It has been told in previous threads the equation followed by Mark V is simply
FSRN = (TNR-TNH)*FSKRN2 + FSKRN1

Now we have 2 GTs in our CPP with both being controlled by Mark V. What i want to ask is that does the same equation holds for all three modes viz. isochronous, part-load and preselect mode? And if so then what are the variables and constants in this equation for different modes and how the goal is accomplished in each mode? Do the equation or significance of variables vary with mode? Kindly help on this...

 

It has been told in previous threads that some early Mark V systems used this very simple relationship for calculating Droop Speed Control FSR.

Most Mark V control systems use another method for calculating Droop Speed Control FSR, and that method is generally called 'Constant Settable Droop Speed Control'. This horribly named version of Droop Speed Control uses load (generator load) as part of the calculation. There is a lot of speculation about why the method for calculating Droop Speed Control was changed, but it had a lot to do with standardizing Speedtronic control schemes when DLN (Dry Low NOx) combustion systems came into production. They both do the same thing, just in slightly different ways.

You need to refer to the Control Sequence Programs (CSPs) for your units to determine which version of Droop Speed Control is used in the Mark V turbine control panels at your site.

Now to your statement/question about modes. Droop Speed Control is a governor control mode, as is Isochronous Speed Control mode.

Part-load (partial load, or part load, "load" being defined as nominal, rated power output which occurs at exhaust temperature control) is not a governor control mode; it's a way of describing whether or not a GE-design heavy duty gas turbine is at some load less than exhaust temperature control (Base Load). If a turbine is operating at less than Base Load (that is, at part load) then it is being operated in either Droop Speed Control mode or Isochronous Speed Control mode.

I've tried to draw it below, with zero load at the bottom and Base Load at the top. Part Load is any load between zero load and Base Load. When operating at Part Load, the unit is operating on Speed Control mode, either Droop Speed Control or Isochronous Speed Control. When the unit is at Base Load, it is operating on Exhaust Temperature Control.

<pre>
Base Load = Exhaust Temp Control
| |
| |
| Speed Control
Part Load = (Droop or
| Isochronous)
| |
| |
Zero Load |
FSNL = ------------- (Full Speed No Load)
</pre>

A GE-design heavy duty gas turbine driving a synchronous generator (alternator) is operated at partial load in either Droop- or Isochronous Speed Control mode. When the unit is loaded (in either Droop- or Isochronous Speed Control) to its exhaust temperature control limit, the control switches from speed control (FSRN) to exhaust temperature control (FSRT). At that point, if being operated correctly, speed control (either Droop or Isochronous) is no longer controlling the amount of fuel being admitted to the unit.

Pre-select Load Control is not a governor mode. It should only be possible to enable Pre-select Load Control when a GE-design heavy duty gas turbine is operating in Droop Speed Control Mode. Pre-select Load Control is a method of automatically adjusting the turbine speed reference (TNR) to maintain a load setpoint. It's possible to manually load or unload a unit to some desired load value using the RAISE- and LOWER SPD/LOAD buttons to change TNR. But most sites use Pre-select Load Control to do what can be done manually.

In my personal opinion, Pre-Select Load Control is misused at many power plants around the world. It's really just a lazy-man's way of operating a turbine. Instead of a thinking, human operator using the RAISE- and LOWER SPD/LOAD buttons to increase and decrease load below Base Load (while on Droop Speed Control) to some load value (say 15 MW, or 25 MW, etc.), Pre-select Load Control is a method for putting a desired load value as a setpoint into the Speedtronic turbine control system and having the Speedtronic automatically raise or lower the load to reach, and then maintain, that setpoint. Which a thinking, human operator could do by looking at the MW meter and using the RAISE- and LOWER SPD/LOAD buttons to get to the same value.

If an operator manually changed load to some desired load value, say, 20 MW, using the RAISE- and/or LOWER SPD/LOAD buttons, and then didn't make any manual adjustments and just left the unit running at that value, the load would not change very much over the course of the day; it would be fairly stable. And in many cases is even more stable than if Pre-select Load Control were enabled with a 20 MW setpoint, because a lot of Speedtronic turbine control systems have poorly "tuned" Pre-select Load Control parameters resulting in hunting and overshooting while trying to maintain the load setpoint (as evidenced by constant RAISE- and LOWER SPD/LOAD indications and load variations above and below the setpoint).

Most plant operators and Operations Managers will <b>NEVER</b> try operating the unit without using Pre-select Load Control while on part load. Why? Because they've never done it that way before, and they don't really understand what Pre-select Load Control was intended for and how it works. Even if Pre-select Load Control is not tuned properly, they somehow believe that if it's not enabled and active that the load will be "uncontrolled" and drift aimlessly. And, that's just not true, but, fear of the unknown is a powerful phenomenon (as is fear of having to explain what went wrong if something "abnormal" were to happen when Pre-select Load Control was disabled!).

A lot of plants also choose a Pre-Select Load setpoint that is deliberately higher than the rated load of the turbine and let the Speedtronic turbine control system go on exhaust temperature control (Base Load) while trying to achieve the elevated setpoint. While this is one method of operating on Exhaust Temperature Control (Base Load), it quite frequently results in lost generation ability if the load setpoint is only slightly greater than the rated load.

Lastly, a lot of plants use Pre-select Load Control while operating at part load on Droop Speed Control where the frequency of the grid to which they are connected is unstable. This is actually contributing to the instability of the grid as Pre-select Load Control actually over-rides Droop Speed Control when it's attempting to respond to grid frequency disturbances, and it does so in opposition to how it should be responding, exacerbating the grid frequency disturbance.

Turning to Isochronous Speed Control for just a moment, one cannot choose a load setpoint on a machine whose control system is operating in Isochronous Speed Control mode. That's because the nature of Isochronous Speed Control is to vary load as necessary to keep the frequency constant.

When Isochronous Speed Control mode is enabled, the turbine speed reference is TNRI (Turbine Speed Reference - Isochronous) and there is very tight control (proportional plus integral) of turbine speed, which is proportional to generator frequency. Droop Speed Control is straight proportional control, while Constant Settable Droop Speed Control has that load element which has some integral control in it.

