# STANDARD DROOP/SPEED CONTROL compared to other LOAD/SPEED CONTROL

#### Isulamu

It is about the "standard droop speed control" and how it can be affected by the FPRG or a bad calibration of the gas valves...etc.

Compared to the CONSTANT SETTABLE DROOP SPEED/LOAD CONTROL, The STANDARD DROOP/SPEED CONTROL has "no FSR value feedback" in its regulation loop. This mode of governor control changes FSR in proportion to speed error (droop) directly under the following equation:

FSRN = FSKRN1 + FSKRN2(TNR - TNH)

where FSKRN1 is the Initially set full speed no load fuel reference as calculated, and FSKRN2 is the ratio of percent change in fuel flow reference to percent change in speed/load reference.

OK! at full speed no load TNR-TNH would be equal to 0. and the FSRN value would be equal to FSKRN1, initially caclulated by the OEM for a particular fuel flow rate. And this value is the opening command of the GCV, so for this particular fuel flow rate it depends on the fuel heating value! also of the P2 pressure refrence FPRG!

Lets have an example: suppose FSKRN1 = 30%, when the machine finishes its start-up sequence normally the TNH (HP shaft speed) would be equal to TNR(turbine speed reference). The FSRN = FSKRN1 = 30%, as the machine was firstly commissioned for sure at this stage everything would be stable, because the 30% value was enough to make the machine stable at FSNL 100% speed, for example with an FPRG (gas ration valve control pressure reference) = 13bar.

Lets suppose we have a problem with the P2 pressure either the FPRG constant value or the SRV valve or even the P2 transmitter. This problem causes the P2 pressure to be 10bar when the machine at FSNL. So when the machine reaches FSNL and FSRN = 30% > we won't be having enough energy to make it stable, and TNH would to tend to drop. When it drops, TNR-TNH would be higher than 0 which will make FSRN higher than 30% to reach the FSNL speed again. We will be having a non stable state of the machine fluctuating at FSNL

Myself I'm having a problem with this mode and how it is vulnerable to any misbehave from the gas valves or P2 transmitter, the P2 set point (FPRG), and even a problem with the fuel injectors because we do not have an FSR feedback value in the loop.

NOTE 1: This enquiry is the same for the base load GCV opening value, at base load. And if the droop value is 4%, so TNR-TNH = 4% and 4% * FSKRN2 + FSKRN1 = FSRN, so there is no FSR Feedback value in the loop, at any load point we will be having an particular opening command of the GCV NO MORE!

NOTE 2: i am comparing this regulation loop to all other load regulation loops for single or double shafts where the FSR feedback value is always present. So we will be adding a certain value the actual FSR until we reach the load reference. So the for a certain load point we won't be having a particular pre-calculated value of FSRN which will make the system control adapt the FSRN value to the present conditions and parameters.

example: load output = 100MW, FPR2 = 12bar, FSRN = 40 %
if the the P2 drops for whatever the reason we will be having - load output = 100MW, FPR2 = 9bar, FSRN = 60 %

So the control system rises the FSRN value to maintain the load output.

#### CSA

isulamu,

Let's just step back and look at this rationally, shall we?

The Expected Fuel Characteristics data is just an estimate--NO ONE EVER CHECKS THIS AGAINST THE ACTUAL RUNNING DATA DURING COMMISSIONING. (Unless there is a serious problem.)

GE-design heavy duty gas turbines don't have fuel injectors, they have fuel nozzles. The fuel nozzles have orifices, and the fuel flows through the orifices. The fuel nozzles are supplied by common manifold(s) (depending on the type of combustion system, conventional or DLN). Orifices can become plugged ("choked") and block fuel flow, BUT this is reflected in the exhaust temperature spread. Also, some fuel nozzles have been shown to leak causing more fuel than would be desirable to flow into the combustor--which would ALSO show up in the exhaust temperature spreads.

You are correct--there is NO FSR feedback in Droop Speed Control. None. Zero. Zilch. Nada. Niente. There doesn't need to be any.

The only time load comes into the "equation" when the unit is operating on Droop speed control is if Pre-Selected Load Control or some form of external load control is active (for units with "pure", straight Droop speed control). And, then, FSRN will be whatever is required to maintain the load setpoint, regardless of P2 pressure or GCV LVDT calibration(s).

The entire reason that P2 pressure is controlled and held constant during rated speed (which implies rated frequency!!!) operation is that by holding the fuel pressure upstream of the GCV(s) constant the flow-rate through the GCV(s) is proportional to stroke (or at least it was before the advent of independent gas control valves as used on DLN combustion-equipped units). That made it possible to have a value of FSR that was proportional to GCV stroke (position). DLN units still have FSR, it's just that FSR is split into what's required for the different fuel nozzle orifices in order to maintain emissions requirements (which is ENTIRELY open loop in most cases).

P2 pressure is set to be high enough so that small fluctuations in P2 pressure won't adversely affect unit operations. In other words, if the gas fuel supply pressure/flow is restricted such that the SRV reaches 100% stroke and still can't maintain the P2 pressure reference (FPRG), then what will happen is that the load will eventually start to drop. (I'm presuming grid frequency is stable at all times in this response.) If the unit is on pure Droop speed control--without Pre-Selected Load Control OR external load control (sometimes called AGC)--the unit will lose load. Because the fuel flow-rate for the given FSR is not sufficient to maintain load because the gas fuel supply is restricted. It's not that P2 pressure is high or low--it's because the gas fuel supply pressure/flow is restricted and is insufficient to maintain normal operating parameters (namely, P2 pressure).

If gas fuel supply pressure were much higher than normal, that would cause the SRV to close more than normal--but as long as it was able to maintain P2 equal to FPRG there would be no operational problems as far as maintaining load, unless the gas fuel supply pressure was so high that the SRV had to go very nearly closed, which would represent a restriction of gas fuel flow-rate.

Constant-settable Droop Speed Control does, in fact, use load in it's calculation of FSRN. BUT, when operating parameters are normal (that is gas fuel supply pressure/flow are normal and fuel nozzles are not choked (restricted)) it operates virtually identically to straight droop speed control--except that it's reference is a load-biased TNR (TNRL if I recall correctly). This was done to prevent load swings during DLN combustion mode transfers, and it has just been carried over to the majority of new units regardless of combustion system. It works to try to help stabilize load during temporary, transient conditions.

And, if a unit with Constant-settable Droop is being operated with some kind of load control active (Pre-selected or external), then it's going to respond primarily to the load setpoint, trying to make actual load equal to the load control setpoint by adjusting TNR as necessary.

Now, with regard to GCV LVDT calibration. The very fact that FSR is NOT used in Droop speed control feedback means that inaccurate LVDT calibrations will not (generally) adversely affect Droop speed control operation. Let's consider the case where the LVDTs were calibrated such that the indicated position is 10% less than the actual position. The speed error is still going to result in the same change of FSRN, which is going to result in the same change in the gas control valve position(s). Same goes for if the LVDTs were indicating 7% less than actual position. The speed error is going to result in the same change in FSR.

If you can describe the problem you are experiencing with unit operation, we can probably try to help understand what might be happening. If you are not experiencing unit operational problems but are having trouble understanding the control scheme(s) be sure you are NOT including the effects of load control (Pre-Selected or external) when you considering data.

And, if you're worried about the effects of grid frequency changes on P2 pressure (because frequency is directly proportional to speed and speed is used to calculate FPRG), you shouldn't be--as long as the gas fuel supply pressure/flow is not restricted. In other words, as long as FPG2 equal FPRG and the SRV is not at 100%, then the unit is going to operate just fine with small changes in P2 pressure caused by grid frequency deviations. If you are experiencing something different, please describe what you are seeing. Actually, post some trend data to a sharing site and post the link to the data here.

