/phase. 1#500 kcm/phase came out VD = 2.5% or 12 volts. You could use the taps to correct this drop.

 

[nvpowerdoc]

This is the order to determine the sizing of everything:

First, size primary conductor to 125% primary FLA of transformer, thus 450A is your required current rating of the wire. 2 parallel runs of 4/0 @ 75 deg C.

Second, check voltage drop, and upsize base on 3% Vd. In this case, 600' primary run requires 2 parallel runs 350kcmil at 75 deg C.

Third, size secondary protection and wire size based on Table 450.3(b) (watch the notes) and 240.4(f) (delta-wye requires secondary protection). Therefore 300kVA @ 208Y gives 833A x 1.25% = 1041A. Secondary feeders must be sized to carry 125% of the secondary FLA of the transformer thus 3 parallel sets of 400kcmil 75 deg C would be correct for this application.

OCPD on the primary side also follows Table 450.3(b), therefore if your were to go as high as the 250% FLA rating as allowed, you do not necessarily increase your wire size too. This may seem counter-intuituive, but wire sizing is based on expected FLA (plus 25%). Commonly on Tx's and motors, the OCPD device is rated much higher than the wire rating. Fault protection (instantaneous current) is provided for through the OCPD's instantaneous protection region. Time Current Curves for OCPD devices are used to ensure that fault current is sufficient to trip a device that is rated higher than the wire size it protects (when a fault occurs). This is very much different than overload protection.

As for the cost of wire, I recommend looking at compact Aluminum. It is much more cost effective, easier to install, AND when properly installed is just as reliable as copper (though remember, all sizes must be adjusted accordingly).

I do not recommend using Taps for correcting voltage drop. Vdrop is Vdrop and must be kept below 3%. Consider the issue that may arise should your utility provide lower than nominal voltage to your facility - how would you provide corrective action at your transformer if you've already used up your taps.

 

[Smart $]

I do not recommend using Taps for correcting voltage drop. Vdrop is Vdrop and must be kept below 3%. Consider the issue that may arise should your utility provide lower than nominal voltage to your facility - how would you provide corrective action at your transformer if you've already used up your taps.

I am of the impression that total VD, i.e. from service to load, is recommended to be no more than 5%. If one permits 3% VD to the primary of a stepdown tx, that would only leave 2% to work with on the secondary side. While keeping under 5% is still an achieveable goal, not knowing all the loads and their circuit distances from the secondary makes me question the support of a 3% VD on the primary side.

 

[hardworkingstiff]

Help me understand please.

Help me understand please.

Bill
I used 361 amps and 600 ft an VD = 2.4% or 11.4 volts with 2# 250 kcm/phase. 1#500 kcm/phase came out VD = 2.5% or 12 volts.

Can you help me understand why the voltage drop is more on a #500kcmil than it is on 1/2 the current on a #250kcmil? In the most simplistic form, VD=I(current of load)*R(ohms in wire). Using the uncoated DC resistance in table 8, 250 is about 50.097% of 500. Using the AC resistance for uncoated copper wires in steel conduits in table 9, 250 is 53.703% of 500. Using the effective Z at 0.85 PF for uncoated copper wires in steel conduit in table 9, 250 is 68.493% of 500.

In all three cases, the resistance in two 250's is higher than in one 500 so it does not make sense to me that the VD would be more in the one 500 cable.

 

[bob]

Can you help me understand why the voltage drop is more on a #500kcmil than it is on 1/2 the current on a #250kcmil? In the most simplistic form, VD=I(current of load)*R(ohms in wire). Using the uncoated DC resistance in table 8, 250 is about 50.097% of 500. Using the AC resistance for uncoated copper wires in steel conduits in table 9, 250 is 53.703% of 500. Using the effective Z at 0.85 PF for uncoated copper wires in steel conduit in table 9, 250 is 68.493% of 500.

In all three cases, the resistance in two 250's is higher than in one 500 so it does not make sense to me that the VD would be more in the one 500 cable.