Your Mark V turbine control system should have been supplied with a document called the Control Specification. There is some brief descriptions of the two types of speed control (Droop and Isochronous) in Section 3, I believe, of the Control Specification.

The CSP, also, has some very good depictions of the control modes using relay ladder diagram means. What I find most helpful when trying to understand these individual schemes is to take a large piece of paper (several, usually) and draw the rungs with a pencil and a large eraser in a left-to-right orientation rather than a top-to-bottom orientation as is shown in the CSP. In doing this, the "interconnections" from one rung to the next and the next and the next can be more easily seen. The several large sheets of paper and the large eraser are usually necessary because as I begin to understand how the rungs are interconnected and work together, I can refine my drawings, and in many cases simplify them, until it's no longer necessary to create drawings because it's easier to understand the CSP after one or two of these exercises.

It's extremely unfortunate that these control schemes are not documented very well, but, in fact they are: in the CSP. It just takes practice in learning to read and decipher and understand the symbology used in the CSP. Which, trust me, is a lot better than what many manufacturers use for depiction of control schemes! People who have experience with other turbine control systems and become familiar with GE systems will usually (reluctantly) admit that after they became accustomed to GE's methods and depictions they prefer it to what they were used to. (Many people, though, will never fully take the time to try to learn all the symbols and methods, preferring to stick with that they knew from their previous experience and just complain about GE's methods and depictions. Heck, even people who've only worked with GE systems never really take the time to try to understand how GE documents their control schemes, preferring to complain because it's not intuitive or easy to understand.)

So, what I'm trying to say is: Have a look at your CSP and the documentation provided with the Mark V. If you have specific questions about individual elements or schemes, ask them here, in separate threads. While related, the questions you asked are very broad and would require much more than has already been written. And, as much as anyone writes, the CSP really "says it all" and is precisely how your machine operates and should always be referred to in any case.

 

Just a quick comment on constant settable droop:

If the DWATT signal is from a single (non-redundant) transducer, then you don't have constant settable droop. If you have redundant watts transducers you PROBABLY have constant settable droop control.
Also, for most DLN combustion systems, true isochronous mode is not allowed - the rapid load changes possible can cause flameout or (worse) flashback.

 

Okay, that solved a lot of other problems i was facing but i still have the same doubt that does the equation is same for all modes of operation with only change being different response of controller to different turbine modes and other (like islanded/parallel)?

for eg. when in parallel and operating in pre-select load: As the system frequency goes down, TNH goes down (instantly?) and so controller, recognizing the mode, varies the TNR to get the same initial FSR. Right? So what in other modes of operation.

Also, can't we calculate droop from the equation itself? and how can we obtain the values of equation variables.

Anyways, that was a very nice explanation, and am really thankful to the control.com for such dedicated moderators and members.

 

When in parallel with other generators and their prime movers on a grid, regardless of whether the unit is at Part Load or Base Load: when grid frequency decreases, turbine speed decreases.

F = (P * N) / 120

Frequency (in Hz)
P = Number of poles of generator
N = Speed of generator rotor (in RPM)

Most GE-design heavy duty gas turbines are directly connected to the generator rotor, even if that's through a reduction gear. So, any change in frequency is immediately reflected in the generator rotor speed and therefore in the turbine speed. That formula defines the relationship, and it's pretty straightforward and there's no time delay involved.

If we're using the very simple equation you originally posted, a change in TNH will cause a change in FSR, right? FSKRN1 and FSKRN2 are Control Constants, so they're not variables. The only two variables are TNR and TNH. And, on most grids (okay, many grids these days--not most) TNH is relatively constant because grid frequency is relatively constant.

Droop Speed Control is about two things, plain and simple. It's about how a prime mover and its generator can operate in parallel with other generators and their prime movers to supply a load larger than most individual generators and their prime movers could supply and do so without causing an upset in the frequency or the power delivery. This is sometimes called "sharing load", because it resembles children sharing toys nicely.

The second aspect of Droop Speed Control is that it's used to try to help maintain grid frequency when it varies from desired. On a properly controlled and monitored grid with many generators and their prime movers connected in parallel, the grid operators anticipate changes in load and vary generation to try to limit frequency changes.

In the event that total generation does not equal load, the grid frequency will start to decrease. Any unit which is operating at partial load in Droop Speed Control will sense the change in frequency (because the prime mover speed will change, from the formula above) and adjust its prime mover's energy input (FSR in this case) to try to support the load and return/maintain the grid frequency.

So, operating stably and steadily while in parallel with other generators and their prime movers and responding to changes in grid frequency to try to maintain grid frequency are the two prime purposes of Droop Speed Control. It's actually insanely simple! And, the fact that one control method serves two vitally important and critical needs is just a tribute to the people who developed the scheme.

It's basically just proportional energy control based on speed: speed reference and actual speed. Actual speed (TNH) is assumed to be relatively constant, so therefore any change in turbine speed reference (TNR) causes a change in fuel flow (FSR) which results in a change in power output of the generator.

Conversely, if TNR is constant (which it usually is when the unit is at steady state conditions while on Part Load), and TNH changes, then FSR changes.

When the unit is on Part Load with Pre-select Load control enabled and the grid frequency changes (which affects TNH), FSR will change which will change load. But, since load is an "outer loop" in Pre-select Load control, it changes TNR to try to maintain the Pre-select Load Setpoint. Which is the *exact opposite* of what one wants to happen when the grid frequency changes!

The concept of Droop on a GE-design heavy duty gas turbine is that when the turbine speed reference is at the specified droop value (104% for a machine with a 4% droop setting, for example) that the power being produced by the gas turbine will be at rated <b>when the unit is in new and clean condition (clean inlet air filters; clean compressor; tight internal tolerances; etc.), and when the ambient conditions (temperature, humidity, and atmospheric pressure) are also at rated</b>.

But since many gas turbines are NOT in new and clean condition and don't have OEM components installed in them and aren't operated at rated conditions (usually they are operated in ambient temperatures and -humidities which are much higher than nameplate rating) the units don't reach the desired droop setpoint before they reach maximum possible power output for the given conditions.

