# GE Speed Control Loop

G

#### gustavo_marcelo

Hi all,

First of all, my sincere apology if this thread somehow had been posted. After using the search facility, I did found a lot of post with regards to Speed Control Loop for GE. Theoretically we were given the loop with:
FSRN = (TNRL - TNH ) FSKNG + FSR.

My inquiries are:

1. Does the FSKNG is the governor droop value?

2. FSR definitions - it was given as the full speed no load value of the fuel, typically 20 % FSR - I don't really understand this statement?

2. What does the unit for FSRN, TNRL,TNH, FSR?

3. How to prove this equation with some numerical value been given to the equation?

4. Does the equation above is also load control equation? Speed droop also means Droop Load control, is it?

5. How does the iso speed control equation looks like?

Appreciate feedback from the expert.

#### CSA

gustavo_marcelo,

Good on you for trying to use the Search function. But, I don't know WHERE you got the equation you posted--certainly NOT from any post to control.com.

Remember: you asked. So, here goes; we're jumping into the pool straight away. The basic equation for <b>Droop</b> speed control in GE Mark* heavy duty gas turbine control systems is:

FSRN = ((TNR - TNH) * FSKRN1) + FSKRN2

(I might have the FSKRN1 and FSKRN2 switched; I don't have any drawings to look at at this writing. But for the purposes of this example we'll leave it as is and call FSKRN1 the "Droop Speed Control Gain" and FSKRN2 the "Droop Speed Control Offset" even though in the Mark* it be the opposite.)

FSRN = Speed Control FSR (Fuel Stroke Ratio) (% of control valve opening, or travel (opening or travel is "stroke" in GE-speak))
TNR = Turbine Speed Reference (% of Rated Speed)
TNH = Actual, Measured High-pressure Shaft Turbine Speed (% of Rated Speed)
FSKRN1 = Droop Speed Control Gain (% FSR/% Rated Speed)
FSKRN2 = Droop Speed Control Offset (% FSR)

The "offset" value is also referred to sometimes as FSNL FSR (Full Speed-No Load Fuel Stroke Reference).

So, you know that it takes a certain amount of fuel to achieve and maintain rated speed (FSNL) in preparation for synchronizing--this amount of fuel flow-rate (or FSR--because fuel flow-rate is a function of valve opening (valve travel; valve stroke)). A typical value for FSNL FSR ("Droop Speed Control Offset") is approximately 20% (some units are a lower; some are higher; but most are around 20% FSR). And this fuel has to be always flowing to keep the machine at rated speed at a minimum.

When the unit is at rated speed (FSNL) and is ready to be synchronized, the Turbine Speed Reference (TNR) is usually (typically) 100.3% and the actual turbine speed (TNH) is also approximately 100.3%--so the difference between TNR and TNH (the "error") between TNR and TNH is 0% (100.3%- 100.3%=0%). Multiply 0% times FSKRN1 (whatever that value might be--we'll get back to that in a minute) and you get 0% FSR; add 20% (FSKRN2) to 0% FSR--and you have 20% FSR. Which just so happens to be the typical value of FSNL FSR for most GE-design heavy duty gas turbines!

Now, for the FSKRN1 value (the Droop Speed Control Gain value). Well, it's calculated like this. We already know that FSNL FSR is 20% (in our example). And, a typical value of FSR for Base Load for a GE-design heavy duty gas turbine in new and clean condition, with clean inlet air filters, and at nameplate ambient conditions (usually 59 deg F at 60% Relative Humidity) and with fuel that matches the specification provided to the turbine packager when it was ordered and built is about 70% FSR. So, to get from FSNL (Full Speed-<b>No</b> Load) to Base Load with a machine in new and clean conditions with clean inlet air filters and at rated ambient conditions with rated fuel means the FSR increases 50% (in our example).

The last little bit of this calculation has to do with the amount of Droop regulation, or the Droop Setpoint, or Speed Regulation of the machine. And for most heavy duty gas turbines (of just about any manufacturer) the typical amount of Droop regulation (or Droop Setpoint or Speed Regulation) is 4%. That means that when the unit is at Base Load in new and clean condition with clean inlet air filters and at nameplate ambient conditions with fuel that matches the specification the turbine speed reference (TNR) will be 104.0% (approximately--maybe a tenth of a percent higher or lower, but pretty darn close). AND, when the unit is synchronized to a well-regulated grid and producing power (ANY amount of power, from 0 MW to Base Load) the actual turbine speed (TNH) will be 100.0%. So, the difference (the speed error) between TNR and TNH when the unit is at Base Load will be 4% (104% - 100% = 4%). Which is equal to the typical Droop regulation, or Droop setpoint, for heavy duty gas turbines of 4%.

So, if we take the amount of fuel required to get from FSNL to Base Load, and divide it by the difference between the turbine speed reference at Base Load and the actual turbine speed at Base Load, we arrive at the value of Droop Speed Control Gain--FSKRN1 in our example. So, ((70% - 20%)/(104% - 100%)=50%FSR/4%Rated Speed=12.5%FSR/%Rated Speed.

Now, let's say we wanted to load the machine to 50% of rated load. Well, that would correspond to 50% of the difference between TNR at Base Load and TNR at FSNL, or (104% - 100%)*0.5=2%Rated Speed. Add 2% to the value of FSNL FSR, 100%, and we get 102%. Plug all of our values into the equation:

FSRN=((102% -100%)*12.5%/%)+20%=((2%*12.5%/%)+20%=25%+20%=45% FSR

Said another way (for the purposes of proof), FSRN at 50% load would be approximately 45% FSR, which is half of the difference between Base Load FSR (70%) and FSNL FSR (20%) plus FSNL FSR ((70%FSR-20%FSR)/2)+20%FSR=(50%FSR/2)+20%FSR=25%FSR+20%FSR=45%FSR.

You can do the same for any proportion of load, and the same proportion of the Droop regulation (or Droop setpoint) value. For Base Load, for example, 4% (which is 104%-100%) times FSKRN1 (12.5%/% in our example here) plus 20% (FSNL FSR) equals 70% ((4%*12.5%/%)+20%).

You can even make a table of the value of FSRN for any value of load. The columns would be FSRN, TNR, TNH (which should be 100%), FSKRN1 and FSKRN2.

The formula FSRN=((TNR-TNH)*FSKRN1)+FSKRN2 is the formula for any straight line (y=mx+b; or f(x)=mx+b). Because that's what it is (the Droop Speed Control formula: a straight line). The "difference" is that the "x" term is actually two variables instead of just one (TNR-TNH). The "x" term is the "speed error", or the difference between the speed reference and the actual speed. And, when the turbine is driving a synchronous generator connected to a well-regulated grid (one with the frequency pretty close to rated) the actual speed (TNH) will be almost always 100%, and fairly stable as well as being constant. So, to change load, one changes the turbine speed reference, TNR, since TNH usually stays constant and is stable. And, the error between TNR and TNH multiplied by the gain and when an offset is added to it defines how much fuel will be admitted to the combustors.

Droop Speed Control is all about how much the fuel flow-rate (FSR in the case of GE-design heavy duty gas turbines controlled by digital GE Mark* turbine control systems) will change for a change in the error between the turbine speed reference (TNR) and the actual turbine speed (TNH). (FSR, Fuel Stroke Reference, is equivalent to fuel flow-rate--because control valve opening is related to fuel flow-rate.)

When a prime mover (ANY prime mover) is driving a synchronous generator synchronized to an AC grid with other prime movers driving synchronous generators the speeds of ALL of the prime movers and generators is what is called "synchronous speed" and synchronous speed is determined by the frequency of the grid. There is a formula, F=(P*N)/120, that says the frequency (F) of a rotating machine (motor or generator) is equal to the number of magnetic poles (P) of the generator rotor, multiplied by the speed of the generator rotor (N, in RPM), divided by 120. Now, if you want to solve the equation for speed (N), it becomes N=(120*F)/P. So, for a grid operating at 50 Hz, a two-pole generator will spin at 3000 RPM; a four-pole generator will spin at 1500 RPM. So, on a 50 Hz grid all two-pole synchronous generators will be spinning at 3000 RPM (that's the synchronous speed for a two-pole generator running at 50 Hz), and all four-pole generators will be spinning at 1500 RPM, and so on for any generator with any number of poles.

When the operator changes load, what is really happening is that--even though the operator is looking at the MW meter (or the kW meter) the prime mover speed reference (the turbine speed reference in the case of GE-design heavy duty gas turbines) is changing--and as the error between the turbine speed reference (TNR) increases or decreases, the fuel control valve will open or close as FSRN increases or decreases. The ACTUAL speed of the machine (the machine's synchronous speed) DOES NOT change (on a well-regulated grid!)--only the speed reference changes. And, so, when the operator increases the speed reference to increase load, the actual speed is NOT CHANGING; the actual speed "lags" or "droops behind" the speed reference. That, I believe, is how Droop Speed Control got it's name. When the speed reference is increased to increase the load being carried by the generator, the actual speed doesn't change (because it's fixed by the grid frequency! which should be stable and constant), and the actual speed "droops" or "lags" behind the speed reference.

