Power grid & generator questions

[essenmein]

That's an interesting feature of governors. Let me talk for a minute about the steam turbine i knew.
How the governor actually works is on the measured difference between measured speed and a reference speed.
The device that tells the valves how far open they should be is called "Speed Changer".
It has that name because it's used to roll the turbine from standstill to synchronous speed.
My turbine's synchronous speed was 1800 RPM.
To start it you might send it a reference speed of ten or twenty RPM. The steam valves would open slightly and it'd accelerate.
You'd send higher and higher reference speeds until it reached 1800 RPM.
Because there's no load on the generator it takes very little steam to roll the turbine and the valves are very nearly closed.

Next you synchronize the generator and connect it to the grid.
Now speed can no longer change because of the "infinite bus" concept.
So if you send it a reference speed of 1801 RPM the steam valves will open slightly .
But speed can't change so the extra thermal energy being carried in by steam goes out as electrical energy.
That's what turbine- generators do -
Mr Turbine turns thermal energy into mechanical energy
and
Mr Generator turns mechanical energy into electrical energy..

Recall the valves are positioned by difference between measured and reference speed.
So if you send it a higher reference speed it'll see more difference and open the valves further causing more kilowatts to flow out of Mr generator.

And that's how the "Speed Changer" becomes a "Load Changer". One device, two names.

What is the relation between speed error and valve position ?
Ours was 100% valve travel for 3% speed error.
So once you've synchronized and connected the generator to the grid , you send increasing reference speed and the valves open.
When reference speed reaches 103% of 1800 RPM, which is 1854 RPM, the valves will be fully open and you're making full power.
If grid should speed up , the speed error will get smaller and valves will close proportionally reaching full closed at 1854 RPM = 61.8 hz.

Now to your question
If a local speed measuring device is off by 1 RPM, it'll indicate 1799 or 1801 when the generator is at synchronous speed ready to synchronize.
>>>No matter, that's the reference speed you'd have to send it to achieve synchronous speed <<<< So it'll position the valves relative to synchronous speed.
Same applies to frequency.

But that's transparent because "Reference Speed" was an analog signal representing speed.
The analog signals for Reference and Measured speeds were hydraulic oil pressures internal to the speed/load changer, not electronic, and were displayed on pressure gauges indicating PSI.
There was an electronic RPM meter for the operators to watch while rolling the turbine up to speed ..
The "Speed Changer" knob doesn't mention RPM it just says "Raise" one way and "Lower" the other. You bump it and watch the RPM meter.
Final speed adjustments you made by watching the synchronizing lights.
After synchronizing you bump the same knob and watch the megawatt meter .
It's that simple.

Of course the governor is more complex than a simple proportional speed/load changer. It has rate of change pf speed compensation, automatic runbacks, and settable limits on valve opening.

I hope that plants the basic idea of a speed/load changer. Nowadays it'll be an electronic computer programmed to do pretty much the same thing.

I hope this isn't clutter. Should be more mathematical in a physics forum, but your question was "How does it work?"

Once you understand how the pieces work individually you can begin to work them connected in your thought experiments.

old jim

Thats cool, so basically absolute speed error is not a problem because you let the infinite bus guide the machine once its connected, then your load controller is just applying tor... ahem, power, by fluttering some valves to deliver more electric power.

I assume once the turbine is turning at the right speed, you bring up the field so you can see your AC output to synch, I'm assuming just set the field current to give you the right voltage before connecting?

So what is the control strategy for the field once connected? Is it essentially just compensating out resistance effects with load or is it more complicated?

 

[essenmein]

(was writing while you posted that video!)

 

[essenmein]

Here's a video about synchronization.
It explains "Syncroscope", a meter that displays the angular displacement between two sinewave voltage phasors.



gross understatement at 1:18
and an actual synchronization about 2:30.

old jim

Statement at 2.59 is interesting... lol.

 

[jim hardy]

I assume once the turbine is turning at the right speed, you bring up the field so you can see your AC output to synch, I'm assuming just set the field current to give you the right voltage before connecting?

correct. You adjust field current and watch the voltmeter.

So what is the control strategy for the field once connected? Is it essentially just compensating out resistance effects with load or is it more complicated?

Not terribly more complicated.

Industrial and residential loads are not purely resistive. Motors in refrigerators and washing machines and steel mills are inductive as are streetlamp ballasts and the huge utility transformers scattered about the grid.
So all power plants supply some reactive current to those loads. We measure it not as amps but as "Vars",
VAR being the acronym for "Volt-Amp Reactive" which is V X I Xsin(angle between them)
Field current controls how much reactive current a generator pushes toward 'infinite bus', or more correctly toward the utility's customers.
Each plant contributes its share under the watchful eye of central dispatcher. That's automated these days, in my day he radioed in requests to pick up or drop megavars.