And, lastly, when a turbine is at Base Load, it can't respond to grid frequency changes like it would when at Part Load. It's being told to put out as much power as it can when Base Load is selected and enabled, and everything else be damned! Speed control is not active, so therefore droop is not active, so frequency response is not possible when at Base Load.

And, yet people think because their turbine reaches Base Load when the turbine speed reference is 103.27% instead of 104% that Droop Speed Control isn't working properly on their unit. And that's simply not true!

Or, they think that when the unit is at Base Load that it should pick up load as it would it if were at Part Load on Droop Speed Control when the grid frequency decreases, so they think that Droop Speed Control isn't working properly on their unit. And that is also simply not true!

Or, they think that because the unit doesn't change load when the grid frequency changes when it's at Part Load with Pre-select Load Control enabled that Droop Speed Control isn't working properly on their machine. And that's simply not true, either!

Many times, it's a combination of the above than makes people think that Droop Speed Control isn't working properly on their machine. And, that's not even true, either!

So, they want to change their droop setting to try to make their turbine behave they way they think it should. When it won't. Because it's not the turbine that has a problem, it's the "thinking" people who think they know how the turbine should operate that have the problem. It's the perception that's flawed, not the reality.

If you post the things from the CSP <b>at your site</b> that are causing doubts and explain what you think is happening or not happening then we can confirm or clarify your analysis, which will help you develop your understanding of how your turbine operates.

 

To answer your question.. Isochronous control is different which tries to achieve target speed all the time irrespective of load.
For part load and preselect load the same equation applies.

For Isochronous the reference is speed. So the equation will be like Raise FSR if TNH<100, Lower FSR if TNH>100.

Little more info on droop loading below..

The equation cited by you means Fuel Required/ FSR = FSRN1(Fuel required for FSNL)+ (TNR -TNH)*FSRN2 (FSRN2 determines the droop setting of the unit).

Typical droop is set at 4%, which means a change in speed of 4% will result in full loading/ unloading of the machine depending on whether speed increased or decreased. TNR is the reference speed equivalent of desired load.

To make you understand more see below equations using example values..
At FSNL

FSRN=20= FSKRN1(20)+ (TNR(100)-TNH(100))*FSKRN2(20)..fuel required for FSNL is 20 and since reference speed and actual speed is same second term is zero.

AT 50% load
FSRN=60= FSKRN1(20)+ (TNR(102)-TNH(100))*FSKRN2(20)...for 50% load reference speed is 100+2%, so TNR 102.

at 100% load
FSRN=100= FSKRN1(20)+ (TNR(104)-TNH(100))*FSKRN2(20)..For full load reference speed is 4% (~ 100%load since droop is 4), so TNR is 100+4.

You can make the droop 5 and back calculate the constant TNKRN2. In this case 5% speed change will result in full load.

Changing the TNR is what you do when you press raise/ lower manually. In preselect load when TNH changes load will change for a moment but then preselect control will bring back the difference again to the preset load.

This is a simple equation used in old units and later the practise was changed to constant settable droop for improved machine response.

At base load the equation changes to the one used for temperature control, ie: FSRT.

Hope this helps.

 

Hmm, That again bring me closer to the object... Thanks a lot to CSA and Rajesh..

Now the question is

1) is there any limit in rate of rise/lower of FSR...in isochronous

2) is the relationship between FSR and TNH linear....

3) If possible, How can we find the slope of the relationship if linear..??

Thanks again and waiting for your reply...

// Hail control.com //

 

The values FSKRN1 and FSKRN2 are Control Constants, meaning they are adjustable, but "fixed" parameters. (USUALLY, a 'K' in the third position of a signal name signifies the signal is a Control Constant in Speedtronic heavy duty gas turbine control panels.) The values don't change during operation of the unit. So the above examples provided by Rajesh aren't 100% correct.

The original values of FSKRN1 and FSKRN2 are calculated when the panel is configured for the unit it will be used to control, and the value of FSKRN1 needs to reflect the average, typical running Full Speed-No Load FSR value when the unit is at FSNL, prior to synchronization.

The actual, running value of FSR (FSRN) at Full Speed-No Load can change with fuels and with ambient conditions and with machine conditions, but FSKRN1 is meant to be an "average, representative" value for the various values of FSR which would be experienced at FSNL. In other words, FSRN at Full Speed-No Load is a value which is related to lots of conditions, including how fuel control valve LVDT feedback (if used) is calibrated, and actual fuel conditions, and machine condition, and ambient conditions. FSKRN1 is sometimes adjusted during commissioning, and sometimes it's adjusted based on changing site conditions, but not usually unless the fuel supply characteristics change substantially.

FSKRN2 is not usually changed, either during commissioning or based on changing site conditions.

If you look closely at the equation eurydice originally posted, it's a very simple linear equation:

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

Let's rearrange the terms:

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

Now, it's more like f(x) = mx +b, where 'm' (FSKRN2) is the "gain" and 'b' (FSKRN1) is the "offset". And the "variable" term, 'x' (TNR-TNH), is really <b>two</b> variables: Turbine Speed Reference (TNR) and Actual Turbine Speed (TNH). When operators are changing load, they are changing TNR, which changes the error between TNR and TNH, presuming a stable grid frequency. When grid frequency (which is directly proportional to TNH) is changing (and TNR is not), the error between TNR and TNH changes. Any change in the error between TNR and TNH will result in a change in FSRN, and *when the unit is being operated at part load without Pre-select Load Control enabled and active* will mean a change in the amount of fuel being admitted to the turbine.

I have two *example* values of FSKRN1 and FSKRN2:

FSKRN1 = 18.4 % FSR
FSKRN2 = 12.5 %FSR/%Speed

Note the engineering units for FSKRN1 and FSKRN2. Any mathematical formula in a Speedtronic needs to have proper engineering units. And, again, FSKRN1 and FSKRN2 are specific and particular to each machine.

Using the *example* values above, one can see that with a TNR of 104% (which would correspond to Base Load for a unit with 4% Droop <b>in new and clean condition, being operated at nameplate rated conditions, with a fuel which is equal to the expected fuel characteristics provided when the unit was configured</b>), FSRN would be equal to 68.4%.

