Generator ATS transitions

[hurk27]

From a resent thread of a problem of OCPD's opening when a generator transition from one source to another, it was stated the cause was motors being "out of phase" or acting like generators "out of phase" with the source?

While I know I have studied this over the years much has been forgotten as much time has passed by since my days at school.

First I have always thought that only single phase generators have the problem of being out of phase (180?) which would double the available voltage of the fault the transition would load into when the ATS switched into it, another words two 240 volt sources transitioned into each other at 180? crossing would effectively put 480 volts into the fault causing the available fault current to be very high.

But the issue I'm not fully understanding I guess is when we have 3-phase source transition back into another 3-phase source we are only looking at rotation and sync, and the most you can be out is 60? which is dead between 120? phase shift of a normal Polly phase system, the transition is not looking at whether source 1 pole A is lined up with source 2 pole A because in a Polly phase system it doesn't care as long as the two sources have the same rotation, in a single phase system it would be a problem because of the fact they would be 180? out.

First we don't have to involve discussion of when we transition two prime mover sources as we can clearly see that the mass of the prime mover is so much more to sync that the two sources would take much longer to line poles up (get into sync) and high current would last much longer resulting in OCPD's opening or failure of one source or the other so this would be another subject.

But with motors I find it harder to understand the problem as first the rotation will or should be the same from one source to the other if nothing has changed, second at the point of 60? being in the middle, meaning 60.9999? the motor would speed up to catch up with sync or at 59.9999? the motor would slow down to sync , while this would produce a surge it should not be at full lock rotor current, from what I figure around 50% of LRA and since the breakers would be designed to handle the full LRA they should not be opening in this situation if sized properly.

So I guess what I'm not understanding is if this problem is more like the problem with transformers where the left over residual circulating current in the transformer windings can cause an out of phase event when we transition from a generator back to the utility too fast, kind of an inrush type problem with an added kick, do motors also have this problem?

so if possible can we discuss this problem to maybe get my mind to wrap around this problem to better understand it?
I guess maybe it is because I haven't experience this type of problem before as I have with transformers so I have never really thought about it until it came up in the other thread?

 

[hillbilly1]

Most of the issues I have had with ocp's tripping, was transformer related. Turned up the magnetic trip adjustment (non ground fault type) on the breaker, and problem went away.

 

[don_resqcapt19]

It is not just rotation, it is phase angle, very much like when you close a breakar to parallel generators. Even with the correct phase rotation more than a few degrees off can be issues. With a motor, it is acting like a generator for a very short time when you transfer between two live sources. Too much difference in the phase angles will cause the motor to draw excessive current. With larger motors, (over 50hp) there are even cases of the motor shaft being broken with a fast transfer between two live sources.

 

[templdl]

Maybe a different way to look at it is when fault current is calculated. When doing so there is motor contribution that must be included into the fault. When this is considered motors do act like generators when they are de-energized as would be if there was a bolted fault that occurs causing a significant voltage dip. The motors co-generate back into the line and contribute to the fault current.
Then take the same motors that are energized and attempt to do an out of phase transfer. The motors are generating a counter EMF out of phase from the normal source when they are disconnected from the emergency source.

 

[Joethemechanic]

My first problem with the generator acting like a motor is that a squirrel cage induction motor can not generate without absorbing reactive power from the line. (or a cap bank)

The second one is something I know from field observations. A squirrel cage motor turning at some percentage of it's rotational speed starts at an amprege much less that starting from standstill (LRA)

I know this because many moons ago I had to come up with a way of starting a D&K #3 alligator shear the size of this Milwaukee or we weren't cutting scrap. (No Scrap= No Money) We had nothing but an already overtaxed 120/240 highleg service and not to mention too small of everything including funding.

Plenty of junk machines and trucks to steal parts from but not much salvage 240 V electrical equipment like discos or magnetic starters,,,,,,,,,,,,

So I made a mechanical "soft start" from an old BW 3 speed box from a Chevy.

