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.

 

[Skogsgurra]

It has been stated by me and many others that the frequency is constantly changing. Your question was about different frequencies in different places in a connected grid.

Are you doing this only to provoke us?

If that is the case, you have certainly made it. But you have obviously not learned anything from it.


Gunnar Englund

http://www.gke.org

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

 

[crshears]

Who said anything about nothing changing? Every power grid changes constantly! When a 100 MW electric furnace kicks in on a 20 GW system, a step load of 0.5% is applied and a corresponding speed change of 0.5% will immediately occur...to which each and every speed-droop-governor-controlled power source on that grid will promptly respond to arrest the frequency decline.

Now take that 20 GW system and parallel it with, say 180 GW of other generation; you will now have a 200 GW system which will change speed by 0.05% in response to that same step load, one order of magnitude less, greatly smoothing out that entire interconnection's frequency excursions. What will now be noticed to change when said step load is applied is the tie line flows to neighbouring systems, while the frequency will barely budge from a nominal 60 Hz. This is the reason why all of the entities in an interconnection will always choose to stay connected whenever possible.

The point is that the frequency of the entire grid and all of the synchronous generators and motors connected to it is [macro] changing at once; the nature of the torque angle between each rotor pole and stator pole, viz., that varies directly as the amount of power transferred, does not permit of any other behavior [ see pg. 289, section 16-4, Electric Circuits and Machines, Sixth Edition, by Eugene C. Lister, Chief Electrical Engineer [Retired], Stanley Consultants Ltd., and Life Senior Member of the IEEE ].

If one were to couple a 300 MW steam turbine to a 100 MW alternator, synchronize it to the local power system at the correct voltage and frequency, then attempt to overdrive the alternator, the machine's torque angle would eventually exceed 90°, at which point the machine's pullout torque would be exceeded as well, and it would either go out of step and overspeed or continue to operate as an induction generator [ibid]...until either its protections or its human minders removed it from service.

Hope this helps.



CR

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

 

[davidbeach]

Ok, yes, somewhat provocative, but for a good reason...

Measurement definitions and techniques matter. There have been a lot of absolutest statements about a single frequency on any system of any size without any definition of how that is determined. There are valid definitions that could make that true under essentially all conditions. There are other, equally valid definitions, such as how relays measure frequency, for which that isn't true.

On a system of sufficient size steady state doesn't exist, at least under certain reasonable definitions. If you were to inquire of every line relay in WECC what the system frequency was you'd get a smear, not a point, nor a scatter plot. The smear is a tight grouping of different values, a scatter plot would allow observation of multiple distinct points. To give the smear/dot/scatter plot two dimensions plot df/dt vs. f.

On reasonably sized systems the smear is so small that it can't be distinguished from a point. On larger systems there's always so much variation that it can never be close to looking like a point.

There's an awful lot of running generation that doesn't have any governor response. Run of the river hydro is run based on water flow and ignores system frequency. A lot of base-load generation is run at constant MW output regardless of frequency. Wind and PV just do their own thing and don't give a whit about any needs of the larger system, they just want some, any, system frequency to track. Those units paid to have governor response are the ones to respond to system frequency deviations; the dispersion of governed units is not uniform across the system. Keep in mind, I'm referring to a 150GW (yes GW) system that includes over 121,000 miles of transmission line. Small, islanded, systems need every unit to be governed but on the big system most aren't.

System frequencies, within a tight but ever shifting band, are always tending toward the same value. But different parts of the system are doing different things, some places df/dt is positive and other places df/dt is negative at the same instant. If I plot our system frequency for the past week I get a very wide, very fuzzy "line" that pretty much shades in the whole band between 59.98 and 60.02 with lots of excursions to and beyond 59.96 and 60.04. But if I plot that on a scale from 0 to 100 I see a line that has an occasional pixel that deviates from a straight line right at 60.

There can be no standing differences in frequency, I've never meant to imply that, but (using a relay definition of frequency anyway) there's also not a single frequency everywhere at all times. At least not one that would plot as a nice, neat, dot.

Using definitions other than what the relays do, to measure frequency it might well all be the same frequency. But if the relays all, system wide, always measured the same frequency the approach to under frequency load shedding would have to be very different than what it is.

Same frequency with different phase angles or different frequencies - it's just different paths from the same starting point to the same ending point, all within well defined constraints. A 50ft view vs. a 35000ft view. Just define the criteria.

If you're looking at motor speed at various locations it is certainly one one frequency. If you're analyzing an underfrequency load shedding scheme it's a multitude of frequencies at a multitude of locations. Plot them all and you get that same smear.

The truth isn't dogmatic, but rather pragmatic. Statements that can't be proven to be anything other than true at certain system scales don't hold as well at much larger scales. Einstein proved Newton "wrong" but Newton still gives perfectly good answers within most conditions. Similarly, within lots of systems "one frequency everywhere" is an accurate answer to any degree of precision possible, but that doesn't mean that it still scales to major interconnect levels.

Sure, at steady state it's all one frequency everywhere, but eventually the system is too large to have a steady state. The major interconnects fall into that too large for steady state category.

