Hello All !!

Could someone, please help me out by explaing the role of FSRMIN & FSRMINN & FSRN?
In many threads regarding FSRMIN FSRMINN and FSRN, engineers have asked the queries as why the mentioned outputs are not working in their sites as they have understood the basics. While I am asking the basic working of the same. Any explanation will be grateful and helpful.

When I refer to the HELP block of FSRMINV2, what I understand is

> FSKMINN & FSKMIND constants comes in line to calculate FSRMINN when L83MIND i.e L94X / L14HS becomes TRUE w.r.t TNHCOR, which is clear.
> FSKMINU constants comes in line to calculate FSRMINN when L83MIND i.e L94X / L14HS becomes FALSE w.r.t TNHCOR.
Q1. What is FSKMINU (FSKMIN FSR corner (FSRSU). How does it contribute to FSRMINN and work?
Q2. Why there is only 2 array value, [1] & [3]?
Q3. Is value of [3] is taken from FSKMIND array values? If so why?

FSRMIN - Minimum Fuel Stroke reference
Q1. Why it is derived from FSRMINN?
Q2. Why this value is alone multiplied with CQTC( corrected air temp factor) ?
Q3. What & when it contribute ?

FSRN- Droop Speed Control FSR
Q1. What is FSRN (Droop speed control FSR) ? I know mathematically it is [(TNR-TNH) x FSKRN2)] +[FSKRN1 + TLC_CMP]
Q2. What is transient load Bias , TLC_CMP? What doe it do?
Q3. In one of the thread, I read that TNR increases when load is increased and reaches 104% (droop set -point) when machine is at Baseload. How and Why? Does it proportionally increase?
Q4. TNH tries to catch up its reference TNR. I know TNH is restricted by due to frequency that has to be maintained (50Hz/60Hz). As far as i understand a Load increase means increase in Rotor current > increase in magnetic field tries to stop the rotor and is in-turn fuel valve opens further to increase energy to maintain the speed. Then why does TNR increases and load increases while the speed has to be maintained at 100%, while the more energy is given by introducing more fuel ?
Q5. Why there is a DROOP set point and why it is required?
Q6. What is role of FSRN when the generator breaker opens?

Thanks
Suds

 

@Suds,

I don't have access to the blocks you are working with, though I will try to look up app code with FSRMINV2 to see if I can help with that.

I have a suspicion that these questions are related to someone working on either a simulator/simulation program, OR someone trying to put together a control system for controlling a GE-design heavy duty gas turbine. I will try to help with the Droop Speed control thing because as others have written on Control.com: Most of the material regarding Droop Speed control in textbooks, reference books and even manufacturers' white papers (and--YES--I'm looking directly at Woodward Governor when I say this) is pure drivel. Why? Because they FAIL to mention all of the conditions for any of their statements.

First of all, do you know and understand what proportional control is? (It's the "P" part of PID control (Proportional/Integral/Differential control).

Second, do you understand the basic formula for a straight line (on a graph of x vs. y)? When I learned the formula for a straight line in the fifth grade (when I was 11 years old) it was expressed as:

y=mx+b

where: y equals the value of the y-axis component of the straight line
m equals the slope of the straight line
x equals the value of the x-axis component of the straight line
b equals the y-intercept of the straight line (ALSO often expressed as the "offset" of the formula for the straight line)

Basically this states that as the value of x changes the value of y will change based on the magnitude and sign (positive or negative) of values of m and/or b.

Sometimes, the formula for a straight line is expressed as f(x)=mx+b, or something similar. But it's all the same.

Proportional control means that as the value of x changes the value of y will change by an amount determined by the values (and signs) of m and b. And, if the value of x doesn't change, then the value of y won't change. There's nothing in the formula that tries to "zero out" the error between x and y. That's proportional control. If the value of x changes, the value of y will change by the same amount every time the value of x changes. It's really important to understand this concept. It's very simple: it's all about a straight line--no exponential stuff, no integral stuff, no differential stuff. It's about what's happening RIGHT NOW (in the present) as the value of x changes--or remains the same (a change of 0).

[Integral control is about past event(s). Differential control is about future events. And both of these types of control are about returning a process variable (speed; temperature; pressure; flow; etc.) to a setpoint (reference) value. If something in the process causes the process variable to drift from the setpoint integral and/or differential control can be used to help the process variable to return relatively smoothly to the setpoint valve, thereby making the process relatively stable. Proportional is about the present (the difference between the process variable and the setpoint at any given point in time). Integral is about what happened when the process variable deviated from the setpoint in the past and how the process variable can be returned to the setpoint smoothly. Differential is about how the process variable will deviate from the setpoint in the future as the result of some change in the process and how the process variable can be used to anticipate the error between the process variable and the setpoint and keep them from drifting apart significantly. Proportional control does NOTHING to return the process variable to the setpoint. wikipedia.org has a very good description of PID control theory, but it is a little heavy on maths for my taste.]

For a practical example of straight line maths, think of a 4-20 mA simulator. For the overwhelming majority of them the mA output of the transmitter is supposed to change in a linear manner as the parameter being sensed changes (pressure; temperature; flow; etc.). And instead of starting at 0mA and increasing to 20 mA as the sensed parameter changes over a specified range, it starts at 4.0 mA and increases to 20.0 mA as the sensed parameter changes over its specified range. Let's say the sensed parameter is pressure, and the pressure of the process can vary from 0.0 to 100.0 psig. This means the output of the transmitter will change from 4.0 mA to 20.0 mA as the pressure of the process changes from 0.0 psig to100.0 psig. Further, this means that for every change of 25 psig in the pressure of the process the transmitter output will change by 4.0 mA, or 25% of the range of the transmitter output. At 0.0 psig, the output of the pressure transmitter will be 4.0 mA; at 25 psig, the output of the transmitter will be 8.0 mA (4.0 mA + 4.0 mA). If the process pressure changes to 37.5 psig, then the pressure transmitter output will change to 10.0 mA (which is 37.5% of 100 psig--the specified range of the pressure being monitored by the pressure transmitter). In other words, the pressure transmitter output changes by a set amount for every change in the pressure being monitored. For every 25 psig change in pressure, the transmitter output will change by 25% of its output range. Whether the pressure increases or decreases. And the pressure transmitter ITSELF does NOTHING to change the process pressure being measured. If we plot process pressure on the x-axis of a graph and the transmitter output current on the y-axis of a graph, for every 25 psig change in process pressure the transmitter will change by 25% (4 .0 mA). AND, the graph crosses the y-axis at 4.0 mA (the y-intercept, or, the offset). So the slope is 4.0 mA per 25 psig--the pressure transmitter output will change by 4.0 mA for every 25 psig change in the process pressure. That's a straight line and its slope (often called "gain") is: m equals 4.0 mA/25 psig, and its offset is b equals 4.0 mA.