We tweak field current to achieve desired VARS out of the machine.
Field current is controlled by the Voltage Regulator. Operator bumps the Voltage Regulator adjust knob while watching the Megavar meter.
As i said that's automated now and centrally controlled by sending raise -lower commands from central to the individual plants.

Voltage Regulator is another closed loop control system. There's a bit of vocabulary to absorb before going there.

You might like this old thread
https://www.physicsforums.com/threads/voltage-and-reactive-power-relationship.876346/#post-5503931

old jim

 

[essenmein]

correct. You adjust field current and watch the voltmeter.
Not terribly more complicated.

Industrial and residential loads are not purely resistive. Motors in refrigerators and washing machines and steel mills are inductive as are streetlamp ballasts and the huge utility transformers scattered about the grid.

Sorry wasn't talking about load resistive effects, more stator resistance as it heats etc.

 

[essenmein]

Then parallel to this the angle between different generator rotor's may also change as one might have a prime mover that applies more power to the generator when experiencing a frequency decrease while some other generators rotor might have a prime mover already going at full capacity or incapable or adding more power so that generator can add much less to the frequency increase than the first one?

This is only true if the generators are the same pole count. A 4 pole generator will spin at half the speed of a 2 pole at the same electrical (grid) frequency.

Thirdly also the angle between each generator's rotor and stator field changes while all of the before mentioned things happen?

Yes, if it was possible to strobe the stator field in the video above then it would be at the average location between the two rotor fields, let's say for example the rotors of the two generator set up in the video are 40deg apart, then the generator rotor field would be ahead of the stator field by 20deg, the motor would lag by 20deg.

If you wanted to get really technical the transmission line between stators would also introduce phase lag between the stator fields.

Whether or not a machine is a motor or a generator can be determined by whether or not the rotor field is lagging or leading the stator field.

@anorlunda does the difference between your example of solar panels and physical rotors and stators is that with an inverter one can control the phase angle better because a semiconductor switch/es are adjustable by electronics versus a large rotor can be adjusted by the help of it's mechanical prime mover and precise timing, or is it not the case?

Both can control the phase angle equally well, its a question of how quickly they get there.

I assume a solar panel inverter's lack of mechanical inertia is simulated by its storage capacitor capacity to supply more power to the grid because sun doesn't change it's strength in the second to minute timeframe, so apart from some stored capacity the energy source itself (sun) is incapable of a sudden increase in power supplied?

Yes and no. Capacitors do store energy, but they are not there to emulate inertia, they can "run" the inverter in case of power loss, but its not going to be much, a few 10's of ms at best. Their primary purpose is to supply a steady DC voltage for the inverter to operate from.

PV Solar power is generally operated at maximum possible power at all times facilitated by various MPPT algorithms, if for some reason the grid didn't want it, you'd be putting it in batteries or some other form of storage. I would expect large scale solar thermal would be back to steam turbines and operate on the grid much the same way any of the other steam turbine would.

In terms of $/W generated and reliability, copper and iron beats silicon hands down.

 

[jim hardy]

Sorry wasn't talking about load resistive effects, more stator resistance as it heats etc.

Think for a moment about real power, ...
a machine's internal resistance must be very low else unmanageable heat would result .
Our stator current was 20 kiloamps.
Stator resistance is designed low as practical so that the machine will be efficient.
We in the plant consider stator resistance to be negligible.
But we do monitor temperature rise of the hydrogen exiting each armature conductor because a change from the machine's normal pattern would indicate trouble. An armature conductor that's broken a strand will run hot because it experiences higher current density. One with a blockage in its cooling flowpath will run hot as well. .

Stator is much more inductive than resistive. Inductance doesn't produce heat - recall reactive current is wattless..

Just as we adjust steam flow to keep megawatts at our fair share of what the customers are consuming at the moment
we adjust field current to keep megavars at our fair share too .
Further, to keep voltage constant where the customers are,
we must raise or lower our voltage as needed to push current through the impedances between us and those customers, namely the transformers and power lines between us. That we do by adjusting field current and again it's done under the watchful eye of central dispatch..

So to answer your question - over the course of a day's operation it's not the machine that changes, it's the load to which the machine is connected.

old jim

 

[jim hardy]

Continuing this question, as I see nuclear power plant use generators that only have two poles on their rotor

Where did you see that ?
Nuke plants are almost exclusively 1800 RPM because of the low steam temperature compared to fossil plants.

en the plant needs to be synchronized to the grid aka "switched online" they first spin up the turbine to the speed which approximates? or is the very speed of the grid frequency and then they can switch it online aka connect the stator windings to the grid step up transformer, but they can only do this at the moment when the rotor B field is in phase with the stator grid field?

