Gas Turbine Free Governor Mode of Operation

N

Thread Starter

niazbibiyana

Hi,

I want to know about difference between droop mode and free governing mode of operation in Gas Turbine. Are they Same or different thing? I am quite clear about turbine droop mode. My curiosity is about FGMO control. Please give an clear idea on this.
 
niazbibiyana,

Where did you see this "free governor mode" reference?

I used my preferred World Wide Web search engine and found this interesting result:

http://freegovernormode.blogspot.com/

I did some other reading of other search results and I believe the intent of this new power generation term is to say:<pre>
DON'T USE PRE-SELECTED LOAD CONTROL!!!</pre>
In other words, let Droop Speed Control do it's thing, and make the operators use the RAISE- and LOWER SPEED LOAD targets when operating at less than Base Load.

Free governor, as I have read, means to remove all external influences which would prevent the governor (Droop Speed Control) from responding in a normal fashion to grid frequency excursions without being over-ridden by any other influence.

Operators and operations management and ownership is going to have to learn how to let the governor do its thing, without using features such as Pre-Selected Load Control to do normal Part Load operation so that units are free (the governors, that is) to respond to grid frequency disturbances to support grid stability.

Pre-Selected Load Control forces the governor to respond exactly opposite to how it should be responding--and that <b>contributes</b> to grid instability, and because when it's responding exactly opposite of how it should be responding it's doing so by "hunting" it's even aggravating grid instability even more.

It's really impossible to say how bad Pre-Selected Load Control is for operating GE-design heavy duty gas turbines, especially during times of grid instability. And operations management and plant management and ownership also incorrectly believe that the output(s) of their unit(s) should be stable during grid frequency excursion--which is completely opposite of how Droop Speed Control works, and needs to work to support grid stability.

I may be wrong--I have been before--but as I am coming to know that in many parts of the world grid regulatory bodies and even governments are "banning" the use of features like Pre-Selected Load Control and even VAr/Power Factor control. Because both over-ride the natural governor tendencies which serve to support grid stability during times of instability. Many of these regulator agencies have only begun to understand how features like Pre-Selected Load Control have and do contribute to grid instability, and as such, are beginning to forbid their use.

This particular OEM offers a paid option, Primary Frequency Response, or something similar, which allows grid frequency disturbances to over-ride Pre-Selected Load Control. But, they're making owners and operators pay for modifying the governors to implement this vital and critical feature.

Again, I may be wrong--I've been wrong before, and I'll be wrong in the future--but from my research this is my interpretation of this new-fangled power generation term (at least it's new in my lexicon/vocabulary and experience).

If you learn something different in your World Wide Web search, please write back to let us know.
 
niazbibiyana,

So, to clarify, when using Pre-Selected Load Control on a GE-design heavy duty gas turbine with a Speedtronic turbine control system the load reference is used to modify TNR. And, when the actual load equals the load reference TNR stops changing. When the operator changes the Pre-Selected Load Control setpoint (reference), then TNR changes until the actual load equals the load reference. And that's all great--as long as the grid frequency is stable.

Since you understand Droop Speed Control, you know that the amount of fuel going into the machine is controlled by the error between TNR and TNH (the actual speed--which is a function of grid frequency when synchronized to a large or infinite grid). When the grid frequency changes TNH changes--which changes the error between TNR and TNH, which changes the fuel flow-rate, which changes the load. BUT, Pre-selected Load Control says, "No, NO, <b>NO</b>!!! The load must remain equal to the Pre-Selected Load Control setpoint (reference)!!!" and changes TNR to change the load, which changes the error between TNR and TNH, which then changes the load again, and so on, and so on, and so on. So, the two (load and speed) start "fighting" each other and driving the fuel flow-rate, and load, crazy.

Which is <b>NOT</b> good for the grid. In fact, the Speedtronic is doing exactly the opposite of what it should be doing to support grid frequency. If the grid frequency decreases, Droop Speed Control will increase the fuel flow-rate--and load--to try to help support grid stability. BUT, Pre-Selected Load Control will see the increase in load and will do the exact opposite--because it's trying maintain load--by reducing TNR to maintain the load.

NOW, just imagine what is happening in the Speedtronic when the grid frequency is very unstable, and is swinging up and down, sometimes pretty violently. And, the Speedtronic is trying to keep up with that--and is also doing the exact opposite of what it should be doing. THAT'S how it contributes to grid instability, instead of maintaining grid stability.