The next thing one MUST understand AND remember about AC (Alternating Current) generation and power systems is that the frequency of the system is EXTREMELY important. And, EVERY synchronous generator synchronized to (connected to; operating in parallel with) an AC power system MUST all be producing power AT THE SAME FREQUENCY. (That's why the process of synchronization is SO CRITICAL!) Multiple synchronous generators operating in parallel with each other (synchronized together) are really like one single synchronous generator and the sum of all the motors and lights and computers and computer monitors and air conditioners and space heaters and tea kettles and hair dryers and vibrators is like one single load to the multiple synchronous generators and the prime movers driving them acting as a single generator and prime mover. It's not possible for one generator synchronized to a 50 Hz grid to be running at 49.6 Hz and another one running at 49.89 Hz and another one running at 50.0 Hz and another one running at 50.27 Hz and another one running at 51.3 Hz. To be operating in parallel with each other EVERY synchronous generator synchronized to that 50 Hz grid MUST be running at the same frequency as every other synchronous generator synchronized to that grid. That's the law--not a legal law, but a physical law. Period. Full stop. (So if the synchronous generator at your site is synchronized to a grid (of any size!) with one or more synchronous generators and your synchronous generator frequency suddenly starts changing or oscillating then every other synchronous generator it is synchronized with is ALSO changing frequency at the same time. Full stop. Period. Has to be. Can't be any other way. That's the facts.)

Please let us know how familiar you are with proportional control, and how familiar you are with the formula for a straight line.

To be continued (with some feedback from @Suds).

 

It’s been explained many times before on Control.com: the speed of a synchronous generator and its prime mover when they are mechanically coupled together and synchronized to a power grid with other prime movers and their generators is “controlled” by the frequency of the system it is synchronized to. Full stop. Period.

When one wants to increase the electrical load of a synchronous generator one does so by increasing the prime mover speed reference. But the synchronous generator and the prime mover speed CAN’T change. TNH is therefore relatively constant, so the only variable in the droop speed equation is TNR. (Ignoring for the moment that TLC_CMP value/variable.)

Most gas turbines have a droop regulation setting of 4.0 percent. This means that when the turbine speed reference increases from 100.0% to approximately 104.0% (on a machine in new and clean condition operating at rated ambient temperature and pressure) the machine will produce rated power. The turbine produces torque which is applied to the generator rotor, and the generator converts the torque into amperes and those amperes are delivered to the power grid the machine is synchronized to and the motors connected to the power system convert the amperes into torque to drive pumps and fans and air conditioners.

But on a well-regulated power grid all of this happens at a relatively constant frequency and machine speed(s). Because the generator has two magnetic fields at work inside it and the magnetic power created by the current flowing out to the power grid is strong enough to keep the generator spinning at a relatively constant speed regardless of the amount of torque applied to it by the prime mover (gas turbine in this thread discussion). In fact, if the fuel supply to the turbine is shut off and the generator breaker connecting the synchronous generator to the grid doesn’t open amperes from other synchronous generators and their prime movers will flow INTO the synchronous generator and KEEP it spinning at synchronous speed.

So, here’s where that straight line formula comes into play. If a machine (the TURBINE—NOT the synchronous generator!) is rated at 100 MW, and the turbine control system has a droop regulation setting of four percent then when the turbine speed reference is approximately 104% (for a machine in new and clean condition operating at rated ambient temperature and pressure) the power output of the machine will be at 100 MW. And, for every 1.0% change in turbine speed reference between 100.0% and 104.0% the power output of the machine will change by 25 MW (one quarter of a possible 4.0% change).

By now you should have realized that when TNR is approximately 104% and TNH is running relatively continuously at 100% this represents a 4% difference, also called an error. Because the machine is trying run at a speed faster than it can so there is an error between the turbine speed reference and the actual turbine. In the GE system, the actual speed is below, or less than, the turbine speed reference. And the magnitude of the error or difference between the reference speed and the actual determines how much fuel will be sent to the prime mover which determines how much torque the prime mover will apply to the synchronous generator rotor which determines how many amperes (how many MW) the machine is producing.

Said in another way—the turbine actual speed is drooping below the turbine speed reference, the actual is less than the turbine speed reference. And the amount of the error (the magnitude of the error) determines how much fuel will be sent to the turbine and therefore how many MW will be produced.

That’s Droop Speed control in a very simple way. And you know what? EVERY synchronous generator and it’s prime mover has a very similar control scheme which allows them all to be synchronized together and act as one huge machine to supply load(s) that are bigger than any single machine could possibly power. AND, do so in a stable way—they SHARE load by using Droop Speed control—or something VERY SIMILAR to it.

That’s enough for now.

To sum up for now, the most important aspect of Droop Speed control is that it allows multiple machines to be connected in parallel with each other (synchronized together) to act as one big machine while doing so in a stable, even predictable manner. The amount of energy flowing into the prime mover—which determines the amount of power produced by the synchronous generator—is controlled in a stable manner which makes it possible to have a stable grid made up of many machines all synchronized together.

And all this happens because the speed of the synchronous generators is “controlled” by the frequency of the grid. And when there are many machines synchronized together on a grid the grid is inherently more stable—especially when all the machines are contributing power in a stable manner.

All because the actual turbine speed can’t change when the turbine speed reference does change. Because the machines are synchronized together—all operating at the same frequency.

And, this has been the way multiple synchronous machines have been operated since the beginning of AC electrical power production. And because there are so many machines of various ages and types of control systems that are STILL operating in synchronism we still do the same thing today—because it still works and for the other aspect of Droop Speed control it is also still necessary (supporting grid operation when the frequency is lower or higher than normal) it will probably continue to be used for a very long time. (Think of how much time and money it would cost to implement a different control scheme for all the existing machines in the world…)

 

It’s been explained many times before on Control.com: the speed of a synchronous generator and its prime mover when they are mechanically coupled together and synchronized to a power grid with other prime movers and their generators is “controlled” by the frequency of the system it is synchronized to. Full stop. Period.

