In our power plant we are in the process to reduce the active power imbalance between grid market set point and actual active power production. The main reasons for these imbalances are the droop characteristic. 4 % droop means that an increase of frequency 4% will mean increase a n increase of load from 0 to 100%, so at TNH of 100.3 you are at full speed no load, and at 104.3% you will be at base load (maximum BASE load).

The question is which constants of Mark VI logic implement the droop speed control? also the droop signal is changing all the time. I guess this is the constant settable droop......Any help is very welcome...

 

You said the droop was always changing. From that I would infer that you have found "the droop" so you should be able to determine which Control Constant(s) affect droop.

Please tell us how much the frequency of the grid the machine is synchronized to is changing and if this is a regular occurrence.

Please tell us, also, if the machine is being operated in Preselect Load Control at Part Load, or if the machine is being operated at Base Load (exhaust temperature control).

Lastly, please provide the signal name for "the droop" that is always changing <b>AND</b> what it is that you think droop does.

 

Hello?

We're waiting for your response to try to understand how to reply.

 

Droop is the misconception that keeps on giving!

There is an active power imbalance? Is the frequency of the system changing or is it steady 60/50hz?

I agree with CSA in all regards.

 

The signal that is changing is the dwdroop. The generator is synchronized at 50 Hz and the grid thus the generator frequency during normal operation changes from 49.90 to 50.10 Hz. The unit operates mainly at Droop mode and partial load, note at Temperature(Base) load. It operates at remote and gets the MW setpoint from DCS.

 

You have neglected to explain what it is you think droop speed control does.

If you look at the block that generates DWDROOP you will see that it multiplies DWATT by DWKDG to derive DWDROOP. So, anytime load changes (that is, any time DWATT changes) DWDROOP changes.

DWDROOP is <b>NOT</b> the droop characteristic of the machine; it is just part of the method of getting load into the equation that compares the turbine speed reference, TNR, and the actual turbine speed, TNH, to derive FSRN, Speed Control FSR. TNRL is the load-biased turbine speed reference that Constant Settable Droop Speed Control uses to get load into the droop function. If you've read any of the myriad droop speed control threads on control.com (and there's no shortage of them), you will see that traditional droop speed control does not have any load term. Constant Settable Droop Speed Control does have a load term.

It's still not clear why there is an "active power imbalance" and when this imbalance occurs. I can guarantee, though, that DWDROOP has nothing to do with the problem you are having.

When there is an active power imbalance, does this occur when the grid frequency is not at 100% (50.0 Hz), or, at other times?

How much is the active power imbalance? Is it 1.0 MW, or 2 MW, or 5 MW or 10.0 MW?

Have you trended the external load control signal, DWATT, the grid frequency (usually signal name SFL1 or SFL2), the turbine speed reference (TNR), actual turbine speed (TNH), and L70R and L70L (the TNR raise and lower logic signals)? Can you post the trend results to some file-sharing site and include a link to the trend in a post to this thread?

Could it just be that the turbine speed reference raise and lower adjustments need to be tuned to prevent some "overshoot"?

It's interesting that you mentioned 'primary frequency response' because GE sells an option called 'Primary Frequency Response'. Do you know if the Speedtronic panel at your site has the GE option called 'Primary Frequency Response'?

But I can pretty much guarantee that changing the signal that calculates DWDROOP is <b>NOT</b> going to solve your problem (which we still don't fully understand). In fact, it will likely cause other unintended problems.

So, help us to understand the problem you are having, and how the Speedtronic panel at your site is configured, and we might be able to make some recommendations.

 

CSA thank you for the reply, the target for me is to understand better the droop characteristic.

Droop speed control, in my understanding is the following: when grid frequency increases then turbine speed increases. this means that the grid power consumption is less than the grid power generation so the turbine (any turbine on the grid) decreases active power to obtain the equilibrium between generation and consumption. When grid frequency decreases then turbine speed decreases. this means that the grid power consumption is higher than the grid power generation so the turbine (any turbine on the grid) increases active power to obtain the equilibrium between generation and consumption.

Active power imbalance is a term of our grid code. So when the grid ask a power setpoint it expects to have it immediately. This is the first imbalance.

The second is the droop response, if grid frequency is not 50.00 Hz but is 50.10 Hz the plant generates less MW than grid setpoint.
If grid frequency is 49.90 Hz the plant generates more power than grid set point. The imbalance occurs at these two circumstances and is calculated every hour.

Our Mark VI Speed tronic has the primary frequency response option installed. The active power imbalance is from 1 to 12 MW. Last for the trend , at the moment i can not make it since i am not on site.

 

Droop speed control is the way that a machine's load is controlled when being operated in parallel with other turbines on a grid. A turbine with 4% droop will change it's load by 25% of the nameplate rating of the prime mover for every 1% error between the turbine speed reference and the actual turbine speed.

A machine operating on a typical grid with a stable frequency is loaded by increasing the turbine speed reference. Because the actual turbine speed (which is directly proportional to the grid frequency) remains relatively constant, the error between the turbine speed reference and the actual turbine speed increases. And that error is used to increase the fuel flow (for a gas turbine). And the increased fuel flow means increased torque production which means increased load taken on by the generator.

Now, when the turbine speed reference is relatively constant at part load and the grid frequency (hence the turbine speed) changes, the error will change. And the increase or decrease in the error will cause an increase or decrease in the fuel/torque/load.

So, droop is how the load is changed when the grid frequency is stable, and it's how the turbine tries to support the load even when the grid frequency is unstable.

Based on what little I know about Primary Frequency Response I would have to say that it's not "tuned" properly or is not configured properly. Because, when properly configured and tuned, it will allow a turbine and generator to respond to help maintain the load in the event of a frequency disturbance even though it has a load setpoint from either Preselected Load Control or from a Remote External source, as in your case.