Now, to switch to Isochronous Speed Control. We have seen in Droop Speed Control that the actual speed does not change when the speed reference changes. In Droop Speed Control the speed reference is 100.0%, and when the actual speed changes the turbine control system immediately changes the fuel flow-rate to the turbine to return the speed to 100.0%, so that the actual speed always remains at 100.0%.

How does the actual speed change, you say, when the grid frequency is constant and stable? Well, let's say a large turbine-generator suddenly trips off the grid. The immediate effect of that will be that the grid frequency will start to drop--and the actual speeds of all the machines synchronized to the grid will also start to drop. BUT, if there is a machine whose control system is set to operate in Isochronous Speed Control the control system will sense the grid frequency change and will immediately add more fuel to the turbine to raise the frequency--of ALL the machines synchronized to the grid!--back to 100.0%. That's what Isochronous Speed Control does--it monitors grid frequency and changes the fuel flow (or steam flow, or water flow--the energy flow-rate into the prime mover) to keep the grid frequency as close to 100.0% as possible. The operator does NOTHING--the turbine control system (the "governor") automatically senses the frequency change and immediately compensates for the frequency change to maintain 100.0% frequency (speed--since speed and frequency are directly related).

Lastly, there should never be more than one machine operating in Isochronous Speed Control connected to a grid. That's because they will fight each other (the outputs of the machines will be unstable--which will make ALL the other machines synchronized to the grid also be unstable (frequency and speed)).

I know you asked several numbered questions. I would like to see you answer the questions based on the information above. That way, we can see if the information was helpful or not.

I actually hesitate--greatly--to explain Droop Speed Control as I have above. For the simple reason that Droop Speed Control is VERY misunderstood, and people can see how to change the Droop regulation (Droop setpoint) and make changes--and that leads to CHAOS. Because, there are LOTS of other things in the Mark* turbine control system that are all based on these relationships all beginning with the Droop regulation--for example, the rate at which the unit is loaded and unloaded when the turbine speed reference is changed. People think to solve some perceived problem (that is, not a real problem!) they need to change the Droop regulation, or the Droop setpoint. And, then other things change which they didn't intend to change--because they don't understand the knock-on effects of changing the Droop regulation. And, they they have problems they didn't count on having, and they blame the Mark* for the problems--problems they created because they didn't understand what they were doing, and because they didn't really have a problem in the first place (that perceived problem, which wasn't really a problem).

So, while I didn't really tell you how to change the Droop regulation (or anybody else for that matter!), DON'T DO IT!!! NEVER change the Droop regulation without fully understanding why you think it needs to be changed, and all of the knock-on effects of changing the Droop regulation. One of the effects of changing the Droop regulation is that when the owners of the power generation equipment at your site signed a contract to be allowed to synchronize to the grid with other prime movers and generators they told the grid operators how much Droop regulation the unit they would be synchronizing to the grid had. And, the grid operators use that value to make stability calculations for the grid--which are VERY important to maintaining and re-establishing grid stability should it ever become unstable. And if anyone changes the Droop regulation without first ASKING the grid operators in some parts of the world the owners of the power plant can be sued, or worse, held criminally negligent (meaning they can go to jail or pay fines). Because in effect, changing the Droop regulation can have detrimental effects on the grid stability. So, it should be obvious--just don't do it (change Droop regulation).

In any case, it should NEVER be necessary to change the Droop regulation of a machine--unless there is agreement with the grid operators of the grid the machine is synchronized to. The grid operators may ask for the Droop regulation of a machine to be changed, or, in some rare case (that I can't even imagine) an owner might ask the grid operators if the Droop regulation of a machine could be changed. But, that's probably the result a misunderstanding of Droop regulation, and the wrong solution to some problem that another, simpler fix would be more appropriate for.

So, let's hear your answers to your numbered questions now that you have read the above. It shouldn't be hard; there should be very little guessing because most of the answers have been given directly; it should only be necessary to reason the answers which haven't been based on the information provided above.

Droop Speed Control is about: How much the energy flow-rate into the prime mover will change for a given change in the error (the difference between) the prime mover speed reference and the actual prime mover speed. It's a way of predicting, using calculations, how much the energy control valve will move for any speed error. Why speed? Because, in the beginning of AC power generation there was only speed. And speed is directly related to frequency. And since frequency is what we'd like to control, and speed and frequency are related (in effect, they are the same--especially if we refer to both in percent (when frequency is at 100% then speed is also at 100%!). And because the first prime mover governors were fly-ball governors and there was no way to get watts or kilowatts or megawatts into a fly-ball governor control mechanism everything was related to speed--because the position of the fly-balls on the governor was directly related to the speed of the shaft the fly-balls were being spun by. And, as governors became more and more sophisticated (and electrical and then electronic and then digital) they still had to synchronize to grids with older governors, so every governor to this day uses speed control. Speed control is important, again, because it's directly related to frequency--AND, we don't want prime movers to overspeed, either, because that can be catastrophic and even deadly.

It's all about how much the energy (fuel for a heavy duty gas turbine) changes when the speed error (or the load error) changes.

#### gustavo_marcelo

The equation was from a notes, based on the block diagram, and I just transferred it into that form. Let me digest your respond first and try to understand it.

#### gustavo_marcelo

Thanks CSA.
Refer below.

"So, to get from FSNL (Full Speed-No Load) to Base Load with a machine in new and clean conditions with clean inlet air filters and at rated ambient conditions with rated fuel means the FSR increases 50% (in our example)."

Q: I thought at the base load (rated load), FSR = 100%.

There is a paragraph you mentioned FSNL =100.3% prior to synchronising.The other paragraph where you were calculating the droop, you mentioned the speed is 104%.

"So, if we take the amount of fuel required to get from FSNL to Base Load, and divide it by the difference between the turbine speed reference at Base Load and the actual turbine speed at Base Load, we arrive at the value of Droop Speed Control Gain--FSKRN1 in our example. So, ((70% - 20%)/(104% - 100%)=50%FSR/4%Rated Speed=12.5%FSR/%Rated Speed."

Q: So, droop value (4%) is not the FSKRN1 as I initially thought?

"Add 2% to the value of FSNL FSR, 100%, and we get 102%. Plug all of our values into the equation:

FSRN=((102% -100%)*12.5%/%)+20%=((2%*12.5%/%)+20%=25%+20%=45% FSR"

Q: Why does here FSNL FSR = 100% and not 20%?

Q: Why does here TNR is 100% before adding the 2%? Is it because we consider after synch, it will be 100% , same as TNH hence no error?

Q: Now the FSR opens more from 20% to 45% when loaded at 50%?

What happened if there is no Offset value, FSKRN2? Can this control works?

Now, to switch to Isochronous Speed Control. We have seen in Droop Speed Control that the actual speed does not change when the speed reference changes. In Droop Speed Control the speed reference is 100.0%, and when the actual speed changes the turbine control system immediately changes the fuel flow-rate to the turbine to return the speed to 100.0%, so that the actual speed always remains at 100.0%.

Q: Does the sentences "In Droop Speed Control....

actually refers to iso, mistyping?

My inquiries are:

1. Does the FSKNG is the governor droop valueqNot the typical 4% droop value? But the calculation to get FSKNG does include the value.

2. FSR definitions - it was given as the full speed no load value of the fuel, typically 20 % FSR - I don't really understand this statement?

Understood. The fuel flow rate to achieve full speed no load value.

2. What does the unit for FSRN, TNRL,TNH, FSR?

FSRN (% FSR), TNRL (% speed ref), TNH (% actual speed), FSR (% FSR)

3. How to prove this equation with some numerical value been given to the equation?

Understood.

4. Does the equation above is also load control equation? Speed droop also means Droop Load control, is it?

Yes, it is.

5. How does the iso speed control equation looks like?

With 100% reference or 100% TNR.

#### Jolek

CSA,

As always, pretty sophisticated process explained in very simple way!

#### CSA

Jolek,

Thanks, but I must have done something wrong for gustavo_marcelo. So, we'll try again. (Doesn't it feel sometimes like this site should be named droopspeedcontrol.com?)

It would seem, gustavo_marcelo, you are not paying attention to the units. And, yes--it is somewhat confusing because all those percents! (% Rated Speed, and % FSR, and % Rated Speed/% FSR).