Put it in low, close the switch, bring the flywheel up to part speed, open the switch, slam it in 2nd, close the switch, speed the flywheel up some more,,,,,,,,,,,,,,,,,,,, Well you get the idea

It took a couple of minutes and some skill to get it up to speed, but once it was there you could cut all day with it

So anyway, by limiting the mechanical load on the motor, we also limited current. Every time I shifted that BW box up a gear the motor's rotor slowed down some and upon reclosing the motor starter current stayed within reason


(BTW ours didn't have all those fancy smancy guards and an air clutch)

 

[Joethemechanic]

It is not just rotation, it is phase angle, very much like when you close a breakar to parallel generators. Even with the correct phase rotation more than a few degrees off can be issues. With a motor, it is acting like a generator for a very short time when you transfer between two live sources. Too much difference in the phase angles will cause the motor to draw excessive current. With larger motors, (over 50hp) there are even cases of the motor shaft being broken with a fast transfer between two live sources.

Paralling two generators is a lot different. AC generators for the most part are synchronous machines. Induction motors are asynchronous machines

 

[hurk27]

It is not just rotation, it is phase angle, very much like when you close a breaker to parallel generators. Even with the correct phase rotation more than a few degrees off can be issues. With a motor, it is acting like a generator for a very short time when you transfer between two live sources. Too much difference in the phase angles will cause the motor to draw excessive current. With larger motors, (over 50hp) there are even cases of the motor shaft being broken with a fast transfer between two live sources.

Ok but how is this any different then in turning off power and reapplying power before a motor has not come to a complete stop which can be a normal function in many operations?

if we haven't changed rotation of the supply the motor will slow down anytime the power is removed and be just as out of sync in the above as it would changing from one source to another when power is just simply restored.

shouldn't the regenerative power from the motor still be a problem when power is removed and restored from the same source while still turning?

again the most a motor can be out of sync electrically is 60? and this should also include when it is acting as a generator? and since the mass of the rotor is the only thing that would determine how long it stays out of sync would be the time it takes the rotor to spin up or down to sync up again.

this is what is baffling me,

I guess I can understand this if the motor was a single phase motor acting as a generator, as it can be 180? out of phase and introduce two opposing voltages against each other when re-energized?

but in 3-phase you would never have opposing voltages other then what would be up to 60? out of phase, and maybe that is what I'm missing is this 60? out of phase voltage from the motor acting as a generator, but it still doesn't explain why we don't see this when we bounce a contactor open for a couple cycles as the motor should still be generating a voltage but out of sync? am I missing something?

I guess what does not seem to make sense is I don't see where changing supply's has anything to do with a motor causing the breaker to trip is any difference to just momentarily loosing power to the same motor from the same supply?

 

[hurk27]

My first problem with the generator acting like a motor is that a squirrel cage induction motor can not generate without absorbing reactive power from the line. (or a cap bank)

The second one is something I know from field observations. A squirrel cage motor turning at some percentage of it's rotational speed starts at an amprege much less that starting from standstill (LRA)

I know this because many moons ago I had to come up with a way of starting a D&K #3 alligator shear the size of this Milwaukee or we weren't cutting scrap. (No Scrap= No Money) We had nothing but an already overtaxed 120/240 highleg service and not to mention too small of everything including funding.

Plenty of junk machines and trucks to steal parts from but not much salvage 240 V electrical equipment like discos or magnetic starters,,,,,,,,,,,,

So I made a mechanical "soft start" from an old BW 3 speed box from a Chevy.

Put it in low, close the switch, bring the flywheel up to part speed, open the switch, slam it in 2nd, close the switch, speed the flywheel up some more,,,,,,,,,,,,,,,,,,,, Well you get the idea

It took a couple of minutes and some skill to get it up to speed, but once it was there you could cut all day with it

So anyway, by limiting the mechanical load on the motor, we also limited current. Every time I shifted that BW box up a gear the motor's rotor slowed down some and upon reclosing the motor starter current stayed within reason


(BTW ours didn't have all those fancy smancy guards and an air clutch)

Interesting point and fix I might add, as I have also worked with fly wheel driven machines from stamping presses to different types of shears, and in many cases having them turned off and on without bringing them to a complete stop never led to OCPD's opening, but if the difference I'm not seeing is the fact that a whole plant of wiring that is what is giving the motors the reactance to operate as a generator where just turning one motor off and on is not, then this could be the difference I'm missing since a transfer switch is turning of and on many circuits and I'm only looking at one?