[Most GE-design heavy duty gas turbines are configured (fuel nozzles; control system) such that when FSR is between <b>approximately</b> 60% and 70% the unit will be at Base Load (exhaust temperature control) <b>when the unit is being operated in new and clean conditions with expected fuels at nameplate rated conditions (ambient temperature, ambient pressure, humidity, elevation).</b>]

Typically configured Speedtronic heavy duty gas turbine control panels will have a lower- and an upper limit on the operator's ability to change TNR: 95% and 107%, respectively. If the question about rate is related to the rate of change of TNR, yes, there are limits to that, too, to try to protect the hot gas path parts from excessive thermal stresses by fast changes in fuel. Those limits are machine-dependent, and I'm beginning to think this post is more about generic Speedtronic turbine control than about any specific turbine.

"Full load" does not occur at FSRN = 100%. Full load occurs when the unit reaches exhaust temperature control <b>for the conditions and fuel it is being operated on at that time.</b> Full load changes during the day as the result of ambient conditions; full load changes depending on fuels; full load changes depending on machine conditions (hot gas path part; axial compressor cleanliness and condition); full load changes depending exhaust conditions (back pressure) and inlet conditions (filter cleanliness).

The amount of load that a Speedtronic heavy duty gas turbine control panel will accept for a given error in the difference between TNR and TNH is expressed by the value <b>FSKRN2</b>." This is the normal operating part of droop speed control. So in the example above, for every 1% change in the error between TNR and TNH a change of 12.5% be seen in FSRN. And, <b>under ideal conditions</b> when the unit reaches approximately 68.4% it will be at Base Load and TNR will be <b>approximately</b> at 104%.

The Control Specification provided with a Speedtronic heavy duty gas turbine control panel lists a lot of very useful information about fuel and load in Sect. 05, Fuel Control; there should be a section titled, "Expected Operating Characteristics" or such, along with a section about "Expected Fuel Supply Characteristics".

But, we're still not clear about which Droop method is being used in eurydice's Speedtronic. And all of the above only relates to the very simple equation that eurydice provided in his original post.

It should be clear from the above, that the relationships established during configuration of the unit are ideal, and the actual values of FSRN and TNR for any particular machine are affected by many, Many, MANY, <b>MANY</b> parameters. I caution, nay, I <b>warn</b>, anyone against changing FSKRN1 or FSKRN2 values without understanding all the implications. Loading and unloading rates (the TNKR1_n values) are related to these values: a *LOT* of other functions (some options; some standard control schemes) are also related to these values. So, changing them based on a partial understanding of their function and relation to other control functions and schemes can, and will likely, result in unexpected and even undesirable behaviour(s).

This is a great question; but from painful personal experience some ill-advised actions have been taken based on an incomplete understanding of the entire control system and how things are related and dependent.

 

There is a(nother) mistake in the reply I originally posted (the moderator(s) graciously corrected the first mistake).

The following sentence:

>The amount of load that a Speedtronic
>heavy duty gas turbine control panel
>will accept for a given error in the
>difference between TNR and TNH is
>expressed by the value FSKRN1.

should read:

"The amount of load that a Speedtronic heavy duty gas turbine control panel will accept for a given error in the difference between TNR and TNH is expressed by the value <b>FSKRN2</b>."

Sorry for any confusion this may cause.

 

So, to try to tie all of this together, let's summarize. (WARNING: Maths used extensively in the explanation below.) And then let's clarify. And, I'm likely going to regret this. Probably in more ways than one.

>>In GE-design heavy duty gas turbine control systems operating on typical fuels, the intent is to design a turbine's control system such that when it's at rated power output (Base Load) the FSR (Fuel Stroke Reference) is <b>approximately</b> 70% when the unit is:

in new and clean conditions;

is being operated at nameplate rated conditions (rated ambient temperature, rated ambient humidity, rated elevation (atmospheric pressure));

and is being operated with fuel(s) matching the expected fuel characteristics provided when the unit and/or control system was built or configured.

(Some machines have design FSRs lower than 60% when at Base Load at rated conditions, and others have design FSRs higher than 70% when at Base Load at rated conditions. But, the typical range is around 70%, sometimes higher, sometimes lower. And it varies from machine to machine depending on configuration, fuel, site, hot gas path internals, and on and on and on.)

>>The power produced by a GE-design heavy duty gas turbine operating on most fuels is generally directly proportional to the fuel flow-rate. The fuel flow-rate is proportional to the FSR (Fuel Stroke Reference), which is a reference for the opening of the fuel control valve and/or the fuel flow-rate through the control valve.

>>The primary purpose of Droop Speed Control on any prime mover governor (control system) is to allow a prime mover and its generator to <b>smoothly and stably</b> "share" in supplying the load of a grid while operating in parallel with other prime movers and their generators. The secondary (some would say primary, depending on their point of reference, but they would be wrong) purpose of Droop Speed Control is to help to maintain grid frequency when it varies from desired.

Presume for the purposes of this explanation that a unit in new and clean conditions was configured such that it produced rated power output at rated conditions with fuel of certain characteristics at an FSR of 68.4%, and that an FSR of 18.4% was required to maintain rated speed with no load (Full Speed-No Load). This means that zero load occurs in a new and clean machine at an FSR of 18.4% and rated full load at rated conditions with expected fuel occurs at an FSR of 68.4%. Again, these are ideal parameters for an example, and are not reflective of any particular machine. In this example, the machine is considered to be in new and clean conditions, and is being operated at nameplate rated conditions on the expected fuel.

Subtracting the rated load FSR (68.4%) from the no load FSR (18.4%) means that a 50% change in FSR will occur when the unit goes from zero load to rated load for the machine in this example. Half (of rated) load would occur at a 25% change in FSR from zero load, or at 43.4%FSR ((50 / 2)) + 18.4 = 43.4).

Now, suppose it was desired to have a droop setting of 4% for this unit. That means that when the error between the turbine speed reference and the actual turbine speed was at 4% that the unit would produce rated power output (full load). So, rated power output is to occur at a turbine speed error of 4%, and at an FSR of 68.4%. Zero load is to occur at a turbine speed error of 0% and an FSR of 18.4%.