When one wants to increase the electrical load of a synchronous generator one does so by increasing the prime mover speed reference. But the synchronous generator and the prime mover speed CAN’T change. TNH is therefore relatively constant, so the only variable in the droop speed equation is TNR. (Ignoring for the moment that TLC_CMP value/variable.)

Most gas turbines have a droop regulation setting of 4.0 percent. This means that when the turbine speed reference increases from 100.0% to approximately 104.0% (on a machine in new and clean condition operating at rated ambient temperature and pressure) the machine will produce rated power. The turbine produces torque which is applied to the generator rotor, and the generator converts the torque into amperes and those amperes are delivered to the power grid the machine is synchronized to and the motors connected to the power system convert the amperes into torque to drive pumps and fans and air conditioners.

But on a well-regulated power grid all of this happens at a relatively constant frequency and machine speed(s). Because the generator has two magnetic fields at work inside it and the magnetic power created by the current flowing out to the power grid is strong enough to keep the generator spinning at a relatively constant speed regardless of the amount of torque applied to it by the prime mover (gas turbine in this thread discussion). In fact, if the fuel supply to the turbine is shut off and the generator breaker connecting the synchronous generator to the grid doesn’t open amperes from other synchronous generators and their prime movers will flow INTO the synchronous generator and KEEP it spinning at synchronous speed.

So, here’s where that straight line formula comes into play. If a machine (the TURBINE—NOT the synchronous generator!) is rated at 100 MW, and the turbine control system has a droop regulation setting of four percent then when the turbine speed reference is approximately 104% (for a machine in new and clean condition operating at rated ambient temperature and pressure) the power output of the machine will be at 100 MW. And, for every 1.0% change in turbine speed reference between 100.0% and 104.0% the power output of the machine will change by 25 MW (one quarter of a possible 4.0% change).

By now you should have realized that when TNR is approximately 104% and TNH is running relatively continuously at 100% this represents a 4% difference, also called an error. Because the machine is trying run at a speed faster than it can so there is an error between the turbine speed reference and the actual turbine. In the GE system, the actual speed is below, or less than, the turbine speed reference. And the magnitude of the error or difference between the reference speed and the actual determines how much fuel will be sent to the prime mover which determines how much torque the prime mover will apply to the synchronous generator rotor which determines how many amperes (how many MW) the machine is producing.

Said in another way—the turbine actual speed is drooping below the turbine speed reference, the actual is less than the turbine speed reference. And the amount of the error (the magnitude of the error) determines how much fuel will be sent to the turbine and therefore how many MW will be produced.

That’s Droop Speed control in a very simple way. And you know what? EVERY synchronous generator and it’s prime mover has a very similar control scheme which allows them all to be synchronized together and act as one huge machine to supply load(s) that are bigger than any single machine could possibly power. AND, do so in a stable way—they SHARE load by using Droop Speed control—or something VERY SIMILAR to it.

That’s enough for now.

To sum up for now, the most important aspect of Droop Speed control is that it allows multiple machines to be connected in parallel with each other (synchronized together) to act as one big machine while doing so in a stable, even predictable manner. The amount of energy flowing into the prime mover—which determines the amount of power produced by the synchronous generator—is controlled in a stable manner which makes it possible to have a stable grid made up of many machines all synchronized together.

And all this happens because the speed of the synchronous generators is “controlled” by the frequency of the grid. And when there are many machines synchronized together on a grid the grid is inherently more stable—especially when all the machines are contributing power in a stable manner.

All because the actual turbine speed can’t change when the turbine speed reference does change. Because the machines are synchronized together—all operating at the same frequency.

And, this has been the way multiple synchronous machines have been operated since the beginning of AC electrical power production. And because there are so many machines of various ages and types of control systems that are STILL operating in synchronism we still do the same thing today—because it still works and for the other aspect of Droop Speed control it is also still necessary (supporting grid operation when the frequency is lower or higher than normal) it will probably continue to be used for a very long time. (Think of how much time and money it would cost to implement a different control scheme for all the existing machines in the world…)

Thank You VERY MUCH !! WTF?, for your beautiful explanation and the efforts taken for it. Today after so many years, the clutter in my mind regarding DROOP has got cleared. I am really happy and have sigh of relief !!

I also have other set of queries in post in terms of GE blocks/terminology. I will be more glad if that also gets cleared.
Thanks once again for the DROOP !!

Regards
Suds

 

@Suds,

Thank you for the kind words. I have read MANY written descriptions and heard MANY explanations of Droop Speed control--from college professors and authors. But, generally, the people writing them don't have any real experience with OPERATING synchronous generators synchronized to a grid. Yes, if you have a single prime mover operating in Droop Speed control driving a synchronous generator powering a load (or loads) and the load increases (for example, because someone starts a pump somewhere on the system the machine is providing power to, the machine actual speed WILL decrease. And if no one does anything and the load continues to increase the actual speed of the machine will decrease. This will cause the frequency of the system powering the load(s) to decrease. If the machine was synchronized with zero load at rated frequency, and the load was increased up to the prime mover's rating, and the machine control system had 4% Droop Regulation the machine speed would decrease to 96% of rated (on a new and clean machine operating at rated ambient temperature and pressure with expected fuel characteristics). But the textbooks and references NEVER say--this only happens for ONE machine supplying a load (or loads) and the operator is unconscious and not paying attention to the machine operation (particularly the frequency). The implication is always this happens for ANY machine operating in Droop Speed control. Which is just plain false. It's not true. If two or more machines were synchronized together and both prime movers were operating in Droop Speed control and the load they were powering increased and the operator was unconscious or not paying attention to the system frequency, the speeds of both machines would decrease. But then there are other issues with the configuration of the machines and/or the operator. When ANY machine is synchronized to a grid supplying power to a load (or loads) with other machines they all act as one machine--one prime mover driving one generator. In almost all cases for medium to large grids around the world most every machine is operating in Droop Speed control. Smaller systems will usually have ONE machine operating in Isochronous Speed control and any other machines synchronized to the system will be operating in Droop Speed control--or some version of Droop Speed control (even Isochronous Load Sharing is just a version of Droop Speed control).

