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.

 

[davidbeach]

Define “Frequency” and specify the means of measurement.

Whether or not different units can have different frequencies depends on definition and means of measurement. If I determine frequency from instantaneous angular velocity of the generator shaft I could find that every unit has a different frequency.

They all have about the same frequency and the maximum variation (assuming a stable system) is small, but they’re not all identical.

Other definitions and measurement techniques will result in different answers for the same system.

 

[marks1080]

Define it purely in electrical terms. forget about the mechanics of the generator. When discussing 'system' frequency that is the best way to look at it.

When looking at the particulars of running and individual unit than you need to look at the mechanical aspects, in which case you're correct - every unit has its own shaft frequency - all of which should be very close to the system frequency, otherwise expect unit damage.

Electrically speaking you will measure the same frequency of any 'system' at any point of the system you choose to measure. Exactly the same the frequency.

 

[HamburgerHelper]

Mark1080,

You should look at the voltage angle difference as being an indicator of import or export of real power. The voltage angle between the sending and receive will be:

Power = (magnitude(Vreceiving)*magnitude(Vsending)* sin(angle)0 / impedance.

Sending real power over any impedance produces an angle difference between the sending and receiving voltages.

If you had two island grids connected by a single piece of transmission, the angle of the two islands would be indicative of who is sending and receiving power on the connected transmission. If one island was operating at a slightly different frequency, it would indicate that the voltage angles were continually advancing or lagging the other island. The frequency difference indicates the net import or export is increasing.

 

[Skogsgurra]

HH, you are completely wrong. Don't spread misconceptions. It doesn't help us and it doesn't help you. There is no frequency difference. There are small variations in df/dt - you may call it Frequency Modulation - but if you measure the frequency with standard frequency measuring devices, you will see the same frequency everywhere in a connected grid.

Gunnar Englund

http://www.gke.org

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

 

[HamburgerHelper]

No, I am not wrong. I overstated how much I thought there might be frequency deviation on the system. There are frequency deviations on the system usually on the order of 1/1000. I sit next to a guy who has worked in generation plants for 20+ years and he has said this much. There is always error on the systems otherwise Balancing Authorities would not be using ACE or LCF to try to balance their areas. No one can guess their loads exactly. I am guessing Sweden is so small and tight that deviations would be harder to notice.

 

[waross]

HH it's time for tough love.
Either you have seriously misunderstood what the guy has tried to tell you or he has given up trying to explain and tells you what you want to hear.

System generator frequencies don't have to be the same all the time as Davidbeach and others pointed out.

You have misunderstood this also.
Yes the frequency may vary, but it won't be at different frequencies at different parts of the grid at the same time.
Any frequency deviations affect all the generators on the grid.
We have tried to explain to you politely and diplomatically that generators tied to the same grid must all run at the same frequency.
Diplomacy and politeness just don't cut it with you.
What part of you are wrong don't you understand?


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

 

[SlideRuleEra]

I worked 22 years on the generation side of an electric utility and have witnessed grid frequency "force" an unsynchronized generator into compliance with grid frequency. On start up, operators would get an outdated manually operated unit as close as possible to synchronous speed and let the grid "lock it in".

The most dramatic was a 280 MW unit that was accidentally motorized from turning gear speed toward synchronous speed... it never made it. Destructive vibration, a hydrogen explosion, and lube oil fire took it out. I did not see this event but participated in the 2 year recovery.

A recent example of a large grid slowly having frequency dragged down has been going on in Europe this year.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[marks1080]

It's easy to get blinders on in this industry. Happens to me (heavily focused on P&C) all the time. It's helpful to take a step back from time to time and look at Power Systems as a discipline all on its own. If you're experience is 30 years of being in a plant I get that the concept of a single Power System frequency is hard to grasp. Hell, this shit is hard to grasp all by itself.

Step back and look at the Power System as a single entity and it really will become very clear that anything but a single power system frequency just doesn't make any sense at all. This has nothing at all to do with what's going on with the mechanical side of any generator, but ALL generators that are connected to a power system will have an output (or input if motoring) frequency equal to the power system. The only discrepancies in measured frequency will be due to the imperfections of the measuring device itself. If you can actually measure two distinct frequencies it absolutely means you have two separate power systems.