Yes, if on "free governor mode" the unit output will vary as frequency varies--<b>it's supposed to vary to help maintain grid frequency.</b> (And people think their unit's frequency/speed/load should be stable during grid frequency excursions (instability)--but it can't, and it SHOULDN'T. This is a fundamental lack of understanding of how synchronous generators and AC power system work to think one or two or three synchronous generators can be stable when grid frequency is unstable. Very serious lack of understanding of basic AC power system fundamentals.) And, if the Speedtronic is being unstable in exactly the opposite way of how it should be "unstable," well, then, it's part of the problem instead of part of the solution.

Pre-Selected Load Control--particularly in parts of the world where grid frequency can often be unstable--is an extremely poor way of operating a GE-design heavy duty gas turbine. EXTREMELY POOR. And, grid regulators and governments are beginning to understand the problem. And trying to prevent it.

By the way, for machines which might use an analog signal (or even digital signal) as a load reference to a Speedtronic turbine control system, sometimes called Remote Load Control, or External Load Control, the effect on TNR is exactly the same as Pre-Selected Load Control--and during grid frequency disturbances. AGC (Automatic Governor Control), if properly implemented, changes TNR while monitoring load, but is NOT the same as Pre-Selected Load Control. (A proper implementation of AGC uses discrete inputs to raise or lower TNR, either via contacts or via a digital method (TCP/IP or similar).

So, operations managers and operators had better get used to not using Pre-Selected Load Control, until the OEM finds a better way of controlling load using a load setpoint--which allowing for proper response to grid excursions. (Again, Primary Frequency Response (another poor choice of function name by the OEM) does this. And, most regulatory bodies refer to this as Secondary Frequency Response.... Ain't this fun?!?!?)

If you learn something different about free governor mode of operation, please let us know. Even if this isn't exactly what free governor mode of operation is or is intended to do--operating GE-design heavy duty gas turbines with Pre-Selected Load Control is not a great operating mode. Especially for regions where the grid frequency is often unstable and really NEEDS large, stable machines to help support grid frequency. Not large machines operating unstably when grid frequency is unstable....

This isn't really a control system problem--it's an operations problem. Operators have to be told to use the RAISE- and LOWER SPEED/LOAD buttons (targets) to change load while monitoring load, and then WAIT to see what happens. AMAZINGLY ENOUGH, when they stop clicking on RAISE- or LOWER the load will stop changing, and will stay pretty much exactly where it stops. Without much operator intervention, and without Pre-Selected Load Control making (sometimes all-too-frequent) adjustments.

They need to learn to do the same without using VAr/PF control, too. Because that's another, very similar, situation for grid stability (voltage stability, which can contribute or detract from, frequency stability, also). Because, the time is coming when there will be fines, and maybe worse, for sites determined to be using (caught) either Pre-Selected Load Control or VAr/PF control. They're great things for operators, but really bad things for grids--where grid frequency tends to be or is unstable.
 
N

niazbibiyana

CSA

Thanx for your quick reply.:D

I have seen in my site, some people from grid site, they tested our 9FA GT in free governing mode (no preselected, no base, no external load). I noticed, while grid frequency changed by 1% out GT output changed by 10% MW.

Our gas turbine normally runs on "Constant Settable Droop response" mode while it runs on base load condition. For your kind information, Bangladesh grid system is not stable and it is not controlled by AGC mode. :p

Could you please tell here more about "Primary frequency response" ?? What is normal condition to activate this mode? Kindly tell also "Secondary frequency response".
 
niazbibiyana,

So, I can hear the wheels of thought grinding away. "If we can't use Pre-Selected Load Control (or External Load Control, or Remote Load Control), how are we to operate the turbine and change load and maintain load?"

Droop speed control uses the error between the Droop Speed Control speed reference and the unit's actual speed to control the amount of fuel flowing into the unit, which controls the torque produced by the unit, which controls the power produced by the generator. And, when the grid with which the unit is synchronized is stable (which most are most of the time--as they should be) then the unit speed is also stable, AND when the Droop Speed Control speed reference is stable then the error between the reference- and actual speed is stable--which means the fuel flow-rate will be stable, which means the torque being produced will be stable, which means the power produced by the generator will be stable. Even when Pre-Selected Load Control is enabled and active and the Pre-Selected Load Control Setpoint is not changing then the Droop Speed Control speed reference is not changing.

Operators for decades used the RAISE- and LOWER SPEED/LOAD switches (before digital control systems and PC-based HMIs) to change load by moving the switch in the RAISE or LOWER position to change the speed reference. And, it didn't take too much effort for the operators to hold the switch in one position or the other to change load on machines which were only rated for 10- or 20 MW.