Presuming a stable grid frequency, which means a stable turbine speed (actual turbine speed), when the turbine speed reference is increased from 0% to 4% the turbine power output will increase from 0 to rated. This is how the power produced by a GE-design heavy duty gas turbine operating at Part Load without Pre-select Load Control enabled is controlled between generator breaker closure (zero load) and full load (Base Load; exhaust temperature control): by changing the turbine speed reference.

For every 1% change in the turbine speed reference (presuming a stable grid frequency/turbine speed), the power output of the turbine will change by 25% <b>of rated</b> for a unit with a 4% droop setting. Again, this is for a new and clean unit, being operated at rated nameplate conditions with fuel of expected characteristics.

In this example, a 1% speed difference represents 25% of the droop setting. This means that a 1% change in speed differential should result in a 25% change in power output. Since 25% of 50%FSR is 12.5%, a change of 1% in the speed reference will result in a change of 12.5% in the FSR.

Let's calculate this another way using the original equation provided by eurydice.

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

For rated load:

68.4%FSR = ((104%Speed - 100%Speed) * FSKRN2) + FSKRN1

For FSNL (zero load):

18.4%FSR = ((100%Speed - 100%Speed) * FSKRN2) + FSKRN1

If we subtract the two equations with two unknowns from each other, we have:
<pre>
68.4%FSR = (4%Speed * FSKRN2) + FSKRN1
-(18.4%FSR = (0%Speed * FSKRN2) + FSKRN1)
-----------------------------------------
50.0%FSR = 4%Speed * FSKRN2
</pre>
Solving for FSKRN2, we get:

FSKRN2 = 50%FSR / 4%Speed = 12.5%FSR/%Speed.

If we plug FSKRN2 into either equation and solve for FSKRN1:
<pre>
68.4%FSR = ((104%Speed - 100%Speed) * 12.5%FSR/%speed) + FSKRN1
68.4%FSR = ((4%Speed) * 12.5%FSR/%Speed) + FSKRN1
68.4%FSR = 50% FSR + FSKRN1
FSKRN1 = 68.4% FSR - 50%FSR = 18.4%</pre>
Or:<pre>
18.4%FSR = ((10%Speed - 100%Speed) * 12.5%FSR/%speed) + FSKRN1
18.4%FSR = ((0%Speed) * 12.5%FSR/%Speed) + FSKRN1
18.4%FSR = 0% FSR + FSKRN1
FSKRN1 = 18.4% FSR - 0%FSR = 18.4%FSR
</pre>
So, the droop setting is a function of the expected rated load FSR since FSR is proportional to fuel flow and fuel flow is proportional to power output. And this is for a machine in new and clean condition, being operated at nameplate rated conditions, with the expected fuel.

Now, imagine the turbine in this example operating at Part Load with Pre-select Load Control disabled and at a turbine speed reference of 43.4% FSR, which corresponds to 50% load when the grid frequency is at rated (100%), and therefore the actual turbine speed is also at rated (100%). If the grid frequency decreases by 1%, then the actual turbine speed decreases by 1%. Which increases the error between the reference and the actual from 2% to 3%. (Instead of the speed error (TNR - TNH) being 2% (102% - 100%), when the grid frequency changes by 1%, the error (TNR-TNH) goes to 3% (102% - 99%.)

[When the grid frequency decreases, it's because there isn't enough generation to support the load, so Droop Speed Control increases the output to try to help support the existing load at that time. When the grid frequency increases it's because there's too much generation to support the existing load load at that time. Load is not a function of the amount of generation; it's a function of the lights and motors and computers connected to the grid. Generation must match load for grid frequency to be stable.]

In this example, because the unit in the example was configured with a 4% droop, when there is a 1% change in speed error <b>for any reason</b> the load will change by 25% of rated. And this is true <b>as long as the error doesn't exceed 4%!</b>

The above should explain, once and for all, how basic Droop Speed Control works in GE-design heavy duty gas turbines with Speedtronic turbine control systems (Mark IV and newer). The equation used above is basic Droop Speed Control. When an operator clicks on the SPD/LOAD RAISE or SPD/LOAD LOWER targets, TNR is being increased or decreased, respectively.

This type of control is referred to as "proportional" control. The amount of power produced is strictly proportional to the error between the turbine speed reference. It is how the fuel is controlled during parallel operation with other prime movers and generators connected to a grid supplying a load. Almost every prime mover governor uses something similar because, as can be seen, it not only allows for stable control of fuel flow from no load to rated load based on simple parameters, it is useful when trying to maintain load on a grid when the grid frequency is not at rated. (The speed error will decrease when the grid frequency increases above rated, which will cause the power output to decrease which is what should happen under this condition.)

Proportional droop speed control does not try to make the actual (actual turbine speed in this case) equal to the reference (turbine speed reference). It relies on the fact that there will be an error because under normal circumstances the grid frequency is stable at 100% and therefore the actual turbine speed is stable at 100%, and the amount of fuel flow is proportional to the error between the reference and the actual. Droop Speed Control relies on the fact that something or someone else is controlling or being done to control the grid frequency (the actual turbine speed). And, when something or someone else is not properly controlling the grid frequency and the turbine is operating at Part Load, Droop Speed Control will automatically make changes to the turbine load in an effort to try to maintain and support grid frequency.

When the unit is operating on Pre-select Load Control, TNR is automatically being increased or decreased to maintain the Pre-select Load Control setpoint <b>regardless of whether TNH is stable or not</b>. In other words, if a turbine is being operated on Droop Speed Control at Part Load with Pre-select Load Control enabled and active if the grid frequency decreases then TNH will decrease which will tend to increase load. But, Pre-select Load Control will then change TNR to try to return the load to the Pre-select Load Control setpoint, which is the exact opposite of what one wants to happen when the grid frequency goes down. It actually exacerbates the problem! (Which is why I argue that operating a unit at Part Load for extended periods of time with Pre-select Load Control enabled and active on grids with unstable frequency is not a proper way to operate a turbine or turbines.)

Many newer GE-design heavy duty gas turbines with Speedtronic turbine control panels use Constant Settable Droop Speed Control, which is very similar but has an integral load component and is for another time and thread. Maybe. But, regardless of the integral load component, load is still increased and decreased based on the error between a (load biased) turbine speed reference and the actual turbine speed. And it will respond to changes in grid frequency similarly (as long as the speed error for a unit with 4% droop doesn't exceed 4%).