Pay attention to the units of the terms and variables. All will be good if you do that. FSNL FSR (% of rated fuel valve opening) is 20% in our example (and for many GE-design heavy duty gas turbines, plus or minus a couple of percent of travel/stroke/position, whatever you want to call it). And, in our example, rated load is 70% FSR (% of rated travel). Yes; one might think it should be 100%, but for heavy duty gas turbines as ambient temperature decreases the unit can burn more fuel <i>at Base Load</i>, and the fuel control valve needs to be able to open more. If it were at 100% of travel at Base Load for a new and clean machine with clean inlet air filters and good fuel and the ambient temperature decreased below nameplate conditions the unit couldn't make any more power, because the fuel control valve couldn't open any further. So, the designers and programmers usually pick an control valve opening of somewhere between 70% and 80% of rated fuel valve travel/stroke/position for Base Load so as not to limit turbine-generator power output when ambient conditions are below nameplate conditions.

When the unit is at rated speed, 100.0% of synchronous speed, FSNL (Full Speed-No Load), there has to be a certain amount of fuel flowing just to maintain rated speed (it takes a LOT of fuel just to spin the axial compressor and generator rotors at rated speed when the unit is not even producing electrical power). That's FSKRN2 (in our example). When the unit is being synchronized the turbine control system usually makes the generator frequency (unit speed) a little higher than grid frequency (for GE Mark* turbine control systems that's usually approximately 100.3% of rated speed). Once the generator breaker closes during synchronizing, the unit speed immediately drops back to 100.0% (if that's what the grid frequency is), and the extra fuel that WAS making the unit spin at 100.3% of rated speed goes to making power. That's 0.3 out of 4 (for a unit with 4% Droop regulation), or 7.5% of rated power output (0.3/4=0.075, which is 7.5%). For a machine rated at 40 MW, that would be (7.5% of 40 MW = 3 MW).

Let's make a table using the values from our example! This table will be for when the grid frequency (actual speed) is at rated (100.0%) and for the conditions specified in the last column on the right (I know; I should have put it on the left, but ... I didn't.)<pre>
FSRN TNR TNH FSKRN1 FSKRN2 Condition
(% FSR) (% Rated Speed) (% Rated Speed) (%Rated Speed/% FSR) (% FSR)
100.0% 100.0% 100.0% 12.5 %/% 20.0% FSNL (Rated Speed)
20.0% 100.3% 100.3% 12.5 %/% 20.0% Synchronizing Speed
32.5% 101.0% 100.0% 12.5 %/% 20.0% 25% Rated Load
45.0% 102.0% 100.0% 12.5 %/% 20.0% 50% Rated Load
57.5% 103.0% 100.0% 12.5 %/% 20.0% 75% Rated Load
70.0% 104.0% 100.0% 12.5 %/% 20.0% 100% Rated Load</pre>
Is there a pattern here? (Yes.) To increase the load on the generator, the operator increases the speed reference, which increases the speed error (I wanted to put a separate column in for speed error--but it makes the table too wide). Droop Speed Control is about the speed error--the difference between the speed reference and the actual speed. ANYTHING that makes the speed error change makes the FSR (the energy flowing into the turbine) change.

Now, let's see what happens when the grid frequency varies while the speed reference remains constant (which is what would happen if the unit were operating at a stable load while the grid frequency was not at rated (TNR would <b>NOT</b> be changing, TNH <b>WOULD</b> be changing), which changes the speed error!!!).<pre>
FSRN TNR TNH FSKRN1 FSKRN2 Condition
(% FSR) (% Rated Speed) (% Rated Speed) (%Rated Speed/% FSR) (% FSR)
45.00% 101.5% 99.5% 12.5 %/% 20.0% Grid Freq. Decrease by 0.5% (Load Increases to 50.0% Rated)
51.25% 101.5% 99.0% 12.5 %/% 20.0% Grid Freq. Decrease by 1.0% (Load Increases to 62.5% Rated)
38.75% 102.0% 100.5% 12.5 %/% 20.0% Grid Freq. Increase by 0.5% (Load Decreases to 50.0% Rated)
32.50% 102.0% 101.0% 12.5 %/% 20.0% Grid Freq. Increase by 1.0% (Load Decreases to 50.0% Rated)</pre>
This is how Droop Speed Control works when there is a grid frequency disturbance. The <b><i>speed error</b></i> (the difference between TNR and TNH) changes, and that changes the amount of fuel (FSR) that will be admitted. It doesn't matter if the speed reference changes, OR the actual speed changes--anything that changes the difference between the two (the speed error) will change the fuel flow-rate. This is simple Proportional control--no Integral. No Derivative. Just pure, straight, unadulterated Proportional control. Simplest there is, really--it's just that instead of one "x" variable, there are two. One (TNH) is usually, and presumed to be, constant and stable (because AC power transmission and distribution grids are supposed to operate at a stable, constant frequency). Sometimes, they don't, but mostly they do. This is "pure" Droop Speed Control--nothing else. (GE uses something else--which I think is the formula you originally provided--called 'Constant Settable Droop Speed Control.' It uses some Integral action from an outer loop, which is load control.

The amount of energy flowing into the prime mover of a synchronous generator is directly related to the amount of power being produced by the generator. Increase the energy flowing into the prime mover, and the generator output will increase. Decrease the energy flowing into the prime mover, and the generator output will decrease. Droop Speed Control--be it "pure" Droop Speed Control, or Constant Settable Droop Speed Control--is how GE-design heavy duty gas turbine load is controlled between 0 MW and Base Load, and between Base Load and 0 MW. It's how almost EVERY synchronous generator prime mover governor controls the energy flow-rate into the prime mover--because with all the prime movers using essentially the same control scheme (Droop Speed Control!) they can all stably share in producing the power required by the load--the sum of all the electric motors and tea kettles and lights and computers and computer monitors and televisions and espresso machines. All of the generators synchronized together act as ONE generator, running at a single frequency, to supply the "load"--which is the sum total of all the electric motors and tea kettles and lights and computers and computer monitors and televisions and espresso machines.

When grid frequency decreases it's because of one of two reasons: Either some generation has tripped off line (or someone reduced the power output from their prime mover and generator--which is the same thing as losing generation), OR some load has been added to the grid and the grid operators haven't responded appropriately yet (to raise generation to match the added load).

When grid frequency increases it's because of one of two reasons: Either some generation has come on line (or someone increased the power output from their prime mover and generator--which is the same thing as adding generation), OR some load has been removed from the grid and the grid operators haven't responded appropriately yet (to lower generation to match the decreased load).

To change a generator's load when the grid frequency is stable and constant, the operator changes the prime mover speed reference, TNR. TNH, prime mover actual speed, stays constant when the grid frequency is constant, and changing TNR changes the speed error--the difference between the speed reference and the actual speed. When the speed error changes, the amount of fuel being admitted to the combustors changes, which change the load being produced by the turbine and generator.

When grid frequency changes and TNR is stable and constant, the speed error between TNR and TNH changes--which changes the amount of fuel being admitted to the combustors which changes the load being produced by the turbine and generator.

Many plant managers and owners and operators think when the grid frequency deviates from nominal (rated) the power output of the machine should NOT change. But that is WRONG! When a synchronous generator is synchronized to a grid with other prime movers and generators, and the prime mover is not at rated power output (a new condition!) and is operating on Droop Speed Control (sometimes called 'free governor mode', or even 'primary frequency response') the load SHOULD change when the frequency changes--to support grid stability. And, grid operators believe and count on units responding appropriately when there are grid frequency disturbances--and when they don't, the grid frequency disturbances can get even worse!

I think this is enough for a Saturday, eh?

G

#### gustavo_marcelo

Thanks CSA.
That's awesome.
By the way may I know which part I understood wrongly so that I can re read it...

#### CSA

guatavo_marcelo,

One more thing you might be missing about AC (Alternating Current) power systems that is extremely important: When multiple synchronous generators are synchronized together on a grid the magnetic fields of the rotors are LOCKED into the magnetic fields of the stators. Thatâ€™s why they all spin at speeds that are called synchronous speeds (related to the frequency of the grid and the number of magnetic poles of each generatorâ€™s rotor. No single machine can spin faster or slower than all the machines.

On a grid with multiple synchronous generators synchronized together there is either one prime mover governor operating in Isochronous Speed Control mode which is continuously changing its LOAD to respond to changes in the number of electric motors and lights and tea kettles and computers and computer monitors and televisions and espresso machines, or the grid operators are using one or more machines operating in Droop Speed Control mode to to respond to changes in the grid load, to maintain frequency. On really large grids (called infinite grids) there is so much inertia from all the generators and their prime movers that the grids are pretty stable and not very prone to large frequency deviations unless there are large disturbances (like large blocks of load suddenly tripping off line, or large generators operating at high loads suddenly tripping off line.