 

[Joethemechanic]

Flywheel powered machinery? For crushing eggs? :lol:

You were the egg plant guy right.

Seriously, let me find this in a textbook

 

[Joethemechanic]

From Electrical Machines Drives and Power Systems by Wildi

ISBN 0-13-367889-X

 

[Joethemechanic]

If you want me to enlarge something, or post pages previous to or proceeding the three I posted I'll try sometime tomorrow. Just let me know.

 

[iwire]

From a resent thread of a problem of OCPD's opening when a generator transition from one source to another, it was stated the cause was motors being "out of phase" or acting like generators "out of phase" with the source?

While I know I have studied this over the years much has been forgotten as much time has passed by since my days at school.

First I have always thought that only single phase generators have the problem of being out of phase (180?) which would double the available voltage of the fault the transition would load into when the ATS switched into it, another words two 240 volt sources transitioned into each other at 180? crossing would effectively put 480 volts into the fault causing the available fault current to be very high.

But the issue I'm not fully understanding I guess is when we have 3-phase source transition back into another 3-phase source we are only looking at rotation and sync, and the most you can be out is 60? which is dead between 120? phase shift of a normal Polly phase system, the transition is not looking at whether source 1 pole A is lined up with source 2 pole A because in a Polly phase system it doesn't care as long as the two sources have the same rotation, in a single phase system it would be a problem because of the fact they would be 180? out.

First we don't have to involve discussion of when we transition two prime mover sources as we can clearly see that the mass of the prime mover is so much more to sync that the two sources would take much longer to line poles up (get into sync) and high current would last much longer resulting in OCPD's opening or failure of one source or the other so this would be another subject.

But with motors I find it harder to understand the problem as first the rotation will or should be the same from one source to the other if nothing has changed, second at the point of 60? being in the middle, meaning 60.9999? the motor would speed up to catch up with sync or at 59.9999? the motor would slow down to sync , while this would produce a surge it should not be at full lock rotor current, from what I figure around 50% of LRA and since the breakers would be designed to handle the full LRA they should not be opening in this situation if sized properly.

So I guess what I'm not understanding is if this problem is more like the problem with transformers where the left over residual circulating current in the transformer windings can cause an out of phase event when we transition from a generator back to the utility too fast, kind of an inrush type problem with an added kick, do motors also have this problem?

so if possible can we discuss this problem to maybe get my mind to wrap around this problem to better understand it?
I guess maybe it is because I haven't experience this type of problem before as I have with transformers so I have never really thought about it until it came up in the other thread?

Hurk, I will be honest I did not read that entire post, I never read long posts.

That said I have some experience with this. Many of our jobs have one or more generators and we will install them test them and all works fine. We turn the building over and all is well until they have a real power failure, the generators pick up the load but when the utility comes back and the ATS transfers back we will have breakers trip.

I go in and add a 1 second transition delay and the problem goes away.

My understanding is that if the ATS closes at the peak of the sine-wave while the fields in transformers and motors are still collapsing that the inrush hits hard and trips the feeder breakers.

 

[Joethemechanic]

How fast are these switches normally?

You can lose a lot of magnetic field in a second.

I can see a transformer doing it, or an induction motor connected across the line with PF correction on it. (reactive supply) Maybe one of those fancy transformers that is supposed to fix the NL PQ issue. (Too much iron in my opinion).

I can also see some powered equipment with either a bad/slow check or unloader


How come nobody has sent an Oompa-Loompa down to see what belongs to these tripping breakers.

 

[iwire]

How fast are these switches normally?

Fast.

You can lose a lot of magnetic field in a second.

Which is likely why when I add just a second delay the issue goes away.

How come nobody has sent an Oompa-Loompa down to see what belongs to these tripping breakers.

Once again in English?

