Reverse fed transformer grounding

[4jo3]

I require some assistance on how to ground a 240x480 to 120/240 single phase transformer. Currently I am proposing to reverse feed a 240x480 to 120/240 single phase transformer because of a long run to combat voltage drop. I would then step down using another 240x480 to 120/240 single phase transformer. My question is since there is no neutral in the 480v run. How would I go about grounding? Can I have an ungrounded system?

Thanks.

 

[electrofelon]

I require some assistance on how to ground a 240x480 to 120/240 single phase transformer. Currently I am proposing to reverse feed a 240x480 to 120/240 single phase transformer because of a long run to combat voltage drop. I would then step down using another 240x480 to 120/240 single phase transformer. My question is since there is no neutral in the 480v run. How would I go about grounding? Can I have an ungrounded system?

Thanks.

Strange, there is nearly the identical scenario in another current thread: https://forums.mikeholt.com/forumdisplay.php?f=93

Yes you need to earth and SDS, or leave it ungrounded and use ground detectors. Check out the exception I mentioned in the other thread and if that applies you can bond at both ends and skip the supply side bonding jumper and only have to run two conductors. If that works, you would earth and bond one "end" of the 480 and make it a grounded conductor, then the exception lets you use that grounded conductor to bond the far end to get your faulty path.

 

[infinity]

Since it's a 240/480 transformer couldn't the center point be grounded?

 

[electrofelon]

Since it's a 240/480 transformer couldn't the center point be grounded?

Yes, but if one wants to use 250.30(A)(1) exception 2 and thus only have to run two conductors instead of 3, then one of the conductors would need to be a grounded conductor. OP didnt say length or size so maybe running an EGC/SSBJ isnt that big a deal, but maybe it would be a big cost adder.

 

[kwired]

I believe there are some instances where you can ground one lead of the 480 and not have to run an EGC (can run just two wires to the other transformer)

But in general you need to run an EGC so you have three wires anyway. You have no neutral loads of any kind so no need for a fourth wire ever.

If required to have an EGC, I'd probably ground the center tap, making voltage to ground 240 instead of grounding an end making voltage to ground 480 volts. On the load end where you step back down leave center tap floating, but connect the EGC to the transformer enclosure.

 

[4jo3]

service is feeding a 200amp breaker. The run is 800 feet.

 

[iceworm]

I have been involved with two of these - out in the country, loooong driveways.

The advice given, two wire, direct burial or plastic conduit, is good.

Something you didn't ask - and perhaps already know the answer:

Light blinks:
How big of a transformer are you getting? 200A, 240V is 48KVA. Some recommend sizing the xfm down to the expected load.
One I helped with, was a 100A service, ~1000', customer bought 2 - 15KVA transformers. House had oil heat, propane cooking, no AC, and a 2hp well pump. Loading was well within 15KVA. And the lights would blink every time the well pump kicked on - and his wife complained.

With incandescents, a 5% dip causes an annoying blink.
Consider sizing the xfm and wire to 3% Vd or less. Yes, it is more money.

Consider 600V. The cost of the transformers should be about the same. For the same wire price, less Vd.

What I don't know:
Incandescants blink annoyingly with a 5% dip. I don't know if this applies to LEDs or curly-cue fluorescents.


the worm

 

[4jo3]

50kVA Eaton which is suitable for reverse feed. Already have the transformers which is why we didn't go 600V. As for Vd I confirmed feeder and branch will not be below 5%.

 

[electrofelon]

Is this a house? What is the ACTUAL current you expect? I would not have done the transformer scheme for this length. Just your no load losses will be $450 per year. Add in transformer costs and load losses, two parallel sets of 250 MCM AL, which give you 5%@ 150 amps, starts sounding pretty cheap.

 

[iceworm]

Is this a house? What is the ACTUAL current you expect? I would not have done the transformer scheme for this length. Just your no load losses will be $450 per year. Add in transformer costs and load losses, two parallel sets of 250 MCM AL, which give you 5%@ 150 amps, starts sounding pretty cheap.

Not saying you are wrong. Just curious where you got your numbers.

I'm seeing high efficiency 50KW xfm no-load loss at 32W
That's about .8KWH/day >> ~300KWH/year.

$1.5/kwh seems is a bit high

Maybe the xfm is not hi-eff

 

[4jo3]

Not a house. Actual Current is around 120amps but could go as high as 150amps. There are also branch circuits. So tack on another 150-200feet. Cost per kWh is about 10cents.

 

[electrofelon]

Im still just not seeing how it is economical to use transformers. Ok no load loss is likely less if these are DOE 2016 transformers; Could be around .25%, so thats 2200 KWH per year, $220 per year. Transformers are $1600 each. Two sets of 350 URD would be about what you need +/- and that is $3 per foot each set, so $4800, plus an EGC but minus what your wire would be at 480 volts (so knock off a grand? How about I take off $800, great deal). so 240V costs the same after 4 years of no load losses. Not included which could swing the numbers:

1, Increased labor for the bigger 240 wire
2. Conduit would add a bunch to the 240 scheme if you arent direct burying
3. Load losses for 480 scheme.

 

[kwired]

service is feeding a 200amp breaker. The run is 800 feet.

That length of a run may not be worth two transformers and extra items necessary vs larger or parallel conductors - especially if they are aluminum conductors.

Next thing is the load. You feeding a 200 amp breaker, but if you never expect to see more than 100 amps of load, you aready have a head start on voltage drop with 200 amp conductor.