Circulating Currents / Gen / DeltaWye

[jimgineer]

The setup that is currently existing involves a pair of 480V generators that feed transformers that step the voltage up to 13.8kV , and in turn feed a piece of switchgear at this voltage. Utility feeds into switchgear, and switchgear feeds a couple buildings, presumably being dropped down to 480 or other distribution level voltage closer to the loads.

The question I have been asked and didn’t have a good answer to involves explaining the tradeoffs in selecting a delta or wye winding on the generator sides of those transformers. It seems like on the switchgear side, delta will be fine, and as I understand it slightly cheaper for both the transformer and the fewer conductors that are pulled (a N/G instead of a separate N and then G). But on the generator side, what goes into the selection of whether the generator needs to supply a neutral to this transformer?

I realize this is a subject that is very deep but I need to develop a much better pros/cons kind of engineering tradeoff understanding of this decision.

Thanks in advance to you smart ones-

 

[davidbeach]

There's a paper on the Beckwith Relay site that talks about this issue, but in general if you have generators connected to transformers with no load (at least no line-neutral load) tapped off between the generators and the transformer the best configuration is delta on the generator side and grounded wye on the load/system side. Then the generators can be high resistance grounded and protected as well as is possible. The utility is likely to require that you appear to be an effectively grounded source while operating in parallel, I know I would, and the grounded wye provides that.

 

[rcw retired EE]

If your generators are the only power source to the 13.8 kV system, a wye winding at 13.8 kV as David sugggests would also provide the best ground fault protection and coordination.

Delta at 480V with a ground fault detection system would be best for the generator connections, as David said, if there are not too many other 480V loads on that bus.

Feeders from the 13.8kV wye transformer only have to be three wire. A neutral wire is not needed. The step down transformers at the loads would be the typical delta wye with delta on the 13.8 kV.

 

[dpc]

Since you have a utility interface at 13.8 kV, whatever you do is going to be something of a compromise. But I agree with davidbeach that making the generator side of the transformer a delta winding is the best option. The utility will be happier with a wye on their side and you can have 13.8 kV ground relaying to keep things sorted out.

The relative cost tradeoffs between delta and wye windings and the costs for a neutral conductor are trumped by the fundamental protection and isolation issues associated with a utility interface to a local generation system.

The serving utility will want to have a major input into the transformer configuration since it will impact their system protection and relaying.

 

[waross]

The delta on the generator side does a better job of distributing uneven currents. Not perfect but better.


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

 

[davidbeach]

The serving utility will want to have a major input into the transformer configuration since it will impact their system protection and relaying.

Yep, what I posted above is what someone trying to connect to our system would be told. The only other reasonable option would be solidly grounded generators and a grounded wye-grounded wye transformer. That would be what you would want if you have to serve line-neutral loads on the 480V system directly off the generators. At 480V it works fine, above that the delta transformer and high resistance grounded generator is the best option.

If your generators are big enough to justify stepping up to 13kV and you have no 480V loads on the generator bus you might want to look at 4160V for the generators. Your circuit between generator and will need only 11.5% the ampacity at 4160V as it will at 480V. 2MW requires a 2770A circuit at 480V but only 320A at 4160V. One or two conductors per phase vs. nine per phase.

 

[prc]

David, I had heard that we prefer delta on Generator transformer LV to avoid zero sequence current circulating in generator stator windings ( stator cannot stand zero sequence currents- Iam told)for a LG fault on HV side.Is it right?

But I have seen star/star GT where generator is used to supply the plant also to simultaneously pump in to grid at times of low demand in plant

As far as transformer is concerned,from economic point both delta and star are the same for this voltage level.

 

[ScottyUK]

Hi PRC,

Using a delta LV winding and HRG on the generator neutral limits the damage that a stator earth fault will cause. With an HRG system limited to about 10A you might get a small burn on the core and need some remedial work to avoid a hot spot. With solid grounding and a fault level of 50kA or so you will probably need a new core or major rebuild plus a full rewind.

Are you confusing the effects of stator zero sequence currents with rotor negative sequence currents? The latter are highly undesirable, and more so on larger machines.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[rockman7892]

Davidbeach

I'm not following why you would need the transformer to be a grounded wye on the primary if L-N single phase loads were needed to be supplied directly off of the grounded wye generator?

 

[davidbeach]

Primary is ambiguous. Please refer to high side or low side.

Grounded wye on the high side makes the generation appear effectively grounded as seen by the utility. To be effectively grounded, the low (generator) side of the transformer needs to be either a grounded wye connected to solidly grounded generators or a delta and any generator grounding scheme. No probably with solidly grounded generator connected to a delta winding, but if you do that you won't be able to serve the L-N connected load from the utility when the generators are off line.

 

[jimgineer]

"Delta at 480V with a ground fault detection system would be best for the generator connections, as David said, if there are not too many other 480V loads on that bus."

Also please explain how the wye on the 13.8 side in this referenced example would be the better option for ground fault protection.

Can someone please describe how ground fault protection would work on a 3W system?

I'm used to in the simplest case, when a 4W system is being used, installing a CT on the bond between the nuetral and ground bus bars (solidly grounded system). I understand this not to work for more 'complicated' setups, we ran into this last week actually.