 

[iwire]

Some info

The disadvantage of open-transition transfer systems
is that during retransfer from the emergency generator
to the utility following restoration of normal power ? or
during system testing ? a fixed open time is needed
to allow the arc developed on opening the contacts to
decay. Typically this time requirement is approximately
five electrical cycles. In applications with a high
percentage of inductive loads (motors, transformers,
etc.), open-transition transfer switches need added
functionality to control the speed of operation of the
mechanical switch and keep the contacts open for
one-half to three seconds to allow time for the residual
voltage in inductive load devices to decay. This feature
is called ?programmed transition? or ?delayed transfer.?
Some other open-transition transfer switch designs
utilize phase angle monitoring between sources to
initiate switching between sources when they are
synchronized to minimize the transfer time between live
sources. These are suitable for small motor applications
where the largest motors served by the transfer switch
are not more than 20 horsepower.

From here http://www.cumminspower.com/www/literature/technicalpapers/PT-7012-ATSselection-en.pdf

 

[jumper]

How come nobody has sent an Oompa-Loompa down to see what belongs to these tripping breakers.

Once again in English?

Translation: Has a service tech checked the loads associated with the tripping breakers?

 

[Joethemechanic]

I was going to say no generator is going to start up and be on line in 5 cycles.

This is happening when going back to generator number one from generator number two


By any chance is generator number 1 still running on cool-down when the load is transfered back to it?


Why are the switches that fast? Is that some kind of hospital requirement?

 

[jcassity]

in my opinion it looks like you've addressed and answered you own question in paragraph 4.
Its hard to discuss this without discussing what you dont want to cover in paragraph 5.

again in my opinion, the utility to gen phase relationship is not going to go away unless you address the commissioning of the genset to include time delay, cooldown, ect.
Another element is we dont know your ats design weather it be a iso-byp or non-iso bypass.

most common ats's are break before make and the expense of a UPS system to aid in this transition time from utility to gen is usually found in designs where you have critical panels that can not miss a beat of power.

You are correct in that 3phase may be a little more forgiving than single phase but if your having a problem with the transitions, thats why gensets have time delays for start up to insure there is not an intermittent outage. There are also settings on the gen where you can instantly turn off the gen output when utility is detected or,, a delay to make sure the gen continues to provide load for a period of time incase the power goes off again. An additional setting to "gen cool down" keeps the gen running with no load which also can span a time to compensate for intermittent power losses.

at the end of the day, your topic is whats happening when both prim and sec power deliveries are at that sweet spot of 60deg out of sync per phase max and how that effects the loads. The answer is to be aware of it and use a tool to fix it cause it isnt going away unless you have equipment to absorb it like a Transector or Rayvoss MOV based TVSS ect with surge counters (or appropriate genset settings) Time delay settings from source x to source y will cure the issue "IF" your loads are not critical/emergency otherwise you need a UPS system to fill in the gap.

I hope i didnt stray from topic, again i dont think you have a problem understanding this, you have an issue you see, i understand your concern and its valid.

 

[iwire]

This is happening when going back to generator number one from generator number two

No, at least not in my case. This is happening when the utility is restored and the ATS shifts back to a very stiff utility supply.

By any chance is generator number 1 still running on cool-down when the load is transferred back to it?

No, not in my case.

When I said more than one generator I meant we may have an emergency and also a optional standby generators supplying different loads.

Why are the switches that fast? Is that some kind of hospital requirement?

I don't think we are talking about the same thing.

I am talking about a typical open transition ATS switching between a utility source and a back up source.

Once any time delay perimeters have been met the switch itself is as fast as it can move. In some cases this is very fast. Unless you program in an additional delay between break and make

 

[templdl]

It has been some 20 years ago be I supplied (3) 2300v autotransformer, reversing starters to AO smith to start 3 of there largest presses. The engineer specified that I provide a zero speed sensor to assure that the press was a zero speed or close to it before the starter was allowed to be reversed and I am with almost all certainty that the sensor did not use proxes or any other external sensing device but relied solely upon the EMF generated by the coasting motor.
I just can?t remember the name or maker of the device.