>Lets suppose we have a problem with the P2 pressure either
>the FPRG constant value or the SRV valve or even the P2
>transmitter. This problem causes the P2 pressure to be 10bar
>when the machine at FSNL. So when the machine reaches FSNL
>and FSRN = 30% > we won't be having enough energy to make it
>stable, and TNH would to tend to drop. When it drops,
>TNR-TNH would be higher than 0 which will make FSRN higher
>than 30% to reach the FSNL speed again. We will be having a
>non stable state of the machine fluctuating at FSNL

This could happen, but, I don't think unit operation would be "non stable." FSRN would just try to increase to make TNH equal to TNR. And, at FSNL a drop of 2 barg in P2 pressure is NOT going to make TNH less than it should be. P2 pressure is calculated to provide proper pressure/flow at Base Load, not FSNL. Gas fuel flow-rate at FSNL is MUCH less than at Base Load. So a small drop, even 2 barg, in P2 pressure at FSNL isn't going to have any appreciable effect on TNH at FSNL.

>NOTE 1: This enquiry is the same for the base load GCV
>opening value, at base load. And if the droop value is 4%,
>so TNR-TNH = 4% and 4% * FSKRN2 + FSKRN1 = FSRN, so there is
>no FSR Feedback value in the loop, at any load point we will
>be having an particular opening command of the GCV NO MORE!

When the unit is operating at Base Load, the reference for the GCV is FSRT. And the Mark* will be trying to flow as much fuel as it can to make TTXM equal to TTRX, and FSRT will be adjusted to keep TTXM equal to TTRX. In this case, the control loop is CLOSED--because the reference is TTRX (exhaust temperature reference) and the feedback is TTXM (actual exhaust temperature). FSRT will be changed to whatever is required to make TTXM equal to TTRX. And load will be whatever it is.

>NOTE 2: i am comparing this regulation loop to all other
>load regulation loops for single or double shafts where the
>FSR feedback value is always present. So we will be adding a
>certain value the actual FSR until we reach the load
>reference. So the for a certain load point we won't be
>having a particular pre-calculated value of FSRN which will
>make the system control adapt the FSRN value to the present
>conditions and parameters.

Yes. Again, FSKRN1 and FSKRN2 are calculated values based on expected fuel characteristics and reasonably good calibrations of GCV LVDTs and a new and clean turbine and ambient conditions matching machine nameplate rating.

I have seen units with poorly-calibrated GCV LVDTs that reached "FSNL" and were not exactly at 100.3% speed (which is the typical TNR for generator drive GE-design heavy duty gas turbines). The unit might have been at 99.87% or 100.45%, BUT when auto-synch was selected TNR was adjusted to make TNH change so that the speed-matching permissive of auto-synch was achieved. Then after that, the value of FSR was not as per the expected fuel characteristics. And, sometimes the unit would be at 103.21% when it reached Base Load, or 104.98% when it reached base load (again because either the GCV LVDTs weren't calibration properly, or the actual fuel didn't match the expected fuel, or because the fuel nozzle orifices weren't the same as when the unit was new)--but the unit operated just fine in all conditions and at all loads, Part Load and Base Load. In fact, you will find that the majority of GE-design heavy duty gas turbines (without ARES, or MBC (Model-Based Control)) have FSRs that don't work out exactly to 4% Droop. And, they run just fine. They aren't 100% perfectly tuned and operating, but can you say that about any car or truck? (Actually, most modern vehicles have O2 sensors in the exhaust and they operate MUCH closer to design that most heavy duty gas turbines, because the O2 sensor feedback isn't used to calculate fuel flow or performance--only emissions.)

Again, small errors in P2 pressure aren't going to impact FSR or GCV position(s). It's when the SRV is fully open and it still can't maintain P2 equal to FPRG, or when it's nearly closed while trying to maintain P2 eaual to FPRG, that problems with unit operation are going to occur. And, that is really only going to happen with the unit is at higher loads, not at FSNL (unless gas fuel supply pressure/flow is severely restricted or excessively high). So, don't worry about P2 pressure differences unless the SRV is at or near full open or is at or near full closed while the unit is at higher loads (probably 40-50% load or more).

Droop speed control is proportional-only control. There is nothing that causes the control system to make the actual speed equal to the speed reference. It is recognized there will be an error, and the error is vital to how the control scheme works. If other parameters aren't all perfec, well, then, Droop speed is not going to be perfect.

>example: load output = 100MW, FPR2 = 12bar, FSRN = 40 %
>if the the P2 drops for whatever the reason we will be
>having - load output = 100MW, FPR2 = 9bar, FSRN = 60 %
>
>So the control system rises the FSRN value to maintain the

I think you're just making too many assumptions about inter-relationships between operating parameters. We don't know a lot about the units at your site (you say you have some two-shaft machines) and we don't know a lot about the fuel(s) being burned, or how much they might vary over time (some units burn waste gases from some nearby process (steel mills; refineries; etc.)). 30% FSR for FSNL is pretty high for a typical natural gas fuel, unless it's pretty bad gas (in my experience).

So, try to look at real data, using comparable terms ("straight" droop and Constant-Settable Droop are similar, but use different terms), and DON'T include any operating data when any kind of load control (Pre-Selected or external) is active when at Part Load. And, Base Load is very different from Droop speed control (it's closed exhaust temperature control, not proportional speed control).

Hope this helps!

#### Isulamu

CSA

Actually like i made the statement in a previous thread, i am working at compression station.

recently i was noticing and making the comparison of two machines running at the same load reference (84% LP speed). however the GCVs were not at the same opening (can't remember the data exactly and i am not on site while writing this post). but i can confirm that this is due to different P2 pressure (the one with higher P2 had a lower GCV opening), note that the FSRN loop have the FSR feedback.

In the same time i have 5001P turbo-generator simulator which i use frequently, and i had the scenario of having the same condition (different P2). i wanted to know how the machine would react. So this led me to study the standard droop regulation loop and this was the origin of my enquiry.

My biggest concern was about reaching the FSNL speed with a P2 lower than expected and without an FSR feedback? i couldn't simulate this in the simulator this is why i posted this thread.

Also the majority of time during loading conditions i was in part load with manual control (imagining the operator is setting the load according to the grid demand).

I agree with you when the unit is being operating at Part Load. it is being operated with Pre-Selected Load Control or external load control that TNR will be adjusted to make the actual load equal to the load setpoint (even if we have a P2 problem). but in the same time we must not forget that TNR is limited with the TNKR3 MAX constant value, and even if we would not reach this value (while adjusting TNR), we may have a disagreement between the droop value and the output load (4% droop will not equal to 100% load).

Also i found this in a control specification OEM document about the settable constant droop control. <i>"This method of speed/load control is applied to units where the fuel stroke reference (FSR) is not predictable as a function of the gas turbine output power due to varying fuel heating values or where fuel is switched between different combustion system injection nozzles.&#8221;</i>

combined with what you said <i>"Constant-settable Droop Speed Control does, in fact, use load in it's calculation of FSRN. BUT, when operating parameters are normal (that is gas fuel supply pressure/flow are normal and fuel nozzles are not choked (restricted)) it operates virtually identically to straight droop speed control--except that it's reference is a load-biased TNR (TNRL if I recall correctly). This was done to prevent load swings during DLN combustion mode transfers, and it has just been carried over to the majority of new units regardless of combustion system. It works to try to help stabilize load during temporary, transient conditions."</i>

<b>Everything seems to be clear to understand now</b>.

#### CSA

Isulamu,

I'm trying to find the right tack to make this work for you.

Droop speed control is about how much the energy flow-rate into the prime mover changes for a change in the error between the prime mover's speed reference and the actual speed.

FSR is a position--a position that is proportional to the valve position (for units with conventional combustors, and units with combined SRV/GCV assemblies). The people who configure the units have a program, called, oddly enough, The Gas Program, that takes the expected fuel characteristics and chooses the correct valves, or valve internals as the case may be, and determines what the flow-rate through the valve(s) will be for a given position. So, you can really say that FSR, while the unit is running at rated speed (or thereabouts, anyway) that the position reference is really a flow-rate reference. And, if the LVDT is even remotely properly calibrated the feedback to the servo-valve output regulator is valve position (stroke). So, the position reference is really a flow-rate reference, and the feedback is similarly equatable to flow-rate.

