Greetings All,

I have a question about sizing a transformer. I've tried looking online and using my understanding of ohms law but I keep getting different answers.

Here's the deal. I need to supply 400amps @ 120V. It will be fairly simple to balance the loads so is there any benefit to setting up the panel as 230/3 phase? (125amp per leg) (Yes I know we'll need a Neutral to get 120V)

If my understanding is correct, using this method will allow me to get away with a 75KVA transformer.

Bottom line, there's plenty of capacity at the 460V main. The guy putting in the transformer is licensed but there's a bit of a language barrier and I'm not certain he understands my question.

Thanks!

Well Mort, I would expect that you might see 120/208 Wye 3 phase Vs 120/240 single phase. The obvious advantage would be that you have the ability to power 3 phase motor loads. I don't see many 400A requirements that don't include motor loads.

That will be 208 3 phase and using my power formula the actual load would be about 45 kva (with 125a per phase) but you need to use 125% of that, so at least a 57 kva transformer. If you want to plan on harmonics and additional load you are still OK with the 75 ... but I am not the engineer here. Maybe Scott will stop by.
If you have a reactive load you also want to make that neutral a couple sizes bigger. (Triplin harmonics)

A 150kVA will give you 416 Amps of 208Y/120, but Mort is looking for 230/3 phase, which is a different animal altogether than the Wye connection we're all familiar with.

He's also going to have a separate zig-zag transformer in order to get a working Neutral, since a Delta connection on a 230/3 phase system is limited to not more than 5% of the transformer rating available for phase-to-neutral loads.

The loads are going to be several computers... no motors and by several, I mean a couple hundred.

What I was getting at was, if I treat the panel like a 208/3ph and run a separate neutral to break each leg down to a 120v circuit, would that allow me to get away with a smaller transformer?

In the end, I need 400amps @ 120v and I want to know if a 75KVA will suffice.

Thanks guys!

Assuming that you what really need is a lot of 120V for the computer loads the answer to your question is no.

A 75kVA transformer will change 90 Amps of 480V-3phase into 208.3 Amps of 208Y/120V-3 phase. There's just no way to draw 400Amps out of a 75kVA transformer...the math just won't allow it.

The next larger standard size transformer is 150kVA, which will change 180 Amps of 480V-3phase into 416 Amps of 208Y/120V-3 phase. You'll probably need to break that down by using a couple of 225A panelboards so you can end up with the plethora of 20A-1p circuits that you will need.
Remember to either run separate neutrals for each circuit or to increase the neutral size to 200% if you wire the branch circuits in a network configuration.

This reminds me of a conversation with a PHDd Electrical Engineer.

He said he had 120 amp service at his home, further questions resulted in 'two (2) 60 amp fuses'. His logic was 120 available amps at 120 volts. I wound up having my guys upgrade his service to 200 amp. He later was bragging that he had a '400 amp service in his home'.

Logic??

I think the OP here is going down the same path....
75 KVA, wye, 120 208 panel with the above logic is 600 amps at 120 volts.

That's a good point, John;
75(a phase)+75(b phase)+75(c phase)+75(Neutral) = 600

I'm having a similar situation here with our IT Department.
They apparently believe that if they plead long enough and hard enough I'll somehow 'find' a way to increase their 100A feed to 200A.

They keep implying that I'm just being mean by telling them that they have to abandon the existing 100A feeder so that they can get a new 200A feeder.

They also (naturally) want to do it for as close to 'free' as possible. Maybe I'll ask them why they want the Electricians to work for free but don't balk about having to pay real money for new computer equipment.

Sometimes folks take the "electricity is just like water" analogy too literally.

Mort,

Where did you get your 400 amps from? Power supply wattage times the number of computers?

Are the computers located in the same room?

Do they all have separate monitors?

What else is going to be powered by the 120V? PC speakers, desk fans, desk lights, copiers and printers, etc. Not to mention the network hardware needed to support 400 computers. Do you have a telephone system too.

Most computers and monitors can run off of 208V or 240V, as well as 120V. It won't reduce the transformer size but it could allow you to run more machines off of a single circuit.