 

[davidbeach]

Ground current will flow and can be sensed. All that is needed to sense the ground current is a CT on each phase, the sum of the three phase currents being the ground current. Having ground current flow to the fault allows easy location of the fault. In a high resistance grounded system or an ungrounded (actually capacitively grounded) ground faults are detected by voltage shifts but can not be readily located.

The most significant benefit of the 13.8kV being effectively grounded is for voltage stability during ground faults. That's why the utility will probably require effective grounding.

 

[dpc]

For a three wire system, the sum of the three phase currents should always equal zero. If it doesn't, there is a ground fault. It's much simpler than a 4-wire system.

Regarding the 13.8 kV side - look at it from the utility perspective. If there is a ground fault on their 13.8 kV feeder, they need to clear the fault. They also do not want to end up tripping their source and leaving your generators supplying their customers. Without a ground source at your generators, it is harder to detect a 13.8 kV ground fault at your facility. With the grounded wye, there will be plenty of ground fault current provided by the delta-wye transformer and they can trip you off the line nice and fast.

 

[rockman7892]

What would be the issue is the low side of the transformer was a solidly grounded wye and the generator was connected as something other than a solidly grounded wye such as a delta or resistance grounded

 

[dpc]

See ScottyUK's comment above. Generators will suffer greater damage from ground faults if solidly grounded. Moving the ground from the generator to the transformer might reduce the duration of the stator ground fault, but the magnitude will still be high if solidly-grounded.

If these 480 V generators are directly serving 480/277 system with 277 V loads, then they need to be solidly grounded. Otherwise, it's much better to use some type of resistance grounding.

 

[davidbeach]

Two differences.

The first is that the low side system would look very different fed only from the utility vs. fed only from the generator.

The second is that when seen from the utility side a grounded wye-grounded wye transformer with the generator not solidly grounded would not present an effectively grounded source to the system. The grounded wye-grounded wye transformer is not a zero-sequence source the way that a delta (low side)-grounded wye (high side) would be and the generator grounding is the zero-sequence source, so it now matters.

 

[prc]

ScottyUK and David. Let me quote from a book on transformers "Power Transformers -Principles and Applications" John J winders Chapter 2 Clause 2.5 (2002 Mercel decker)

The delat-wye transformer connection is used for connecting generators to transmission systems because of two very important reasons.

First generators are provided with sensitive ground fault relay protection.The delta -wye trf is a source of ground currents for loads and faults on the transmission system,yet the generatorground fault protection is completely isolated from ground currents on the primary side of trfs.

Second rotating machines can literally be shaken apart by mechanical forces resulting from zero sequence currents.The delta connected winding blocks zero sequence currents on the transmission system from the generator.

Unquote

I wish to get your expert opinion on the above view.Do you agree.

I wish to correct myself.When export /import of power is involved I have seen star/delta generator trfs feeding to grid-not star/star with which it will be impossible.

 

[ScottyUK]

prc,

No argument that heavy imbalance on the loading can cause problems. The reason is magnetic assymetry which occurs in the air gap due to the interaction of unevely loaded phase conductors in the stator, which leads to potentially very powerful forces acting to move the rotor out of its normal rotational path. Those forces ultimately act on the machine bearings and will show up as a vibration signature. To be honest it's not something I have spent a vast amount of time looking at because large machines always have a delta wound GSU transformer with an HRG scheme and it has not been of huge practical significance to the utility class machines I work with, certainly not in the way that negative sequence currents can affect our machines. The link below might be useful - I bookmarked it a while ago as 'interesting' and it has some explanatory narrative in the first few pages.

http://www.itee.uq.edu.au/~aupec/aupec03/papers/008 Oliquino paper.pdf

----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[davidbeach]

I don't have any argument with that. Everything I've seen to date has dealt with the machine ground fault aspect of connecting to a delta winding, but the mechanical issues of imbalanced loading makes sense. Put the delta in for its grounding benefits and you get the other benefits for free.

 

[jimgineer]

I’m still not completely seeing – we’ve used NGR (also termed NER on these boards I’ve noticed) which by my current understanding is a form of high resistance grounding on the generator. High resistance/impedence grounding, as well as solidly grounding, as well as low resistance/impedence grounding are all a way of operating the generator as a wye source. Please correct me if I am wrong. But the main difference that I have noticed when we choose to do it this way, is that the generator would be supplying a switchboard/gear directly, where a neutral needs to be derived (for example 277/1 loads are needed like someone else mentioned). The alternative that we have been discussing involves if the generator is feeding a transformer on the low side, and then this is being stepped up to a MV and then going to a switchboard, etc.

To summarize what I do think I understand; Ground fault is easy to detect when a 3W (delta) feed from a generator is used, because ground fault current = sum of phase currents which will equal 0 in a no fault condition.

Now comes the 87 (differential current) detection. How does this tie in with this discussion? I understand this to be a CT on the ground conductor at the generator, and this current is compared to the ground current in whatever the generator is feeding. But in the case we have been discussing where there is a transformer in between, would the 87 relay function be useful? How is this related, and am I off base in trying to relate this to the selection of the grounding method of the generator.

Thanks in advance for access to your wealth of knowledge.