Now, Droop speed control is proportional control only. And, all it does is change the energy flow-rate reference based on the error between the prime mover's speed reference and its actual speed. Look at the engineering units for FSKRN1 and FSKRN2. The gain's engineering units are % FSR/% Speed, and the offset's engineering units are % FSR. So, the speed error is in % Speed (TNR and TNH are both in percent speed), and when multiplied by the gain (which is % FSR/% Speed) the % Speed "cancels out", and you're left with % FSR. And when you add the offset, which is also in % FSR, you get: % FSR (or just FSR, for most of us).

Droop speed control is about how much the energy flow rate changes for a given change in the speed error. The offset (FSNL FSR) is how much energy it theoretically takes to maintain approximately 100% rated speed. And when TNR is 100.3% (which is usually the off-line speed setpoint) and TNH is 100.3% (which it will usually be when the generator breaker is open, so before and during synchronizing), the error is zero. Zero plus FSNL FSR equals FSNL FSR. When auto synch is enable, then the process of making the generator speed a little greater than grid frequency begins--and when the generator breaker closes: <b>WHAM!</b> The unit instantly slows down to grid frequency. BUT, TNR remains the same (for an instant) and because it's higher than TNH is (when the generator breaker closes and slows the unit down) the difference (the speed error) gets multiplied time the gain and that extra fuel--which <b><i>was</b></i> making the unit spin a little faster than grid frequency immediately gets converted to amperes, which is how the unit takes on load (produces electrical power).

Now, when the operator wants to increase the load, he watches the MW meter BUT he clicks on RAISE SPEED/LOAD. Those words mean: Increase the turbine speed reference, and if the unit if off-line (the generator breaker is open) the actual speed will increase (there's nothing stopping it from increasing in speed!). OR, if the unit is on-line (the generator breaker is closed) the fuel increase can't make the unit speed increase (the grid frequency is holding it from increasing) but the generator senses the increase in torque coming from the increased fuel and converts the torque to amperes--which is what makes the MW meter increase its reading. (The opposite happens when the operator clicks on LOWER SPEED/LOAD.)

And, just because the value that's being calculated is FSRN (Speed Control FSR) doesn't mean that FSR is simply a valve position. It's NOT--it's really a fuel flow-rate reference, that's in the form of a valve position.

Now, when P2 pressure is at rated, or even when it's near rated (it's <b>NOT</b> an exact value, there's really a range of P2 pressure the unit can operate very comfortably in!) the flow through the gas control valves is, for all intents and purposes, proportional to stroke. Normally, a single valve whose stroke (position) is proportional to the flow-rate through the valves is VERY expensive (or at least it was a few decades ago when GE heavy duty gas turbines were first being produced). BUT, by using a valve upstream of the gas control valve to maintain a constant pressure upstream of the gas control valve (in this case the SRV), the gas control valve can be a less expensive valve and still provide the functionality of the flow-rate being proportional to the stroke (position) of the valve.

I'm not going to get sidetracked into "what is the exact range of allowable P2 pressure for a particular turbine or turbines"--suffice it to say that unless the SRV is 100% open and can't open any further, as long as the P2 pressure is nearly equal to the P2 pressure reference then the unit will operate without much problem. If the gas fuel supply pressure/flow is restricted for some reason and the SRV has to go full open and still can't maintain P2 pressure well then at some point something has to give ("fail")--and that is probably going to be flame in the combustors. This is the classic "Loss of Flame" trip--insufficient fuel to maintain flame in the combustors.

However, given that there are parameters which protect the turbine (unless they are defeated!) from serious under-speed (under-frequency) or over-speed (over-frequency) the speed changes which result from grid frequency disturbances are NOT going to cause problems for the gas control valves.

That's why you hear about this "free governor mode" of operation. Which is nothing more than Droop speed control. GE has something to "allow" load control (Pre-Selected or external) to properly respond to grid frequency disturbances, and this can be considered "free governor mode."

Now, gas turbines have a dirty little secret. When they are operating at Base Load, producing as much power as they can while synchronized to a grid with other prime movers and generators and the grid frequency changes they DO NOT respond the way they should. When the grid frequency decreases the axial compressor speed decreases (and the IGVs are full open at Base Load), and this causes the exhaust temperature to increase. So, the Mark* says, "Hey! I have to reduce fuel to keep the exhaust temperature from increasing!? Which has the effect of reducing the power output of the gas turbine--at precisely the time when it should be INCREASING its output! But, it can't--or it will trip on exhaust over-temperature. (Don't worry; GE has a fix for this, too!) And the opposite happens when the unit is at Base Load and the grid frequency increases. This causes the axial compressor speed to increase, which causes the exhaust temperature to decrease so the Mark* adds fuel to turbine to maintain exhaust temperature--which increases the load being produced at the precise instant it should be decreasing load! (That GE fix I just mentioned also takes care of this aspect of the problem at Base Load, too.)

Remember, from FSNL to Base Load a GE-design heavy duty gas turbine-generator unit operates on Droop speed control. Load is changed by changing the fuel flow-rate, which is done by looking at the difference between the actual speed and the speed reference. And even if the result of the calculation is valve position (FSR), it's really fuel flow-rate--which is what Droop speed control is all about: how much the energy flow-rate into the prime mover changes for a change in the speed error between the speed reference and the actual speed.

There's a lot to think about and comprehend. However, it's NOT rocket science--and there are quite literally tens of thousands of prime movers and generators around the world which use Droop speed control. And there are thousands of GE-design heavy duty gas turbines that also use Droop speed control--and they all work just fine, even if they're not perfect (with their digital turbine control systems). P2 pressure is only about making FSR (valve position) proportional to fuel flow-rate. It's much easier to measure valve position than it is to measure natural gas (or any other gaseous fuel) flow-rate (think about it, you need some pretty sophisticated sensors, or at least some sophisticated calculations (that use square roots!) to calculate gas fuel flow-rate). And when GE heavy duty gas turbines were first being designed those sensors (which are readily available and relatively cheap today) were expensive and required a LOT of calibration adjustments. But, a simple LVDT measuring valve stroke (position) that is proportional to fuel flow-rate was much easier and cheaper to implement.

Constant-settable Droop speed control just uses a load bias to modify TNR. And, it does correct for load variations--which is something I've never really been a fan of (I understand why it was implemented, I just don't think it's necessary today). But, in the end, it still accomplishes the same thing: changing the fuel flow-rate for a change in the speed error (even if one of the components of the speed error is load-biased!). Constant-settable Droop just adds a level of complexity that doesn't seem warranted today (it did when DLN was in its infancy--that's for SURE!). It's also useful for fuels that have a continually varying heat content, or multiple fuels from different sources.

Anyway, it's a lot to think about. The FSR "loop" is, in fact, closed. The servo-valve output regulator looks at the difference between the valve position (for the gas control valve(s)) and the actual valve position(s) and adjusts the current to make the actual position equal to the position reference. And, we know--valve position is proportional to fuel flow--rate (energy flow-rate). The Droop speed control loop is NOT closed--it's not supposed to be closed. It's only function is to determine how much the fuel flow-rate will change for a change in the speed error (that's the function of the gain parameter--the one that's multiplied by the speed error, and the one that determines the Droop setpoint, or Droop characteristic). For units with 4% droop, the fuel flow-rate should change by 25% for each 1% change in speed error. (That's the fuel flow-rate ABOVE FSNL--which is the load-producing region of unit operation!)

If this doesn't do it for you, I don't know how else to try to explain it. It's not rocket science. Sure, there are some complexities (P2 pressure being one of them--but under normal circumstances, or even under allowable grid frequency excursions, P2 pressure changes are NOT going to adversely affect unit power output). Droop speed control is about how much the load changes for a change in the speed error.

And it's that way because back when electricity was first being introduced to the world, the ONLY feedback to the prime mover governor was speed. There wasn't any kind of sensor that could convert watts to spring tension--and governors all used springs back then. There wasn't even any hydraulics in the beginning. And, the real beauty of Droop speed control is that it still works today even with digital control systems (governors). And, it allows today's prime movers and generators to be synchronized with yesteryear's prime movers and governors--because they all "understand" speed! Speed and frequency are proportional (they are directly related)--they always have been and they always will be in an AC power system.

