synchronize generators

[coulter]

...Another thought...The load can go up or down during the transfer, everything is automated. Several 150+ HP motors (and other lower hp motors) can stop and/or start during the transition (if load sharing, not an outage). In a simulated/real outage those motors will shut down, restart on a time delay. Could that sudden drop/rise in demand have an effect? Most of the big motors (blowers) are on a 20 minute delay, so I don't think they factor in with an outage, some are on a one minute delay (pumps).

Oh yeah. When the gens are coming on-line to a dead bus, there is no problem. When attempting a parallel/transfer to back to utility all of the changing loads are on line and are dealing the synchros a fit. The synchros are sending commands to move the gens to match the utility and the loads are moving the gens around at the same time. Add in the different characteristics of the two gens and the synchros have a hard time getting the gens to stay in the sync window long enough to issue a CB Close command.

I suspect that with only one gen on-line the master (utility) synchro has an easier time moving the gen to match the utility.

carl

 

[coulter]

... and you have about 12 operating modes, plus kw share and var share.) lost me there. ...

The PLC has to keep track of how the gens are operating. Here is a likely example:
G1 coming on-line to dead bus
G2 coming on-line to dead bus
G1 coming on-line (paralleling) to live bus (utility or gen doesn't matter)
G2 coming on-line (paralleling) to live bus (utility or gen doesn't matter)
G1 running, off-line, G2 paralleling to G1
G2 running, off-line, G1 paralleling to G2
G1 on-line/loaded, sync and parallel to utility
G2 on-line/loaded, sync and parallel to utility
G1 and G2 on-line/loaded, sync and parallel to utility
G1 on-line/parallel to utility, base load gen kw, base load kvar
G2 on-line/parallel to utility, base load gen kw, base load kvar
G1 and G2 on-line/parallel to utility, base load each gen kw, base load kvar
G1 and G2 on-line/paralleled, kw share gens, kvar share gens

I know that this all sounds really repetative, but the PLC has to keep track of every available mode and provide the inputs/outputs to the equipment. Unattended automatic operation is non-trivial.

... It is a Square D transfer switch if that matters? ...

Probably. But I don't know what is in the transfer switch:
PLCs?
Synchronizers?
Circuit Breakers?
Dead Bus Relays?
Generator Protection relays

My only suggestions are likely things you have already done:
Get the schematics and take every opportunity to go over the operating scenerios with the vendor reps.
Ask if there is a detailed design document, or scope document. Those will tell you how the system is suposed to work. If they are available, but not updated, get the boss to insist these be current.
Also get the boss to insist that you have a complete current set of redlines.

... The PLC programs are, ummmm, what is that word? Anyway, we can't access the program.

I think the precise engineering term is &^(@*^$@%!. Bummer you can't see the programming. That makes it hard.

... The factory guys tweak, it works, they leave, minor changes in load, utility, etc. it is out of the parameters. ...

Depending on how the installation was engineered and the equipment was supplied, you may have five vendors/factory reps to deal with:
New generator vendor/factory rep
Old gen vendor/factory rep
PLC programmer (probably works for the Package Assembler)
Synchronizer vendor/factory rep
Package assembler (new equipment)
Whoever specified the gen Protective Relay Settings

The PLC guys don't understand the gens or synchronizer programming.

The gen guys know about the required synchro outputs their gen will require, but they won't know about synchro programming modes. No extra cost bonus: The gen guys may well be engine guys, not generator guys.

The synchro factory reps won't show up at all. They are only available by phone, usually several time zones away.

Normally the one you want to beat until the system works as per plan is the Package assembler - the one that supplied the equipment and said it would meet your requirements. But if they have been paid, you're screwed.

The fix will require a PLC programmer, generator guy (or two), someone that understands the synchros, all working together. If the package assembler won't do it, then your management is going to have to make some decisions about hiring someone that can pull it together.

carl

 

[coulter]

... And in those situations the phase angle, Hz, the voltage and the RPM's of the engines themselves are fairly far apart from each other. ...

Here is where we have a difference in our understanding of the concept of a "near zero impedance bus". What you are suggesting just can't be - doesn't fit any of the known Laws of God and Physics.

... Different issue: There may be a phase angle difference in the rotor positions - generally refered to as "power angle". This has to do with the physics/mechanics of a generator developing power - the rotor phase has to lead the stator phase. Think of it as the rotor magnetic field is pulling the stator field around. Exactly.... If it is doing this it is not sharing the load, and technicly not really paralleled...

Just so we don't have a mis-undrstanding, I'm going to define the Rotor Phase Angle as the rotor position measured from the voltage output zero crossing. This rotor position has nothing to do with the phase angle of the voltage zero crossing. That is set by the bus (other generation)

Your are correct in saying that if the RPAs are not the same, the load share is likely not the same. The RPA is related to the Power Angle.

However, when the gen cbs are closed to the bus, they are technically paralleled

I'm starting to get repetative, so there's really no more I can say.

Whoops, there is one there question I would ask you:
Supposing you were correct and the two paralleled generators were running at a dfferent frequencies. By paralleled, I mean both gen CBs are closed to the bus. This would mean we would have significantly different voltage between the G1 terminals and the G2 terminals as the two voltage wave forms swept by one another. Since the two gens terminals are connected to a near Zero Impedance Bus, what would limit the current? We have a difference of potential and a near zero impedance, looks like the current is near infinite. In your mind, what would limit the current?

carl

 

[e57]

Carl there is no impedance on the buss there is an impedance on both ends of it - the windings of two generators. Or in this case one generator, the secondary and primary windings of a transformer through it fields, then another generator. These are not zero impedance items are they? It's more like a handful of rubber bands linking the two engines than some sort of solid electromagnetic-mechanical linkage as you make it sound to be.

