Generator operating frequencies. 27

[HamburgerHelper]

I have thought about this for awhile and I don't understand it. So, let's say you have generation spread out over a large region and there is a disturbance in the system that causes one of the generators to be at a lower or high frequency than the rest of the generators in the system and the controls don't work to bring that generator back up to normal frequency. What happens? I have a hard time understanding this because in my mind if a generator is operating at a different frequency than the rest of the system, that generator or island around the generator is effectively isolated from the rest of the system from a power flow perspective. The rest of the grid is going to try to motor or add generation to it as the phase angle of the different frequency generation slips around the rest of the grid. I just have a hard time grasping why a generator can operate for example at 59 hz while the rest of the grid is humming along at 60 hz.

 

[ScottyUK]

HH,

I suspect that minimum frequency data in that report is measured on a single-cycle basis, so during significant swings in load angle there will be isolated cycles which are slightly short or slightly long giving anomalous frequency readings. The report itself alludes to angle changes in the comment on page 6 relating to the trip event.

I'm not completely in agreement with the author's use of the term 'minimum frequency' to describe these single-cycle events.

 

[HamburgerHelper]

ScottyUk,

Go read the other reports. There is ringing across the system and other events that don't resolve themselves after a quick changes in generation or load. It is interesting reading even if you don't agree with me. There are bunches of cases where the generators keep on oscillating in respect to each other minutes after a quick change. In my opinion, there is likely more odd stuff that happens on the grid that previously went unnoticed or was assumed didn't happen due to not having synchrophasor data, time stamped voltages and currents over large regions. As hard as wind generation can ramp in ERCOT, (+/- 1 GW 10 minute ramps), it isn't very surprising to me that Dr.Grady watches the system closely during high wind penetration.


I brought this whole thing up because I saw relaying that was set for very low under frequency tripping. So low, I had a hard time understanding how the system could get there without being islanded or the whole interconnect being in real trouble. Now I believe that large block changes in load or generation could sag the frequency quickly to beat the generator ramp rates and cause it to under frequency trip without the whole system going down. The oscillography I saw when a load got switched in likewise was not felt the entire region. Maybe, this isn't a profound question but just something I hadn't thought much about before that generation can co-exist at different frequencies ,often very slightly different frequencies, so long as the deviations don't produce an angular difference to cause stability issues. So you can have odd things, like generators ringing for minutes without them causing any issues in the system.


I do find it kind of curious the definition of frequency is somehow the crux of this. If I said the phase angles of the system voltages will sometimes change relative to each other, everyone would be in an agreement. If I said that the frequencies of each generator will be different relative to each other during a change, some I guess will disagree. This is all while most of us are using microprocessor relays that operate using subcycle fourier filters to screen out the fundamental and various harmonics.

 

[waross]

Grid frequency is a measure of the number of cycles of alternating current in one second.
Each cycles lasts approximately 1/60 seconds or 0.01667 seconds.
Short term measurements of frequency in the order of cycles may be reported as a frequency excursions.
Let's go back to the line shaft analogy:
Imagine that the line shaft drives the various loads through sheaves that have resiliant hubs, so that as the torque increases the resilient hub will allow as much as one quarter of a turn lag as the load/torque on that sheave is increased from zero to full load, the angle of lag increases up to one quarter of a turn.
Now someone applies very sophisticated instrumentation to each sheave to measure the instantaneous speed of each sheave.
He will be able to prove that as the load is increased or decreased the speed of the individual sheaves changes in relation to the speed of the line shaft.
"I just have a hard time grasping why a generator sheave can operate for example at 59 RPM while the rest of the grid line shaft is humming along at 60 RPM."
It may be technically correct to report phase shifts as frequency excursions but it is misleading.
Your synchrophasor data is valid data but please put it into perspective.
Look at the duration and magnitude of the frequency shifts that the synchrophasor reports and calculate how many actual degrees of lead or lag that it relates to.
How many times have you seen accuracy reported as"accuracy plus or minus x.x% plus or minus one count"?
There is a loose analogy here where you are extracting a digital value from analog data.
Not wrong but very misleading.
Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[MikeHalloran]

<tangent>
My Grandfather had a business card designating him as
Superintendent of Phase Change
for what was then Niagara Mohawk
in upstate New York.