When a machine is operating in Droop Speed Control mode and is synchronized with multiple other generators and their prime movers changing the energy flow-rate into any single machine has little effect on the speed of that machine, because, in effect, that machine is trying to increase the speed of ALL the machines, because the interaction of the magnetic fields in each machine makes all of them operate like a single machine operating at one frequency (because they are all operating at a single frequency!).

Watch the machines at your plant when they are synchronized to the grid and loads are being changed; specifically watch the speeds of the machines to see if they are changing. Yes; when the machines are being started and stopped when they are NOT synchronized to the grid their speeds do change when the ful flow-rates change. BUT, once the generator breaker closes and the unit is synchronized to the grid changing the fuel flow-rate DOES NOT change the machineâ€™s speed. The magnetic field interaction in the generator prevents the speed from changing and the generator converts the extra torque to amperes in the stator, which means the power being produced by the generator increases.

Thatâ€™s what generators do: they convert torque to amperes. Which is the opposite of what motors do: they convert amperes into torque. Generators really â€œdriveâ€œ motors. Thatâ€™s what electric transmission and distribution systems do: They transmit torque from one place (where the prime movers and generators are) to many other places where it can do work, using wires.

On an AC system, that is done by producing enough work to keep the grid frequency stable and only producing as much power as the load(s) need. Any imbalance results in grid frequency deviations and disturbances. Itâ€™s a balancing act, and every generator and prime mover plays a role in that balancing act, and the larger the prime mover and generator the more important the role.

But, Droop Speed Control lets some other machine or some other entity (like the grid operators) control the frequency by balancing generation and load. Thatâ€™s why itâ€™s called Droop Speed Control, because even though the speed reference is telling the machine to speed up (to load the machine) it canâ€™t speed up. And the difference, the speed error, is what controls the energy flow-rate into the prime mover. And because the speed canâ€™t change when the energy flow-rate changes the generator converts the torque that would otherwise make the speed increase into amperes in the generator stator which increases the power being produced by the machine.

So, forget about the speed of an individual machine changing when the energy flow-rate into the prime mover changes, that doesnâ€™t happen (to any appreciable extent). That only happens on small grids that arenâ€™t being operated properly. Thatâ€™s what control systems are for, to automatically control load, and frequency. And grid operators help with that, too.

Look at the speeds in the first table as the speed reference (and load) changes, and go watch what happens to the speeds of units in your plant when units in your plant are loaded and unloaded. Because the speeds wonâ€™t change very much, if at all, when they are synchronized and producing power. Not if the frequency is stable.

Hope this helps!

#### CSA

gustavo_marcelo,

>By the way may I know which part I understood wrongly so
>that I can re read it...

Yes; let's try to correct some of the misunderstandings.

>Q: I thought at the base load (rated load), FSR = 100%.

No; as was explained, Base Load FSR is less than 100%--because the fuel control valve should never be fully open under any circumstances. FSR means Fuel Stroke Reference; 'Stroke' is GE-speak for position, or travel. 100% stroke would be fully open--the valve would be up against the fully open mechanical stop, and that would not be a good thing.

>Q: So, droop value (4%) is not the FSKRN1?

No. GE almost NEVER actually shows the Droop setpoint/regulation value--because almost everyone would be changing it because of misconceptions and misunderstandings about what the Droop setpoint/regulation value does.

So, we said FSKRN1 (in our example) was the no-load to full-load FSR divided by the Droop setpoint/regulation (the engineering units for FSKRN1 are "% FSR/% of Rated Speed). The units for Droop setpoint/regulation would be % of Rated Speed (4% of Rated Speed in our example, and for most GE-design heavy duty gas turbines--and for most heavy duty gas turbines around the world!). So, if you want to know what the Droop setpoint regulation is for your machine (using the GE "pure" Droop Speed Control formula, with FSKRN1 and FSKRN2) you would divide the no-load FSR (20% in our example) to full-load FSR (70% in our example), or 50%, by the value of FSKRN1 (from the Mark* Control Constants Display), 12.5%/% in our example. That would be 50%/12.5%/%, or 4%. The Droop setpoint/regulation for the machine in our example would be 4%: ((70%-50%)/(12.5%/%))=4%. The units for Droop setpoint/regulation are % of Rated Speed.

>Q: Why does here FSNL FSR = 100% and not 20%?

I made a boo-boo here. Instead of FSNL FSR I should have said FSNL Speed--the Full Speed-No Load speed is 100%. My sincere apologies for any confusion; I am not the best proof-reader of my writing--especially on a Saturday.

(Actually FSNL Speed and Full Speed-Full Load Speed when a unit is synchronized to a grid are exactly the same: 100% of Rated Speed.)

>Q: Why does here TNR is 100% before adding the 2%? Is it
>because we consider after synch, it will be 100% , same as
>TNH hence no error?

I'm not quite sure I understand what the problem here is. When the unit is at FSNL, the Turbine Speed Reference (TNR) is 100%. And, when the unit is synchronized to a grid operating at rated frequency at any load the actual Turbine Speed (TNH) is 100%.

In other words, under normal circumstances (when the grid frequency is constant and stable at rated frequency) the actual Turbine Speed (TNH) is always 100% of Rated Speed. So to change FSRN, since TNH isn't changing and can't be changed, one has to change TNR. At FSNL, or even 0 MW, TNR is 100% of Rated Speed. To increase FSRN, either TNR or TNH has to change--and TNH doesn't (normally) change. To change load, one has to change TNR, which is what clicking on RAISE SPEED/LOAD, or LOWER SPEED/LOAD does: changes TNR. Which changes the difference between TNR and TNH: the speed error. Remember: Droop Speed Control is about how much the energy flow-rate into the prime mover (FSRN, and FSR in our example) changes for a change in the speed error. Pure and simple.

>Q: Now the FSR opens more from 20% to 45% when loaded at
>50%?

Yes; when the unit is loaded to 50% of rated load, FSR will increase from 20% to 45%. That means the fuel control valve will open from 20% position to 45% position--and that will increase the fuel flow-rate into the combustors to increase the unit load (the electrical power output--because the speed can't change when it's synchronized to the grid, the generator converts the torque <i>that would be increasing the speed</i> into amperes. That's what generators do: they convert torque to amperes. Amperes are transmitted by wires to motors (and other devices that do work), where they are converted back to torque (or other useful work). That's what electric power transmission and distribution systems are for: transmitting and distributing torque from places where torque is available to places where is it desirable--using wires. In the early days of electrical systems, the most abundant sources of torque were hydro turbines, located sometimes very far from towns and cities where people worked--and where they needed torque for textile mills and flour mills and such. So, by using wires and hydro turbines and generators, the torque could be converted into amperes and transmitted long distances over wires where they were connected to motors at the textile and flour mills, and do useful work.

>What happened if there is no Offset value, FSKRN2? Can this
>control works?

No; the control cannot work without an offset value. Again, it requires a certain amount of energy just to keep a prime mover and its generator spinning at synchronous speed (100% speed; FSNL; rated speed)--and without that energy the unit can't be synchronized. (Well, it can, but it will cause a LOT of physical and electrical damage to the generator, switchgear, the coupling between the generator and the prime mover, and maybe even to the prime mover. It's not pretty!)

>Q: Does the sentences "In Droop Speed Control....
>actually refers to iso, mistyping?

No. When one changes the speed reference, TNR, (actually TNRI for Turbine Speed Reference-Isochronous for GE Mark* turbine control systems--which ultimately changes TNR) of a governor operating in Isochronous Speed Control mode one will change the actual running frequency of the system the Isochronous unit is synchronized to. The purpose of Isochronous Speed Control is to maintain a desired speed (frequency!) setpoint regardless of the load on the machine.

I have tried in the past to use a bicycle analogy for AC power generation and control, and it doesn't seem to be very well accepted. But, I'm going to use it here--because it is completely analogous to this explanation. A bicycle is a means for using the energy of one's body to move a person or people and/or some amount of goods from one place to another, and it can be done at various rates (speeds).

Now, if you pass your home and you throw the 5 kg of potatoes out of the basket to your sister while continuing to ride the immediate effect will be for the bicycle speed to increase--but your mind, sensing the increase in speed, will decrease the pressure your feet are applying to the pedals to return to and maintain the desired speed.

Electrical power generation is exactly the same--again, because it's all about transmitting energy from one place to another at a specific frequency--and speed is directly related to frequency. (I don't know why people don't like the bicycle analogy--because it gets even better!)

Now, let's say you were riding a tandem bicycle, and you had your best friend riding and pedaling also. And, you two were trying to maintain a desired speed on your ride on a flat road. Let's say your friend can't see the speedometer you can see. And, let's say he suddenly starts pedaling a little harder. The immediate effect will be that the speed of the tandem bicycle will increase--but you, seeing the change in speed, will have to decrease your pedaling pressure to return the tandem bicycle speed to the desired speed. YOU are again acting as the Isochronous governor of the bicycle. If your friend doesn't keep applying the same pressure to his pedals you are going to have to counter that by changing the pressure you are applying to your pedals to maintain the desired speed.

