Distributed Generation 3

[dwlawrence]

I am assessing the impact of a 1600 kW synchronous generator located approximately 15 kM out from a 13.8 kV substation. There is no regulation on the feeder and the utility maintains a fixed tap that maxes the voltage at the sub out at about 106% which results in 101% at feeder end at peak load. At minimum load, feeder end is around 104%. When the generation is added the voltage at the generator is around 108% at peak load and 110% at minimum load. I can get it down by absorbing more VArs (don't get me going on vars and Vars and VArs)at the generator but that seems crude.

The feeder conductor is a mix of 1/0 ACSR and 3/0 ACSR and there is one section that's #2 ACSR which of course is the root of the problem but reconductoring is not a preferred option.

Does anyone have any ideas for mitigation? I was thinking about a feeder regulator at the substation but I've seen voltage profiles for regulators that are all over the place with high voltage at low loads and not a lot of apparent intelligence. Are they any smarter now? If I could set it to sense if the generation was on or off and boost or buck accordingly this might work.

Comments?

 

[davidbeach]

If you can stay well away from the capability curve limits, then absorbing VArs is probably the best and easiest means of voltage control. It works, don't knock it.

 

[dwlawrence]

David

That simple huh? My background is distribution where we did not have distributed generation so this DG is all new and adjusting to the notion of increasing VArs and increased losses just doesn't sit well. Guess I'll just have to get with the program.

If you got paid for all the great advice you give people on Eng Tips you'd be a rich man.

Thanks for the advice and taking the time.

 

[davidbeach]

I'm not sure how well the gen set controls will do at voltage bucking, but many base load generators spend the wee small hours of the night bucking voltage to keep the system voltage down during light loading, produce watts and absorb VArs. But stay comfortably away from the capability and stability curves of the machine.

 

[dpc]

I've seen this outcome when doing power flow studies for generation added to distribution feeders.

I agree with David Beach that I don't see a big problem with absorbing vars (leading pf for the generator) but obviously the power factor must be kept within the reactive power capability curve of the generator. There are some limitations in this region. It may be necessary to use a fairly sophisticated voltage regulator that will allow operation on voltage control but with under and over excitation limiting.

Adding regulators at the substation may not help all that much. Power has to be pushed from the generators to the substation and there is a certain voltage gradient that goes along with that.

 

[redfurry]

Generating VARs will work fine. The only downside is it will increase losses on the feeder.

Regulation will also solve the problem. The best place to put voltage regulation is about half way along the feeder. Setting it to regulate to about 102% may work quite nicely (will boost voltage if generator is not there and buck if is there). You'd have to model it to pick a good setting and also ensure there are no voltage change issues if the generator suddenly drops out.

Ian Dromey -

http://www.dromeydesign.com

 

[QBplanner]

Absorbing Var under system normal is not suggested. The reason for that is because of the Machine maintaining synchronism concerns.

You could check the Var limit of the generator and leave certain margin.

 

[ScottyUK]

Not suggested? Why not? An underexcited generator importing VArs is an entirely reasonable mode of operation. It's slightly unusual because the vast majority of loads consume VArs and thus require that the generator export VArs, but provided the machine remains within its capability curve there's no reason why it shouldn't be done. Any half-decent AVR should have a limiter which will prevent the field being reduced far enough to hit the stability limit.


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

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

 

[catserveng]

I have done a lot of portable and temporary DG systems into large and small utilites. Really didn't like operating in a leading power factor much, but with a properly monitored and protected systems, I have seen thousands of operating hours on DG units at reasonable leading power factors with no ill effects.

As said above, know the capability curve limits of your machines, also be aware of system voltage stability. Many digital voltage regulators have underexcitation protection, you just may need to watch the system in operation and see where the actual limits of stability really are. If your system is soft or can be unstable you may want to look at loss of sync protection, stator temp monitoring. At a smaller Alaska utility the protection people added a form of rate of change protection so if things started getting unstable, they tripped DG units, got things settled down, and reclosed them back on.


Easiest way to find out if unit is getting unstable, look at the field output of the AVR, preferably with a scopemeter. I have found I can find stability problems early on looking at the excitation long before it gets bad enough to affect the system voltage.

Hope that helps.

 

[ScottyUK]

Hi catserveng,

Two different stability issues there: one is the AVR control loop stability which I think is what you've been looking at. The other is the generator stability limit beyond which there is a fair chance of the generator breaking out of sync during an external electrical system fault. The simplest way of thinking of it is that the magnet created by the field is not strong enough to resist the torque from the engine and the transient torque from the fault. With a stronger magnet - higher field current - it takes more torque to break out of sync. Obviously there's more engine torque at higher power output which is why the capability curve allows a slightly more leading PF at low output than at full power giving that part of the capability curve the characteristic shape.


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

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

 

[dwlawrence]

I want to thank everyone for their input. I am now newly converted to the notion of operating the generator underexcited and the absorption of VArs as one possible mitigation measure and a bi-directional regulator halfway along the feeder as another.

One thing to keep in mind with some of the latter comments is that this thing is not being dragged in on a flatbed and connected and if things don't work out, ho hum it can be dragged away. The impact study needs to predict the outcome because someone is going to spend big dollars installing this thing running on methane from land fill and the utility has to be able to live with what happens when it connects.