Sorry; I haven't had an opportunity to get to some app code with FSRMINV2 but I will keep trying to find some time in the next couple of days.

You wrote:
" Q1. What is FSRN (Droop speed control FSR) ? I know mathematically it is [(TNR-TNH) x FSKRN2)] +[FSKRN1 + TLC_CMP] "

As has been written many times on Control.com (so many times I wonder if the site should be called "DroopSpeedControl.com"...) the basic formula for GE-design heavy duty gas turbines is:

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

FSRN is one of several FSR calculations that go through a selection to actually tell the fuel control valve(s) how much to open (or in the case of liquid fuel how much to close) which determines how much fuel (energy) is sent to the turbine. FSRN is compared to the other FSR calculations and when it’s appropriate FSRN is the controlling calculation that drives the fuel control valve(s). The long name description for FSRN should be something like SPEED CONTROL FSR. And it’s usually the calculated FSR that’s controlling the fuel control valv(s) from FSNL, during synchronization and from the instant the generator breaker closes all the way to Base (or Peak or Peak Reserve) Load—regardless if the machine load is being done manually by operators or automatically by Pre-Selected Load Control or AGC control or by some reference from a DCS or external source—Droop Speed Control is active under normal operating conditions between FSNL and Base Load. (The area between generator breaker closure and Base Load is generally referred to as "Part Load.")

I'm not familiar with the new term "TLC_CMP" but you mentioned it is a Transient Load Bias, which probably means that unless there is some kind of unusual disturbance on the grid to which the machine is synchronized the value of TLC_CMP is probably zero (or some fixed value that doesn't change unless there is some kind of disturbance). This is all just a SWAG (Scientific Wild-Arsed Guess) and for now I'm going to ignore that term as it's some kind of bias and we're really only interested in the basic Droop Speed control formula.

I mentioned the formula for a straight line (y=mx+b) previously. Remember: it's the value of 'y' that we are trying to determine with the equation as the value of 'x' changes. (Again think of a graph of’x’ versus ‘y’ with ‘x’ on the horizontal axis and ‘y’ on the vertical axis.) The value of 'm' term is the slope (steepness) of the straight line and is often referred to as the "gain" of the equation--how much the value of y will change as the value of x changes. The value of 'b' is often referred to as the offset of the straight line (the point at which the straight line intercepts the y-axis). If you're familiar with 4-20 mA transmitters, they also have two basic adjustments: gain (sometimes called "span") and offset. That's because, generally, we want the output of the transmitter to change linearly (in direct proportion to) with the change in the parameter being sensed/measured (pressure; temperature; level; etc.). And that makes the output of the transmitter a straight line--a straight line that changes at a constant "rate" as the value of the parameter being sensed changes. (This was all described above, and it can be found in MANY places on the World Wide Web.)

Let's equate the terms of the basic Droop Speed control formula with the terms of the formula for a straight line:

FSRN = y (the value of Droop Speed control FSR (Fuel Stroke Reference--essentially the value of fuel to flow to the turbine)
(TNR-TNH) = x (the value of the x-axis component of the straight line the ERROR between TNR and TNH)
FSKRN1 = b (the "offset", or the y-intercept of the formula for FSRN/a straight line)
FSKRN2 = m ( the slope of the FSRN "curve"/straight line (even though it's a straight line it's often called a curve))

The BIG difference between a transmitter's output and the Droop Speed control FSR is that the 'x' value is actually two variables: TNR and TNH. We know TNH varies as the machine is started and stopped--when the generator breaker IS NOT closed, and also very slightly during synchronization. And we know that when the generator breaker IS closed, the value of TNH is relatively constant because the speed of the generator rotor (AND the turbine) is fixed (controlled) by the frequency of the grid with which it is synchronized AND which should be relatively constant (at 50 Hz or 60 Hz, plus or minus a few hundredths of a hZ (0.0n hZ)) on a well-regulated grid. If we plot FSRN versus (TNR-TNH) on a graph it will be a straight line rising as (TNR-TNH) increases.

We know that when we want to increase the load on the machine (when the generator breaker is closed) what ends up happening is that the turbine speed reference increases (and when we want to decrease the load on the machine the turbine speed reference decreases). [This happens EVEN THOUGH the operator is looking at the MW meter/value!!! Or, the machine is operating at some load reference (Pre-Selected Load Control, or AGC, etc.). As the operator is clicking on RAISE- or LOWER SPEED/LOAD while watching the MW meter/value] what's happening in the Mark* is the value of TNR is being increased or decreased, respectively. Most operators, and even their supervisors, don't know this is happening "in the background"....] With TNH being "fixed" (relatively constant because, as it should be, the frequency of the grid the machine is synchronized to is relatively constant) when TNR changes the ERROR BETWEEN TNR AND TNH CHANGES. The gain (slope) and offset (y-intercept) values are constant (literally--they are Control Constants!) and not changing, so the only thing that changes when TNR changes and the error between TNR and TNH changes is FSRN--by amounts equal to FSKRN1 and FSKRN2 as per the Droop Speed control (FSRN) formula. (The engineering units for FSKRN1 is percent FSR; and often FSKRN1 is called the "FSNL FSR" because it usually is the FSR at FSNL, on a new and clean machine operating at rated ambient temperature and atmospheric pressure and with expected fuel characteristics (BTU; methane content; ULV; OHV; etc.).

It's that simple--really, it is, just that simple. And because FSKRN1 and FSKRN2 are fixed values of Control Constants they don’t vary (change). There is no polynomial term, no exponents, making the value of FSRN have a shape OTHER THAN a straight line rising up as the value of the error between TNR and TNH increases (when the machine load is being commanded to increase). [NOTE: The 'x' axis value of speed is alternately TNH or TNR--it's TNR below FSNL (100% turbine speed) and it's TNR above FSNL when the generator breaker is closed.