Now let's again say someone you are passing tosses a sack of rice in the basket on the front of the tandem bicycle; this time it's a 20 kg sack. Well, the immediate effect of the addition of the sack of rice will be to cause the tandem bicycle speed to slow down. If your friend keeps pedaling with the same pressure you are going to have to increase your pedal pressure to return to and maintain the same speed. And, if your friend also increases his pedal pressure at the same time, then the tandem bicycle speed might actually exceed the desired speed setpoint, and you will have to decrease your pedal pressure to return the tandem bicycle speed to the desired speed. In effect, he can't see the speedometer and he can't see what effect his change in pedal pressure has on the speed of the tandem bicycle. If he doesn't pedal consistently, then you are going to have to continually change your pedal pressure to control the tandem bicycle speed.

It's exactly the same with a power generation and transmission and distribution system. NO difference. It' takes a certain amount of pedal pressure to maintain a desired speed. Add some load, and it will take more pedal pressure to maintain the same speed. Take some load away, and it will take less pressure to maintain the same speed. When electric motors or espresso machines or televisions are turned on that increases the load on the grid (the transmission and distribution system). That in turn has the immediate effect of causing the grid frequency to drop (it might be a VERY small drop on a large or infinite grid--but it DOES drop!), and if there is a machine with its governor operating in Isochronous Speed Control mode it will detect the frequency change and increase its energy flow-rate into the prime mover to restore the grid frequency to normal. It does so very quickly, and automatically--without a human having to do anything! All of the machines which are operating in Droop Speed Control can be likened to your friend on the tandem bicycle just keeping a constant pressure on the pedals while the load in the basket changes--and you, as the "Isochronous governor" have to change the pressure you apply to the pedals to return to and maintain the desired speed.

A tandem bicycle has more than one crank, and those cranks are synchronized together. To maintain a constant speed the pressures applied to both those cranks have to be controlled. Ideally, one of the riders would be applying a constant pressure at all times regardless of the load in the basket or the headwind or the tailwind or the crosswind, or the steepness of the road's elevation changes (all of which are going to cause a change in speed), and the other rider would be adjusting the pressure from his feet to maintain the desired speed. If one rider needed to "rest" a little the other rider would have to make up the difference in order to maintain the desired speed. And the bicycle is transporting ("transmitting") the rider and any contents of the basket (or bag or backpack(s)) of the rider(s) from one place to another--doing useful work, using the energy from the rider(s).

(I still say this is a PERFECT analogy for an AC power generation and transmission and distribution system. I fail to understand why it isn't used more in texts and classes....)

>1. Does the FSKNG is the governor droop value Not the
>typical 4% droop value? But the calculation to get FSKNG
>does include the value.

We discussed this above, didn't we??? Look--this is DANGEROUS--telling how GE factors the Droop setpoint/regulation into the Mark* turbine control system! Very, VERY DANGEROUS!!! Why? Because there are a LOT of people who THINK they know what Droop is and does and think that by changing the Droop setpoint/regulation they can correct some <b>PERCEIVED</b> problem with their turbine operation or plant operation. And, they're <b>WRONG!</b> Very, VERY WRONG!!! There are so many ways changing the "Droop" setpoint can have unintended consequences in the rest of the turbine control and operation--and people just don't understand that, and they don't try to or even want to understand the knock-on effects. They THINK they know and understand Droop Speed Control, or that by just knowing the Droop setpoint/regulation value they will know and understand what Droop
Speed control is and does. I FAIL to understand why people (most EVERYONE) insist to know what the Droop setpoint/regulation of their turbine is, and how to change it. Like knowing this information is going to help them understand how Droop Speed Control works. IT'S NOT--going to help understand anything. And changing it is only going to have unforeseen and unintended consequences, which will ultimately be blamed on the Mark* (because it's the roof of all evil, according to many people, anyway).

So, if I can ask anything of anyone who has bothered to read this far: DON'T CHANGE FSKRNn (it's FSKRN1 in our example, but I think it's actually FSKRN2 in a real Mark* turbine control...). Don't change FSKRN1 or FSKRN2--unless you understand exactly what you are doing and why and what all of the knock-on effects are! (The knock-on effects are too many, and too complicated, to go into here--at ANY time. Ever.) Forget about the Droop setpoint/regulation. It should almost ALWAYS be 4% by convention around the world for just about EVERY heavy duty gas turbine made by any manufacturer. Knowing the value for your machine, or any machine, isn't important--and it WILL NOT help you or anyone understand Droop Speed Control. Not now; not ever. And, by explaining it as I have, I have probably caused at least one person somewhere in the world to think they can now change FSKRNn and fix some <b>PERCEIVED</b> issue with the turbine, or somehow increased performance or efficiency or power output (which changing the FSKRNn will NOT do!). There are extremely few real issues, if any, that changing the Droop setpoint/regulation will solve. Regardless of what anyone says or writes--it's just not the solution to any problem (unless there is something extremely unusual about the power plant or its proximity to other power plants or some extremely unusual fuel characteristic (which almost never happens, either). Just don't do it.

Droop Speed Control is about how much the power output of the unit will change for a given change in the speed reference. In other words, for a given change in the speed reference, the speed error will change. And that change will cause a change in the energy flow-rate into the generator prime mover. That will cause the torque being produced by the prime mover and transmitted to the generator rotor to change. And because the generator is LOCKED into synchronous speed because it is synchronized to a grid and can't change it's speed the extra torque is converted to amperes in the generator stator--which increases the electrical power being produced by the generator.

In our example (a machine with 4% Droop setpoint/regulation), every 1% change in the speed error will cause a 25% change in the load being produced by the unit (turbine and generator). Since 4% represents 100 of rated load, a 1% change in speed error represents a 25% change in load (1/4=0.25, or 25%). That's what Droop Speed Control does--it adjusts the output of the machine be a predetermined amount for any change in the speed error. And, the speed error is the difference between the speed reference (TNR in our example) and the actual speed (TNH in our example). So a change in TNR <b>or</b> TNH can cause a change in the speed error. (Usually, it's a change in the speed reference--because, usually, the actual speed is constant and stable if the grid the unit is synchronized to is constant and stable.)

Look at the tables in the reply above, especially the upper table. If the operator is told by the supervisor to produce 20 MW (and the machine is rated for 40 MW) the operator will click on RAISE- or LOWER SPEED/LOAD until the MW meter reads 20 MW. 20 MW represents 50% of rated load. For a machine with 4% Droop setpoint/regulation (as in our example), if you or the operator or his supervisor looked at the value of TNR in the Mark* when the unit stabilized at 20 MW it would be pretty close to 102%. That 2% comes from the fact that the unit would be at 40 MW when the speed reference was at 104% (which is rated speed plus the Droop setpoint/regulation), and when added to 100% (rated speed) the FSR (FSRN when Droop Speed Control is active). Clicking on RAISE SPEED/LOAD increases TNR; clicking on LOWER SPEED/LOAD decreases TNR. Changing TNR changes the speed error (the difference between TNR and TNH). Changing the speed error causes the FSR to change (FSRN actually changes, which feeds into the MIN SEL block and becomes FSR).

Droop Speed Control is about how much the load of the unit will change for a change in the speed error. The speed error changes when EITHER the speed reference changes OR the actual speed changes. And, for a machine with 4% Droop setpoint/regulation, every 1% change in the speed error results in a 25% change in the load. For a machine with 5% Droop setpoint/regulation, every 1% change in the speed error will result in a 20% change in the load. That's what Droop Speed Control does--and how it works. Full stop. Period. Nothing more and nothing less.

Now, why do they use speed--and not load? Because in the very early days of prime mover governors it was impossible to get load (Watts, kW, MW) into the governor. Governors were completely mechanical devices, and they only responded to changes in speed. And for AC power generation we know that speed is VERY important--it is directly related to frequency--and frequency is very important! So, all the control parameters were related to speed--because that's the only variable early governors "understood." (Look up "flyball governor" on the World Wide Web.) And, for many decades speed was the only variable governors understood--because governors were basically mechanical in design and construction. It wasn't until the advent of analog electrical systems that it became possible to integrate the value of the load of a prime mover and governor into the governor. BUT, new governors still had to interact with the existing mechanical governors as they were used on new prime movers and generators. And, speed was--and is--STILL very important for any AC power generator prime mover. Even if the speed of a generator can't be changed while it is synchronized to a grid, the designers of governors know that and understand that and they actually use that fact in the Droop Speed Control equation.

>2. FSR definitions - it was given as the full speed no load
>value of the fuel, typically 20 % FSR - I don't really
>understand this statement?