My Dad explained that it wasn't really 'phase' they were changing,
but grid frequency, as a bunch of smaller grids were integrated,
and not coincidentally changed from 25Hz or 50Hz to 60Hz.

When I began to understand what that meant,
I asked about instantaneous frequency
and Dad explained that the system guaranteed
the correct number of cycles in a day,
but not necessarily the correct number of cycles in a given minute.

One of the unobvious operations associated with the 'phase' change
was replacement of affected customers' clock motors.
For reasons of his own,
perhaps just for the amusement of a curious youngster,
Gramps had accumulated several large boxes of useless
but perfectly good Telechron clock motors.

</tangent>

Mike Halloran
Pembroke Pines, FL, USA

 

[waross]

Hi Mike
I worked for a time in a very old facility in Toronto. We decided to reactivate a part of the plant that had been in disuse for years. I did a survey of the equipment in place.
I found a number of motors still bearing the tags indicating that they had been rewound from 25 Hz to 60 Hz as part of the Canadian harmonization with your Grandfather's project.

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

 

[davidbeach]

The line shaft may not be the best analogy. Over a large interconnection there can be inter-area oscillations such that one area is slightly slower than the other at one moment. The slower area speeds up and the faster area slows down. Both overshoot. The pattern continues. The line shaft would have a fair amount of torsional flexibility.

Over a long period both areas/ends have the same number of accumulated cycles/revolutions but at any instant they may not (probably won’t) be the same if measured over a sufficiently short time with a sufficiently high accuracy.

Measuring the interval between every positive zero crossing will give a result with more variability over the whole system than measuring the interval between 60 positive zero crossings and calculating frequency from that.

 

[waross]

"The line shaft would have a fair amount of torsional flexibility."
This was my attempt to address that issue David. "sheaves that have resilient hubs"
I think that we are in complete agreement.

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

 

[ScottyUK]

I just want to observe that this has been quite an interesting discussion - one of the better threads on Eng-Tips.

HH -

Being from a generation background I'm familiar with how our machines interact with the larger system; I haven't worked in transmission so I haven't seen the same effect on an inter-area scale although I can see how it could occur in a large system with long interconnectors. In the UK with its geographically small and heavily interconnected grid I'm not sure I would ever see it.

Our relatively small grid can see the effect of a major event such as loss of a large generating station. When the 4GW coal-fired station at Drax tripped in the early 2000's, taking out about 8% of the overall system infeed, we saw the effect right across the country with a significant drop in grid frequency being the result. Our turbine shaft speed measurements tracked the frequency as it fell and slowly recovered.

 

[crshears]

Agreed, Scotty; good thread.

So you can have odd things, like generators ringing for minutes without them causing any issues in the system.

What is generator ringing? I am unfamiliar with this terminology.



CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[Skogsgurra]

Answer: Swinging back and forth (around pole angle zero degrees) after a shock load change.

Re best thread: It could have been an interesting thread if it hadn't started with the question "..one of the generators to be at a lower or high frequency than the rest of the generators in the system.."

That still is not possible. Transient pole angle deviations, yes. Different frequencies, no.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[crshears]

Ah! That! I had that happen once when I did a "bad" unit synch; the three units being put on line did this wum-wum-wum thing at gradually decreasing intensity until it disappeared altogether. My instructor said that was "for sure a rough shot; but if it stays in, it's good."

Thanks, Skogs.

It could have been an interesting thread if it hadn't started with the question "..one of the generators to be at a lower or high frequency than the rest of the generators in the system.." That still is not possible. Transient pole angle deviations, yes. Different frequencies, no.