Now, the last thing to consider here is the effect of operating a turbine in not new and clean conditions, and not at rated nameplate conditions (ambient temperature greater than nameplate; ambient humidity greater than nameplate; atmospheric pressure less than rated). And, for the moment let's just forget the 'not new and clean conditions' part of the equation. When the ambient temperature and humidity are greater than rated (and remember: many older machines are rated at a 15 deg C, 59 deg F, ambient!), a unit with 4% droop will reach exhaust temperature control (Base Load) at TNR <b>LESS THAN</b> 104%, and the load will be less than the nameplate rated value. But, in general, for every 1% change in TNR the unit power output will increase by <b>approximately 25% OF RATED</b> and that <b>still corresponds to a droop setting of 4%!</b>

A machine with a 4% droop setting in generally new and clean conditions being operated at ambient temperatures less than nameplate rated will usually reach exhaust temperature control (Base Load) at TNR <b>greater than</b> 104%, and the load will be greater than the nameplate rated value. But, in general, for every 1% change in TNR the unit power output will increase by <b>approximately 25% OF RATED</b> and that <b>still corresponds to a droop setting of 4%!</b>

A machine in not new and clean conditions, with a high inlet filter differential pressure, and a dirty compressor, and increased tolerances in the axial compressor and/or turbine sections, will not be as efficient and the change in load for the change in TNR will be slightly (sometimes more than slightly depending on the severity of the condition(s)) less than specified values. A unit's power output with 4% droop might only increase by 23% for a 1% change in speed error, but <b>does NOT</b> mean the unit is not operating with 4% droop. It just means it's being operated at less than optimal conditions.

So, all you people who are right now getting up run to your operator interfaces to look at TNR and FSR and power output, and finding that the unit is at 103.27% TNR at Base Load with an FSR of 54.6% and starting to try to calculate the droop setting of the machine: <b>STOP</b>. Your calculations will be meaningless. The only droop calculation that means anything at all is the one that's done by observing the change in power output <b> as a function of rated load</b> versus the change in speed error. For a machine with a 4% droop setting the power output should change by <b>approximately 25% of rated</b> for each change of 1% in the difference between the turbine speed reference (TNR) and the actual turbine speed (TNH). That's presuming the unit is in new and clean conditions and the fuel is approximately what it was expected to be when the control system was configured.

Remember: Droop speed control is what allows the prime mover and generator to smoothly and stably control power output when connected to a grid in parallel with other prime movers and their generators.

Droop Speed Control is only partially about the change in power output versus the change in speed error, and that portion is related to rated power output not actual power output under less than rated conditions. And rated power output generally only occurs at rated conditions on a machine which is in new and clean condition. And most of the machines out there are from from new and clean condition, and are not being operated at rated conditions, and don't have fuels which are quite what was expected when the turbine was initially built and/or configured. It's a fact of reality.

And, if a turbine's power output is unstable, it's <b>NOT</b> likely the result of a problem with Droop Speed Control! <b>Do not go there!</b> All of this discussion about Droop Speed Control is only to explain how a GE-design heavy duty gas turbine with a Speedtronic turbine control panel is loaded and unloaded between zero load and exhaust temperature control. If a unit has been operating properly for weeks, months, years, decades, and suddenly it becomes unstable, it's not because of a problem with Droop Speed Control.

In all the years I've been working on turbines, I've only seen GE control engineering change a droop setting once. Once in more than two decades. Again, this is not an explanation to be used to troubleshoot stability problems, it's just an explanation of how load is changed.

I just know I'm going to regret this. I need to go find a beer, or four.

 

yaah thanks, no confusion about that.....and thanks for the longggggg reply....and very well explained.

Anyways... now, this means we cannot define the exact behaviour of the Mark V, especially for the isochronous mode. For other modes, at least we have the idea that how should the machine behave ideally(i.e. the behaviour of controller not considering the response of turbine and generator)... but no clue about isochronous mode!!!...We know now that TNH will be increased till it reaches defined TNR value or some difference..

but whats the relation for change... like droop???
Dead End !!!!!!!

 

eurydice,

Hopefully more people are reading this than just you, which is why there was so much detail and so many caveats. As I've said, I've lived to regret every time I've explained Speedtronic droop speed control, and probably the only difference this time will be I won't be on the receiving end of a frantic phone call telling me that the unit is unstable or doesn't load/unload as quickly (or loads/unloads more quickly) or it can't be synchronized, and it worked fine before but now it's now working and they need somebody back immediately.

And that somebody arrived to find that the droop parameters were changed. But nobody on site changed them; it must have been that field service representative who was here before!

Bollocks!

Isochronous speed control is proportional-plus-integral control, very tight proportional-plus-integral. It changes fuel in response to frequency (actual turbine speed) changes. No operator intervention required, because any operator intervention is just going to change the speed (frequency) setpoint.

In general, the limits of isochronous speed control are 99.83% to 100.17% of rated speed, and even tighter than that. TNRI (Turbine Speed Reference-Isochronous) is usually set to 100.0%, and any operator changes will increase or decrease that setpoint. And the Speedtronic just changes fuel (FSRN) to make the actual speed equal to the reference, which should be 100.0%. Because Isochronous Speed Control is all about very tight speed control.

I'm finished with this thread. So, any more discussion of basic speed control, droop or isochronous, is finished at this point. A LOT has been written on control.com in the past, and with this thread it should be over.

I wish you luck in your modeling endeavor. I think you need to find some basic texts on governors and speed control and give some very serious thought to what you just wrote.

You need to have a look at a Mark V CSP and Control Specification.

 

> Hopefully more people are reading this than just you, which is why there was so much detail and so many caveats. <

YES, at least one other person was reading. Many of my questions were answered. Thanks.

> I'm finished with this thread. So, any more discussion of basic speed control, droop or isochronous, is finished at this point. A LOT has been written on control.com in the past, and with this thread it should be over. <

One follow-up question? Regarding, machines that fire (or possibly co-fire) two fuels. My particular interest is natural gas and syngas produced by an IGCC plant.

Will the constants FSKRN1 and FSKRN2 be constant or vary depending on the fraction of which fuel is being fired?