The value of FSRN will change by the amount specified by the “gain” term value (FSKRN1), which has the engineering units of percent FSR per percent speed (the speed error between TNR and TNH). For a machine with 4% Droop Regulation (Control) that means the fuel flow-rate when the generator breaker is closed will change by 25% of the loaded fuel flow-rate. It you want to see that you need to look at the EXPECTED FUEL FLOW CHARACTERISTICS in Sect. 05.nn (05.1n for liquid fuel, and 05.2n for gas fuel). You can also plot the values of expected fuel flow characteristics on graph paper for a better "view" of what's happening. NOTE: There are expected fuel flow-rates for starting and for loaded operation. NOTE: The EXPECTED FUEL FLOW CHARACTERISTICS are just that: EXPECTED. They are calculated values based on the expected fuel supply characteristics given to the packager during the ordering and purchasing phase and are based on "iso" conditions: turbine nameplate rated ambient temperature and atmospheric pressure and elevation, AND actual fuel supply characteristics do vary over time as do ambient conditions (temperature and atmospheric conditions). Compressor cleanliness, IGV LVDT calibration, IGV cleanliness, turbine inlet air filter cleanliness, internal clearances, turbine nozzle- and bucket conditions ALL also affect the actual fuel flow characteristics. SO discovering this "hidden" gem of information IS NO GUARANTEE of output for anything other a machine in new and clean condition, operating at rated ambient and atmospheric pressure with the fuel supply characteristics given to the machine packager during the ordering and purchasing phase. In other words, in general, they are just a "guideline" and not gospel or any part of any performance/efficiency guarantee. Full stop. Period. End of this part of the discussion.

If the machine were operating at 50% of rated load, TNR would be 102% speed. However if the generator breaker were to suddenly be opened (tripped or manually opened) and the Mark* didn't decrease fuel flow VERY quickly the machine speed would increase WELL ABOVE 102% because the Droop Regulation (control) setting is just for controlling fuel flow--NOT for controlling actual speed (TNH). (The Mark* DOES detect a generator breaker open condition at loaded operation and QUICKLY changes TNR to be 100% (sometimes 100.3% in order to prevent a turbine overspeed trip.) Looking at the Expected Fuel Flow Characteristics values one can see that fuel flow increase much more when the generator breaker is closed (and actual speed isn't changing).

One more thing I want to set the record straight on: As load is increased it is true that the current flowing in the generator rotor (the "field") does change as the machine is loaded--but it does so to maintain a relatively constant generator terminal voltage. Synchronous generator output terminal voltage doesn't usually vary much--between =/-5% of generator nameplate voltage rating. The reason generator terminal voltage must increase as generator load (stator current) increases the strength of its magnetic field increases--a lot. And that causes the generator voltage to decrease slightly because of the interaction between the two magnetic fields (the one on the generator rotor and the one on the stator). So, the generator rotor current doesn't increase the generator rotor current increases--the generator rotor current increases to keep the generator terminal voltage relatively stable. Generator rotor current controls generator terminal voltage--and generator reactive current flow (sometimes called "reactive" power--even though it does no real work, it does make it possible for real work to be done...) and if nothing is done to increase the generator terminal load increases (as the generator stator current increases because the torque being applied to the generator rotor is increasing (TRYING to make the generator spin faster as TNR increases which increases the ERROR between TNR and TNH). That extra torque (resulting from the fuel flow increase to the turbine) CAN'T make the generator spin faster and the generator converts that torque into generator stator CURRENT. The basic formula for electric power is WATTS=VOLTAGE*CURRENT, Generator terminal voltage (which doesn't change much (less than+/-5% usually, out of thousands of volts!) so to make more WATTS (or MW) the current value of the power formula has to increase and that's what the generator does when it's converting torque that's trying to make it spin faster than it can into generator stator current. Even though generator terminal volts are variable (usually only by no more than +/-5%) AND generator stator current can vary, if one wants to make more WATTS (or MW) one has to increase generator stator amps. And to do that, one increases the fuel flow to the turbine which increases the torque applied to the generator which TRIES to make the generator rotor spin faster than it can and the generator converts the "extra" torque (above what's required to maintain FSNL) to generator stator current--which causes the WATTS (or MW) of the machine output to increase.

Okay; enough for today. (And probably tomorrow, too.)

CAUTION: This is all GREAT stuff (as you have noted)--understanding how Droop Speed control works, but it's just the tip of the iceberg--there's a LOT more to Droop Speed control (because of what happens when TNH actually changes as it does during many grid disturbances!!!) and I can assure you that the Mark* turbine control system is very well tested and proven. And that's one of the really great things about the Mark*: They have basically kept doing the same thing (with newer and newer electronics) for decades--since the advent of the Mark* IV in the 1980's, and even before (though they used different analog electronics and methods of depicting the control methods and sequencing). So, as some others before you have done--don't let his knowledge go to your head. It should just relieve some nagging questions (I really hate to call them "doubts") and fill in some blanks for you (and possible others reading this thread) and correct some of the really bad explanations of Droop Speed control we've all heard and read--and start thinking that this problem or that problem can be attributed to a fault in the Mark*. It would be highly unusual if it were the case, and I would suspect some people would have been "tweaking" Control Constants. This explanation is also NOT intended to give anyone the ability to change Droop Speed control Control Constants--because they affect MANY parts of turbine operation, control and protection. Unless you really understand all of the knock-on effects of changing Control Constants--and in particular THESE Control Constants--JUST DON'T DO IT. I'm happy to help fill in the blanks and try to help people understand how these basic control schemes work, but the information one can glean from these explanations and descriptions should NEVER be used without a much deeper understanding of the turbine control and protection schemes and philosophies of the GE Co (or GE Vernova as they're now called I think) before changing Control Constants without the review and assistance of a qualified, knowledgeable person or persons.

Tchau!

 

@Suds,

FSRMIN is one of the several FSRs that are compared to select the proper value of FSR for the current operating conditions. It has several functions, which makes it a real bear to properly adjust and sometimes it does need to be adjusted (like when fuel supply characteristics change).

The primary purpose of FSRMIN is to prevent the flame from being extinguished should conditions result in FSRMIN being the selected value for fuel control when the machine is operating. For example, let's say the machine was running at 48.9MW when a generator protective relay tripped the generator breaker--not the turbine, just the generator breaker. Further, lets say the machine was rated at 111.7MW, meaning it running just below 50% of rated power when the generator breaker was opened (tripped).