Think about that.
What do you mean by "stator grid field" ?
With nothing connected to the machine terminals, armature current is zero so there's no armature MMF to make a stator field.
Only field in the stator is therefore that of the rotor.

You must get terminal volts in phase with grid volts else there will be violence when the two connect..

Ok so one more question, in reality then multiple things happen all together, firstly as the grid frequency changes a little due to various factors like large loads etc, the real physical rpm of the generators change by some few fraction of an rpm which might correspond to the 0.1 (that's a lot) or so about change in electrical Hz?

remember it slows down only briefly and just enough to move a few degrees relative to nearby machines.

Then parallel to this the angle between different generator rotor's may also change as one might have a prime mover that applies more power to the generator when experiencing a frequency decrease while some other generators rotor might have a prime mover already going at full capacity or incapable or adding more power so that generator can add much less to the frequency increase than the first one?

Yes, one with its valves already wide open can't pen them any further.

Thirdly also the angle between each generator's rotor and stator field changes while all of the before mentioned things happen?

That angle is called "Power Angle" and is proportional to power being delivered, and to field strength.
You have multiple things interacting - voltage regulator, governor, nearby machines., and customers...
Imagine a mobile hanging from your ceiling,

hit one and they all respond.

 

[essenmein]

Think for a moment about real power, ...
a machine's internal resistance must be very low else unmanageable heat would result .
Our stator current was 20 kiloamps.
Stator resistance is designed low as practical so that the machine will be efficient.
We in the plant consider stator resistance to be negligible.

For power plants with more emphasis on efficiency than space, that's a nice luxury!

For us its all about transient power from small machines, so stator resistance is significant, you can watch the control terms change as the stator heats up (the current controllers compensating for the changing iR term).

But we do monitor temperature rise of the hydrogen exiting each armature conductor because a change from the machine's normal pattern would indicate trouble. An armature conductor that's broken a strand will run hot because it experiences higher current density. One with a blockage in its cooling flowpath will run hot as well. .

Stator is much more inductive than resistive. Inductance doesn't produce heat - recall reactive current is wattless..

Just as we adjust steam flow to keep megawatts at our fair share of what the customers are consuming at the moment
we adjust field current to keep megavars at our fair share too .
Further, to keep voltage constant where the customers are,
we must raise or lower our voltage as needed to push current through the impedances between us and those customers, namely the transformers and power lines between us. That we do by adjusting field current and again it's done under the watchful eye of central dispatch..

So to answer your question - over the course of a day's operation it's not the machine that changes, it's the load to which the machine is connected.

old jim

I went and looked at the other thread, and reconciled it with how I look at machines, (analog come DC-DC now inexplicably working on inverters)

If I were to paraphrase what adjusting the field does once running, it adjusts the BEMF vector to compensate for the voltage vector introduced by the stator inductor and changing load current/phase to keep the stator power factor at 1 (it's most efficient operating point).

So in DQ terms, what you are doing as a generator is aiming for negative iQ (ie negative torque producing current with positive rotation, this flips if spinning the other direction), with no iD, however since iQ will create a voltage vector perpendicular (and leading) due to iwLQ (LQ being the q axis stator inductance), this requires you now adding a voltage in the D axis to do that, in addition to the Q axis voltage vectors (BEMF and iR), since you cannot change the terminal voltage, the L, or the stator current, all you can do is change the BEMF, by twiddling the field current.

 

[jim hardy]

If I were to paraphrase what adjusting the field does once running, it adjusts the BEMF vector to compensate for the voltage vector introduced by the stator inductor

Yes. Please look up 'armature reaction' to get a handle on the standard terminology.

...and changing load current/phase to keep the stator power factor at 1 (it's most efficient operating point).

NO! We operate at lagging power factor nearly all the time. That's because the load is rather inductive as i explained.
Please don't speculate that's what was wrong with the earlier thread..
Go back and check your basics before asserting.

 

[jim hardy]

guys this thread is getting wild again.

I think we've laid a good foundation. I'm going to bow out and let you self study for a while.

Practice with a generator connected directly to infinite bus. Learn synchronous impedance and short circuit ratio.
Then add impedance between your generators and infinite bus.
You'll be well on your way.

old jim

 

[essenmein]

Yes. Please look up 'armature reaction' to get a handle on the standard terminology.NO! We operate at lagging power factor nearly all the time. That's because the load is rather inductive as i explained.
Please don't speculate that's what was wrong with the earlier thread..
Go back and check your basics before asserting.