I had hoped we had covered this in prior responses--FSR is the reference position for the fuel control valve. (Again, 'Stroke' is GE-speak for position, or travel--whatever you want to call valve position or travel or ... ) There are FSRs for FSNL, for 13% of rated load, or for 39% of rated load, or for 81.2% of rated load--they are all determined from the Droop Speed Control equation. And, yes--the Droop Speed Control equation uses an FSR (FSNL FSR--20% in our example) as part of the equation. That is the position of the fuel control valve that keeps the machine running at synchronous speed (100% of rated speed)--and it always has to be flowing when the unit is running and synchronized to a grid and producing power. It can't be avoided, or subtracted--without affecting the load being produced by the unit (the prime mover and generator). The servo-valve output regulator adjusts the amount of servo current being applied to make the fuel control valve's actual position (stroke; travel) to be equal to the reference position (the FSR).

FSR is not just the amount of fuel required to achieve FSNL. It's the amount of fuel required to achieve any speed or load of the unit. It just so happens that the amount of fuel required to maintain synchronous speed (the FSR at FSNL) is part of the GE "pure" Droop Speed Control equation.

Look at the tables, gustavo_marcelo. If I had put the CONDITION column at the left side of the tales, would that have been more helpful--next to the FSRN column? FSRN changes for every condition--not just FSNL, but for every load change (the upper table) or frequency change (the lower table).

>2. What does the unit for FSRN, TNRL,TNH, FSR?

>FSRN (% FSR), TNRL (% speed ref), TNH (% actual speed), FSR
>(% FSR)

You have two Questions Number 2....

>3. How to prove this equation with some numerical value been
>given to the equation?

>Understood.

Good.

>4. Does the equation above is also load control equation?
>Speed droop also means Droop Load control, is it?

>Yes, it is.

Again--good.

>5. How does the iso speed control equation looks like?

>With 100% reference or 100% TNR.

Yes; very good!

This isn't easy, but it's really not complicated, either. You're not going to pick it up from one reading of the threads you have started. It's the RARE individual who understands Droop Speed Control (any version of it!) in one or two readings. There are textbooks and references and governor manuals which DO NOT properly and accurately describe Droop Speed Control--because the all say that the speed of the unit will change when the load changes. And, for machines synchronized to grids (almost all of them!) the speed doesn't change when the load changes. And the descriptions by those educated authors fail to mention that little detail (actually, it's NOT very little--it's <b>HUGE!</b>) So, don't expect to become any kind of expert (newbie or experienced) from a few days of trading posts on World Wide Web forums. It ain't going to happen. Slow down. Think it over. And, I (too often) let typographical errors and mistakes get through to the responses I post--sorry about that. It happens. I can't afford a proof-reader (I can't afford very much since I get paid the same amount you paid for posting your question(s) to the forum(s): Zero).

So, just slow down. Re-read the posts over time (a few days; a few weeks). And, slowly things will become more and more clear. "Learning is finding out what you already knew." (Richard Bach, '<i>Illusions</i>') Think about that for a minute. When someone explains something to you, or you come to an understanding of something for yourself, doesn't it seem like, "YEAH! Exactly--that makes perfect sense! I knew that--I just hadn't thought of it in those terms or like that before!" Now, I'm <b>NOT</b> saying I have explained things so you have completely understood them--yet. I'm just hoping that I've added to your critical thinking and consideration of this, and that in time, you will come to think or say, "YEAH! Exactly!" But, it ain't going to happen in a few days, my friend.

That's why I also add some history of Droop Speed Control--and why the control system designers use speed in the control of the prime movers. It all goes back to what was available in the beginning of AC power systems--and speed was it. That's all there was. And, it turns out: It was MORE than enough!!! It was all that was (and still is!) necessary, really. Because speed and load are related, too!

Droop Speed Control is about how much the load of a unit will change for a change in the speed error. Full Stop. Period. Nothing more, and nothing less. (Well, actually, it's a lot more--and that's the part that confuses people.) Stick to the basics, and you will come to know Droop Speed Control--and even be able to explain it to others so they can understand it.

And, don't ANYONE go changing FSKRN1 or FSKRN2--unless you fully understand why and what the knock-on effects will be. Because there WILL BE knock-on effects. And it should NEVER be necessary to change the "Droop"--and if you do change it without informing the grid operators you may even be held civilly or criminally liable for fines or worse. Grid operators rely on machines having the Droop setpoint/regulation they expect--and they get pretty angry when they trace grid frequency problems to improper Droop setpoint/regulation. JUST DON'T DO IT. Now now. Not ever. NEVER.

Finally (and I do mean finally), I think one of the things that confuses many people is that Droop setpoint/regulation is a number like 4 or 5, when were talking about % of Rated Speed. Well, I can't exactly explain that; it's really more of a gain, as we have discussed before. Because, if the unit was synchronized to a grid and operating at a TNR of 102% and someone opened the generator breaker and the FSR (45% in our example) didn't change--the unit would trip on overspeed. Without the grid to hold the unit speed to rated (100%) 45% FSR would be so much fuel the speed would exceed 110% of rated, and the unit would trip on overspeed. Same for 75% of rated load (103% TNR) and Base Load (104% TNR). The value of Droop setpoint/regulation is really unit-less (it's a ratio)--but that confuses a lot of people even more... So, just remember: the Droop setpoint/regulation is just about how much the energy flow-rate into the generator prime mover will change for a change in the speed error. That's it! That's all it is. It's just a number that is used to control the energy flow-rate into the machine--and to give designers a means to predict and/or calculate how much the load will change for a change in a parameter. And that parameter is speed--again, because in the beginning there was only speed in governors. And, that' all that was--and still is--necessary.

Now, back to my Sunday.

Buffalo nickel

#### Jolek

Hello,

How Droop Control handles Summer/Winter output margin? For instance, we have about 20 MW difference between hot and cold seasons. That means Base Load may change a lot and there are the same 4% for Droop SP Regulation.

#### CSA

Hi, Jolek,

Thanks for the question; I hope I understand it correctly.

What it means is that when the axial compressor inlet temperature is at the value on the nameplate of the gas turbine, and the inlet air filters are clean, and the axial compressor and IGVs are clean, and the unit has pretty new hot gas path hardware (combustion liners, combustion transition pieces, turbine nozzles and buckets) that TNR at Base Load (presuming Base Load is enabled and active!) will be approximately 104% (and also presuming the unit does not have Peak Load) and that the electrical power output will be approximately equal to the gas turbine nameplate rating.

When the ambient is colder than nameplate rated, the electrical output will be higher than gas turbine nameplate rating (when the unit is at Base Load!). And, TNR will be slightly higher than 104% (the amount depends on how much colder the axial compressor inlet temperature is than the gas turbine nameplate).

When the ambient is hotter than nameplate rate, the electrical output will be lower than gas turbine nameplate rating (when the unit is at Base Load!). And, TNR will be slightly lower than 104% (the amount depends on how much warmer the axial compressor inlet temperature is than the gas turbine nameplate).

The "transition" from Droop Speed Control to Exhaust Temperature Control occurs when FSRN goes above FSRT AND the IGVs are at their maximum operating angle. If that occurs before or after TNR reaches 104% (for machines with 4% Droop) doesn't matter. (I'm presuming the Process Alarm 'Back-up Exhaust Temperature Control Active' is not being annunciated!) A machine with dirty inlet air filters and dirty IGVs and axial compressor and "old" hot gas path hardware will also reach Base Load before TNR reaches 104% when the axial compressor inlet temperature is less than gas turbine nameplate. This isn't caused by Droop Speed Control--it's caused by the air flow through the machine not being at rated, and by the internal clearances of the turbine section not being at rated. High exhaust duct back-pressure can also cause performance problems (causing Base Load to occur before TNR reaches 104% when the axial compressor inlet temperature is less than gas turbine nameplate).

Again--this isn't a Droop Speed Control Problem. It's purely an issue with getting enough air into and through the machine with no "leaks" in the turbine section (caused by higher-than-specification internal clearances, which is what happens as turbine nozzles and buckets wear, and shroud blocks wear, and seals wear). Droop Speed Control is still doing it's thing--it's just that the exhaust temperature for the given axial compressor discharge pressure is not at rated for machine and/or ambient conditions. Droop Speed Control is working just fine.

In other words, the gas turbine can only make rated power when it's in a new and clean condition, and the axial compressor inlet temperature is at or very near the rated temperature on the gas turbine nameplate, and the fuel being burned is at or near the specification used to build the machine, and the exhaust duct back-pressure is at or below rated. And the rating for a machine with 4% Droop occurs when all of these things are met. When any of them are not, then the unit will not be at 104% TNR, and that's not the fault of Droop Speed Control--it's the conditions are either preventing the unit from making rated power when the unit is at 104% TNR, or the conditions will allow the unit to make more than rated power and TNR will be more than 104%.

Hope this helps!

