Yet another gen grounding question 4

[alehman]

My client has a system with two 2M DG sets, 12.47kV on a paralleling board. Each is grounded via an 18 ohm resistor. We are proposing to add a third - the gear is set up for eight ultimately. My concern - 18 ohms seems like a rather low resistance value, esp. for what could be eight units in parallel. Should I be concerned for the potential damage if there is a winding ground fault? I'm trying to decide if we should recommend increasing the grounding resistance.

GF protection is 87G and 51N. Windings are 2/3 pitch.

Thanks.

 

[dpc]

Zigzag grounding transformers are virtually extinct in the US. I've put one in in 35 years and I've only seen a few others, most of them relics.

As far as zigzag CB trip - you NEVER want to trip the zigzag off the line. If you trip it, it should be the last thing on the system to go.

 

[alehman]

dpc, thanks for the link to that presentation. One or two of the articles referenced therein were also very helpful.

The jist I get is that for external ground fault contribution less than 1200 to 1600 A or so, the damage from the internal contribution is typically greater than that from external sources. Above that level of external fault current, external sources may become the greater contributor. All dependent on clearing and decay times of course.

 

[dpc]

I think that the damage level will depend on the method of local grounding used. If you have ground differential, the external fault should be gone within 6 to 8 cycles. But there are a lot of smaller generators in service out there without ground diff.

The main impetus behind the new IEEE Working Group recommendations was the complaints of the insurance companies who were getting tired of paying to restack stators on generators that were protected based on the current IEEE protection recommendations. The low resistance grounding methods, while a big improvement over solid grounding, were still allowing significant stator damage.

I recently did some relay settings for a new 13.8 kV system at a large plant with lots of large motors on it. Their grounding resistor is sized for 1000 A. I recommended that this be reduced to 100 A, but the design consultant said that 1000 A was their "standard", although they couldn't really explain why. Old ways die hard in this business.

I think just increasing the resistance for the low-resistance grounding method would be a big improvement for many installations. But this also means maybe replacing relays, CTs, etc.

 

[alehman]

Yea, I was assuming 87 protection. Not to get off on another topic, but my client has a propensity to connect large (comparable to one generator kVA) transformers downstream of the ATS, so they sometimes end up with long delays on generator 87 to avoid nuisance trips. Not good.

Old ways die hard in this business.

 

[slavag]

Hello Alehman.
What is delay on 87G? Only right choice of CT's avoid nuisance trip. Don't agree with any delay on 87 protection
( if it's possible)
Dave, 1000A it seems so much.
Up today in my area used 300-500A ( resistors for 10sec only), but I think your 100A solution is more better. For newer relay isn't problem.
Yerstaday, we check some 22kV SWG with newer relays. On the 400/5A CT for curios we inject primary 4,4.1,4.2A. Relay show those currents 1:1.
Regards.
Slava

 

[alehman]

CT's are another item I'm looking at on this project. The existing drawings are not clear. One source says the existing are ANSI class C20, which seems very low. Obviously need to do a burden calculation once I know for sure.

 

[slavag]

C20, I don't remember correlation between ANSI and IEC CT types. I see, you have,very "fun" project and customer.
Actally once, we need organizaied FAQ with correlation between ANSI and IEC for CT, VT and wiring size.
Regards.
Slava

 

[davidbeach]

C20? That not a CT; thats a current sensor. ;-)

A generator is not the place for low quality CTs. For slavag, a C20 CT will support a standard burden of 0.2[Ω] at 100A secondary without saturation, but there is also an assumed X/R in that and on a generator it is highly likely that the X/R will be considerably higher.

My thoughts:

Available ground fault current is much too high, but unfortunately all too common.

Hybrid grounding as mentioned by dpc could save you a restack, but at 2MW may not be considered economically viable.

Generator differential with good CTs doesn't need to wait for anything else, but you have to know that the CTs can never saturate.

Full differential (6 CTs) is better than ground differential (4 CTs) and the more ground resistance (otherwise a good thing) the more poorly ground differential will work, particularly with the breaker open.

For the single resistor between a common neutral bus and ground, you could add a neutral disconnect switch, key interlocked with the generator breaker. That way when a generator is off-line for servicing you could disconnect it from the neutral once you have also locked out the breaker.

With what I've learned over the years, for that size generator, if starting a new installation, I'd buy the less expensive 480V generator and use those savings (and probably a bit more) on the purchase of a unit transformer. There is a 12.47kV generator installation I designed once that I was sorely disappointed when the project was canceled. The longer it's been since then the more I'm glad it was never built, because it would have had solidly grounded generators.

I'm now on the other side of the table and it is my job to tell customers that want to connect their generation to our system that it has to look effectively grounded at the point of common coupling. Above 480V (or 600V where that is used) solidly grounded should never be used, low impedance grounded should be avoided if at all possible (much easier for new installations than existing) and high resistance grounded coupled to a unit transformer is the gold standard. The hybrid grounding systems are a vast improvement over the low resistance grounded systems.

 

[alehman]

Thanks David. I agree on the CT's, but I must go through the exercise of proving it to my client.

 

[slavag]

Alehman, David.
Maybe for the maintanance used individual neutral switch for each generator?
Regards.
Slava

 

[alehman]

I found this tool to be useful in calculating CT performance.

http://www.nxtphase.com/documents/CT_SAT_Calculator - IEEE Version.xls

mikeholt.com has a good tool for calculating wiring impedance.

 

[QCE]

Quote 'Above 480V (or 600V where that is used) solidly grounded should never be used, low impedance grounded should be avoided if at all possible (much easier for new installations than existing) and high resistance grounded coupled to a unit transformer is the gold standard. The hybrid grounding systems are a vast improvement over the low resistance grounded systems.'

In USA and Canda every large hydropower generator I worked with used a transformer and resistor for grounding the generator neutral. In Europe and Asia I have never seen a transformer used - only a resistor. Why is this?