Although I agree you do have your facts straight, I must politely disagree with your sentiment; to me what made this thread interesting was the arguments brought to bear in support of the argument, however incorrect the premise was.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[Skogsgurra]

Done that, too!
A 40 MW hydroelectric unit, pressed button when synchronoscope was at noon. No-one had told me that the breaker took a couple of seconds to close. So I was at tea time when it finally happened. But no wum-wum. Only a Big Bang and then silence.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[HamburgerHelper]

Crshears,

The initial question of 59 hz or slip frequency was off by a magnitude. I only had relay settings as reference as to what could be allowed. By post four, the question became "why am I under the impression different frequencies can exist on the system?" Any post going back to "59 hz" was basically beating a dead horse since that likely would be too much of a deviation unless the whole grid is slowing down as well. Maybe, that is why the relay settings were so low. Maybe, to give plenty of margin during blackstart when trying to balance load and generation. The only fault of the initial post is the magnitude, the rest is valid.

 

[HamburgerHelper]

http://prorelay.tamu.edu/wp-content...ection-to-Meet-NERC-Reliability-Standards.pdf

Fig. 06

No trips for various regions to not exacerbate grid problems.

That answers my relaying question. Each interconnect has no trip zones for generation as part of PRC-024 with very wide bands to prevent tripping of generation from making problems worse.

 

[VTer]

I am late for this discussion but I would like to offer my interpretation for my own understanding. If I am not correct I would expect the experts on this site to chime in and correct me and hopefully I can take this opportunity to learn something.

It is my understanding that there can only be one system frequency. What I believe the misunderstanding for HH is the machine rotational frequency vs system electrical frequency. The system electrical frequency is the "infinite bus" since it characteristics do not change regardless of the power supplied or consumed by any device connected to it.

Once the generator is connected to this infinite bus, the stator frequency is fixed. If there is a disturbance in the system, the disturbance will alter the internal magnetic interaction in the machine and cause an increase or decrease in the machine internal power angle. This will result in acceleration or deceleration of the machine rotor only, which is explained by the swing equation. However, this does not mean that the machine terminal frequency has changed. It may be beneficial in this case to separate the two and think of the stator as fixed and unmovable part once connected to a large power system and rotor as controllable part that needs to increase or decrease its speed during disturbances to maintain stability before major damage.

As you also noted HH, it is true that the angle between terminal voltages is largely responsible for real power flow but that does not mean that the frequency is different between the two. In fact, there cannot be any real power flow if the frequencies are not the same.

One simple analogy that may be beneficial is thinking of yourself on a bicycle connecting to a large moving train via a rubber band connection. If you do not want to lose the train than you need to make sure you are at the same speed and moving in the same direction once the connection is made otherwise the band will break and you lose the train. Once connected however, you are not going to change the speed of the train with your bike. If the train starts to move up the hill at its constant speed you need to start to pedal harder since the rubber band will begin to stretch and potentially break. If the terrain makes your bike increase speed than you need to slow down before you crash into the train. But in any case, no matter how hard or slow you pedal you will not move that train.

I hope this help clear at least some of the confusion. Take care!

"Throughout space there is energy. Is this energy static or kinetic! If static our hopes are in vain; if kinetic ù and this we know it is, for certain ù then it is a mere question of time when men will succeed in attaching their machinery to the very wheelwork of nature". û Nikola Tesla

 

[davidbeach]

Nope. The infinite bus is a reasonably useful high level approximation during reasonably steady conditions but does not actually exist.

Again, it all goes back to how "frequency" is defined, there are many possibilities, from instantaneous angular velocity of the generator shafts to averaging a multitude on positive zero crossings. The tighter the measurement, the greater the variation across a large interconnection and the more different "frequencies" within that interconnection. Within a stable system there are limits to how much variation there can be and how the df/dt at various locations must relate to each other. But there are variations. Small, sure (no, not 59Hz in a 60Hz world, there in lies madness), but real none the less.