You might want to get a more experienced engineer to look over your assumptions and calculations.

Good luck with the project.

I really wish Scott would stop by because the square root or 3 comes in here.

Here you go:
Amps = 1000 × kVA / (√3 × volts)

Amps = 1000 × 150KVA / (1.73 × 208V)
The 208Y/120V output of a 150kVA transformer is 416.7 Amps.

Keep in mind that this is does not take into account any reduction due to harmonics generated by the solid state switching power supplies in the computer equipment.

John, you hit the nail right on the head man. Everything I was taught about electricity suggests that what you are saying is correct and I have spent the last 20 years of my life believing in that logic.
Bear in mind, I only had 3 months education in electrical as that's all that's needed for hvac.
I've lost recent arguments with people that have degrees that made me question my beliefs.

Will you do me a favor and read through this thread and tell me what you think?
http://www.contractortalk.com/f5/am...t/#/forumsite/21253/topics/254793?page=1

Larry, I appreciate your concern. We've done testing and determined the average computer we're using with monitor pulls 3.3 amps x 100 rigs = 330 amps. We're using 400 amps for good measure.
Networking gear is on its own circuit.
For the record, I originally thought we needed a 150 but someone I work with convinced me of the flawed logic above.
Thanks

Mort;
Always remember that once the undersized 75KVA transformer failed, it would be considered YOUR fault and the person who convinced you to use a smaller transformer (so it could be done cheaper) will suddenly disappear and simply leave you swinging in the wind.

It's the same false logic that tells you that beating the train at the crossing should be possible ...just as long as you have a good reason for going around the gates.

Mort, I'm slow to be critical ... but if idiots could fly, that link would take you to an airport!

Not only is the use of multi-pole breakers to control separate circuits common, the NEC actually requires it in several circumstances.

As for this thread, the math for a transformer is pretty simple. The only 'jokers' in the deck are that you use the 'nominal' voltage in the calculation, and that 1.73 figure in three-phase calculations.

Not enough transformer? Transformers are pretty tolerant of abuse. If you're really overloading one, it will be reflected in a severe voltage drop.

Harmonics are pretty much a non-issue, in that it is the practice today that every circuit have its' own neutral. This has the effect of over-sizing the neutral, which takes care of the higher amp loads caused by the harmonics.

A 'normal' common spec would be 200 amp, 3 phase, 480 volt feeder for primary of a 150 KVA transformer, with the secondary (at 400 amp, 3 phase, 4 wire; 400 amp OCP) feeding an 84 circuit panel setup (If needed).

There seems to be a pretty strong movement in NFPA to not allowing shared neutrals for single phase circuits anymore.
It works fine on the initial installation but when breakers get rearranged to create a 2-pole or 3-pole space so a new breaker can be added to the panelboard the odds are strongly against things ending up on the corect phase.

The recent changes are more about keeping the whole circuit bundled in some way in the panel. As long as the grouping is made apparent on single phase multiwire circuits there is no problem and they do have superior voltage drop characteristics.
The problems come up in 3 phase wye with harmonics. It is fairly well understood and people are doing things to mitigate it these days.

I'm not trying to be difficult... But this formula suggests that at 120 volts a 75KVA would be rated for 625 amps.
Can you help me understand wye (pun intended) we have to use the 208/3 formula?

I'm not sure how you got 625 Amps.
The secondary of a 75kVA 3 phase transformer is 208 Amps at 208Y/120 Volts (kind of a weird coincidence, huh?).

There are 2 possible ways to connect the 3 transformer windings in a 3-phase transformer; one has each winding connected to the next and forms a triangle when drawn (this is the Delta connection) and the other has 1 end of each winding connected together and forms a 3-pointed star when drawn (this is the Wye connection).

The square root of 3 comes about because the ratio of the magnitude of the voltage vector between phases and the magnitude of the voltage vector between any phase and the common point (Neutral) is equal to the square root of 3.