The Mark* is going to recognize the very sudden loss of load and very sudden rate of speed increase and the opening of the generator breaker and it's going to reduce fuel to a very low level while still trying to maintain flame in most of the combustors (not all of them perhaps, but most of them). It does this by selecting the current calculated value of FSRMIN which was determined using empirical data for similar machines running on similar fuels under similar conditions and is supposed to prevent the flow of air through the axial compressor and into the combustors from extinguishing all of the flame in the majority of the combustors--essentially tripping the turbine (it will actually be tripped by LOSS OF FLAME, a condition that occurs when flame is lost in a majority of the combustors with flame detectors but no emergency condition was detected prior to the loss of flame (such as low-low oil pressure, or high-high oil temperature, or high-high vibration, etc.)).

Some--but NOT ALL--machines may have certain contractual conditions which require the machine to be capable of re-synchronizing to the grid in a short period of time (say, 2 minutes) after a breaker open event (such as described above). This is usually necessary for power systems (grids) that are highly loaded and don't have much spinning reserve available if a machine trips off-line suddenly as in the example. So, FSRMIN has to be set/adjusted to allow the machine to maintain flame while not tripping on overspeed or underspeed and be resynchronized to take some or full load to help maintain the grid stability or to return stability to the grid.

The issue with the way FSRMIN was used to do this is that FSRMIN has also been used for several other control functions (including starting and shutdown, if I recall correctly) and so adjusting FSRMIN to maintain flame on a breaker open event can have negative implications on starting and shutdown. It's a very tough balance and it takes several tests to determine the proper value (which can't really be calculated in the field without a LOT of data for which there usually aren't proper sensors to gather) and not result in tripping the machine on LOSS OF FLAME or preventing proper acceleration or achieving a proper fired shutdown. If it's not understood that FSRMIN has many uses it can be a mess, a real mess, and a costly mess (in terms of lost generation and testing, which requires fuel and usually results in a few trips of the turbine).

I don't recommend spending a lot of time researching FSRMIN without a really good need to know. It's NOT just how it's calculated in the application code, it's how it's used in other parts of the application code and sequences and processes. If you believe the machine is having problems related to FSRMIN I would strongly recommend having a knowledgeable, qualified person come to site and work with you to gather information and make a plan for possibly making some adjustments. I have had to have factory engineer support on a couple of occasions that required some very intricate sequencing changes to implement in order to keep all the functionality of FSRMIN working properly. As I say, it's been "extended" to some functions it probably shouldn't been extended to. It usually works fine for the majority of machines out there, but there is the odd combinations of factors that can require some creative and intricate modifications to implement without knock-on effects.

Hopefully this will help you with your desire to understand speed control and various FSRs (FSRMIN, FSRSU, FSRSD, FSRN, FSRT, FSRACC--just find where the signal FSR is written to and you should find it's the output of a MINIMUM SELECT function that looks at multiple calculated values of FSR and chooses the lowest (minimum) value and writes that to FSR to control and protect the machine). Again, if you believe there are issues possibly related to the value of FSRMIN, please try to explain them here. BUT, be prepared--I and/or others will probably ask for Trender data...! Because we have learned that using anecdotal information (the operator reported this or that was happening but they have no actual, recorded, actionable data to support the claim(s)) can be a long, deep and disappointing rabbit hole to go down. Actionable data is the BEST data to have and to analyze and try to explain or find to be a possible cause of a reported problem. (And know this, too: MANY reported problems with GE-design heavy duty gas turbine operations are perceived problems--meaning that someone (a technician, an operator, a supervisor, even the plant manager or the site owner) firmly and strongly believes that the machine should do THIS or THAT and it's NOT doing this or that (or SHOULDN'T do this or that and it is doing this or that). An excellent example of this is people thinking that when the grid is having frequency instability problems that the machine(s) at THEIR site should and MUST remain stable while the grid frequency is unstable. And that's just NOT TRUE, as we have already agreed (we DID agree on that, correct???). These people believe--without really giving it much thought, actually, or without really understanding some of the very basic fundamentals of AC (Alternating Current) power generation--that their machine can run at a different frequency while synchronized to a power system/grid with other machines (which we have agreed is not possible--because we DID agree on that, right???). And, so they scream and curse and make a field service person travel to site to fix this "problem" which usually ends in a one or more shouting matches because one or more people just BELIEVE they are correct and won't admit that they were wrong. (Ask me how I know this....))

 

Some kind of important corrections:

"[NOTE: The 'x' axis value of speed is alternately TNH or TNR--it's TNR below FSNL (100% turbine speed) and it's TNR above FSNL when the generator breaker is closed." It should have read, " [NOTE: The 'x' axis value of speed is alternately TNH or TNR--it's TNH below FSNL (100% turbine speed) and it's TNR above FSNL when the generator breaker is closed."

"The value of FSRN will change by the amount specified by the “gain” term value (FSKRN1), which has the engineering units of percent FSR per percent speed (the speed error between TNR and TNH). " It should have read: "The value of FSRN will change by the amount specified by the “gain” term value (FSKRN2), which has the engineering units of percent FSR per percent speed (the speed error between TNR and TNH)."

There are other needed corrections, but these jumped out at when I re-read it today. If you have other issues let me know. It's too late for me to correct the original post (I only get 24 hours or so to do that), but I can post corrections to the thread which help some people.

 

@WTF? , Again heartful thanks for taking so much Efforts and TIME for explaining, and the dedication to write even after reviewing the post after posting it . My respects !!

y=mX+C itself has cleared and made me corelate mathematically !!

And as mentioned by You, "this only the Tip of the Iceberg", and it is. Understanding this basic thing was important and this will make dive more deeper and definitely I will come with more queries.

My apologies for delayed reply.

Thank You
Suds

 

@Suds,

Thanks again for the kind words; they are very much appreciated.

I would like to finish up with a couple of final things about "basic" Droop Speed control before we move on to the "other" function of Droop Speed control. If you still have questions about FSRMIN and the other FSRs (you can see them on the FSR bargraph display on the HMI) please ask them here, or in another thread as this one has become mostly about FSRN (Speed Control FSR).

The amount of electric power made by a synchronous generator and its prime mover (in this thread a GE-design heavy duty gas turbine) is directly related to the amount of energy (fuel) that flows into the combustors. As we have seen in our discussion about basic Droop Speed Control between FSNL and Base Load (CPD- or CPR-biased exhaust temperature control--FSRT) the amount of fuel flowing into the machine is a function of the error between the turbine speed reference (TNR) and the actual speed of the machine (TNH). We know that under normal operating conditions when the machine is synchronized to a grid with other synchronous generators and their prime movers that the grid frequency is stable and while it varies slightly above and below the rated grid frequency it remains very close to rated (50 Hz or 60 Hz) so TNH remains very close to the machine's rated speed.