Newer machines, with digital control systems and PC-based HMIs, use mouse-clicks on the RAISE- or LOWER SPEED/LOAD targets on the HMI display(s) to change load manually. For large machines (150-250 MW or more) that can take a lot of clicks to change load from no-load to Base Load. And that can result in RSI (Repetitive Stress Injury) for many operators.

Pre-Selected Load Control can be used when changing load (raising or lowering) to enter the desired load setpoint, and then once the unit reaches the desired setpoint and stabilizes at the desired value the operator can click once on RAISE- or LOWER SPEED/LOAD--which cancels/aborts Pre-Selected Load Control. Clicking once on RAISE- or LOWER SPEED/LOAD does not change the Droop Speed Control speed reference by very much but it does stop Pre-Selected Load Control from making any further adjustments to the Droop Speed Control speed reference. Since the actual speed is stable (as long as the grid frequency is stable), and the Droop Speed Control speed control reference is stable--the fuel flow-rate into the unit will be stable, the torque being produced by the turbine will be stable, and the power output produced by the generator will also be stable <b>at the desired load value</b> achieved by using Pre-Selected Load Control.

This is what I believe is being called "free governor mode." In this mode, if the grid frequency changes then the governor (the digital turbine control system) will sense the speed change and will adjust the fuel flow-rate in order to support grid stability. If the grid frequency decreases, the error between the Droop Speed Control speed reference (which is stable) and the actual speed will increase--which will increase the power output of the generator which helps to support grid frequency when the grid frequency decreases. The opposite happens when the grid frequency increases above rated; the error decreases and the power produced by the generator decreases to help support grid stability.

All of this (fuel flow-rate changes in response to grid frequency changes) ONLY happens when the unit is operating on Droop Speed Control. When the unit is operating on Base Load exhaust temperature control, frequency changes will cause the actual speed to change which will affect the power output of the turbine--but only because the air flow through the machine is changing (because the axial compressor speed is changing), not because the fuel flow-rate is changing in response to grid frequency changes. AND, the opposite of what the grid regulators want to happen happens. In other words, when the grid frequency goes below rated when the unit is operating on Base Load exhaust temperature control, the generator power output goes down; when the grid frequency goes above rated the generator power output increases. (This is the dirty little secret of gas turbines--and some OEMs offer special programming to do the opposite to help support grid frequency, which some grid regulators/governments are demanding.)

Pre-Selected Load Control can be used to change load--thereby preventing RSI for the operators. But, once the unit stabilizes at the desired load, the operator needs to cancel Pre-Selected Load Control to return the unit to "free governor mode"--and the way to cancel Pre-Selected Load Control is to click once on either RAISE- or LOWER SPEED/LOAD.

But, Pre-Selected Load Control <b>should NOT</b> be used for long-term unit operation. Droop Speed Control works just fine for that--trust me (though no one ever does--they just can't conceive of the idea that the unit will maintain a load setpoint without Pre-Selected Load Control enabled and active; but prime movers have done this for decades without any form of Pre-Selected Load Control whatsoever). Pre-Selected Load Control can be used to change loads, but it should not be used for maintaining a desired load.

I think I can hear the wheels of thought slowing down and grinding a LOT less--but, Operations Managers and Operators will still have to be convinced. And that may take some serious "persuasion." Free governor mode requirements (when they are understood by operators and their managers and plant management and plant ownership) may just be the persuasion I've been trying to find for decades.

When the unit is on 'BASE LOAD - TEMPERATURE CONTROL' it <b>is NOT</b> on Droop Speed Control (Constant Settable, or otherwise). FSRN (Droop Speed Control FSR) is driven up out of the way of FSRT (Exhaust Temperature Control FSR). This can be seen on the FSR Display, and when Base Load is enabled and active FSRN is maintained approximately 3% higher than FSRT--and FSRT is the lowest value of FSR, and is THE FSR. When the unit is operating on Base Load-Exhaust Temperature Control and the grid frequency changes, then the actions noted above will occur.

The air flow through the axial compressor will change (it will decrease if grid frequency decreases; it will increase if grid frequency increases). And, this will have an effect on exhaust temperature as well as CPD (and CPR), and this will cause the machine's output (the output of the generator) to change--it will decrease if the grid frequency decreases, and it will increase if the grid frequency increases.

Both actions in this case are exactly the opposite of the grid personnel are expecting--but, again, that's how heavy duty gas turbines operate and grid personnel are coming to know and understand this (even if operators and operations management and plant management and -ownership aren't). It's kind of like having Pre-Selected Load Control enabled and active when the unit is operating at Base Load-Exhaust Temperature Control.