Enough for today (actually for this month!).

#### CSA

Isulamu,

The plot thickeneth! I completely missed that you wrote you were working at a compressor station; I do remember you writing there were single- and two-shaft machines at the site--and wondered, but didn't specifically ask. My bad. But, at least now I'm beginning to understand (I thinketh!).

Anyway, so, it's been MANY years since I worked on a two-shaft machine (as in a couple of decades), but I DO remember that once the LP speed gets to near rated the HP shaft speed is NOT controlled when the unit has variable second-stage nozzles. LP speed is the primary control loop, and the HP is "just along for the ride." The second-stage nozzles control the "split" of the energy being produced by the HP shaft, some of which goes to drive the axial compressor (which is connected to the HP shaft), and the remainder goes to the LP shaft--to produce pressure or flow, whatever the reference is that's coming from the compressor control scheme (even if it's in the Mark VI/VIe). I don't recall is FSRN is HP shaft speed or LP shaft speed, but it's certainly not Droop speed control. That's really only for generator-drive applications. Compressor-drive applications are either controlling the gas compressor discharge pressure or the gas compressor discharge flow-rate, or some value which is biased by one or the other (a "hybrid" reference of sorts).

The SRV is STILL controlling P2 pressure as a function of HP speed, which is NOT fixed, but variable in a two-shaft machine. It's variable in a range, because the axial compressor (driven by the HP shaft) has a range of speed where it can safely and efficiently operate. And the P2 pressure reference offset and gain (FSKRN1 and FSKRN2) are calculated to produce sufficient P2 pressure to make sure there is flow into the combustors (the fuel gas pressure downstream of the GCV is higher than CPD AND there is sufficient flow to maintain flame and produce torque--enough torque to drive the axial compressor, and also to drive the LP shaft which is driving the gas compressor).

The second-stage nozzles are modulated (controlled) to keep the HP shaft operating in the proper range when the LP is being controlled by the plant operating reference (gas compressor discharge pressure and/or flow).

The FSR when the LP shaft is being controlled by the plant reference is derived from the amount of fuel required to make the required plant variable (pressure and/or flow) equal to the plant reference. And, the second-stage nozzles are modulated to keep the HP shaft speed in the proper range. (Yes; fuel DOES affect HP speed, too, but in the typical GE-design two-shaft control scheme it's secondary because the REAL output of the unit is the gas compressor pressure and/or flow.)

So, the P2 pressure of a two-shaft machine IS going to be variable--while the P2 pressure of a single-shaft machine at rated speed/frequency is NOT going to be variable, it's going to be constant (as long as grid frequency is stable and at rated).

I'm fairly confident the Operations & Service Manuals provided with the two-shaft machines have an explanation of how the control scheme works for a two-shaft machine. And, it's not exactly the same as for a single-shaft generator drive application, which uses Droop speed control for Part Load operation (from FSNL to Base Load). Again, the high-points of GE-design heavy duty two-shaft gas turbine control are that the primary fuel control is used to control the LP shaft speed which is driving the gas compressor which is producing pressure and flow, and the second-stage nozzles are used to control HP shaft speed. Both do use the same P2 pressure control formula (FPRG=(TNH*FPKGNG)+FPKGNO), but because the HP shaft speed of a two-shaft machine is NOT FIXED P2 pressure WILL NOT be constant during loaded unit operation, as it is for a single-shaft generator drive. (NOTE: Some two-shaft GE-design heavy duty gas turbines are also used for generator drive applications. I worked on one even before I worked on compressor-drive two-shaft machines, but I could only spell Droop speed control and didn't really have any clue how it worked or why and I was more interested in how the unit got up to speed (the start-up sequence) than how it controlled LP shaft speed and load and frequency. (And, the generator was driving the main propulsion engine of a ship, so it more than likely operating in Isochronous speed control, anyway...!)

Hope this helps! All GE-design heavy duty gas turbine control schemes are NOT the same (especially when we're talking about single- and two-shaft machines and generator- versus mechanical- (compressor-)drive units). But, I don't recall ANY GE-design heavy duty gas turbine that didn't have an SRV for gas fuel control that used HP shaft speed as the reference for P2 pressure.

One more thing before I sign off this thread: It's entirely possible that the LP shaft speed of two compressor-drive units of the same Frame size can be identical, but the FSRs and HP shaft speeds can be different (they shouldn't be drastically different, but they can and most likely will be different). That's because of a LOT factors: time since last maintenance outage; condition of turbine inlet air filters; cleanliness of axial compressor; condition of gas compressor; type of gas compressor; calibration of second-stage nozzle LVDTs; calibration of IGV LVDTs; calibration of GCV LVDTs; and so on. Every two-shaft machine isn't going to produce exactly the same amount of torque at exactly the same amount of speed for exactly the same amount of fuel flow at exactly the same ambient conditions. Same goes for two generator-drive units of the same Frame size....

#### Isulamu

CSA

I can see that our discussion and your explanation has expanded to make everything clear to anyone who reads this thread.

>Now, when P2 pressure is at rated, or even when it's near
>rated (it's <b>NOT</b> an exact value, there's really a
>range of P2 pressure the unit can operate very comfortably
>in!) the flow through the gas control valves is, for all
>intents and purposes, proportional to stroke. Normally, a
>single valve whose stroke (position) is proportional to the
>flow-rate through the valves is VERY expensive (or at least
>it was a few decades ago when GE heavy duty gas turbines
>were first being produced). BUT, by using a valve upstream
>of the gas control valve to maintain a constant pressure
>upstream of the gas control valve (in this case the SRV),
>the gas control valve can be a less expensive valve and
>still provide the functionality of the flow-rate being
>proportional to the stroke (position) of the valve.

I am already aware of that and we have discussed this subject in a previous thread: https://control.com/thread/1510863192

>I'm wondering if this isn't your problem that you're
>struggling with.... When the grid frequency decreases (in
>other words TNH decreases!) what happens is that turbine
>load should go up--because of Droop speed control. When TNR
>is constant (which it will be when the load is constant and
>stable) and TNH goes down, the speed error increases--which
>cause FSRN to increase. This should cause the generator load
>to increase--but MOST plants and plant operators and
>technicians believe that THEIR turbine-generator load should
>remain stable when the grid frequency is off rated (in this
>example it's low). BUT, that's NOT what's supposed to
>happen--the prime movers and generators synchronized to a
>grid should ALL (those that are not already at full power
>output!) increase their load to try to help maintain grid
>stability. BUT, if anyone is operating their turbine with
>any kind of load control while it is at Part Load then the
>load control is going to try to counter (act against) the
>Droop speed control. Load control (Pre-Selected or External)
>wants to keep the load constant--regardless of grid
>frequency. But, in reality, when there is a grid disturbance
>the prime mover governor SHOULD change the load to help
>support grid stability. Those units operating in load
>control actually make the grid frequency problem WORSE!

I was able to contribute and make it clear for others also

My problem like i mentionned in my previous reply <b>was</b> making the analogy with the LP load control in double shaft heavy duty GT and the generator load control (Standard droop load control) especially when we have some problems with the P2 pressure, and like i said <b>was</b>, now i can confirm that issue is gone.

>Now, gas turbines have a dirty little secret. When they are
>operating at Base Load, producing as much power as they can
>while synchronized to a grid with other prime movers and
>generators and the grid frequency changes they DO NOT
>respond the way they should. When the grid frequency
>decreases the axial compressor speed decreases (and the IGVs
>are full open at Base Load), and this causes the exhaust
>temperature to increase. So, the Mark* says, "Hey! I have to
>reduce fuel to keep the exhaust temperature from
>increasing!? Which has the effect of reducing the power
>output of the gas turbine--at precisely the time when it
>should be INCREASING its output! But, it can't--or it will
>trip on exhaust over-temperature. (Don't worry; GE has a fix
>for this, too!) And the opposite happens when the unit is at
>Base Load and the grid frequency increases. This causes the
>axial compressor speed to increase, which causes the exhaust
>temperature to decrease so the Mark* adds fuel to turbine to
>maintain exhaust temperature--which increases the load being
>produced at the precise instant it should be decreasing
>load! (That GE fix I just mentioned also takes care of this
>aspect of the problem at Base Load, too.)