It's also desirable for the machine to produce electric power at a steady and stable rate except when the amount of power it produces is being increased (raised) or decreased (lowered). Since the amount of power being produced between FSNL and Base Load is a function of the error TNR and TNH and since TNH is stable under normal operating conditions to change load (raise or lower; increase or decrease) TNR must change. However, when the load is to remain at some desired value (say 50MW out of a machine rated at 100MW) this means that TNR must remain stable so that the amount of fuel flowing into the machine remains stable and the electrical power out of the machine remains stable at the desired value. So, except when load is changing (raising or lowering; increasing or decreasing) the value of TNR can be seen to be stable and not changing, which means the error between TNR and TNH is stable and not changing which means the amount of fuel flow is stable and not changing which means the electric power output of the machine is stable and not changing. And THIS is VERY important for machines synchronized to a grid with other machines--that their electric power output is stable. This means the machine (prime mover and generator) is a stable participant in "sharing the load" of the grid with the other machines (prime movers and generators) also synchronized to the same grid. This is very important--stable load and sharing "nicely" in providing the necessary power to supply the load(s) connected to the grid. And it happens because BOTH TNH and TNR are "constant" and not changing (unless the operator or some other control system is changing load). TNH is stable and constant because the frequency of the grid the machine is synchronized to is (should be) stable and relatively constant (it varies slightly above and below the rated frequency of the grid but usually just hundredths of a Hz on a large grid/power system). And TNR is stable and "constant" because the machine is being operated at some desired MW setpoint for some period of time until it's necessary to change load or stop the machine at which time TNR will be decreased (lowered) in the process of stopping the machine.

It's VERY important to understand this concept of normal operation at Part Load (between FSNL/generator breaker closure and Base Load): TNR is raised or lowered (increased or decreased) to change load, but when the machine is to be producing power at a particular value TNR is not changing. When a machine with 4% droop regulation rated at 100MW is to be producing power at 50MW TNR will be at approximately 102% when actual fuel characteristics closely match the expected fuel characteristics used to build the machine--and TNR will remain at approximately 102% as long as the electrical output remains at 50MW.

Just as a prelude to discussing the "other" aspect of Droop Speed control, try to imagine what would happen if a machine rated at 100MW is producing 50MW with a TNR of 102% and the frequency of the grid the machine is synchronized to starts changing by a large amount (tenths of a Hz)--which means the error between TNR and TNH will change. Not because TNR is changing, but because TNH is changing. And, if the error between TNR and TNH is changing that means the amount of fuel being sent to the machine will be changing which means the electrical output of the machine will be changing--not because of anything an operator did, but because the error between TNR and TNH is changing. And in this case TNR is not changing but TNH IS changing. Again, Droop Speed control is about the change in fuel (which changes the load) as the error TNR and TNH changes. FSRN=((TNR-TNH)*FSKRN2)+FSKRN1. FSKRN1 and FSKRN2 are Control Constants which means they don't change as the machine is operating. If EITHER TNR or TNH changes FSRN will change which will change the amount of fuel flowing into the machine which will change the electric output of the machine.

Try to imagine what would happen if the grid frequency was changing very quickly, going above and below 100% by several tenths of a Hz. In the case of a 50 Hz machine and grid, a change of 0.5 Hz corresponds to a 1% change in speed! If the grid frequency is oscillating between 49.75 Hz and 50.25 Hz the error between TNR and TNH is changing by a total amount of 0.5%--which is 12.5% of rated power (in our example of a machine rated at 100MW which was operating at 50MW before the grid frequency started changing the output of the machine will change by plus-or-minus 12.5MW (for a total of 25MW!) as the grid frequency changes by plus-or-minus 0.25 Hz.

There's NOTHING an operator can do to stop this changing of the load. Even enabling Pre-Selected Load Control with a reference of 50MW can't stop it from happening. AND--remember if the machine at your site is experiencing grid frequency changes EVERY machine synchronized to the same grid is ALSO experiencing grid frequency changes. In fact, the grid operators WANT this to happen because it actually helps to try to keep the grid frequency from drifting further and further away from rated grid frequency. (Of course, they don't want the grid frequency to be oscillating above and below grid frequency--that's another phenomenon that is dealt with using another method--not an easy method, by the way.)

Forgetting for a moment an oscillating grid frequency (one that is changing and is not stable) when the grid frequency starts to decrease from rated that means that TNH starts to decrease. Let's say a machine rated at 100MW was operating at 50MW (a TNR of Approximately 102% before the grid frequency decreased and that TNH decreases by 0.5 Hz from 50 Hz, this means TNH will decrease from 100% to 99%. This will mean the error TNR (at 102%) and TNH (now at 99%) will INCREASE from 2% (102%-100%) to 3%, which means the electric power output of the machine INCREASES by 1% to 75MW. In fact, ALL machines operating on Droop Speed control will see their electric power output INCREASE by an amount proportional to the 1% change in the error between their machine speed reference and their machine's actual speed.

[IMPORTANT NOTE: This ONLY occurs for machines operating on Droop Speed control (so for GE-design heavy duty gas turbines machines NOT at Base Load with Base Load control enabled and active) and not already at or near rated electric power output. Machines AT Base Load with Base Load control enabled and active and already producing near rated electric power output cannot respond appropriately or similarly as machines operating on Droop Speed control. And if a machine is very near rated power output (say 90MW or 95MW out of 100MW) it's power output can't increase to more than rated power output (unless the machine has some special algorithms/code in the control system). However ALL machines synchronized to the grid WILL experience a decrease in actual machine speed and generator frequency at the same time! It's just that machines already producing rated output or near rated output can't respond to a "large" grid frequency disturbance as machines operating on Droop Speed control at part load might be able to respond. BUT, they will ALL experience the SAME frequency deviations. And MOST will also experience electric power deviations of some magnitude.]