Take a look at this pdf from Jefferson transformer and see if this helps...sometimes the vectorial math involved in electricity gets confusing.

http://www.jeffersonelectric.com/s3/site/catalog/basics.pdf

Better late then never I guess...that link has my head in a tail spin!!

That said, what's the math that concluded 625 amps @ 120volts??

1000 x 75 / 120

It's single phase so I didn't multiply x 1.73

The math for single phase is different than the math for multiphase.

If you had a single phase transformer you wouldn't use the square root of 3 at all.

Well, it's also shown in the table you linked on page 1.4

That's 625 Amps for a single phase 75kVA transformer.
For a 3 phase 75kVA transformer it's 416 Amps.

Isn't your question about a standard 75kVA 3-phase transformer? A 75kVA single phase transformer is an oddball special order.

I didn't. 1000 x 75 / 120 = 625

I think part of the confusion is the fact that the loads I am stepping down for are 120v but the primary is 460/3.

I think I need to look at it from the panel perspective instead?

Generally what I use for loads like yours is a 3-phase transformer. 460V 3phase 3wire primary and 208Y/120V 3phase 4wire secondary.
That will feed a 208Y/120V 3phase 4wire panelboard that will be filled with 1-pole circuit breakers that will then feed the branch circuits to the individual 120V loads.

Single panelboards generally come with either 125A or 225A bus in them, the usual setup is a 75kVA transformer feeding a 225A panelboard. If you need a larger transformer you can get a larger panelboard, but you'd probably be better off breaking the loads up so you could use 2 regular 225A panelboards.

With most PC and electronic loads wanting 1 (or only a few) loads on each individual breaker; you'll generally run out of breaker positions before you run out of power.

Our intention was to use 1 dedicated circuit to 5 users. We've done this in the past via extension cords and power strips without issue.... just not on this scale.

I see you are also from Chicago. I hope this isn't out of line but if you're interested in the job, I'd love for you to come take a look.

PM me for details?

If you're going to break it up by cubicle you should think about having a receptacle for the PC and separate general receptacles.
I've seen too many situations where the computer circuit ends up with the clock, printer, desk fan, and everything else under the sun on it.
If you're going to the trouble to deliver clean power for the PCs you should do whatever you can to keep people from piling on more loads.

In one office we provided a blue-colored duplex receptacle for the PC and monitor and required that they be plugged in DIRECTLY. If people can't use power strips it tends to keep them more honest about what they plug into those circuits.

I would be interested, but my boss keeps me pretty busy where I am; thanks for the invite though.

It's not set up like that at all. Basically, the panel is going to be on the back wall of a basketball court and the only time it will be used is for our (or another) special event. This will be for semi annual gaming events where people bring their own computers. The receptacles will most likely be connected via 1900 box right to the panel and we'll use (properly sized) extension cords. There will be ONE outlet per circuit so... if anything is on it, we'll know.

The job is near Chicago Heights if you change your mind =)

So we had a guy come out and take a look. We decided to go with a 225kva for 600 amps.

We're going to split this up in to 3, 200 amp panels wired up for multi wire branch circuits.

Should I be concerned about uneven loading or am I thinking too much?

If you don't use larger neutrals when sharing circuits you may have some transformer heating, but if the transformer is already oversized you shouldn't have anything to worry about.
In the interest of storage between use some of the exhibition panels that I've done is mounted the panelboards on wheeled stands and put twistlocks on the same stand.
The various cords had twistlock plugs on them so they could be disconnected and stored (rolled up) on the same stands like garden hose would be.

If I understand you correctly, you are talking about powering up a gymnasium full of PCs. If that is the case, I would suggest placing the three panels in three different corners of the space. That way you avoid a river of extension cords all coming to one location. If you are also using cabled network connections, same suggestion, spread them out to all corners.

If you can, find a local company that does temporary shows like fairs and conventions or a theater and movie equipment rental business and go see how they set up temporary power distribution.

The previous suggestion about placing the panels on movable stands that also double as cable storage is a good idea.

Alternatively, install large power receptacles at the three locations, put a mating plug on a cord that connects to the 200 Amp panel. Now the power panels and associated cables can be rolled up and stored all together.