In the same time after imagining the situation(lets take the case where the grid frequency drops), when the compressor speed decreases means HP shaft speed TNH and the compressor discharge pressure CPD will decrease, which means the speed biased temperature setpoint TTRXB and the CPD bias to isothermal will <b>increase</b> which i imagine will either not reducing fuel flow rate because like you said when the axial compressor drops in speed the exhaust temperature will increase (as TTXM increases still equal to TTRXB) or allowing the turbine to add more fuel flow rate (to make TTXM equal to TTRXB) which will increase load output.

So could you explain more the situation and i hope this is not the fix you were talking about because i am translating one of the recent application code of GE nothing more.

#### CSA

Isulamu,

The "fix" that GE sells allows the unit to over-fire when it's at Base Load AND the grid frequency decreases--BUT, they add a timer that keeps track of the time the unit is over-fired and that has to be factored into maintenance outage planning. By over-firing it is meant that the exhaust temperature control is allowed to exceed the Base Load rating for a short period (because it's not allowed to happen indefinitely, either!).

And, the opposite happens when the grid frequency increase--instead of producing more power the exhaust temperature reference is reduced, but this isn't usually counted against maintenance outages, though I think it's also time-limited.

It's just a couple of "tweaks" to the exhaust temperature control scheme that will allow the unit to operate as if it were in Droop speed control in order to help support grid stability during brief grid frequency disturbances.

I'm really at a loss to understand your other question. Most generator drive applications do not use a speed bias, but many two-shaft units do. A LOT of people think TTRXB means "Base Load Turbine Exhaust Temperature Reference", but as you noted it really means "Speed-Biased Exhaust Temperature Reference." When axial compressor speed decreases on a single-shaft generator drive application, the flow through the axial compressor will decrease. This does two things: it reduces CPD, AND it causes the exhaust temperature to increase (if fuel were held constant). When the unit is at Base Load the combination of decreasing CPD AND increasing exhaust temperature will cause the exhaust temperature reference to change. Since the exhaust temperature reference curve has a negative slope, it will cause the exhaust temperature reference to increase but the actual exhaust temperature is already tending to increase (because of the reduced air flow through the unit). So, the net effect is that the power output of the unit will decrease--because of both the decreased air flow (mass flow) through the unit AND because the fuel will also be decreased to limit the exhaust temperature increase and protect against exhaust over-temperature. This is what happens at Base Load, not on Droop speed control. And only for single-shaft generator-drive applications. (The negative slope of the exhaust temperature reference curve can be very confusing for many people; it's counter-intuitive--but that's the way gas turbines work!).

Again, if the fuel were held constant the exhaust temperature would increase simply by virtue of the decrease in axial compressor speed. So, it's necessary to reduce the fuel to limit the exhaust temperature increase, and because the air flow through the machine is decreasing the air flow (mass flow) is decreasing which is also decreasing the power output of the machine.

When ambient temperature increases it decreases the density of the air flowing through the axial compressors (running at a constant speed). This causes CPD to decrease which causes the exhaust temperature reference to increase, but the power output of the machine goes down. The fuel flow also decreases very slightly, but the majority of the power loss is attributed to the decrease air/mass flow, not so much the fuel flow decrease. Again, this only happens when operating at exhaust temperature control.

Hope this helps!

#### gustavo_marcelo

Now I am getting a bit confuse. apology for that.

We have droop, constant settable droop (CSD), preselected load( PSL), FGMO and Primary Freq Response (PFR).

Droop = PSL = FGMO = PFR.

How does the above differs with CSD?

#### CSA

gustavo_marcelo,

Good on you, mate, for clarifying your use of abbreviations and keeping my confusion to a minimum.

Droop Speed Control and Free Governor Mode of Operation (FGMO) are, I believe, the same thing. Or at least, FGMO is intended to allow a prime mover to respond to grid frequency disturbances as if Droop Speed Control was active and not being "influenced" by any outer loop, such as Pre-Selected Load Control.

Primary Frequency Response (PFR) is a mode of control that is enabled when Pre-Selected Load Control is ALSO enabled that allows the unit to properly respond to grid frequency disturbances as if only Droop Speed Control was active. (Like what GE did with that? They require Pre-Selected Load Control to be active--which is an outer loop to Droop Speed Control, and then to allow the unit to respond as if only Droop Speed Control was active they require a site to purchase and also enable Primary Frequency Response--which is like an outer loop to Pre-Selected Load Control--in order for the unit to respond to a grid frequency disturbance as if only Droop Speed Control was active! AND, that privilege has to be paid for by purchasing the option of Primary Frequency Response!!! Isn't capitalism great?!?!?!)

Constant Settable Droop (CSD) is equal to Droop Speed Control--it just uses a load bias to help stabilize load when fuel may not be homogenous (not have a stable BTU content) or when fuel is being "staged* (switched between nozzles during combustion mode transfers). They seem to have now pretty much standardized on using CSD (which was previously clarified!) on just about every GE-design heavy duty gas turbine.

If you want to use equal signs to indicate similarity, then:

Droop Speed Control = CSD = PFR = FGMO

PSL is an outer loop to Droop Speed Control on GE-design heavy duty gas turbines (either Droop Speed Control or CSD), and really should not be used during continuous operation but only for changing load--unless Primary Frequency Response was purchased and enabled.

Isn't this fun?

I'm curious why you chose not to abbreviate Droop Speed Control as DSC? ;-)

G

#### gustavo_ marcelo

> I'm curious why you chose not to abbreviate Droop Speed Control as DSC? ;-)

Lol..... .
So from you explaination when PSL is active, it does not necessarily comes with PFR...hope you still remember the abbreviation..
DSC - use a speed as a feedback.
PSL and CSD - uses load as a feedback ( does the speed feedback still being used)? The link below shows the speed and load arw both meausred as feedback.

https://www.bhelpswr.co.in/Technica...EEDBACKS/constant_settable_droop_design_s.htm

When we said use a load as a bias, how does this bias works to adjua2t the fuel to maintain the current load?

#### CSA

gustavo_marcelo,

>When we said use a load as a bias, how does this bias works

As was told, we can't post images to droopspeedcontrol.com. The image of Constant Settable Droop Speed/Load Control from a GE Control Specification was posted to the IGTC website. Review the image. DWATT is the load signal (feedback) from the generator. It feeds into the block that calculates DWDROOP, which is then subtracted from TNR to produce TNRL, that is then used as the minuend for the speed error calculation (the difference between TNRL and TNH) that feeds into the integral function of Constant Settable Droop (at least that's how I think it works; again, that whole scheme has always been like trying to speak Greek to me--I know it works, but I'm not sure exactly how. (I think that's only the second time in my life I've been able to use 'minuend' in a sentence--actually maybe the third!)

In the image referred to above, N is the speed feedback (TNH) which is used to calculate the speed error. DWDROOP has to be a value that is proportional to the Droop setpoint/regulation value and the rated load of the gas turbine.

I know that, in essence, Constant Settable Droop Speed/Load Control uses a speed error--it's clearly shown. It's just the other input to the speed error calculation is a load-biased speed reference signal--TNRL. The bias (modifier) is how the actual load, as a function of the rated load and the Droop setpoint/regulation value is brought into the equation.

Study the image. And compare it to the formula you posted for "pure" Droop Speed Control. One is straight proportional control using a speed reference and speed feedback. One loop. The other, it seems to me has an inner loop and an outer loop--one that biases (modifies) the speed reference signal that is used in the innner loop. And, there's an integral function in there. Maths have never been my strong suit.