 

[Mbrooke]

@David- Very well said and I think this ties many posts here together, even those disagreeing.


And while I don't want to get in the middle of the slaps so to speak: I personally enjoy hearing both sides to this argument. I don't see anything elementary or proactive. Rather I am learning and thinking about this in ways I never thought about in depth.


Also, if this forum had a rating system, I would give this particular thread 5 stars.

 

[Skogsgurra]

Sorry, David. The OP did not have such subtleties in mind when he put his question. He just couldn't understand how there could be different frequencies, in a very general way. He also gave an example that no-one can stand behind. Quote: "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"

This whole damn thread has confused more than it has helped. For me and, luckily, a few other engineers, it didn't do more than stir up emotions. For many others, engineers or not - I really don't know what to think, it cannot have done any good either. But those participants seem to value fine phrasing more than actual facts and understanding. Are you all under influence of Mr. Trump? Has it gone that far? If so, I feel very sorry.

Gunnar Englund

http://www.gke.org

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

 

[Skogsgurra]

OK, I got somewhat carried away. But, still, there is just one frequency across the whole system. And someone that presumed that my examples from Sweden shows one single frequency because it is such a small grid should be aware that the distance between the two end-points in the Scandinavian grid is like Miami, FL to Ottowa.

I cannot understand what is meant by "I don't see anything elementary or proactive". It is the "proactive" I don't get. I understand proactive as in proactive maintenance. But cannot understand what it means here.

Gunnar Englund

http://www.gke.org

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

 

[Mbrooke]

My mistake- I meant to type provocative but auto correct failed me again lol.

 

[Skogsgurra]

Ah, that makes a lot more sense! I was really confused there.

Gunnar Englund

http://www.gke.org

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

 

[Mbrooke]

I could certainly understand!

 

[crshears]

davidbeach wrote: There's an awful lot of running generation that doesn't have any governor response. Run of the river hydro is run based on water flow and ignores system frequency. A lot of base-load generation is run at constant MW output regardless of frequency.

David, in my experience the turbines of ALL nuclear units I've ever encountered [ and in my world they all operate as base-load plants ] have governors, not valve positioners; and the way it's been explained to me is that when such a unit is in "turbine follows boiler" mode, the turbine governor's speeder gear is automatically and integrally adjusted by steam pressure so as to hold it steady, while the action of the governor itself provides proportional response to system frequency.

In the run-of-the-river plants I know of, all but the very smallest of hydraulic units have governors, most of them manufactured by Woodward. Speeder gear is adjusted to alter the unit water flow, but again the action of the governor itself provides proportional response to system frequency.

All the conventional [ meaning not combined-cycle ] thermal plants I worked in had turbine governors.

Not having worked in combined-cycle plants, I cannot speak to what types of governors or input control such might have.

In my world there is therefore a much larger percentage of governed generation on the system than in yours.

CR

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

 

[crshears]

Just re-read this entire thread; what a ride!

Thought I'd comment on some of the things said along the way...

HH wrote: This I thought was kind funny. Kosovo has been running area imbalances and has been slowing down the clocks in Europe this year by as much as 6 minutes between January and March.

Having become cognizant of how important definitions are to this thread, here's my response:

This is a result of an entire power grid's or interconnection's common macro-frequency; if one control area neglects its responsibilities, fails to load up its generation, and instead relies on its neigbors to carry it through peak periods, that entity's ACE [ area control area ] will float into seriously negative territory, that entire interconnection will be under-generated, its frequency will drop below standard, and for the duration of that period its time error will accumulate. Do this enough times in a row, and a cumulative system time error of six minutes is not at all beyond the realm of possibility.

HH also wrote: I asked the generation guy what was the point of setting the underfrequency setting and if it was to protect the generator. He didn't believe that the generator would be damaged if operated at low frequencies and that it might be to protect the customers still connected to the grid during the frequency sag.

As it's been explained to me, neither of the foregoing is correct; the cited primary reason for employing UFLS is to save the interconnection by shedding load in response to declining frequency so as to crudely arrest the generation/load mismatch that has developed.


CR

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

 

[HamburgerHelper]

The ERCOT fundementals training manual that is used as operator reference and for training for the ERCOT System Operator Certification Exam has this in it.

http://www.ercot.com/content/wcm/training_courses/109608/ERCOT_Fundamentals_Manual.pdf

ERCOT System Operators are in the least told to look changing angles as frequency deviations and that changing power flows require generators to have a "relative accelerations". I think that if they are mentioning small frequency differences during changing power flows, they probably have frequency readings out to several decimal points.


I got a guy to pull up the RPMs on two generators that were being dispatched that were 400-500 miles apart. The two were readings were a few thousandths or hundredths (I can't remember. It was Friday and I had to pick up my daughter) of an RPM apart but I couldn't get them both timestamped at exactly the same time before I had to go. I don't know how accurate the RPM data that I saw was but I know that someplace precise RPM or frequency data is collected and used for grid controls to maintain tie line and countrol area power flows. I have access to this RPM data and it probably would be interesting to watch a unit RPM ramp up and down for the day.