So, this is a little bit about the "other" aspect of Droop Speed control--actually, a very important aspect of Droop Speed control. One that grid operators and regulators rely upon to help support grid stability during times when the frequency is much less or much more than rated. MANY people confuse the two aspects and the importance of one or the other, but because the error determines the change in energy flow-rate into the machine and the error can change when EITHER TNR or TNH changes and because when either TNR or TNH changes the other variable (TNH or TNR, respectively) is relatively stable (or should be!) the error will change which will change the electric power output of the machine. Droop Speed control is all about the error between two variables: If one variable changes with respect to the other the error will change which changes the energy flow-rate which changes the electric output of the machine. When the electric output of the machine is stable, neither variable (TNR or TNH) is changing and the energy flow-rate into the machine is stable and the electric output of the machine is stable.

Since almost EVERY machine synchronized to any alternating current grid or power system anywhere on Planet Earth uses some form of Droop Speed control almost EVERY machine's reaction to changes in the error between the (speed) reference or the machine's actual speed can be calculated and understood. AND, most every machine will respond "share" in providing the electric power necessary to supply the load(s) which are connected to the grid/power system the machines are synchronized to. It's been this way since almost the beginning of AC electrical power production. And consider this--all of the machines synchronized to the same grid and using Droop Speed control at Part Load (less than rated power output) "communicate" with each other by sensing the changes in grid frequency! They don't require separate networks and wires to do that--it's all done by sensing the machine speed (generator (grid) frequency) when they are synchronized together. It's downright ingenious!

And, it's also very misunderstood and most electric machine operators don't know what is happening when they are watching the MW meter/display as they are changing the machine's electric power output, or when the machine's electric power output is stable at some desired value. And even supervisors and power plant managers and power plant engineers and power plant owners don't really understand Droop Speed control, or be able to explain it to others so they can understand it. It's just accepted--like closing the light switch on the wall will cause the electric light in the room to be illuminated.

Droop Speed control is all about the error. And in a GE-design heavy duty gas turbine and synchronous with a GE Mark* turbine control system that error is used to control the amount of fuel entering the machine through FSRN. A stable constant power output of a GE-design heavy duty gas turbine-generator occurs when neither of the two variables, TNR or TNH is changing. When the machine is synchronized to a well-regulated grid TNH is relatively stable and constant. But when TNH is changing (as grid frequency changes) even if TNR is stable and constant the error will change so the electric power output of the machine will change as grid frequency changes (and there's NOTHING an operator or control system can do about that!) when the machine is running at Part Load on Droop Speed control. AND, when the frequency of the grid the machine is synchronized to is changing, EVERY MACHINE on that grid will experience the same frequency changes and deviations--because they are synchronized to the grid with other machines. That word--synchronized--is very important. VERY important, and very misunderstood and overlooked. Remember: Multiple synchronous generators and their prime movers synchronized to a grid/power system act as ONE SINGLE generator because they are synchronized together. No machine can run at synchronous speeds faster or slow than the grid frequency permits. NOT ONE. Not possible. Not no how; not no way. No.

Food for thought.

 

Great info here, thanks.

one question though, it was mentioned above that selecting “pre-select load” wouldn’t interfere with the machine’s droop response to grid frequency disturbances.
Why is that? I assumed that pre-select load would be looking at DWATT and trying to maintain a specific generation level regardless of grid frequency whereas “load-clear” would be inferring a generation level based on speed reference error, allowing the machine to react to both setpoint changes and grid frequency disturbances.
Wouldn’t pre-select load fight the typically expected droop response by ramping the machine back to its original generation level on an over-frequency grid when other, droop enabled machine would be cutting back?

thanks

 

@jeff3300,

I don't know where you read that but that's completely incorrect. Pre-Selected Load Control, when selected and active DOES interfere with the machine's response to grid frequency disturbances. [UNLESS the machine uses one of GE's workaround software methods--which don't always work well, either.]

What happens when the frequency of the grid/system the machine is synchronized to starts to deviate from the nominal frequency setpoint is that when selected and active and the machine IS NOT at Base Load Pre-Selected Load Control WILL change the fuel flow-rate to try to make the actual load stay equal to the Pre-Selected Load Control reference/setpoint. For example, if the frequency starts to drop then TNH will also start to decrease--but TNR doesn't change. This will increase the error between TNR and TNH and the fuel flow-rate--and load--will start to increase, which is exactly what SHOULD happen. BUT, Pre-Selected Load Control says, "No, No, NO!!!" when the actual machine load increases and it then reduces TNR to try to make the actual load return to the Pre-Selected Load Control setpoint. Which is the exact OPPOSITE of what should be happening during a frequency decrease. And the opposite happens if the grid/system frequency increases--the actual load will start to decrease (because the error between TNR and TNH decreased when the grid/system frequency increased) but Pre-Selected Load Control will increase TNR to make the actual load remain at the Pre-Selected Load Control reference--again exactly the OPPOSITE of what should happen during a frequency increase.

Now, imagine what happens if the grid/system frequency is oscillating above and below nominal--then Pre-Selected Load Control will be doing the exact opposite each time the grid/system frequency deviates from nominal. And THAT will make grid frequency deviations worse instead of better.

Pre-Selected Load Control should ONLY be used to change load, say from 23.5 MW to 40 MW, or from 67 MW down to 15 MW. Once the desired load is achieved the operator should then click on either RAISE SPEED/LOAD or LOWER SPEED/LOAD to cancel Pre-Selected Load Control. TNR will remain stable at the new load and so will load--remain stable at the new value magically without Pre-Selected Load Control keeping the load stable during steady-state operation. The use of Pre-Selected Load Control is SO PERVASIVE that most operators WILL NOT even try operating the machine without it, they're so afraid the machine will start oscillating load and they might lose their job. Even many Operations Supervisors and Plant Managers are terrified of operating the machine without Pre-Selected Load Control selected and active.

Finally, many grid regulators are beginning to understand this little "feature" of GE Mark* turbine control philosophy called Pre-Selected Load Control, and it's my prediction (and fervent hope!) they will ban the use of Pre-Selected Load Control during steady-state operation. Which will force GE to fix their dirty little secret problem once and for all. And grid stability will be much the better for it. (Already, many grid regulators DO NOT ALLOW the use of voltage control or VAr or Power Factor control--unless there are contractual reasons that obligate the use of one of them. Why? Because they can exacerbate grid/system voltage instability issues. It's only a matter of time until Pre-Selected Load Control suffers the same well-deserved fate.)