I'm trying to work through some numbers for an explaination of Constant Settable Droop Speed/Load Control. But, also have a day job--one that pays my bills and provides a little extra for a beer now and then. And I like my infrequent beers. And my creditors like it when can pay my bills.

I also usually find that if I understand why a question is being asked, I can usually find a way to answer it that satisfies the questioner. I don't understand what about Droop Speed Control has not been explained. With some history thrown in as a bonus to try to help clarify words and fundamentals. Until such times as someone else can help with explaining Constant Settable Droop Speed/Load Control, or I can work through my understanding so I can explain it to other people so they can understand it. If you can help me understand what part of Droop Speed Control you are having trouble understanding and why you need to understand it it's possible we can help clarify things.

Let's try this. Since the beginning of AC power generation, transmission and distribution the way that the energy flow-rate has been controlled into the prime mover of a synchronous generator has used the actual speed feedback from the prime mover and generator and a signal that is proportional to speed. And for multiple units synchronized together they all had to have the same mode of governor control to be able to work together as one generator powering a load (loads) that are bigger than any single generator could power. And, as we've learned, generators synchronized together are all operating at the SAME frequency--as one big generator. So, it makes sense they all use the same basic control scheme: speed. And speed error. And because frequency (speed) is so fundamental to AC power systems it's a fine scheme. Pretty darn near perfect, actually. Constant Settable Droop Speed/Load Control is probably a good technological improvement on simple Droop Speed/Load Control. It's unfortunate that the OEM in this case doesn't do a much better job of documenting their control schemes--not only for their Customers, but for their field service personnel.

If you could get access to a Mark VI HMI and Toolbox Trend Recorder, or at least look at some real numbers for the signal in the image referenced above, it's highly likely things would become clearer and clearer. And if you could also look at how the same scheme is accomplished for the other turbines at your site, things would probably become even clearer.

To be quite frank, gustavo_marcelo, if you can't get access to an HMI and Toolbox your learning curve is going to be quite steep and difficult. Especially given that there is so little documentation available from the OEM. And, if you don't have any training in control loops (PID control loops, in particular) delving into them is going to be even more difficult. (I don't have any way of knowing what your experience and training is; that's one of the bad things about these anonymous forums. I'm not criticizing; just saying we don't have any knowledge one way or the other. It might help if we did.)

I don't have access to a running GE-design heavy duty gas turbine so I can't get real data. It took me literally decades (seriously) to arrive at my current understanding of "pure" Droop Speed Control and to be able to at least try to explain it to others. I tried very hard to learn this in university from my professor and the text- and reference books, but I also eventually came to know that those were not properly defining and stating real-world conditions--and that people more educated than I did not truly understand Droop Speed Control in the "real world" and so could't explain it properly in their texts and references. But, I didn't quit, and I kept leaving it for a while, and coming back to, and leaving it and coming back to it, and finally I could see what was missing and bring together all of the fundamentals and principles and things started becoming clear. And clearer. And clearer. I also had the advantage of commissioning some of the biggest and best machines in the world, with tons of real-world data and tools like Trend Recorder.

I learn a LOT from responding to these threads on control.com, and I've been able to refine my explainations over time (I even thought I had gotten over using the bicycle analogy; was it not helpful at all?). But then I'm continually failing to understand how people can grasp that electric motors convert amperes into torque, but those same people can't understand how generators convert torque into amperes. They know that generators provide the amperes to to the grid that get connected to electric motors. But these same people also think their generator runs at variable speed when synchronized to the grid.

It's a wonderful world, isn't it? We are all so lucky that machine designers have been able to create machines that can be "operated" and work so well by people without really understanding how they work. But, then the same can be said of automobiles and a lot of automobile operators (drivers), too, right? ;-)

It does seem, though, that such expensive and critical machines should have better documentation and well-trained operator and technicians. doesn't it? But, this is the real world we live in.

#### gustavo_marcelo

Thank you for the wonderful explaination, realy appreciate it.

#### mariselvan

Hi CSA ,

Thank you for sharing your great knowledge with everyone. This unbelievably great for those who work with MK series controllers and GE machines as the all the notations and nomenclature are from GE system. I saw so many threads for DRROP SPPED control but I could not find anything in general to explain about DROOP SPPED and FREQUENCY RESPONSE of turbine (Primary control) . Can you please advise if there a such thread which can be useful for other OEM turbine end users too . I would like to understand this DROOP SPPED and FREQUENCY RESPONSE of turbine (Primary control) .Thank you

#### CSA

mariselvan,

I do not know of another thread on Control.com for other OEM's turbine control systems.

I can say that Droop Speed Control is what allows generators and their prime movers made by ANY manufacturer to all be synchronized together. That means, that whatever the notations and nomenclature are for other OEM control systems they all have to operate the same. And they have had to do that since the beginning of AC power generation. That's how newer prime movers and generators can be synchronized to older, existing prime movers and generators--because they ALL follow the same principles of Droop Speed Control.

Some prime mover control systems use load instead of speed--but it's all the same in the end. (I say this because I have been told, and I have seen some systems that do this--but I am NOT familiar with any of them, not as familiar as I am with GE-design heavy duty gas turbine control systems and philosophy. So, if you want details about this comment--I probably shouldn't have put it in the thread; sorry. I can't provide them. And, I believe that while it is true--in some form--that this doesn't really conform to proper Droop Speed Control philosophy, and under the "wrong" conditions can actually make grid instability worse.)

Droop Speed Control is a universal concept--used around the world.

Droop Speed Control has been covered SO MANY times on Control.com, sometimes I think the site should be called DroopSpeedControl.com...! So, there are LOTS of threads, over almost two decades.

When you say "frequency response" what exactly do you mean? Are you trying to understand how a large ("infinite") grid controls frequency, or are you trying to understand how a smaller grid with only a few, or two or three, prime movers and generators controls frequency? Are you trying to understand why the load of the prime mover(s) and generator(s) at your site fluctuates, possibly wildly, when the grid frequency is not stable? Do you believe the load of the unit(s) at your site should be stable when the grid frequency is unstable? Are you trying to understand how it might be somewhat possible to have the load(s) of the unit(s) at your site remain relatively stable when the grid frequency is not stable?

Droop Speed Control is about how much the energy flow-rate into the generator prime mover changes for a given change in the frequency of the grid the generator is synchronized to. For example, if the grid frequency changes by 1% for a machine with 4% Droop, then the load of the prime mover and generator will change by 25% of rated. If we keep everything on percent of rated speed and percent of rated load, it makes everything much easier. The energy flow-rate into the generator's prime mover can be changed in one of two ways: by the operator changing the machine speed reference or by the frequency of the grid the generator is synchronized to changes. The Droop Speed Equation is about the error between the actual speed and the speed reference. If the operator wants to increase the load being produced by the unit, he/she increases the speed reference. On a stable grid the unit can't actually increase it's speed--so the additional energy flowing into the prime mover gets converted by the generator to amperes, which increases the watts (kW, MW) out of the generator. In this case, the grid frequency is stable (as it should be) which means the generator and prime mover speed are also stable, and by changing the speed reference the difference between the speed reference and the actual speed changes which changes the energy flow-rate into the prime mover which change the power flowing out of the generator.

When the unit load is stable (the operator is not changing the unit speed reference), if the grid frequency changes the generator--and prime mover--speed will change. And that will cause the error between the two to change, which will cause the energy flowing into the prime mover to change which will cause the power flowing out of the generator to change.

It's all about the difference between the speed reference and the actual speed. Change either of them while the other is stable, and the energy flow-rate into the prime mover will change which will change the energy output of the generator. Period. Full stop.

And, because all prime movers and generators subscribe to this control philosophy (which is genius, actually!) they can be synchronized together to produce more power than any single prime mover and generator could every produce--because when they are synchronized together they are acting as one prime mover and generator.