 

[spciesla]

The most dramatic was a 280 MW unit that was accidentally motorized from turning gear speed toward synchronous speed... it never made it. Destructive vibration, a hydrogen explosion, and lube oil fire took it out. I did not see this event but participated in the 2 year recovery.

That sounds like a really bad day at the plant followed by two years of hell. Ouch!

 

[wroggent]

This is how I think of it - Imagine a perfect system where the loads, generation, voltages, currents, and impedances never change. Also imagine the grid is at 60Hz and a connected generator is running at 60.0000000001 Hz and they start out with their voltages exactly in phase.

The period of one cycle for the grid is 1/60

The period of one cycle for the generator is 0.016666666666639 seconds

One degree at 60Hz is equivalent to (1/60)/(1/360) or 1/21600 seconds

For one degree (at 60Hz) of phase difference to accumulate between the grid and generator, the grid and generator would have to run at their respective frequencies for:

(1/21600)/(1/60-1/60.0000000001) = approximately 1.70068 billion seconds, or almost 54 years, and even longer before stability becomes a concern - but it eventually would.


But who cares - we are talking about the real grid not this imaginary one and such precise measurements aren't possible or meaningful since things are constantly changing. Different machines will approach new steady state conditions after system changes at different rates depending on their inertia, but these are transient phenomena and I don't think it makes sense to consider them in the steady state frequency. Again, it all depends on how you define and measure the frequency.

I think everyone (including those arguing) is saying the same thing, actually.

 

[Skogsgurra]

Given the fact that grid frequency is constantly changing at least +/- 0.1 Hz, I really cannot see that such an academic example would bring clarity to the thread. And I also disagree that everyone in this thread are saying the same thing. Some know and some think they know. And some have buddies that have told them lies. How to measure frequency is well defined in many text-books. There are, to my knowledge, just two ways of doing it.

There cannot be opinions on this matter. Only facts. Let's put an end to this thread now. Or kill it.

Gunnar Englund

http://www.gke.org

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

 

[ScottyUK]

Perhaps try looking at this from another perspective:

If measured with sufficient accuracy the angular velocity of any individual generator at a precise instant may not exactly match that of all other machines on the system. For example as a generator's output increases, its load angle relative to the system also increases. For the load angle to increase there has to be a transient increase in angular velocity, i.e. for a finite period the machine is 'faster' than the system until it reaches its new stable load angle for the increased load. As a generator reduces load the load angle relative to the system also reduces and for a finite period the machine runs 'more slowly' than the system until it reaches a new load angle.

Even when machines have a stable power output, changes in load and load-flow on the system itself will cause minute variations in load angle of individual machines. As the load angle changes then by definition the machine has to undergo a transient acceleration or deceleration relative to the system in order to reach the new load angle.

During faults the machines closest to the fault can see significant swings in load angle as the AVR works to keep the machine in synchronism, with the machine initially accelerating and pulling ahead of the system. For acceleration to occur the instantaneous speed of the machine has to be higher than that of the system, but for the machine to stay synchronised the AVR then has to boost the field sufficiently that the acceleration is checked and is followed by a period of deceleration relative to the system. If the machine goes past its stability limit then it breaks synchronism and will pole-slip at a frequency faster than that of the system until hopefully the protection operates and trips the machine.

I don't personally believe that the transient acceleration and deceleration associated with changes in load angle equates to the machine experiencing a change in frequency, but I can see how the confusion might arise.