Another REALLY POOR use of Pre-Selected Load Control is to input a load setpoint that is higher than the unit can achieve and let the unit go on Exhaust Temperature Control. This <b>IS NOT</b> Base-Exhaust Temperature Control, because there can be interaction between FSRN and FSRT and FSRN can be negatively affected by Pre-Selected Load Control during grid frequency excursions.

I believe that regulatory bodies refer to Secondary Frequency Response as the same thing that GE refers to as Primary Frequency Response.

Again, and finally--Pre-Selected Load Control should only be used for changing load. Once at the desired load setpoint, Pre-Selected Load Control should be disabled. AND, when it's desired for the unit to operate on Base Load-Exhaust Temperature Control, then the operator should select BASE LOAD and let the turbine control system do it's thing, and get Droop Speed Control FSR out of the way.

Hope this helps!
 
Dear CSA,

That was excellent and very comprehensive.

I have had similar discussions with a few operators in the past but it fell on deaf ears because the usage of pre-select load has become virtually a norm in many power stations just like p.f/Var control, and many operators as well as grid personnel are very much unaware of the effects on the grid. B

oth the pre-select and the p.f/Var control has rendered many control room operators very inactive because once a turbine generator is started, synchronized and these pre-selections are made and the load becomes stable at the desired pre-selected values, many operators go virtually into a trance. I have walked into a control room where the alarm "DWATT TOO LOW TO SUPPORT TNR-TNR LOWER" has knocked off pre-select load and the speedtronic has issued a lower command and the load has dropped from 100MW to a very low value with the operator completely unaware.

I am very happy you have stressed on this severally.

I hope it will transcend down very well.

More grease to your elbows!!!!

However, I have worked on a site where the grid frequency is extremely erratic and unstable.You could suddenly have your load swinging between 110MW to as low as 20MW and up, several times in a day.

What effect would this have on a heavy duty gas turbine in a "free governor mode" operation.

Thank you.
 
kwaku,

Thanks very much for the kind words.

Certainly, load swings are not good for the turbine. They result in thermal stresses on the hot has path components, as well as physical stresses on the compressor and load couplings, as well as on the Accessory Gear and components.

But, there's not much can be done--if the unit trips off line because of excessive frequency- or load swings the grid is going to be less stable.

If you're referring to E-class turbines, they are pretty robust machines--especially if they don't have conventional combustors. F- and FA-class turbines really don't like off-frequency operation (especially the compressors!)

(By the way, if you like any post (by any responder) you can use the Thumbs Up to signify your like.)
 
Dear CSA,
Most grateful for the quick response.

>(By the way, if you like any post (by any responder) you can
>use the Thumbs Up to signify your like.)

Very well noted.
 
N

niazbibiyana

CSA
That was very brilliant explanation I ever seen. Really impressive. I am telling 3 scenarios. Could you please give explanation?

Scenario 1 - Few days ago. GT load was preselected at 197 MW. Frequency was 49.8 Hz. Suddenly in jumped at 50.6 Hz at very short time. GT load suddenly reduced to 155 MW. Although TNR was increasing.

Scenario 2 - Again some days ago. GT load was at 200 MW. Frequency was 50.9 Hz. Suddenly it goes down to 49.75 Hz. Again GT load jumped to 220 MW (NOT BASE LOAD). Although TNR was decreasing.

As I have told you, out GT always runs at "Constantly Settable Droop" mode and GT continuous runs at preselected load (Not free governing mode). According to definition of droop, response is ok.

Scenario 3 - During base load (230 MW), I noticed, while grid frequency slowly increased (50.50Hz to 50.55 HZ), GT output also increased slowly (235 MW). While Hz slowly low, output also low.

Here, it seems not droop control. If compressor air flow increase or decrease, FSRT also should follow this. I mean also in base load condition output MW also should act like droop. Right?

But during very fast swing of frequency, GT output acts like droop characteristic. Although I already know how this system work, I could not trace why machine load output like this, during base load.

Waiting for your detailed brilliant explanation.

 
niazbibiyana,

When a GE-design heavy duty gas turbine with a Speedtronic turbine control system is operating anywhere between zero load and Base Load (CPR-biased exhaust temperature control) it is operating on Droop Speed Control ("straight" droop speed control, or Constant Settable Droop). When the unit is operating at Base Load (CPR-biased exhaust temperature control), it is <b>NOT</b> operating on Droop Speed Control (straight or Constant Settable).