Genius. Really. Simple. Break it down into it's components: actual speed and speed reference. Actual speed, when synchronized to a grid with other prime movers and generators, is a function of grid frequency (F=(PN)/120). The speed reference is a signal which is controlled by the operator, using the RAISE SPEED/LOAD or LOWER SPEED/LOAD buttons. On a stable grid, the unit speed (frequency) isn't changing. So, by changing the speed reference one changes the error between the actual and reference speeds and that changes the energy flow-rate into the prime mover which changes the amperes flowing out of the generator. Which changes the watts (kW, MW) flowing out of the generator.

Nothing more. Nothing less. Simple proportional control.

And, again, no matter what the notations or nomenclature is/are--it all has to work in the same way in order for units from any manufacturer to be syncyhronized together and to stably produce electric power.

Hope this helps!

#### PungentReindeerKing

When you say "frequency response" what exactly do you mean?
I don’t have much to add as the explanations on this and other threads are literally better than the GE documents, but I may be able to clear this up. Some OEMs label droop speed control “frequency response.” It’s exactly the same thing. But you can dig through their documentation, their HMIs, and their code and not find a single reference to droop. It’s all “frequency response” verbiage.

CSA

#### mariselvan

Hi CSA ,

Your explanation also genius to make the things easier to understand with real life examples such as straight line equation and calling it things on % . I was asking about Primary Frequency Control (PFC =Governor response for individual turbo-gen) . And in the upper level there is also AGC (Automatic Generation Control) which is pretty much controlled by the community grid operator. Plant receive load demand in the form of pulses to our central plant controller and that in turn direct the GT to respond for the demand.

On PFC , Now I would like to understand how the speed reference and MW relationship works in real word. Say for example if my machine is with the base load of 200 MW , 50 Hz and droop value of 4% . From this specifications , I understood from the online documents that load would be 0 for the frequency of 52 Hz and 200 MW for 50 Hz. Wonder how does it works after this ? 100 the professional gain ? I don't see a controller rather ramping up/down the values with positive and negative gradients with multiple selection logics to add this response value to the final demand value. I would wait to understand before we go into AGC.
Thank you

#### CSA

mariselvan,

Please be extremely careful--and critical--of on-line documents (and textbooks and reference books) discussing droop speed control and various off-frequency operations. MOST of the time they do not state all of the conditions for their examples. MOST of the time they are presuming the generator set (prime mover and generator) are operating in Droop Speed Control while NOT SYNCHRONIZED to a grid with any other generator sets--which just doesn't happen in the real world. MOST authors of these papers, which extremely intelligent, have very little--if any--real world operating experience in power plants. They only have theoretical knowledge and mathematical knowledge and they try to use that to explain concepts without properly defining/stating all of the conditions for the proclamations and statements they are making. So, just be aware that a lot of documentation (even from some major power generation equipment manufacturers) can be very lacking, and so, misleading if you read them literally (as if every word was 100% true and correct for any situation).

A prime mover can't effectively make any more power than it's nameplate rating (there are some exceptions to that statement, such as certain ambient conditions being significantly different from stated nameplate rating, or when unique logic/sequencing is employed in a particular unit's control system). So, when you're talking about what happens when the grid frequency is not at rated, it needs to be understood that the prime mover's nameplate rating basically sets the upper limit.

And, the amount of load a prime mover and its generator can accept is basically limited by the amount of power the unit was producing at the time the grid frequency starts to deviate from rated. For example, if a unit rated at 200 MW is operating at 180 MW when grid frequency decreases from rated, it can only take on (accept) another 20 MW of load--for a total of 200 MW of load--no matter how low the frequency drops below rated. Or, if the same unit is operating at 100 MW and the grid frequency starts to increase above rated, it can only shed (lose; decrease its electrical power output) 100 MW (because anything more than that is reverse power and there are usually protective relays to prevent that from occurring).

Here's a principle of AC (Alternating Current) power generation, transmission and distribution that is almost NEVER explained or discussed or considered. For a generator to produce rated frequency it has to be accelerated to the generator's synchronous speed by it's prime mover. (For a two-pole synchronous generator, to operate at 50.0 Hz, the prime mover has to spin the generator rotor at 3000 PRM--that's the generator's synchronous speed.) And, to be synchronized to a grid the prime mover must sustain the amount of power required to keep the generator spinning at its synchronous speed. Once the generator is synchronized to the grid with other generators and their prime movers the prime mover must keep producing the amount of power required to maintain synchronous speed--or, the generator it is driving will start drawing power from the grid, called reverse power. That's because, if the prime mover can't produce enough power to maintain rated generator frequency (which is directly related to speed) because the generator set is synchronized to the grid the grid will supply the power required to keep the generator spinning at synchronous speed. All the time the generator is synchronized to the grid its prime mover MUST be producing enough power to keep the generator spinning at its synchronous speed AND to produce electrical power to supply to the grid. If the amount of power being produced by the prime mover drops to the point that it is only sufficient to maintain generator synchronous speed then the power output of the generator it is driving will be 0 (watts; kW; MW). And, if the power being produced by the prime mover continues to drop below 0 (watts; kW; MW) then the generator still has to keep spinning at its synchronous speed (because it's still synchronized to the grid!) so the grid will supply power to the generator ("reverse power") to keep it spinning at its synchronous speed.

A prime mover, therefore, is producing power to do two things: Maintain the generator at its synchronous speed, AND to produce electrical power. For any prime mover, it takes a certain amount of energy (steam for a steam turbine; fuel for a gas turbine; water for a hydro turbine) to reach and maintain the generator's synchronous speed. And that amount of power must be continually produced while the generator is synchronized to a grid with other prime movers and generators in addition to the amount of power required for the generator to produce electrical power to supply to the grid.

Let's discuss how grid frequency deviates from rated under most conditions. If the total load on a system (grid) is, say, 1000 MW, and the generators and their prime movers synchronized to the grid are producing 1000 MW and the grid frequency is at rated, and a generator set producing 100 MW suddenly trips off line, the immediate effect of that is for the grid frequency to drop. NOTE that the load didn't change--only the amount of generation changed, from 1000 MW to 900 MW. SO, the remaining generation has to still supply 1000 MW of load, but it doesn't have the amount of torque previously available to produce 1000 MW of power to supply the 1000 MW of load--so the power required to maintain 1000 MW of load has to come from somewhere.

Initially it comes from the power required just to maintain rated speed (synchronous speed). But, if nothing else were to happen, eventually, the grid frequency would just continue to decay (decrease) and that wouldn't be good.

BUT, when there are prime movers which are NOT operating at rated load but ARE operating in Droop Speed Control, the governors (control systems) of the prime movers will sense the increase in the error between the speed reference and the actual speed and will increase the energy flow-rate into the prime mover in order to help maintain the load while preventing the grid frequency from decreasing excessively. AND, some of the energy which was being used to maintain rated speed (synchronous speed) ALSO gets used to help keep the grid frequency from excessively decreasing. In this preliminary example, no operator or grid regulator did anything to increase the power output of any generator synchronized to the grid--all that happened was that there were some generator prime movers which were not at rated output, and they increased their output in response to the decrease in frequency (speed) and that was enough to produce enough power to keep the grid frequency from decreasing excessively--but the grid frequency remained lower than rated. And, again--some of the energy which was being produced just to maintain rated frequency (speed) was also converted to energy to supply the 1000 MW of load. In this example, there were enough generators operating at less than rated power output that were able to increase their power output by Droop Speed Control action to increase the total power being produced by all of the generators back to 1000 MW--but nothing more than that.

So, the load--1000 MW--didn't change, but the grid frequency did. The amount of generation, which had suddenly dropped to 900 MW, increased to 1000 MW because of Droop Speed Control action, but the amount of energy increase was not sufficient to get back the grid back to rated frequency (speed). The amount of power being produced was only equal to the amount required to power the load (1000 MW) but not enough to return the grid to synchronous speed AND produce 1000 MW.

In this example, ALL the prime movers were operating in Droop Speed Control (there was NOT a unit operating in Isochronous Speed Control), and NONE of the operators at any power plant lifted a finger to click on a mouse or push on a button to change the energy flow-rate into any of the prime movers. All of the changes to the prime movers driving the generators synchronized to the grid were were the result of automatic Droop Speed Control action--responding the change in speed error between the speed references and the actual speed (frequency).