So everything needs to be known ahead of time including mitigation. I think between the two mitigation methods outlined above this proposed installation can be made to work.

Again. Thanks.

 

[ScottyUK]

Most of my comments are from the utility generation world and trust me there is no way one of those generators would be going anywhere on a flatbed.

Have you considered getting some modelling work done? There are quite a few good software packages out there but I strongly advise that you get an experienced consultant to do the modelling rather than try to do it yourself.


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

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

 

[dwlawrence]

I meant more figuratively rather than literally.

Actually I am using software and that's what alerted me to the problem and I'm pleased to report that when I model the suggestions above they work.

It goes something like this. If utility reduces taps to get their voltage down to 103% then at min load at generator

PF = 0.98 V = 109%
PF = 0.95 V = 107.9%
PF = 0.9 V = 106%
PF = 0.98 w reg V = 107.2%
PF = 0.95 w reg V = 105.9%
PF = 0.9 w reg V = 104.9%

PF is overexcited importing VArs

This is a bit of an oversimplified summary but the point is there's a solution and with some more tweaking of the parameters it looks like it can be made to work out.

 

[ScottyUK]

Conventional notation is that overexcited = exporting VArs for a generator. Pleased that the model is agreeing with what we expect to happen in the real world.



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

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

 

[QBplanner]

if you read notes carefully, not suggested under system normal.
Not not suggested all the time.

"The other is the generator stability limit beyond which there is a fair chance of the generator breaking out of sync during an external electrical system fault."
Not necessary an external fault, a generator can still be O.O.S due to small disturbances depends on the characteristics of the machine ,interconnection network and the rest of the system.

In the utility I work, all the major generators (above 200mw voltage control buses)are all not suggested absorbing var under system normal.

 

[QBplanner]

if that happens (absorbing Var even under system normal condition)during design stage, we will typically install reactors to do the work instead of relying on the generators.

Since I work on tranmission side, there may be rules different from DG due to the different roles in the system.

 

[QBplanner]

ScottyUK:
machine remains syn with the system is not simply a generator problem but related to the system. transiently stable after a fault does not mean the machine can remain syn.after minutes. diffferent countries have different network topologies. Some ar estrong some are not.I believe what you said based on what you experienced but maybe something you havenot seen and have not experienced.
I moved fro mone place to the other and I saw different designs, networks and events.
Have you seen people design a 500km single 13xkV line system with machines at the end. Have you seen seen 180km 2xkV network? Maybe you do but I did not in the past but gained a lot after I experienced.

 

[slavag]

QBplanner.
You are right, all depend
OP say about 1600kW generator, absorb kVar isn't problem.
of course according to gen stability curves.

For example, one of our member story to us about Spain, where he must in the night absorb Var's.

BTW, 500km of 13kV...isn't picnik, good luck to you.
Best Regards.
Slava

 

[rmw]

Scotty,

Not so fast there. The generators from your former life (at least the ones that were CT driven) were among the largest that your favorite generator company every built. Plenty of their stuff could be and regularly is carried on platform trailers (drop deck type naturally). I think the same regarding size is true of the Circle W gas turbines you used too, or at least true at the time they were built. But there are some real petite little CT turbogenerator packages out there-nothing like the monsters that you lived with. I think I could just about fit the engine out of a GE LM 2500 in a large pickup-said tongue in cheek, naturally.

I had an opportunity to visit Falcon Works a couple of years back when one of your (formerly your) exciter armatures was in their shop and when I saw it, it was staged in the final assembly area along with some of their new product and it was every bit as large as some of the whole complete generators that they were in the process of shipping. I had to ask what that big ugly black thing was because of its unproportionality to the other stuff I was seeing in their shop. That is when they told me that it was from Teeside.

On another visit to Falcon Works, one of Teeside's rotors was there and I had just visited another generator shop on the continent and had observed a 500 MW hydrogen cooled rotor and your air cooled rotor wasn't discernably much smaller than that 500 MW hydrogen cooled rotor. That rotor by itself probably weighed as much as some of the smaller generators that your favorite generator company builds.

I once observed one of their (crated) generators in the Houston area on a drop deck type flatbed trailer high centered on a railroad track. Cute. Glad it wasn't my problem.

rmw

 

[ScottyUK]

Hi rmw - I honestly don't have a massive problem with Brush, but I can't pretend that the Brush machines were anywhere near the build quality of the Westinghouse machines at Teesside, nor of some of the older Parsons, GEC, and English Electric machines I'd been periodically involved with prior to joining Teesside. You touched on the likely reason for some of the problems: they were pretty big for air-cooled technology of the time, and large in terms of the experience at Loughborough. Regarding transport I can remember in the early 80's seeing one of the 750MVA sets leaving Heaton works, probably bound for Drax. It was very impressive, particularly if you recall how tight the roads are around the old Parsons works. So certainly they move via road, but not on a regular flatbed!

QBplanner,

I appreciate your comments regarding transmission systems. My experience is UK-based where we have a geographically small and heavily interconnected grid, and I sometimes forget that other grids can behave very differently to ours. Thanks for the information.


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

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