 

[waross]

Very interesting.
That has been explained several times and now we have a definitive answer.
This explains how a change in power flow will cause a change in power angles.
This in turn will lead to a slightly different frequency only during the time that the power flow is changing.
Where I have a problem is reconciling this with your original statement.

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.

To most of us, "humming along" implies a steady state condition at 59 Hz, not a slight variation to about 59.8 Hz lasting for a few seconds.
Steady state versus transient event. One size does not fit all and one explanation does not fit all.

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

 

[waross]

Governor action. When a unit is in base load configuration, the control could be set to deliver the required output and locked in position.
That would work until the unit lost its load for some reason. Then the locked control (fuel valve, steam valve, water gate, etc.) would cause a runaway.
Instead many base load generators are controlled by governors in droop mode. At 5% droop, they are relatively insensitive to minor frequency variations but become active and limit the speed to 105% or less on loss of load.
With a large change in loading such as CR mentions, 0.5%, all units will initially rspond, and then the swing set(s) will over the next few seconds act to correct the frequency back to 60 Hz.
From a controls perspective in relation to a PID controller, base load sets use Proportional and swing sets use Proportional-Integral.

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

 

[HamburgerHelper]

Waross,

That question was based on underfrequency setpoints I saw that were very low and oscillography for frequency deviation during a block load switching in that I knew wasn't felt across the interconnect. You keep bringing it back to this point of 59 hz after I have already stated earlier that this isn't the question anymore and the magnitude was off. The forth post is the more general question. You are banging away on a point that has long been conceded and I have already said this much and any mention of deviations have included statements that it happens and is allowed provided angular deviations aren't large enough to cause instability.

If you want to bring up steady and transient state, I would be of the opinion that the grid is always changing and never in steady state. Generator controls, AGC, and dispatch are continually trying to limit frequency and generstion and load deviations and maintain tie line flows. These do a pretty good job because sizeable deviations are only seen during block changes. Everything would be deadballs 60 hz steady state only if powerflows in the system quit changing. No changing loads. No changing generation. No changing system configurations. That isn't the case.

 

[marks1080]

HH you were (not sure if you still are) implying that changes in system frequency could mean that the system may have different frequencies at different points in the system. This is not the case. Frequency is not constant, but that sure doesn't mean a system with a dynamic frequency isn't stable or steady-state. Minor changes in system load flows (as load changes through the day for example) or minor dynamic changes in frequency do not meet my definition of a transient state. Transients are caused by switching, faults, equipment failures and major load flow swings (not the normal everyday stuff).

 

[crshears]

I'd agree, or maybe even tighten it up a bit, and state, to use NERC-speak, that if there isn't any type of bulk electrical system contingency that requires the system to be re-prepared for the next contingency within 30 minutes, steady-state conditions exist...and please take note that I didn't write "static state".

To rephrase, I would exclude planned [ and sometimes even unplanned ] switching from the criteria that invoke a transient state; switching of static capacitors or even in some case high voltage circuits for voltage control purposes, for example, may cause observable voltage bumps on the system that will nevertheless not come anywhere near the qualifying threshold for declaring a contingency.

Similarly the synchronizing, loading, unloading and disconnection of generating resources in the perpetual pursuit of matching supply to demand involves switching that is, while always executed with vigilance against the unexpected, quite routine.

Likewise, as the load on a higher-voltage autotransformer approaches the point where it would be overloaded by a recognized contingency, nobody panics; either a re-allocation of load onto other available sources of supply, or the designation of specific generators as "must-run" as required to alleviate "the ping" is undertaken as a matter of course.

CR

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

 

[ScottyUK]

crshears,

The UK grid requires that a generator stay on indefinitely at 47.5Hz, maintain a minimum of 95% of rated output at 47Hz, and must remain connected for a minimum of 20 seconds if the frequency dips below below 47Hz.

Under-frequency protection on generating plant serves a couple of purposes: on gas turbines the compressor output drops away quite quickly as the speed falls, so as speed drops the machine will struggle to maintain output. The 95% rated output at 47Hz is a tough demand on a GT - it can be met by roasting the arse off the power turbine through over-firing the engine. On a steam set there are often shaft critical speeds not too far below sync speed, and you can't safely operate too close to them without risking damage to the machine.

 

[marks1080]

crshears - yes that is a better way to look at it.

I spent 5 years at a nuclear site. They used the word 'transient' for very different, and much more scary reasons. Basically, if something 'unexpected' happened within the nuclear side of things they would announce a 'system transient'. Absolutely no reference to the electrical side of things.

We always wanted to run away when those announcements happened, but they usually would lock the plant down and we wouldn't be able to leave.

 

[crshears]

Hey Scotty,

Fair enough; but I was referencing HH's mention of UFLS, which is Under Frequency Load Shedding. The generator under-frequency tripping you allude to is indeed another animal entirely; the 60 Hz steam plant I worked in had UFGT @ 57.5 Hz with a time delay of 10 seconds, and @ 57.0 Hz with no intentional time delay.

CR

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