One last question, do you run the generator with lagging power factor on the stator by choice or due to system limits?

 

[jim hardy]

One last question, do you run the generator with lagging power factor on the stator by choice or due to system limits?

Ohm's law. We have to supply what is the load.
A load with non unity power factor draws current at non unity power factor..
700 megawatts and a hundred or so megavars is not atypical for each of my two units. We ran at constant 700 megawatts because nuke fuel is so cheap, just varied megavars.
Plants all share both real and reactive load .

 

[essenmein]

Ohm's law. We have to supply what is the load.
A load with non unity power factor draws current at non unity power factor..
700 megawatts and a hundred or so megavars is not atypical for each of my two units. We ran at constant 700 megawatts because nuke fuel is so cheap, just varied megavars.
Plants all share both real and reactive load .

Yeah I got a bit ahead of myself mentioning load current/phase. I shouldn't have added those as I was still looking at it from infinite bus perspective to understand why changing field current affects power factor when for automotive alternators it controls output power.

 

[jim hardy]

Ok so many things are getting clearer for me,

great. Build from there.

your freshman physics class taught that work = force X distance and power = force X velocity. Power also = torque X angular velocity..

Ii hope I said everything correctly.

Words are difficult and that's the beauty of math it's unambiguous.
In my day in US we used English units where
horsepower = 2π X Torque(in ft-lbs) X RPM / 33000
and it's pretty obvious that 33000 is the product of units conversion factors 550 ft-lbs/sec/hp and 60sec/min
SI equation is cleaner Power(Watts) = Torque(Newton-meters) X ω(radians/sec)
i think your car analogy will work.

@Jim since you've worked in a nuke plant, may I ask what was the typical day to day output power in percentage of the plant's generator's (in electrical MW) with respect to their maximum rated capacity and also how that corresponded to the used thermal capacity of the reactor also in percentage?

We ran at nominal full power. The reason is economics, nuke fuel was far cheaper than natural gas or oil.
Reactor power you measure by a heat balance on the plant - basically it's
reactor power = (feedwater flow ) X (enthalpy rise between inlet and outlet of boiler) + (heat lost through pipe isulation)
and we kept that as close to 100% of licensed power as we could.
Obviously feedwater flow and temperature are the two most important measurements for that calculation . We installed dedicated sensors and programmed the plant computer to keep track.

Our turbogenerator was slightly oversized to accommodate an eventual licensed power uprate. So it's reactor power and steam side efficiency that limit electrical output. While capable of slightly over 760 megawatts we made typically 720 dependent mostly on temperature of condenser cooling water.

reactor has a given margin from minimum to maximum thermal output at which it can operate and probably there is a spot within that margin at which it operates with maximum efficiency (fuel burnup versus output power) I assume that most (any?) reactors are most efficient if run at 100% of it's limit?

Think about that. The reactor is just a big water heater . What heat it makes all goes into that water.
Since it takes about 40 megawatts to run all the pumps and heaters and lights in my plant, it's the plant surrounding the reactor that enjoys an efficiency gain from full power operation.

Also if the nuke plant had a (hypothetical) generator that could be run at 100% reactor output all the time without that affecting VARs or other grid factors that are determined by the control of turbine flow would that be beneficial in any way to reactor efficiency or would there be no difference?

Not to the reactor. But to the overall plant, yes. And that's how we ran, the term is "Base Loaded". Fossil plants track changing load demand.

It gets more complicated when one considers stator attached to grid which has inductance and capacitance added to the resistance, so now let's consider that the waveform at the stator terminal connection (from the grid) has it's current lagging it's voltage by some 45 degrees or so, which means reactive power is flowing in the grid, can someone explain me in simple detail how increasing excitation current will affect this and in which way?
Isn't it also important at which power angle the rotor is when the excitation current is increased, doesn't that determine in which direction will the phase angle difference go?

i think that's answered in the old thread i referenced earlier

https://www.physicsforums.com/threads/voltage-and-reactive-power-relationship.876346/#post-5503931

see also the three threads linked in post 3 of that thread

 

[anorlunda]

This thread has become too rambling. That destroys its archive value for future people who might read it. I hate to be overly strict, jumping on every minor off-topic comment, but in this case we are way beyond minor and into major.

I'm going to close this thread for now. I may also have to do some cleanup, deleting off topic stuff to preserve the archival value of the thread.

@girts and @essenmein please see your inbox.

 

[jim hardy]

Thanks @anorlunda