While they may be 'technically' parallel - not truly operating in parallel if not sharing the load 50-50. Since we are talking about engine driven generators with a finite amount of power/torque and easily effected by their load, and if paralleled to each other unequally it is something you can hear in the engine, feel if near them, and see if you put on a scope.

In keeping with the entertaining (at least for me) fruit analogy and some crude sketches, (So don't take the too seriously) Lemmie show you what I mean. For the sake of discussion we'll leave the tomato - the impedance bearing transformer out of it. And since all 3 phases will act the same if also equal, we'll only talk about one as an example. (Say a few poles per winding)

Below is an 'Apples to Apples' combination - in parallel -equally loaded. Both engines bog down just a little under the peaks of the positive and negative alternations. They require the same amount of torque for similarly sized engines to do so - and most importantly they do it at the same time. And as the poles move away - the required torque is less - the shaft moves just a little more quickly through this portion of it rotation, and again they do this equally.... The result on the buss is a nice relatively clean waveform. And it is a close match to the relatively clean waveform of the utility - in this case the pomegranate. :wink:

In the Apple - Orange scenario things are much different. One generator is always going to be in the lead in terms of time. The apple is much larger and more powerful in terms of capacity and available torque required to drive the windings into the positive alternation - AND - it is being assisted by the other generator, in this case an 'orange'. So the apple takes the lead easily to its peak, right after that generator reaches peak and is heading towards it negative alternation - the orange reaches peak, and is now holding most of the load, and requires MUCH more torque from its engine for this portion of the rotation of its shaft - keeping it slower than the other 'apple generator, fortunately it too is headed for it negative alternation and has a chance to speed up - but not enough. Because now the Apple generator is already starting into is negative portion - and the cycle continues.... But the combined result on the buss is slightly different. (Forgive me I can't draw this very well - is more a rounded square wave than a nice ac waveform) Two peaks, and nearly no time when the buss is zero - because the negative alternation is already started on the leading generator. And during this time when both outputs are over-lapping there is a potential difference between the two generators - through the zero impedance buss to the higher impedance windings of each - and therefore current will flow between the two if not under load, and if under load will percentage of current shift from one to the other.

More importantly what the transfer switch sees in voltage difference between the utility (Pomegranate) shown in black as a nice AC wave form, and what is on the buss in red as a distorted waveform, is very little time where there is not a voltage difference between the two... And not enough time to make the switch.

 

[don_resqcapt19]

e57,
If that were correct, how is the output the thousands of generators connected to one of the US interconnects not approaching a flat line with the overlap of the various sine waves?
Don

 

[e57]

Don - the Op's issue is he can not transfer or load share with the utility power with both on-line - only if one or the other is on-line.

they will not transfer if both are online.

Can they ever work together?

:roll: Do you have a reason on why that might be?

Remember we are talking about fairly decent sized (quite possibly deisel) engines - not the Hoover Dam with the full force of the Colorado...

 

[coulter]

hockeyoligist2 -
Did you ever find out what was causing the problems? I'm still curious.

carl

 

[hockeyoligist2]

hockeyoligist2 -
Did you ever find out what was causing the problems? I'm still curious.

carl

Well, Carl, The factory guys fiddled around, changed some of the programming on the PLC's and declared it fixed. I don't know what they actually done. I was at another plant that day.

It transferred and everything went smooth last week when Duke Power done the weekly load share transfer. All I can do is keep my fingers crossed!

 

[hockeyoligist2]

Update! After more than a year of trying to get these suckers to work together, they announced this morning that the were putting $$$$ in the budget to hire a 3rd party (or is that 6th party?) to try and fix it...............They could have dumped the Cat, installed a matching Onan for the same price that they are paying the 3rd party! Much less all of the money already spent! I'm glad we have such brilliant management!

On the bright side, insurance didn't go up and we got a 4.5% raise!!!

 

[hockeyoligist2]

Update

Update

We have the "third party" on site this week. They are saying, dump the Cat, installed a matching Onan ! LOL

 

[electricalperson]

this is a pretty interesting thread. hopefully you find a solution to the problem. how much more money would it be to just get a matching generator and ditch the cat? is it worth messing around with a 10 year old generator when you could install a new one that will work for sure?

 

[hockeyoligist2]

this is a pretty interesting thread. hopefully you find a solution to the problem. how much more money would it be to just get a matching generator and ditch the cat? is it worth messing around with a 10 year old generator when you could install a new one that will work for sure?

MY FEELINGS EXACTLY! It's not my decision.....Unfortuantly

 

[coulter]

... is it worth messing around with a 10 year old generator when you could install a new one that will work for sure?

Work for sure??

I suspect having two matched gens is not going to solve all of the problems. It will make the PLC issues easier to deal with since the programming will be the same for both. And it likely will make programming the synchronizers easier - again both will be the same.

But it is not a magic bullet that will make synchronizing with the utility any easier.

I'm not being pesimistic: With the same gen capacity, and the same loads moving around, those issues are still the same - new gen or not.

carl