As described previously, when the unit is on Base Load the control scheme uses an exhaust temperature limit as it's reference--and dumps in as much fuel as possible to try to keep the actual exhaust temperature equal to the exhaust temperature limit/reference. It no longer looks at Droop Speed Control; Droop Speed Control is biased up out of the way of Temperature Control, the usual difference that is maintained between FSRN (Droop Speed Control FSR) and FSRT (Exhaust Temperature Control FSR) is 3% FSR. As FSRT increases or decreases, the turbine control system raises or lowers FSRN to keep the 3% differential constant.

And, as previously described, the unit behaves differently when operating on Base Load, exhaust temperature control, than when it's operating at Part Load on Droop Speed Control. And the behaviour you reported seems to match that different operating characteristic. When operating on Base Load exhaust temperature control and the grid frequency/unit speed is lower than normal the air flow through the axial compressor is lower than normal which means the exhaust temperature would tend to be higher. But, the turbine control system senses the increase in actual exhaust temperature and reduces the fuel to keep the actual exhaust temperature equal to the exhaust temperature limit/reference. Conversely, when the grid frequency is higher than normal the air flow through the axial compressor is higher than normal and this tends to cool the actual exhaust temperature, but the turbine control system increases the fuel to raise the actual exhaust temperature to keep it equal to the exhaust temperature limit/reference.

So, when operating on Base Load exhaust temperature control and the grid frequency increases and the grid regulators want the gas turbine to reduce load to support grid frequency--the turbine control does exactly the opposite and increases load (because it increases fuel flow to keep the actual exhaust temperature equal to the exhaust temperature limit/reference). And the opposite happens when grid frequency decreases while the unit is operating on Base Load exhaust temperature control--instead of increasing load to support grid frequency the turbine control system reduces load because it has to reduce fuel flow to keep the actual exhaust temperature equal to the exhaust temperature limit/reference. Again, this is the "dirty little secret" of gas turbines--AND it explains why the unit does NOT behave as it does when it is operating on Droop Speed Control.

The OEM does make a <i><b>temporary</i></b> mode of exhaust temperature control that monitors grid frequency and will allow a little "over-firing" when the grid frequency drops from rated and a little "under-firing" when the grid frequency increases from rated. But, this really only applies to grids that are, for the most part, very stable and only occasionally experience grid frequency excursions. From what you've described and what's been published in trade journals and other industry-related papers about the grid frequency in South Asia it doesn't seem like it would be good to implement this control scheme for the units at your site. Again--F- and FA-class machines were originally designed to be Base Load machines, and primarily for stable grids. The world has changes, and most F- and FA-class machines have been installed where they do not run at Base Load for any appreciable period of time, they are started and stopped--sometimes more than once per day, and they have been installed in places where they haven't done much to support grid frequency. Or, perhaps I should say, they haven't been as effective as hoped in supporting grid frequency. And, their off-frequency operating characteristics have been exposed.

So, it's incorrect to say the unit ALWAYS operates on Droop Speed Control. Yes; Droop Speed Control is always being calculated--but when Base Load is active and the Status Field on the Main HMI Display reads 'Exhaust Temperature Control' and the IGVs are fully open (when the grid frequency is at or very close to rated!) then the unit is operating on CPR-biased exhaust temperature control. Droop Speed Control FSR IS being calculated, BUT it is biased to be greater than exhaust temperature control FSR--and since they are inputs to a MINimum SELect block and the lowest input value of FSR becomes FSR to the fuel control valve(s) Droop Speed Control FSR is NOT in control when CPR-biased exhaust temperature control is enabled and active.

That is, unless there is something VERY UNIQUE about the control schemes used in the Speedtronic turbine control system at your site. And, since GE-Belfort is responsible for all Frame 9 GE-design heavy duty gas turbines it is entirely possible that there is more than one thing which is unique about the turbine control schemes at your site. They deviate from standard practice and proven control schemes just about whenever possible--because they can. BUT, what has been written in this thread applies to the overwhelming majority of GE-design heavy duty gas turbines around the world. And has for decades.

Can we drop this subject now?

Please.
 
N

niazbibiyana

CSA,

So kind of you... Yes its totally clear. Actually I have learned many many things from this forum. Especially from you. Because, I dont have other source about GE-9FA GT. Thats why I sometimes bother you !! :p :p
 
Hi

We have added a special mode which is described by GE as a mix of droop and Preselect. We have a button on the graphic that is called Preselect/Droop and this is what operators always select.