 

[HamburgerHelper]

If you want to say that the entire system is always dead balls the exact same, that isn't true. Any time load or generation increases, you will see the voltages lead or lag when compared to whatever you want to use as reference. Dispatch couldn't change if voltages couldn't change angles compared to whatever reference. The amount that frequency deviates due to just dispatch isn't much and if you do a back of the napkin calculation with ramp rates and maybe a 90 degree swing in voltage angle (going from load to generation), you can see the order of the frequency change. There has to be frequency change as you add generation or load or the voltage angles couldn't change. If you want to look at something more abrupt, take a look at when load or generation is added or tripped out in blocks. The entire system doesn't see the same frequency excursions as those right next to the event. When there were power swing issues up as I mentioned earlier up by Duluth, the rest of the eastern interconnect wasn't going goofy. In my opinion, generators can do whatever they want frequency-wise as long as they never deviate so much angularly that they start slipping poles. The angle is the issue not frequency. You run fast or slow, you are ok as long as you fix yourself before you go too far. I believe if someone created a contour map with the frequencies of every bus, scaled like from 59.999 to 60.001, someone would be able to clearly see where loads and generation were changing.

 

[marks1080]

Frequency changes all the time HH. But it changes equally throughout the grid... because... there can only be a single system frequency in a single system. This has NOTHING to do with any single generator angular anything at all. You can set all the units in your grid to all run at different output frequencies if you so choose, but you still only will measure a single system frequency, and end up with a bunch of destroyed generators. I'm going to refrain from adding comments about harmonics into the discussion, because it's just going to confuse things further.

I've seen such frequency 'maps' that you're referring to for the entire north eastern seaboard. There is one system frequency. The only deviations you will see are measurement errors, and I've never personally encountered measurement errors, only exactly the same frequency readouts throughout different parts of the grid... well except when the grid starts to split into smaller grids... than you see the freq. readouts deviate, but that's because it's no longer a single system.

THE ENTIRE SYSTEM FREQUENCY IS ALWAYS DEAD BALLS THE EXACT SAME. If you're going to claim otherwise please provide references.

 

[rcw retired EE]

A couple comments:
Frequency/speed as a turbine setpoint causing confusion.
Some turbine governors use frequency or turbine rpm as the control set point. On the Woodward unit we enter a 3600 rpm setpoint for a 60hZ 2-pole generator to run at synch speed, no load. The internal voltage of the generator and the system voltage are in synch and at 0 degrees apart. To load up the generator you increase the "speed" setpoint, the throttle opens, more steam goes into the turbine, the internal rotor angle increases and power flows to the system. If the unit was set for a 4% droop, the 100% load setpoint would be 3600/1.04= 3744 rpm. Shaft and generator would still be spinning at 3600 rpm and 60Hz, but the control system is calling that throttle position 3744 rpm (62.4Hz).

Hamburger Helper mentioned that load decreases as frequency decreases. That is true if the load is predominantly induction motors driving pumps, fans and other load whose horsepower requirements vary with speed/frequency. The motor slows down and the pump/fan output drops. If load is mostly computer power supplies, servers, electric heat, vehicle chargers and variable speed drives, load stays constant as frequency changes. The individual power conversion modules adjust to maintain the load output. If the power system stability models are based on outdated historical data when industrial motors were the major load, the simulations will overestimate the load drop during a low frequency event.

20-30 years ago, the system was partially self-correcting. During an event, as frequency dropped due to generation/load imbalance, load would drop, helping the correction. The steam driven turbines and hydro units would maintain their MW output until the governors reacted and increased MW to rebalance load/generation. Selected steam turbines were operated at less than 100% throttle to be able to increase output quickly.

Compounding this load vs generation imbalance due to constant loads is the preponderance of combustion turbines in today's generation mix. A combustion turbine's MW output varies with the square of the turbine speed. In the first seconds after a loss of generation event when frequency drops, load stays relatively constant while combustion turbines' outputs drop, exacerbating the problem. Governors eventually kick in and boost generation to pull frequency back but that initial dip is more drastic than a decade ago. (Ignoring other advances in system control).

When coal and nuclear driven steam turbines were the major power source, system operators could count on the droop response of the governors to correct a frequency dip. Many of the steam turbines operating today are combined cycle units running at "valve wide open" to turn every bit of steam from the combustion turbines' exhaust heat into MW. With the throttle already at 100% there is no way to quickly increase steam turbine output in response to a frequency drop. The gas turbines have to increase fuel input, increasing the exhaust heat flow to make more steam to increase the steam turbine output. Only the combustion turbines have a true droop response to a frequency drop.