#### CSA

guatavo_marcelo,

"Pure" Droop Speed Control uses Proportional-only action (the P in a PID control function) to control the amount of fuel flowing to the combustors of a gas turbine driving a generator synchronized to a grid with other generators and their prime movers. When the difference (the error) between the reference and the actual values of the control function changes the amount of the control output of the PID control function changes. BUT, with a Proportional-only control function, there is nothing to drive the error between the reference and the actual to zero.

Here's an attempt at graphically describing Proportional-only control:<pre>
|
TNR |
|
101.10% | _________
| |
101.00% |___|
|
-------------->
time

|
FSRN |
|
33.75% | _________
| |
| |
| |
| |
32.50% |___|
|
-------------->
time

(DWATT) |
|
11.0MW | _______
| /
| /
10.0MW |___/ |
-------------->
time</pre>The three "graphs" show that for a 0.1% change in TNR, FSRN will change by 1.25%. (This follows our example of 4% Droop equals a 50% change in FSR; every 1% change in TNR will result in a 12.50% change in FSRN. So, a 0.1% change in TNR will result in a 1.25% change in FSR.) The actual turbine speed, TNH, WILL NOT CHANGE (because the unit is <b>synchronized</b> to a grid--which means the speed (frequency) is stable and constant (at 100% and not changing). There's a short time lag between when the fuel flow-rate increases, the fuel is combusted (burned) and the hotter combustion gases flow through the turbine section to produce more torque which the generator converts to stator amperes which makes the load (signal name DWATT; units MW) increase. In the example, the unit is rated at 40.0 MW--so every 1% increase in TNR will result in a 10.0 MW increase; a 0.1% increase in TNR will result in a 1.0 MW increase.

A PID controller looks at a reference (sometimes called the "Process Setpoint") and uses feedback (the actual value, sometimes called the "Process Variable") and changes the Control Output to affect a change in the actual value if necessary. In a Proportional-only controller, there is nothing to drive the difference (the error) between the reference value and the actual value to zero--to make the actual value equal to the reference value. The controller output (FSRN in our example) is a function of the difference between the reference value and the actual value (the feedback). (One doesn't even need to use a PID controller if one is using Proportional-only control, which is how GE accomplishes "pure" Droop Speed Control--without a PID controller.) A simple subtraction and multiplication and addition is all that's necessary, and that's all they use.

Constant Settable Droop Speed/Load Control uses Proportional-plus-Integral (P+I, or PI) control. AND, GE use Transfer Function blocks in Mark VI and Mark VIe application code to accomplish the Proportional-plus-Integral control. Transfer Functions are way beyond my maths abilities, and the Item Help provided in Toolbox and ToolboxST for Transfer Function blocks is completely unintelligible for anyone but a maths expert or a control design expert. The Transfer Function block GE uses can also perform many different functions in addition to serving as a PID controller depending on how they are configured (what values (parameters) are passed to (supplied to) the Transfer Function). And I don't believe the exact step-by-step process for changing load is necessary to understand what happens when load is changed.

Here's my attempt at graphically describing the full formula for Constant Settable Droop Speed/Load Control in a Mark* turbine control system:<pre>Simplified Constant Settable Droop Speed/Load Calculation

(DWATT*DWKDG)=DWDROOP--------
\___________/ |
| |
| |
(Transfer Function) |
n.0 Second Time Constant +|- + +
TNR-----O----TNRL---------O-----(*FSKNG)-----O-----FSRN
-| +|
| |
| |
TNH------------------------ |
|
FSR--------
\___________/
|
|
Proportional plus Integral
Control???
(Transfer Function)
n.0 Second Time Constant

DWKDG Calculation:
(Droop Setpoint)/(Rated Load)=DWKDG 4.00/40.0=0.100%/MW</pre>The table below describes the <b>results</b> of changes in TNR for a unit using Constant Settable Droop Speed/Load Control--after everything has stabilized. We're using the same unit: FSNL=20% FSR; FSNL-Base Load FSR Change: 50%; Droop Regulation is 4%; and Base Load is 40 MW.<pre>
Assumptions:
Unit Rated Power Output: 40 MW
FSNL FSR: 20%
DWKDG: 0.100%/MW

FSRN TNR DWATT DWKDG DWDROOP TNRL SPEED ERROR TNH
20.0% 100.0% 0.0MW 0.100%/MW 0.00% 100.0% 0.000% 100.0% FSNL
20.3%* 100.3% 0.0MW 0.100%/MW 0.00% 100.3% 0.300% 100.3% Synchronizing
20.3%* 100.3% 3.0MW 0.100%/MW 0.30% 100.0% 0.000% 100.0% Breaker Closure
32.5%* 101.0% 10.0MW 0.100%/MW 1.00% 100.0% 0.000% 100.0% 25% Load
45.0%* 102.0% 20.0MW 0.100%/MW 2.00% 100.0% 0.000% 100.0% 50% Load
57.5%* 103.0% 30.0MW 0.100%/MW 3.00% 100.0% 0.000% 100.0% 75% Load
70.0%* 104.0% 40.0MW 0.100%/MW 4.00% 100.0% 0.000% 100.0% 100% Load
*These FSRN Values are assumptions based on experience</pre>
Here's an attempt at graphically describing Proportional-plus-Integral control for Constant Settable Droop Speed/Load Control for a step-change in TNR:<pre>
|
TNR |
|
101.10% | __________
| |
101.00% |___|
|
-------------->
time

|
TNRL |
|
100.10% |
| |\
100.00% |___| \_______
|
-------------->
time

|
FSRN |
|
33.75% | _______
| /
| /
| |
| |
32.50% |___|
|
-------------->
time

(DWATT) |
|
11.0MW | _______
| /
| /
10.0MW |___/
-------------->
time</pre>I believe what happens is when the initial change occurs in TNR the Proportional portion of the controller "jumps" FSRN (a step change in FSRN)--but not immediately to the value required to make the load to increase to the desired value. The Integral portion of the controller takes care of the last little bit of increase in FSRN to achieve the desired load increase. That's what I've attempted to show in the graph of FSRN versus time above--the vertical increase in FSRN represents the Proportional change in FSRN and ths "slanted" portion of the graph represents the Integral portion of the change.

As load (DWATT) increases, the value of DWDROOP is going to increase. That is going to cause the value of TNRL to decrease. As can be seen from the table, when a load change is complete, TNRL returns to 100%. BUT, even though the error between the load-biased reference (TNRL) and the actual speed (TNH) is driven to zero, the controller output (FSRN) doesn't decrease to zero--and that is in some way related to the result of Proportional-plus-Integral control.

When grid frequency changes while TNR is stable, TNH will change, and we know that will cause a change in FSRN also--just like in "pure" Droop Speed Control.

I'm constrained by the keyboard characters when trying to "draw" graphs and charts since we can't post images to control.com, but I think the concepts are well-represented above. Again, I can't fully and exactly describe what's happening during each scan of the application code (40 msec time slices), but I can very accurately describe the results of a load change when FSRN stabilizes (in the table). Under normal operating conditions when the grid frequency is at rated and when the fuel make-up is stable and the unit is not transitioning combustion modes the table above accurately describes what happens (the end result of a load change when fuel and load stabilize after a load change).

Proportional-plus-Integral control action is what causes FSRN to change as TNRL changes, even though eventually the error between TNRL and TNH is driven to zero under normal operation. I just can't describe precisely what happens, and exactly how the Mark* does it using Transfer Functions. (To complicate matters, I believe the drawing of Constant Settable Droop in the GE Control Specification description of Constant Settable Droop Speed/Load Control is incorrect. (Yet another inaccuracy.)) I'm not entirely sure my diagram above is 100% correct, but it's better than the one in the Control Spec--and more complete than the formula in your original post of this thread.

So, I think we've accomplished our stated objective--for BOTH "pure" Droop Speed Control (which is effectively load control), and Constant Settable Droop Speed/Load Control (which includes a load biased turbine speed reference signal in the loop). You wanted units (Engineering units) for values used in the partial formula you originally posted; we completed that formula and provided Engineering units for all the values. We also provided values for various conditions (loads) as examples of what happens as load is changed, or frequency changes. I want to again STRESS how very, Very, VERY STRONGLY it is recommended NEVER to change any Control Constant associated with Droop setpoint/regulation without fully understanding all of the effects of the change AND the need to notify the grid operators/regulators the unit is synchronized to of the change.

Thank you, gustavo_marcelo, for the question and the learning I had to do to be able to answer your question.

#### gustavo_marcelo

I should be thanking you for all the details .But sometime I am no sure whether to ask or not.

Now I can see Constant Settable means "continuously and automatically been changed" based on loading. At first I thought the droop is fix then what difference it makes with the pure droop then.

With regards to PSL, I believe it also have a feedback loop from Watt, just that the mechanism on how it able to hold the watt at constant value even if the speed decrease or increase is unclear to me.