The logic basically recalculates the MW's required as a result of 4% droop (100-TNH)/4% x 100MW (base load) and adds it to the operator entered set point and then this value becomes the TNR or Preselect set point that enters the Load control block where the set point and DWATT are compared and the Raise/Lower commands are issued. Its purpose I think is to ensure that the machine WILL NOT go back to a preselect set point (if we only had preselect button) which we do not want should frequency change as this will not give us SOR and TOR.

Have any of you seen this? It was implemented by GE. We always debate the merits of it here on site.

If we turn of Preselect control, and say we at 40MW, should the frequency drop to 49.5hz the load will increase by 1% or 25MW so we now at 65MW. However when the frequency returns to 50Hz the machine will not go back to 40MW . Can send on a picture as you guys might be interested and am particularly interested in what CSA has to say on it.

Many Thanks.
 
fudge,

GE has at least two versions of a mode they affectionately refer to as 'Primary Frequency Response.' It's a "special" mode of Pre-Selected Load Control that senses changes in turbine speed (frequency) and allows the turbine control system to respond in the appropriate manner (increasing load when the frequency decreases, and decreasing load when the frequency increases). You may have one of these versions of this special Pre-Selected Load Control modes.

This is, or was not, typically provided with new units or retrofit (upgrade) turbine control systems; it has usually been a purchased option (getting the Customer to pay for their oversight).

If you can post your picture to a web-hosting site (for example, www.tinypic.com) and then post the URL for the location here on control.com that would be great.
 
Mr CSA,

what is the difference between isochronous mode and this special preselected mode if both sense grid frequency fluctuaion and change the load setpoint upon that?
 
Isulamu,

Primary Frequency Response can only be used when the unit is at Part Load in Droop Speed Control. (Pre-Selected Load Control mode is used when the unit is in Droop Speed Control mode, though some sites use it to stay in Base Load--which is another poor usage of Pre-Selected Load control.) Droop Speed Control mode is used when a unit is synchronized with other units on a grid, not when a unit is providing frequency control for a small, islanded grid or captive power plant (when there is no external Power Management System or external load control system).

Isochronous Speed Control mode is used to provide frequency control for a small, islanded grid or captive power plant (usually, and when there is no external Power Management System or external load control system).

Primary Frequency Response over-rides Pre-Selected Load Control when the prime mover's actual speed changes--which is what happens during grid instability. It essentially lets a unit operating in Pre-Selected Load Control act as if it was in "pure" Droop Speed Control mode. (Pre-Selected Load Control over-rides Droop Speed Control mode. And, Primary Frequency Response over-rides Pre-Selected Load Control. Complicated? Yes. But it would not be necessary if people didn't mis-use Pre-Selected Load Control.)

Neither Isochronous Speed Control nor Primary Frequency Response change the load setpoint. Both Isochronous Speed Control and Primary Frequency Response use speed setpoints and modify speed setpoints--not load setpoints.

Droop Speed Control actually allows the actual prime mover speed to be less than (to "droop below") the prime mover's speed reference. And the energy flow-rate into the prime is controlled by the error between the speed reference and the actual speed. And when the actual speed is stable (as it is during stable grid frequency), changing the speed reference changes the error between the reference and actual. When the unit is running at steady load and stable frequency, the turbine speed reference is stable and the actual speed is stable. If the grid frequency changes the actual speed changes--but the turbine speed reference does NOT change. This, however, DOES change the error which changes the energy flow-rate. Just as if someone had changed the speed reference while the speed (grid frequency) was stable--changing the error which changes the energy flow-rate.

A unit with 4% Droop will be at rated power output (the prime mover will be at rated power output) when the prime mover speed reference is at 104%. So a prime mover rated at 20 MW will be at 20 MW (on an iso day) when the prime mover speed reference is at 104% and the actual speed is at 100% (when the grid frequency is at 100%). The error between the reference and the actual is 4%--which is equal to the Droop setpoint: 4%. If the speed reference is 2%, when the actual speed is 100%, the unit will be at 50% of rated output.

If the speed reference of that same unit is 102% and the speed is 100%, the output will be 10 MW. If the grid frequency changes by 1% (let's say it decreases by 1%) the error between the speed reference (102%) and the actual speed (now at 99%) increases to 3%--which causes the prime mover's power output to increase to 75% of rated, or 15 MW--while the grid frequency is at 99%.

Droop Speed Control is all about the error between the reference and the actual--BOTH of which can change. Under normal circumstances, the actual speed is constant (when grid frequency is stable and at rated). So, when the operator wants to change load--even though the operator is looking at the MW meter while changing load--the operator is really changing the prime mover's speed reference which is changing the error which is changing the energy flow-rate into the prime mover which changes the power output of the generator (when the operator is clicking on RAISE- or LOWER SPEED/LOAD). And, when Pre-Selected Load control is active while on Droop Speed Control it is changing the turbine speed reference to make the actual load equal to the Pre-Selected Load Control reference--just like an operator does, but automatically.