When Palo Verdi trips the Arizona area slows down more and faster than the Pacific North West. No place is still at 60Hz, or what ever the pre-event frequency was determined to be, but there are a range of frequencies across the interconnection. The South West and the North West probably start oscillating against each other for a while. Each slower and then faster than the other as things come into tighter and tighter stability. As long as none of those frequency changes are enough to push anything out of stability all is good. The ending frequency isn't the starting frequency. It can take seconds or minutes to come to a new equilibrium.

Don't confuse stable with quasi-stable. In a "stable" system, defined as one in which the machines don't lose synch, the system frequency isn't actually stable; it moves all the time and different parts move differently. The system frequency is quasi-stable, it keeps tending back to the mean and doing so with sufficient strength that everything holds together. Lots of calculations can safely ignore the actual frequency and instead assume a fixed system frequency and a variable rotor angle. An extremely useful fiction, makes lots of simulations solvable. But just because it's useful, just because it makes the calcs much easier, doesn't mean it's true. Close enough for all purposes if the system hangs together and if the system falls apart it doesn't matter.

Define frequency as the average number of positive zero crossings per second measured over the past hour on a sliding window, and sure the whole system has the same frequency everywhere to a phenomenal degree of accuracy. Define frequency based on instantaneous angular velocity of the shaft over the past 4 milliseconds and with enough resolution you'll find that every machine on the system has a different frequency, including units sharing the same GSU.

Nobody's been willing to define what frequency is. Is it like the definition of pornography from one former supreme court justice "I know it when I see it" and nobody can actually define it? Or are we willing to arrive at a consensus definition, including measurement techniques and time resolutions?

Much of the back and forth here to fore in this thread could be reduced if we could place one statement in one measurement bin and a different statement in a different measurement bin. We argue like we know what the other posters are talking about, but maybe we don't. Maybe the "there's only one frequency" crowd is is total agreement with the "there's lots of frequencies" crowd because they're talking two different thing. "Well if that's what you mean by frequency, well, then perhaps you have a point...".

We may still be the blind men arguing about how an elephant is to be perceived. It all depends. Well, for whether or not there's one precise frequency or a bunch of different frequencies that all smudge around a common value it just depends. It's both, or it's neither. Or, it's somewhere in between, it just depends.

 

[waross]

David, should we be talking about frequency error times time?
The greater the frequency error, the less time it may persist without dire consequences.
You will know the exact figure better than I but I suspect that any frequency error that persists until the phase angle difference at the point of common coupling is greater than 90 degrees will be in danger of pole slip.
If we are to consider frequency on a cycle to cycle basis then we must also consider the resulting angular displacement and the time limit until angular displacement becomes too great.
Respectfully

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

 

[Skogsgurra]

David, your attempt at philosophy is wasted in this thread. You just created cunfusion and even misled Bill. David, say aftewr me:

There cannot be different frequencies at two ends of a good conductor - no matter how long it is!​

​

The vector's angles can vary. But never the frequency.


Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[waross]

I'm okay Gunnar. Just having trouble expressing myself.
As the vector angles are changing, the period of the cycles will be either a little more or less. Any instrumentation that determines frequency by measuring the period of each cycle and expressing the reciprocal as frequency will report a short time (very very short time) frequency deviation while the angles are actually changing.
I would characterize this as a phase shift mis-reported as a frequency difference.

How about?
When the frequency is determined by taking the reciprocal of the period of a cycle, there may be small errors in the reported frequency while a vector angle is changing. These errors will be cancelled out by similar errors of opposite sign as the vector angle returns to its original position. The magnitude of the error will be proportional to the degrees of the vector angle shift and inversely proportional to the duration of the shift.



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

 

[waross]

The vector angle is the thing.
An incoming set may be several cycles off the grid frequency and still sync successfully when connected to the grid. (After which the set frequency will equal the grid frequency.)
If the vector angle is too great, bad things happen.
All of the examples of different frequencies are a snapshot of one instant in time. If that condition continues for more than a handful of milliseconds, bad things will happen.

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