Connecting Generator in Double Delta 3

[IBRCAN]

I have a customer with a 100 kw 3 phase Stamford generator, and he wants to re-connect it for single phase 240/120V operation. I realize this will de-rate the available power output to 67% of rated kVA.

What I am not sure about is the connection drawing states that if 120V (L-N) is needed from the genset, then the neutral point must not be earthed. Please refer to the attachment. Does anyone know why this is? This installation will fall under the Canadian electrical code, if that makes any difference.

 

[xnuke]

That's because N isn't your neutral point in the configuration they show, W is the neutral point. U and N are your L-L hot phases. Earth W. If you earthed N, you wouldn't have a standard split-phase output, you'd have a 120 V L-N from W to N and a 240 V L-L from U to N, both with the same phase angle. Also, if you look at wires 6, 7, & 8, you'll see the AVR connects wires 7 & 8 from N-U to ensure you get a regulated L-L voltage, and 6 is connected such that 6 and 8 are across a single phase winding V5-V6. If you grounded N, the AVR would be connected with 6 on the wrong delta corner.

xnuke
"Live and act within the limit of your knowledge and keep expanding it to the limit of your life." Ayn Rand, Atlas Shrugged.
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[waross]

Exactly as pointed out by xnuke;
"N" is the neutral point for the star or wye connection.
When reconnected in double delta, "W: becomes the neutral. Be sure to locate and remove any grounding jumpers presently connected from "N" to ground.
While the KVA drops to 2/3 of the original rating the PF goes from 80% to 100%. (With a standard generator rated at 80% PF.)
The new allowable kW at 100% PF will be about 83% of the original kW rating.


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

 

[edison123]

A very interesting connection. Need to study it. lps anyways.

Muthu

http://www.edison.co.in

 

[waross]

Hi Muthu.
The double delta connection has become standard among gen-set makers to convert from three phase to single phase.
Single phase gen-sets are generally only built in sizes up to about 15 KVA for four pole machines and up to about 25 KVA for two pole machines. For larger sizes it is common to supply a reconnected three phase set. The same engine is used so that the machine will be able to deliver full kW if it is reconnected for three phase use. This allows the set to be uprated from 80% PF to 100% PF when reconnected for single phase use.
For single phase loads applied to a delta connection I find it helpful to consider the virtual transformer formed by an open delta.
Think of it as a virtual transformer in parallel with a real, single phase transformer. If you look at the voltage drop under load of the virtual transformer you will find that the terminal voltage and the regulation is equal to the real transformer.
The connection acts as if two, equal, transformers were connected in series. The same current flows in all three transformers.
For a resistive load, the current is in phase with the real transformer. That puts the current through the two transformers making up the open delta at 50% PF due to the phase angles. One will be at 50% leading and the other will be at 50% lagging.
The effect is the same whether the delta is formed by transformers or by a delta connected generator.


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

 

[IBRCAN]

Thanks xnuke and waross...learned something new...its you guys who deserve the LPS.

 

[waross]

My favourite connection for reconnecting older sets is the zig-zag connection.
You start with two leads for L1, two leads for L2 and two leads for L3 plus 6 leads for N.
Identify the pairs of leads on the neutral and remove one pair of leads. Connect that pair of leads to one of the other pairs of line leads and insulate and forget.
Now you have two line connections and a neutral connection left. No matter which pair you have removed from the neutral and which line leads you have connected it to, you will have center tapped single phase power from the remaining line leads and the neutral.
Even if this is a very old machine with wire numbers unreadable or missing, this is a quick and easy conversion. No numbers needed.
With jumper straps and a diagram for a double delta connection, I would probably go double delta in your case.

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

 

[JohnMcNutt]

Onan was always into double delta, and Stamford is the successor of Onan, sort of anyway.

Bill, I am not sure if I read it correctly but it seems you are claiming the same current flows in the in-phase and out-of-phase windings. Through empirical observation I found this to be not true. The out of phase windings in series carry 1/3 of the current and the in phase winding carries 2/3.

However, when connected in this fashion, the total ohmic heat dissipation in the total winding package is less than if they are connected in dog leg (zig zag).

I also go for dog leg if I have no manufacturers instructions.

 

[waross]

Consider:
You have two in phase windings and four out of phase windings.
The out of phase windings are connected in pairs to give two sets of dog leg windings.
Each dog leg pair develops 120 Volts. If you draw the vector sketches of the voltage drops you will see that the dog leg winding has the same voltage regulation under load as the in phase winding.
The currents are the same in all of the windings.
An old friend used to say, when the theoretical does not agree with the field measurements, something is wrong.
In this case there is too much theoretical evidence contradicting your empirical measurements not to mention other empirical measurements.
I suspect an issue with your measurements.
For the same load, the heat loss is the same for either double delta or zig-zag.
Full load at a 2:1 current ratio would result in overloading due to the square factor of the I[sup]2[/sup]R law.

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

 

[JohnMcNutt]

I've got a double delta machine to commission this week or next. Nobody has laid a finger on it since it came from the factory. I'll try to remember to measure the individual windings.

The prior measurement method was a known good Fluke amp clamp at appropriate points in the connection box. Agreeing with the display on the load bank. Near full load, resistive.