The problem comes in when, while the unit is operating on Pre-Selected Load Control the grid frequency (actual speed) changes. In this case, Droop Speed Control wants to change the load--but Pre-Selected Load Control does not--the two actually fight each other, which contributes to grid frequency instability. So, Primary Frequency Response over-rides Pre-Selected Load Control to allow Droop Speed Control to change load as necessary to help support grid frequency stability.

Isochronous Speed Control DOES NOT allow the actual speed to be different than the speed reference. It is VERY tight speed control--and only ONE unit on any grid (large or small) can be operating in Isochronous Speed Control mode at the same time--otherwise multiple units operating in Isochronous Speed Control mode will fight each other to control the frequency--and that's NOT pretty and results in wild load- and frequency swings, and, usually, blackouts. (There is a special mode called 'Isochronous Load Sharing' but it's really just a de-tuned mode that, for all intents and purposes, is Droop Speed Control.)

It's complicated, but it's not that complicated. If operators and their supervisors understood what's supposed to happen during grid frequency instability and how Droop Speed Control actually worked, they wouldn't use Pre-Selected Load Control--at all. In fact, it's not going to be too long before grid regulators are going to prohibit the use of Pre-Selected Load Control for most power plants, because it can cause worsening of grid frequency instability.

Hope this helps!
 
Yes, exactly what i wanted to confirm. thank you very much indeed Mr CSA

PS: when i said load setpoint, i meant the energy flow-rate into the prime mover.
 
fudge,

From the notes in the photos you sent, it appears the code (which is a different version of "primary frequency response" than I've seen before--so, that makes it three known versions of PFR (Primary Frequency Response) that exist that I know of) will work when Pre-Selected Load Control is enabled while the unit is at LESS than Base Load (which it will on Droop Speed Control).

So, if the unit is at 40 MW (which is Part Load, which is less than Base Load--so it's on Droop Speed Control) with Pre-Selected Load Control enabled and active AND the grid frequency changes by 1%, then the unit should increase its load by 1% (25 MW per your example) until such time as the frequency returns to normal. The unit load will increase as the frequency decreases, and will decrease as the frequency returns to normal.

OR, if the unit was just operating on "straight" Droop Speed Control (Pre-Selected Load Control was NOT enabled and was NOT active) at 40 MW, it would behave just as above with the code in your Mark VI--it would increase its load as the frequency decreased, and decrease its load as the frequency returned to normal.

Pre-Selected Load Control tries to over-ride Droop Speed Control--which is active when the unit is at less than Base Load, regardless of whether Pre-Selected Load Control is enabled and active or not. Let's try this. Here's an attempt at a diagram that shows when Droop Speed Control is active:<pre>
0 MW BASE LOAD
100% TNH 100% TNH
| |
|<--------Droop Speed Control------>|
| |
|<----Pre-Selected Load Control---->****</pre>
Droop Speed Control is ALWAYS active between 100% TNH (FSNL/0 MW), and Base Load. <i><b>When enabled and active,</i></b> Pre-Selected Load Control over-rides Droop Speed Control--but Droop Speed Control is still active.

[Actually, Droop Speed Control is still active at Base Load--but it is higher than Base Load, so the Minimum Selector, which chooses the lower of all the inputs for fuel control, chooses Base Load (exhaust temperature control). And, the **** next to the Pre-Selected Load Control line represents plants that use Pre-Selected Load Control to operate the machine at Base Load (exhaust temperature control--which is also wrong, but is often done. The Minimum Selector function will always choose the minimum value for fuel control, be it Droop Speed Control or exhaust temperature control.]

Droop Speed Control uses the error between TNR & TNH to control the amount of fuel. TNR is the reference (setpoint), and TNH is the actual (process). As long as TNH is stable--and it will be when grid frequency is stable, and if TNR is not changing then the error will be stable and the amount of fuel flowing will be stable.

When the operator clicks on RAISE- or LOWER SPEED/LOAD, they are changing TNR--which changes the error between TNR and TNH, which changes the fuel flow-rate.

When Pre-Selected Load Control is active, the operator is abdicating the responsibility for controlling load and is letting Pre-Selected Load Control change TNR to change the error to change the fuel flow-rate. Pre-Selected Load Control looks at the Pre-Selected Load Control reference/setpoint, and compares it to the actual load (usually the signal DW) and changes TNR in order to make the actual load equal to the Pre-Selected Load Control reference/setpoint.