 

Most newer units these days will use something (terribly inappropriately) called 'Constant Settable Droop' which factors load (MW) feedback into the equation. The equation used by the originator is straight proportional droop speed control. (Only the name is terribly inappropriate; the function works very, very well.)

I tried to be very, very clear about that particular formula: it is an older formula, which is more than adequate for most every prime mover on the planet. But it is not the only method of droop speed control, though all of them produce the same results: Allowing multiple prime movers and generators to produce power in a stable manner and to respond to disturbances in the grid frequency to try to maintain grid frequency (which is one of the most important aspects of an alternating current grid).

Constant Settable Droop Speed Control will take into account the different heating values of the two fuels, and the droop setting will remain the same.

A fuel splitter algorithm can also be used if there are separate gas valve arrangements for the two fuels. I'm not familiar with the current IGCC designs, but I believe on some of the early design the two fuels each used their own gas valve and manifold and nozzle arrangements so as not to avoid mixing the two fuels.

In the IGCC sites I was privileged to have visited the syn gas system was well purged (meaning it had a lot of purge valves, double block-and-bleed valves, and instrumentation (limit switches, solenoids, air pressure regulators and actuators, pressure transmitters, flow-rate transmitters, and pressure- and differential pressure switches). All in an effort to prevent mixing the two fuels in the manifolds and nozzles.

I believe those control systems used a fuel splitter-type algorithm to transfer between fuels, and the systems were designed such that the heating value of the fuels flowing through the control valves and nozzles were roughly equal for the same FSR. In other words, at, say, 70% FSR on natural gas the calorific content of the fuel was the same as 70% FSR on syn gas.

But, it's been many years since I've worked on a syn gas unit, and the systems have evolved greatly since then, I believe.

 

Thanks for the information provided in the thread. But we are presently facing a problem that when a Frame 5 machine is taken in isochronous mode, it starts decreasing its load both in gas & HSD fuel and its gcb gets tripped on reverse power.

As per the information provided, I checked fskrn1 and fskrn2, the value of fskrn1 is constant at 12.5% but the values of fskrn2 for other 2 machines is 21.5 and 21.9% resp. Also, TNHI in the same problematic machine was at 100% in comparison to other machines at 100.13% & 100.17% resp.

Kindly help me out...
Thanx

Awaneesh

 

Awaneesh,

The definition of Isochronous Speed Control Mode is frequency control. If, when you switch a governor to Isochronous Speed Control Mode it reduces its power output to zero, or less than zero, then that means, by definition, that the frequency of the system/load with which it is connected is greater than nominal.

The reason the frequency is higher than nominal is that there is either:

An excess of generation for the amount of load (the total of motors and lights and computers and computer monitors being supplied by all the generators and their prime movers).

OR

The load (total of motors and lights and computers and computer monitors) has reduced to less than the output of the generators and their prime movers are currently producing.

(The two statements are really just two ways of saying the same thing: Load (the total number of lights and motors and computers and computer monitors) does not match the output of the prime movers and the generators supplying the power to the load.

We don't know enough about the circumstances of how you are operating your unit--the total number of prime movers and generators it is synchronized with, and how they are being operated (by some kind of Power Management System, or manually by human operators, or ???). But, if an Isochronous governor reduces its load to zero, or below, then it has nothing to do with FSKRN1 or FSKRN2. It is only because there is too much generation for the load, or the load has reduced below the amount of generation.

Too many people believe that when they switch a governor to Isochronous mode that everything is then "automatic" and the operators don't have to do anything at that point. Say, for example, that there are three units, two in Droop mode and one in Isoch mode, each producing 5 MW of power and the system frequency is 50.0 Hz, what that means is that the load (the total number of motors and lights and computers and computer monitors) is 15 MW and three units are all producing 15 MW of power at rated frequency.

As the number of motors and lights and computers and computer monitors decreases, then what would tend to happen is that the frequency would start to decrease. But, the Isochronous machine will automatically decrease it's output to keep the frequency at rated. So, if the load (the total number of lights and motors and computers and computer monitors) dropped to 13 MW, then the two Droop units would remain at 5 MW each, and the Isoch unit would reduce its power output to 3 MW.

If the load (the total number of lights and motors and computers and computer monitors) continued to reduce to 10 MW and the human operators (presuming there is no automatic Power Management System controlling the units in this example) did nothing the two Droop units would each continue to produce 5 MW and the Isoch unit would reduce its power output to 0 MW.

If the load (the total number of lights and motors and computers and computer monitors) continued to reduce below 10 MW and the human operators continued to do nothing then the Isoch unit would trip on reverse power. But, if the human operators reduced the load on one of the Droop units to, say 2 MW, before the load (the total number of lights and motors and computers and computer monitors) reduced below 10 MW, then one Droop unit would be producing 2 MW, the other Droop unit would be producing 5 MW, and Isoch unit would be producing 3 MW.

Now let's say the load (the total number of lights and motors and computers and computer monitors) started increasing. If the load (the total number of lights and motors and computers and computer monitors) and the human operators did nothing then the Isoch unit would increase it's power output to 5 MW, one Droop unit would be producing 2 MW, and the other Droop unit would be producing 5 MW.

For the purposes of this example, let's say the maximum power output of each of the three prime movers and generators in this example is 6 MW. If the load (the total number of lights and motors and computers and computer monitors) the load continued to increase to 13 MW and the human operators did nothing then the Isoch unit would increase its output to 6 MW, one Droop unit would continue to produce 2 MW and the other would continue to produce 5 MW.

If the human operators continued to do nothing and the load (the total number of lights and motors and computers and computer monitors) continued to increase to 14 MW, the Isoch unit output would remain at 6 MW--because it's at rated power output and can't produce any more power--and the two droop units would increase their power output to a total of 8 MW (the split of the power increase would depend on FSKRN1 and FSKRN2), but the frequency of the system being powered by the three generators would decrease to something less than 50.0 HZ.

If the human operators then adjusted one of the Droop machines to bring the frequency back to 50.0 Hz )say the machine that had been operating at 2 MW) the Isoch unit would still be producing 6 MW, the one Droop machine would return to producing 5 MW, and the Droop machine that had been producing a little more than 2 MW when the frequency was less than 50.0 Hz would now be producing 3 MW.