(To be continued)

#### CSA

(Continued)

NOW, what should happen in this case is that the grid regulators should notice the decrease in frequency and they likely have the ability to directly send a signal to one or more power plants not operating at rated power to increase their power output to return the grid to rated frequency. How do they do this? AGC (Automatic Governor Control; Automatic Generation Control). Certain power plants have agreed to be remotely controlled by the grid regulators who have the ability to raise or lower the power output of their plant to help control grid frequency.

OR, the grid regulators call one or more power plants and ask them to increase their plant's output in order to return the grid to rated frequency. (What's happening is that the energy to maintain rated frequency which was reduced when the 100 MW unit tripped is being replaced. The actual total electrical power being produced by all the generators synchronized to the grid isn't increasing--it remains at 1000 MW, but as the loads of some units are increased by increasing their speed references the error between the speed references of the other plants and the actual speed reduces, which reduces their outputs back to what they were before the 100 MW unit tripped.

The opposite happens when a large block of load, such as a factory, or a cement plant, or a refinery, or a large section of the grid suddenly gets separated from the grid (a large circuit breaker in a substation opens, or an air-break switch opens, etc.) and the amount of load on the grid suddenly drops in relation to the amount of power being produced. Now, there's too much electrical generation, and in this case Droop Speed Control Action reduces the power outputs of those units it can (down to 0, and hopefully not less than 0), in order to try to prevent the grid frequency from increasing out of control. In this case, the grid regulators have to reduce the load(s) on one or more power plants in order to get the frequency to decrease back to rated.

Does this make sense? It may not at first, but keep studying it, and keep thinking it over. Don't have doubts--because there's nothing here that's false. It's all true, it just may be over-simplified by me in an attempt to get concepts across. Is it "complicated?" At first, maybe, but it's really not. People just have to know--and remember--that it requires a certain amount of energy just to get to and maintain the synchronous speed of every synchronous generator synchronized to a grid and supplying the loads connected to the grid. AND, another thing which most people never really think about--when multiple prime movers and generators are synchronized together they are all acting as one generator. So, that means, they all have to be producing electrical power at the same frequency which most people just can't seem to grasp. That word--synchronism--in all its forms is a VERY POWERFUL word, and term. It's not just any old word, it has real physical properties associated with it when referring the AC power generation, transmission and distribution. And, not just one, but several.

Another way the grid frequency can be affected is when large motors are suddenly started or stopped. Or, when large loads are suddenly re-connected to the grid (say, that cement plant that tripped off-line, or that circuit breaker in the substation is re-closed, or that refinery is re-connected to the grid). Also, all during the day in most places in the world, the electrical load on the grid changes throughout the day. In the morning as people rise from their sleep and turn on their radios and/or their televisions and their computers and computer monitors and tea kettles and coffee makers and lights and air conditioners (or heaters), the load goes up. The grid operators have to know this and be responding to this, because if they do nothing, the grid frequency will decrease, and keep decreasing. And, later in the day, as people return from their jobs and cook their meals and then retire to bed, the load on the grid will decrease--which, if the grid operators are not responding attentively, will cause the grid frequency to rise. And they will use AGC, or something similar, to control grid frequency by changing the load(s) of power plants in the system (synchronized to the grid).

TANSTAAFL - There Ain't No Such Thing As A Free Lunch

Not in anything. Certainly not in AC power generation. It takes energy to get to and maintain rated speed/frequency (synchronous speed). And, that's important to remember and understand. And, it's a HUGE balancing act.

People who believe--firmly--that their power plant should remain at a constant load and speed/frequency when the grid their power plant is synchronized to is experiencing frequency disturbances just do not understand how AC power systems work--including the generators they operate. The amazing thing is: Most people operating power plants do not. It's the control systems that work silently in the background that do most of it. When the control system doesn't behave as they think it should when the grid is experiencing problems, well, they blame the control system (the prime mover governor, or the prime mover's control system). They don't understand how the governor/control system works, OR how Droop Speed Control works and is intended to work.

And, let's remember, this is just ONE aspect of Droop Speed Control--responding to changes in grid frequency. The other aspect of Droop Speed Control is how the output of the generator set is changed when the grid frequency is normal and not having problems. Droop Speed Control is all about the speed error--the difference between the speed reference and the actual speed -(which is a function of frequency). Change either one--the speed reference or the actual speed (frequency)--with respect to the other, and the output of the prime mover and generator changes. The operator changes the speed reference, which changes the speed error, to change the generator output as desired. When the frequency of the grid the generator is synchronized to changes which changes the actual speed, which changes the speed error, the output of the generator changes. It's all about the speed error. Full stop. Period. And, the speed error changes when EITHER the speed reference OR the actual speed changes with respect to the other term.

AGC is just another way to change the speed error.

And, one more time--no prime mover can produce more than its nameplate rating (under most circumstances, anyway), and no less than 0 (watts; kW; MW). If the generator the prime mover is driving is producing less than 0 (watts; kW; MW) while the generator is synchronized to a grid with other prime movers and generators then the generator will be acting as a motor and will start spinning the prime mover. For some prime movers, such as gas turbines, this isn't such a bad thing (it's a waste of energy, but it doesn't hurt the prime mover very much as long as it's still burning fuel). For other prime movers (steam turbines and reciprocating engines for example), it's a VERY BAD thing for the generator to be spinning (driving) the prime mover. That's why there are reverse power protective relays.

That's about it. Hope this helps! Again, mull this over. Nothing above is false (so there should be no doubts). It may be an oversimplification, but it's intended to convey concepts. Keep thinking about it. Try to find ways to prove what was written above, rather than ways to disprove it. It will slowly become more and more clear (it took me more than 20 years--and I'm STILL learning!).

One last thing (which should go without saying, but shouldn't be overlooked): When a prime mover and generator is synchronized to a grid with other generators and prime movers and is producing a constant power output the speed error is not changing. The speed reference is not changing, nor is the actual speed. The speed error is not changing. So, when a grid frequency excursion occurs while a prime mover and generator are operating at a stable power output while operating at Part Load (so, not a rated power output) and are operating on Droop Speed Control, what happens is the actual speed changes (because the grid frequency changes--remember that part about all generators synchronized together are actually operating as one generator!) because the grid frequency changes. Which changes the speed error. Which changes the power output, up to the prime mover's nameplate rating (and hopefully not below 0). Again, it might seem a little complicated because I'm listing all of the required conditions for my statements--but it's really actually required to do so in order to properly explain these concepts (not like most texts and references and on-line stuff does--the authors are just trying to sound smart in as few words as possible, and in the process, they are not properly explaining what are really rather simple concepts but sound more complicated when all of the conditions are listed).

#### CSA

PFC is a dangerous term to use these days when referring to proportional, Droop Speed Control, I am presuming your usage to mean proportional Droop Speed Control. Not Pre-Selected Load Control. Not PFR (Primary Frequency Response). Basic proportional Droop Speed Control. (Proportional meaning the energy flow-rate into the prime mover of a generator set changes in proportion to the change in the prime mover's speed error.)

Why, because some OEMs (Original Equipment Manufacturers) are using this term to refer to slightly different aspects of Droop Speed Control.

#### PungentReindeerKing

Another great explanation. In case the concept isn't already confusing enough, units can (and often are) be subject to multiple layers of droop speed control that may or may not be coordinated well based on regional grid requirements, grid code, and operational requirements. There's AGC, the plant or power block level, enhanced transient stability (ETS) implemented at the unit level that biases the setpoint from the AGC, and DCS systems with load frequency control (LFC) implemented at... whatever level they decide on that day. These layers "should" be coordinated--meaning they should all work together to achieve the same result and leverage the unit/block/plant appropriately to return the grid frequency to what it should be. They are not always coordinated, and can have different droop/deadband settings, and even different speed reference parameters.

Long story short, even with a comprehensive understanding of the concepts... a complete and total understanding of everything CSA said... there's no guarantee that it's going to work that way in real life.