What sort of droop response can we count on from a solar panel or wind turbine? Zero.

This creates some interesting issues for system operators.

 

[Mbrooke]

I will chime in. After reading reports on many blackouts and watching voltage collapse simulations in power world, it is possible to have differing frequencies in an unstable system with interconnected with AC lines. Maybe not 59 vs 61 HZ, but something like 0.005 is possible.

 

[waross]

Mbrooke; For how long? Or, for how many angular or electrical degrees?

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.

It can't for more than about 1/4 of a cycle, or 4 milliseconds.

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

 

[crshears]

[edit: entire post re-worked]

As I'm only a power system operator/controller and not an engineer, and as math is not my strong suit, I've had to dumb this down for myself and think of a hugely finite grid - like the Eastern Interconnection - as one honking great matrix of line shafting arranged in loops and spurs, with all of its tapped [synchronous] generation and loads solidly connected to it by either solid shafting or through solid gearing, meaning with no intentional slip built in.

The speed of one great honking line shaft matrix like this will definitely fluctuate with imbalances between generation and load, but the entire matrix’s speed – from Key West to Kashechewan, from Sioux City to Sydney - will fluctuate at the same time and by the same amount. There will of course be varying degrees of twist in the various shafts as load and or generation are applied and removed, but the speed of all connected components remains the same. The only way for there to be a difference in speed is if a separation occurs somewhere, either by shifting one or more of the intervening gearboxes out of gear, or if a shaft breaks in one or more places.

If you wish to think of a generator's [or motor's, or grid portion's] speed as "changing" while the amount of twist in its connecting shaft varies, more "power" to you; but I for one don't view it that way.

As to the mechanical analogues for the various components attached to the system, induction generators can be viewed as connected to the matrix by a fluid coupling, with the result that the only way power can be transferred from this generator to the matrix is by increasing its energy input so that a difference in speed develops between the machine and the matrix, else a power transfer will not occur.

The same rule applies to induction motors but in reverse as the speed of an induction motor will deviate from the speed of its supplying grid in direct variation with the applied load.

Phase shifters [ aka quadrature boosters ] are like spline boxes consisting of two concentric cylinders with left- and right-hand slots resembling the “rifling” of a gun barrel cut into them; by means of an axially shifted hydraulically driven pin and collar, the angular relationship between the two sides of the device can be adjusted. Such adjustment alters the amount of energy transmitted via that particular path by varying the torque, but has no effect on overall matrix speed.

In a station tying together two grids operating at different frequencies, all but one of the frequency changers is commonly equipped with stator shift gear to facilitate the sharing of load between the machines.

As has been noted previously, solar and wind farms are like gate-position-controlled sources of energy input to the matrix, and as such have no speed droop capability associated with them, therefore contributing no stability to the system, which is why these resources are the last to be permitted to re-synchronize to the grid during power system rebuilding and restoration following a grid collapse.

CR

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

 

[HamburgerHelper]

If anyone wants to see real synchrophasor data, Dr.Grady puts out reports on synchrophasor data on mostly ERCOT and a little on the other two interconnects.

https://web.ecs.baylor.edu/faculty/grady/Texas_Synchrophasor_Network.html

The 2017 report is note worthy because wind generation in ERCOT got up to 45% of the total for a few days in January. Aside, from the one unit trip, you can see the effects that a large penetration of wind generation ramping up and down has on his synchrophasor voltage, angle, and frequency measurements at his sites in ERCOT. The frequency measurements differ at each site, mostly during when wind generation is ramping up or down or something else like a unit trips that abruptly change power flows. Other reports and studies make note of things like loss of a DC tie, mysterious frequency ringing, generator trips, and frequency variations during the super bowl in each of the interconnects. It is worth checking out just to see real synchrophasor data.

 

[Skogsgurra]

There are people with three arms (not talking about Zaphod B. now) but that doesn't prove that humans in general have three arms.

Gunnar Englund

http://www.gke.org

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