Human operators can't directly change the power output of an Isoch machine--unless they change the power output of one of the Droop machines the Isoch machine is synchronized with. The Isoch machine will adjust its power output to try to maintain nominal frequency (50.0 Hz in this example).

If an Isoch machine's power output is approaching zero--or approaching rated--and human operators (or an improperly configured automatic Power Management System) does nothing, then the Isoch unit will open its generator breaker on reverse power as the frequency starts to increase above nominal--or it will not be able to produce any more than rated power if the frequency starts to decrease below nominal. The operator must take some action (or a properly configured automatic Power Management System must take some action) to try to change the load on one or more Droop units to keep the power output of the Isoch unit above zero and less than rated--or the frequency will increase above nominal or below nominal.

And that is how it works.

It has nothing at all to do with FSKRN1 or FSKRN2.

 

Dear CSA
Thanx for ur reply. The description provided by you are basics of isochronous mode.

Actually , I should tell u how we operate and then the problem will become more clear.

We are running 2 Frame 5 gas turbines, 1 in droop mode and 1 in isochronous mode to supply the end user loads, adjust the droop machines load if load on isochronous machine starts increasing to near rated output value.

We recently had a Major Inspection on Machine-1 and it was coming in isochronous mode before the inspection and maintaining frequency of system. After inspection, when we tried to take the machine in isochronous mode, it started decreasing the generation load, while the end load remained constant.

Say, for example, We took trial at End user load of 15 MW, with Machine -1 at 5 MW in droop mode and Machine-2 in isochronous mode at 10 MW, maintaining frequency of the system.

Now ,to take trial of machine-1 , we put Machine 1 in isochronous mode and machine-2 in droop mode. As soon as it is done, the load started decreasing on machine-1, with end user load remaining constant. We maintained frequency by raising speed of machine-2 by governor.

Meanwhile, the machine-1 continued to shed its load until its breaker trips on reverse power.

As soon as the breaker cuts, we put the machine-2 in isochronous mode, it maintains frequency adequately.

Also, the machine-1 runs perfectly alright in Droop-Preselect Mode with all parameters in normal range.

Thanx in advance
Awaneesh

 

Awaneesh,

What is the frequency when Machine 2 is controlling frequency in Isoch mode and you bring Machine 1 on and switch Machine 2 to Droop Mode and Machine 1 to Isoch mode?

What is the Isochronous Turbine Speed Reference of both units--Machine 1 and Machine 2? (The typical Speedtronic signal name for Turbine Speed Reference-Isochronous is TNRI.)

What version of Speedtronic turbine control panel do each of the two units have?

It's not likely the problem is mechanical; the feedback for both Isochronous- and Droop Speed Control Modes is actual turbine speed: TNH (in percent). In Isoch mode, the reference is TNRI (in percent); and in Droop mode the reference is TNR (in percent). In both modes the Speedtronic tries to make the actual turbine speed (TNH) equal to the Turbine Speed Reference-Isochronous (TNRI) when operating in Isochronous mode.

The problem is still not related to FSKRN1 or FSKRN2. Either the speed feedback scaling has changed for Machine 1, it's wrong for Machine 2, or the Isochronous Turbine Speed Reference for the two machines is not 100.0% (for nominal rated frequency).

We also need to know what the Load Gear Box nameplate rating is for each unit--Machine 1 and Machine 2. How many RPM on the turbine side equals synchronous speed (3000 or 3600 RPM (for 50 or 60 Hz, respectively)) on the output shaft of the Load Gear?

Has Machine 2 ever been successfully operated in Isoch Speed Control mode prior to this problem?

We may have more questions after we learn the above information.

 

Awaneesh,

I mis-spoke in my earlier post. In Isoch mode, the Speedtronic tries to make TNH equal to TNRI. And it is usually very successful at doing so. If TNH differs from TNRI by more than approximately (typically) 0.17% then fuel will be adjusted as necessary to make TNH equal to TNRI. In this situation, TNH should remain fairly constant (+/- 0.17%). And all generators and turbines synchronized with this Isoch unit will have the same frequency and the same frequency variations. So, in Isoch mode TNRI should be 100.0% and TNH should be 100.0%, +/-0.17%.

In Droop mode, the Speedtronic uses the difference between TNR and TNH to control the amount of fuel being admitted to the turbine. When a turbine is operating in Droop mode, that particular turbine is <b>NOT</b> controlling frequency/speed and under normal circumstances TNH will remain constant--if the Isoch unit it is synchronized with maintains a constant frequency. So, as TNR is increased the error between TNR and TNH will increase and the fuel flow-rate to the turbine will be increased. As TNR is decreased the error between TNR and TNH will decrease and the fuel flow-rate to the turbine will be decreased.

In Isoch mode, the reference (TNRI) remains constant and the fuel flow-rate is adjusted to make TNH remain extremely close to TNRI.

In Droop mode, when the frequency is stable, TNH remains constant (controlled by the Isoch unit) and TNR is increased or decreased to raise or lower the amount load being carried by the Droop unit.

When you are synchronizing Machine 1 to Machine 2 and putting Machine 1 in Isoch mode, are you putting Machine 2 in Droop mode before or after you put Machine 1 in Isoch mode?

 

CSA

When only Machine-2 is running, it maintains a frequency of 50 Hz. When we start & synchronise Machine-1(keep it in droop mode), the frequency remains 50 Hz.As soon as we put machine-1 on isoch mode, it starts shedding its load, with overall frequency of system going down. But we maintain it by raising speed of machine-2 thru governor.

TNRI for machine-1: 100 and for machine-2: 100.1354

Control system for Machine-1 is Mark-VI and machine-2 is Mark-VIe.

Load Gear rating is 5100 rpm/3000 rpm for both machines with a frequency of 50 Hz(3000 rpm).

We have been regularly running machine-2 on isoch mode prior to this problem and are still running machine -2 in isoch mode. But Machine-1 is not coming in isoch mode now. Before Major inspection, it was coming in isoch mode.

Also, we have shut down Mark-VI for welding works during shutdown, can it have any relation to the problem..??

Awaneesh