Stiff
Good question. If you look at the DC resistance of 250 and 500 CU you will
see that R of 250 (0.0515) is about twice that of 500 (0.0258). However when you are using AC voltage there are two effects that take place.
One is called SKIN EFFECT and the other is PROXIMITY EFFECT.
The following site give a fair explanation of both effects.

http://www.generalcable.co.nz/Technical/10.3.2.1.pdf

Basically when you use AC Voltage the current tends to move toward the outside surface of the conductor leaving the center of the conductor carrying less of the current, thus a higher resistance. It is more noticeable on the larger conductors. The site below shows a table of the AC/DC ratios of
conductors. Notice that the smaller conductors are less effected.
AC resistance of 250 = 0.00588 and 500 = 0.00297. X of 250 = 0.00487 and 500 = 0.00458. The impedance
of the 250 = 0.00763 for 1 conductor. Z for 2 conductors is about 0.00763/2 = 0.003815. Z 1-500 = 0.0546.
Table 9.
http://www.okonite.com/engineering/ac_dc_ratios.htm

The AC/DC ration of 250 is 1.005 and 500 is 1.018. On short circuit runs
this effect is small but your circuit is 600 ft so it has more effect.
Note that while the 500 kcm meets the VD requirements, it does not match the proposed 500 amp breaker.

 

[bill@usps06492]

nvpwrdoc
It's my impression that the ocdp on the secondary, per 450.3b can be set to a max of 125%, and that would be 1041amps. On the primary my ocpd could be up to 250%, 362a x 250%=905a. so if these are max values, then using the lower raed ocpd, as long as I provide for in XF rush current, and a continous XF load then I should be okay.

Sizing at 250%&125% puts me into article 240.4c that states that ocpd rated over 800amps will have conductors that are rated equal to or greater than the ocpd device.
bill

 

[nvpowerdoc]

Bill:

Primary OCPD should be set to 125%. Unless you have a really strange breaker or fuse, inrush should not be a problem. You should check your manufacturer's breaker or fuse Time Current Curve against the inrush value of the transformer (8xFLA @ 0.1Sec typically) to confirm proper expected operation.

As for the secondary OCPD, you are correct in stating the requirements of 240.4c. However, note that if you are to select a 1000A breaker, three parallel runs of 400Cu @75 Deg is appropriate as each run is rated at 335A.

All of this works because your expected continuous load is 833A secondary. Though this seems shy of 1000A by 41A (based on the 125% calculation) it is the prudent solution. I have seen applications where a 1200A secondary OCPD was used (with 1200A feeders), and in fact I discussed this with my AHJ and it was found to be acceptable - its just not how our organization would do it.

 

[kingpb]

I think a point that needs to be made here is the importance of also properly selecting the OCPD's. It is difficult to select a device, especially if it does not have adjustable settings, to be able to ride thru the inrush, and still coordinate with the transformer damage curve. This is especially true with delta-wye transformers that may be feeding a lot of non-linear loads.

I see many people talking of using the 250% on the high voltage side of the transformer. This is a perfect example of how the NEC allows this (Maximum) setting, but it could very likely result in improper device coordination, leading to transformer damage. Looking at damage curves show that it is rarely possible to use that high if a setting.

On the other hand, a huge amount of emphasis is consistently put on cable sizing, yet when you look at the cable damage curves, it becomes apparent that cable damage will not usually be an issue.

Now, as far as transformer taps, someone mentioned that it is best not to use them as part of your original design, but rather to make use of taps during final field adjustments, start-up, and checkout. This is sound advice. Unfortunately, taps have a tendency to be forgotten and consequently never used when everyone is trying to meet their deadline and get final sign off and payment. This can lead to client dissatisfaction and avoidable return trips to troubleshoot equipment operational issues.

Sizing cable, selecting OCPD's, adjusting taps, is all part of a well designed and coordinated system. So, try to look at the system as a whole, and give each piece it's deserved attention.