One transformer or two?

[fsck]

The electrician planner has spec'ed a 112.5kva 480 delta:208/120 wye transformer. This is to handle the house and pool loads. This is a challenge but allegedly not impossible to also feed from the limited standby power.

I was debating in my head the pros & cons of instead having say a ~75kva, and a ~30kva ; with the 30 going to the designated loads needing backup power. We'd disconnect the 75 from the 480 when the grid fails.

What are the pros and cons of each? My thoughts:

One transformer:
[ul]
  [li]lower initial cost[/li]
  [li]lower losses in normal ops[/li]  
[li]lower installation costs[/li]
[/ul]




One big, one small:
[ul]
[li]far easier to start from standby[/li]
  [li]better line regulation[/li]
  [li]less % of standby power lost in xfmr[/li]
[/ul]

The 15KVA of inverter is only 7.5% of the the transformer's rating. Is that asking for trouble in terms of voltage regulation of the output?

I can't recall which losses in a transformer are a percentage of load, and which are not, but at least the [~2% rated] core losses are a noticeable of the standby power.

Is there some other issue I've not thought of? [I'm sure there is....]

 

[davidbeach]

How about posting one-line diagrams of the two options? Seems to me that you put the transfer between normal and standby (what is the standby source?) downstream of the transformer and don't worry about it.

 

[waross]

I agree with David. One transformer add the transfer switch for the critical loads on the 120/208 Volt system.
One large transformer will often have less voltage dip than either of two smaller ones for motor starting loads.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[fsck]

Here's a very simplified one.

The backup is 15KVA of "Sunny Island" inverters, backed up by an on-demand 25KVA generator, and [not-shown, and daylight only] 20KVA of PV-driven grid-tie inverters.

 

[fsck]

Re: motor starting loads.

The bigger ones [not shown] will be 480, tapped off at the transformer. Alas, some will be single-phase 208; because we can't get compliant ones in 3 phase, period.

 

[waross]

Your one line doesn't show any 480 Volt loads.
Most of us would try very hard to move the standby closer to the transformers.
If that can't be done, you may have trouble handling the transformer inrush.
I would investigate this operation sequence.
1 A> When PGE fails, the transfer switch connects the generator to the feeder and the transformer.
1 B> The load shed contactor opens.
2> The generator starts with the transformer connected.
3> When the generator voltage and frequency are established the inverters are sync'ed and connected.
You may need additional load shedding until the inverters are online and for loads that don't like to be ramped up.
10 HP may be the maximum HP that you can start on the generator.
Will the inverters be able to sync and supply power with the standard generator no load frequency of 61.8 Hz?
Before you go isochronous know that isochronous control responds to block loading (or unloading) with a droop or frequency error, and then slowly corrects the frequency.
Droop control (61.8 Hz at no load) is transparent to most users and is very common and dependable.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[fsck]

Your one line doesn't show any 480 Volt loads.

I said simplified. A 480 panel will T off at the transformer primary. Those loads will also be shed by interrupting their starter control
circuit.

Most of us would try very hard to move the standby closer to the transformers.
If that can't be done, you may have trouble handling the transformer inrush.

{Moving} has major issues of its own. Inrush is an issue I've explored in another thread. Several solutions have popped up, a reduced-voltage start & a TSR.

I would investigate this operation sequence.
1 A> When PGE fails, the transfer switch connects the generator to the feeder and the transformer.
1 B> The load shed contactor opens.
2> The generator starts with the transformer connected.

The goal is to start and run from the inverter, with the generator starting later only as needed.

Will the inverters be able to sync and supply power with the standard generator no load frequency of 61.8 Hz?

The inverter supplier claims that the standby inverters will start and run, the PV-fed grid-tie inverters will sync up and contribute, and further, when the standby inverter starts the generator, they will track the generator sync with all contributing.

Generator or grid-tie: the backup inverter shall send any excess supply to recharge the standby bank.

 

[davidbeach]

When you say PGE do you mean PGE (Portland General Electric) or do you mean PG&E (Pacific Gas & Electric)? Or something altogether different?

 

[fsck]

PG&E, aka pge.com; the location is near Santa Cruz.

They'll put a HV:480/277Y transformer in place to feed the illustrated meter.

 

[waross]

Yes, I understand that standby equipment may not always be located in the most efficient place electrically.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[fsck]

It has to be collocated with the PV/grid-tie system, or that inve$tment is useless during the day+ long blackouts.

Needless to say, you can't put the PV array in the middle of the forest.