#### gustavo_marcelo

I should be thanking you for all the details. But sometime I am no sure whether to ask or not.

Now I can see Constant Settable means "continuously and automatically been changed" based on loading. At first I thought the droop is fix then what difference it makes with the pure droop then.

With regards to PSL, I believe it also have a feedback loop from Watt. just that the mechanism on how it able to hold the watt at constant value even if the speed decrease or increase is unclear to me.
Somehow the control has the ability to maintain the fuel constant hence the output regardless what is happening at the utility side.

#### CSA

gustavo_marcelo,
<pre> TNH------------------------------------------
|
+ + -| +
L90PSEL-------o--------TNR-------o-----TNRL-----0-----X-----O------>FSRN
-| -| + | +|
| | | |
DWATT-------- | FSKNG-------- |
| | |
| | |
DWKDG-------X------DWDROOP------ FSR---------</pre>The above is a highly simplified diagram of how Pre-Selected Load Control is an outer loop to Constant Settable Droop Speed/Load Control. When Pre-Selected Load Control is active, it compares the actual load (DWATT) to the Pre-Selected Load Control reference (L90PSEL--one of the FEW analog signals which begins with an 'L'! (and NO analog signal name should begin with an 'L'!)) and then adjusts TNR based on the error between the actual and the reference. TNR, in Constant Settable Droop Speed/Load Control is then modified by a load bias (DWDROOP) to derive TNRL, which is compared to TNH and multiplied by a gain (FSKNG--which is NOT the Droop regulation/setpoint) and added to FSR to derive FSRN: Speed Control FSR. Which is then fed to the MIN SEL function to choose FSR which is fed to the fuel control valve signal.

So, when grid frequency goes down (in this example) the immediate effect on the above loops is for TNH to decrease. (Remember ALL generators will experience the same drop in frequency--they are <b>synchronized</b> together on the same grid--so the speeds of all of the generators will also decrease!) So, the Constant Settable Droop part of the loop will sense that the load-biased speed error (the difference between TNRL and TNH will increase and that will cause FSRN to increase which will cause DWATT to increase.

HOWEVER, the Pre-Selected Load Control portion of the loops above will see that DWATT increased above L90PSEL--and will drive TNR down, which will cause TNRL tend to go down--but the DWROOP function will cause TNRL to increase. It's all a big mess--with the undesirable result that DWATT will oscillate, TNR will oscillate, FSRN will oscillate, and the grid frequency will no stabilize. Imagine this happening on MANY machines--not pretty; NOT pretty at all!

You can, gustavo_marcelo, make your own table of signal values--using the ones above as examples. It will become very obvious, very quickly, what happens when Pre-Selected Load Control is active when the grid frequency deviates from rated.

Next, you're going to ask how Primary Frequency Response works. And so to head off that question, when enabled Primary Frequency response will sense the change in TNH (because of the change in grid frequency) and will basically "suspend" Pre-Selected Load Control until the grid frequency returns to near normal at which time Primary Frequency Response will then allow Pre-Selected Load Control to become active again.

Pre-Selected Load Control is an outer loop to Droop Speed Control (either "pure" Droop Speed Control, or Constant Settable Droop Speed/Load Control).

And that feeds into Droop Speed Control (when not at Base Load on exhaust temperature control).

Hope THIS helps!!!

#### CSA

gustavo_marcelo,

I really think that what I haven't been saying--but have been expecting people to just intuitively grasp and understand--is that the gas turbine is loaded and unloaded (at all times) by increasing or decreasing TNR, an analog value that is usually between 100% and 104% when the unit is synchronized to a grid and producing electrical power.

Whether the load is being manually raised or lowered or monitored to maintain a constant value, or whether Pre-Selected Load Control or some other automatic load control is being used, TNR is always being used to cause a change in the fuel flow-rate if necessary. The "pure" Droop Speed Control formula is very clear about that:

FSRN=((TNR-TNH)*FSKRN1)+FSKRN2

There's no other way to cause fuel flow to change. Fuel flow-rate in GE Mark* heavy duty gas turbine control systems is proportional to (related) to fuel control valve opening. And the two are linear; as one FSR changes fuel flow-rate changes linearly.

Let's say that at 20% FSR (FSNL; 100% or rated speed and zero MW), the fuel flow-rate was 1.428 #/sec, and that at 70% FSR (100% of rated speed, and Base Load (rated load)). That would mean that it required 1.428 #/sec of fuel (which occurs at 20% of fuel control valve opening) to maintain rated speed and 0 MW, and that to reach Base Load (100% of rated speed and 100% of rated load) it would take an additional 3.572 #/sec of fuel flow. Now, if we wanted the unit to have 4% Droop, we would divide 3.572 #/sec by 4 to arrive at 0.893 #/Sec of fuel flow per percent of Droop.

At TNR of 101%, the fuel flow-rate would be (1.428+0.893) 2.321 #/sec, and that happens at a fuel control valve opening of 32.5%. At TNR of 102%, the fuel flow-rate would be (1.428+(2*0.893)) 3.214 #/sec of fuel flow, at a fuel control valve opening of 45%. At a TNR of 103%, the fuel flow-rate would be (1.428+(3*0.893)) 4.281 #/sec, at a fuel control valve opening of 57.5%. And, at a TNR of 104%, the fuel flow-rate would be (1.428+(4*0.893)) 5.000 #/sec, and a control valve opening of 70%.

What this "says" is that for every 1% change in the speed error, the fuel flow-rate is going to increase by 0.893 #/sec above the FSNL fuel flow-rate of 1.428 #/sec.

Said another way, for every 1% change in the speed error, the fuel flow control valve opening is going to increase by 12.5% above the FSNL fuel control valve opening of 20%.

To change the speed error, <i><b>one has to change TNR.</b></i> And, one can do that by clicking on RAISE- or LOWER SPEED/LOAD, OR by using Pre-Selected Load Control, or External Load Control, or Remote Load Control, or AGC, or whatever scheme. They all have to modify (increase or decrease) TNR to change the fuel flow-rate to change the load. There's no other method in the Mark* for changing fuel flow-rate and load. Just TNR.

Anything that modifies TNR is going to cause a change in the speed error, which is going to cause a change in fuel flow-rate, which is going to cause a change in load.

Anytime the load is stable when the unit is synchronized to a grid, TNR is also stable. If TNR changes, the fuel flow-rate will change, and the load will change. So, to make the load stable TNR has to be stable. (I'm presuming the grid frequency is stable.) Pre-Selected Load Control, when it's trying to maintain a desired load, is trying to make TNR stable. But, often it's not tuned very well, and causes small oscillations in load and fuel flow-rate. (If the RAISE- and LOWER SPEED/LOAD buttons are toggling from green to orange and back to green and then orange and then green, that's the result of a poorly tuned Pre-Selected Load Control function that is constantly adjusting TNR to try to return DWATT to the desired value. Some places, the adjustments are not very often; others, it's continuous, and that's not really good for the turbine or auxiliaries (the fuel control valve actuators and seals in particular).)

But, ALL load changes are done by adjusting TNR. That's the only method available. If that wasn't clear, I apologize. And, when it's desired for the load to be stable then TNR has to be stable.

TNR is a number that is incremented or decremented. It is stepped up or stepped down. It is ramped up or ramped down. And, it can be anything between 100% and 104% (sometimes even higher than 104% when it's a really cold day, or the axial compressor inlet air is cooled, and the machine is in a new and clean condition).

Constant Settable Droop is also a function of TNR--it's just that TNR is "biased" or modified in order to have a tighter, smoother control over load using the load signal. But, even if the unit uses Constant Settable Droop, if Pre-Selected Load Control is active then TNR is being controlled (stepped up; stepped down; held constant) by Pre-Selected Load Control. And that feeds into the Constant Settable Droop function to produce FSRN--just like "pure" Droop Speed Control. The both "start" with TNR. And changing TNR changes the speed error, which changes the fuel flow which changes the load.

I certainly hope this helps clear up any confusion about how fuel and load is controlled for a GE-design heavy duty gas turbine with a Mark* turbine control system. Load--either manually or automatically--between zero load and full load is controlled by changing TNR which causes the speed error to change which causes the fuel flow-rate to change. You can use fuel flow-rate or valve opening (valve stroke; valve position--whatever you want to use or call it). Load is proportional to fuel flow-rate.

Look at the tables provided, and look at what happens when TNR changes (from one row to the next). Droop is just a number that someone long ago decided was a good way to base how much to change the energy flow-rate into a prime mover by or with. Droop defines how much the energy flow-rate into a prime mover will change for a change in the speed error. That's it. Nothing more mysterious or magical than that. It's all about the speed error.

#### gustavo_marcelo

Thanks for the wonderful explaination CSA