This is all well and good--<i>as long as the grid frequency is stable.</i> When the grid frequency varies from nominal, TNH changes which changes the error between TNR and TNH--and that changes the fuel flow-rate which changes the load (DW). BUT, Pre-Selected Load Control detects the change in load caused by the change in frequency and TNH and and causes TNR to change in the opposite direction to counter the load change. Which is exactly, 100% what should NOT happen.

So, GE has tried various methods of "tweaking" Pre-Selected Load Control so that operators and their management don't have to manually change load by clicking on RAISE- or LOWER SPEED/LOAD and continue to use Pre-Selected Load Control to operate the turbine.

The really sad part is that when the load changes in response to the grid frequency change (as it should) operators and management <i><b>incorrectly believe</i></b> the load should remain constant--and so they complain that their turbine is swinging load because they think it should remain stable during grid frequency disturbances. They say, "That DARNED control system isn't working properly!!!" when in reality, it's their thinking that isn't correct.

If your code works for your site, that's great. If your site experiences grid frequency excursions and your management is happy to let the operators use Pre-Selected Load Control, then the code you have (which GE also calls "Primary Frequency Response," PFR,in its other forms) is what is needed to allow the unit to properly respond to grid frequency disturbances to help support grid frequency until such time as the grid regulators can restore the grid to nominal frequency. There's not one way this has to be done ("tweaking" Pre-Selected Load Control during frequency excursions).

By the way, what is SOR and TOR, please?

As was said before, I believe it's not going to be long before grid regulators all over the world are going to limit or deny the use of Pre-Selected Load Control without some kind of "primary frequency response." The name "primary frequency response" implies that the governor (control system) is going to respond to frequency changes regardless of what control mode the unit is in <i>when it's operating at less than Base Load.</i> Many places in the world are telling power providers the produce more than 50 MW (approximately) that they can no longer use VAR or Power Factor control. And, that's very much like Pre-Selected Load Control for system voltage and stability during grid problems. When grid regulators finally understand how Pre-Selected Load Control works--without "primary frequency response"--they are going to put a stop to Pre-Selected Load Control without primary frequency response. Which is what the whole primary frequency response movement is about--allowing generator prime movers to properly respond to grid frequency disturbances.

Try this (if your management will permit it): If the load setpoint for your plant at some time is, say, 40 MW (to use the value from your previous example), use Pre-Selected Load Control to get the unit to 40 MW, and then click on RAISE- or LOWER SPEED/LOAD, which will disable Pre-Selected Load Control. And wait. And watch. And wait. And watch. You will very likely be very surprised to see that (as long as the grid frequency is stable!) the load will not change by very much, if at all. (And any change will be the result of grid frequency variances.) It might be exactly 40 MW, but I'll even bet US Dollars that even when Pre-Selected Load Control is enabled and active with a 40 MW setpoint that it still bounces around from approximately 39 MW to 40 MW (depending on how well Pre-Selected Load Control operation has been tuned). So, you just might find that it will hold very steady at 39.2 MW, or 40.5 MW, with very little fluctuation like you normally see when Pre-Selected Load Control is active.

If possible, used Trend Recorder (in Toolbox) to monitor load (DW), turbine speed reference (TNR), actual turbine speed (TNH), FSR (fuel stroke reference), and fuel flow (usually FQT, or something similar). Watch L0R and L70L (the two logic signals which cause TNR to decrease and increase, respectively), also, and record SFL (Station Frequency-Line). Do this for as long as management--and the operators--will permit. (They will be SWEATING, and probably swearing) because they just KNOW that the turbine is going to trip if it's not on Pre-Selected Load Control holding the Pre-Selected Load Control reference/setpoint. They just <b>KNOW IT!</b>

Then, while Trend Recorder is still running, turn on Pre-Selected Load Control and leave Trend Recorder running for same length of time as Pre-Selected Load Control was disabled. And compare the results. Look at the stability of DW (again, presuming frequency was stable), when Pre-Selected Load Control was enabled and active versus when it was disabled. I think everyone will be VERY surprised.

But, I'll wager US Dollars, the results won't change anyone's mind--they will STILL believe that the unit HAS TO BE operated using Pre-Selected Load Control at all times. That it will trip or worse if Pre-Selected Load Control is not active and enabled at all times (even when the unit is operating at Base Load!).

Hope this helps. It would be great if you could run the little test suggested above, and record the data requested, and then share it with us. Really great.

Thanks!
 
Top