So, the problem isn't the droop characteristic of the machine; it's most likely in the configuration or tuning of the Primary Frequency Response option. Because Primary Frequency Response will allow the turbine load setpoint to vary in response to grid frequency disturbances when operating on load control.

 

CSA, so according to your analysis we should leave the droop characteristic of the GT the same (meaning the constant FSKRN1= 15%) and see the primary frequency response. This is done to the DCS. In details GT sends hardwire the signal tnhdip_dcs to DCS , tnhdip_dcs is derived if you take out from 100% the signal TNHSYS. Afterwards in DCS the signal tnhdip_dcs is multiplied by 75 and reduced from the net load feedback of the plant load controller. Is that correct?

 

I was under the impression you had the GE version of 'Primary Frequency Response', but it appears you do not. Actually, it's not clear exactly what you have, except some interface to the DCS.

I'm suggesting the problem is <b>NOT</b> the droop characteristic of the Speedtronic.

Without understanding exactly how 'Primary Frequency Response' is accomplished it's virtually impossible to say. This does not sound like the 'Primary Frequency Response' that GE sells. This sounds more like someone's attempt at tie-line power control using the DCS to send RAISE & LOWER commands to the Speedtronic. The problem could be that the loading/unloading rate that's being used are not fast enough to respond to the DCS signals. (And, there are also limits as to how fast the loading/unloading rates can be made to load/unload the machine.) Another name for these kinds of systems is "PSM" (Power System Monitor), and in my experience, most of them are marginal or just don't work at all.

And tie-line power control to try to control load during frequency excursions is going to be very, very difficult to tune to be made to work properly. Why isn't the utility better at controlling the grid frequency excursions? And, why isn't your site complaining about the grid frequency excursions causing the power imbalances?

I'm also not clear about the two machines with different droop characteristics, and it's not clear if one of the machines handles frequency response differently than the other.

Do the two machines have the same rating (I'm specifically talking about the turbine Base Load rating)?

Do they both have Mark VI controls?

Again, it's just not clear what you have nor what you're asking. Which makes making recommendations very difficult--except to say that if you change your droop characteristic (or what you believe to be your droop characteristic) be prepared for a lot of unanticipated knock-on effects, including differences in the loading/unloading rates.

Of this I am certain.

 

You are right. There should be a constant settable droop ranging from says 0-12%. During the old days we could simply find it near turbine panel. We just have to turn the knob to the percentage set point that we require. Today, it won't be that easy to fin it. Modern controllers have become more sophisticated. Recently we have changed it for our diesel units. We have to call the supplier to change it for us. There was no more knob that we can use to set it like the old days.

Anyway, I think you should change it from says 4% to 6%, not the other direction. It may or it may not work. If it doesn't work please share with us about what you see.

Please check whether you have legal right to do so. Normally this set point manages by grid operator.

 

Namatimangan08,

It is irresponsible to suggest changing droop settings (droop characteristics) and <b>then</b> checking to see if it's legal or permitted by the grid regulator. Changing droop settings (droop characteristics) is not to be done lightly, and the cavalier manner in which you are making these recommendations is multiple threads is suggesting it's an easy and painless thing for anyone to do.

And that's just not true. Droop affects stability and loading/unloading and response to frequency deviations--whether connected to a large grid with many other prime movers and generators, or when operating with only one or two other prime movers and their generators in an "island" configuration.

The prime movers at your site may be operated in a manner that is completely different from the way that prime movers are operated at your site.

If you have experience, as you say, with GE Speedtronic turbine controls then you should know that the droop setting (droop characteristic) is a Control Constant, which is subject to being ramped up or down when the turbine is running to make changes using the Control Constant Adjust Display or Toolbox or ToolboxST. This means that making adjustments (should) requires a password to access the application, and changing the value will be a relatively slow process, in order to try not to introduce any instability into the operation of the turbine. If someone makes a change for a temporary period, then that change needs to be "un-made", at some time.

The effects of any change may not be immediately obvious, and can have unintended consequences on many other operating parameters.

Please re-consider making these multiple recommendations to change droop settings (droop characteristics) without making the proper admonitions <b>BEFORE</b> making the recommendation.

And please remember, there are tens of thousands of prime movers operating around the world that operate just fine day in and day out without making any changes to the droop setting (characteristic). Electrical grids are designed for prime movers with droop settings (droop characteristics) in a particular range, and machines are programmed to operate, and protect the prime mover, with certain ranges of droop settings (droop characteristics) in mind. Suggesting people change droop characteristic to solve a myriad of problems is irresponsible.

In all my years or working on power generation equipment (now approaching three decades), I have only been directed by Engineering <b>ONCE</b> to change droop from 4% to 5%, and that was only done after consultations with the utility and coordination with the droop characteristics of other power generation on that particular site.

I have, however, been to multiple sites that have changed the droop setting (droop characteristic) and have experienced many unintended consequences, and in <b>every case</b> I have returned the droop to the original value and solved the "problem" with other solutions after fully understanding what was required or needed.

We welcome your experience and contributions, but please--when it comes to recommending changing droop settings don't do so lightly when responding to threads here on control.com.

 

> It is irresponsible to suggest changing droop settings (droop characteristics)
> and <b>then</b> checking to see if it's legal or permitted by the grid
> regulator. Changing droop settings (droop characteristics) is not to be
> done lightly,

---- snip ---

> We welcome your experience and contributions, but please--when it comes
> to recommending changing droop settings don't do so lightly when responding to
> threads here on control.com.

Then you CSA for your comment. I take note.

FYI I don't post something that I don't know and I don't believe in. More important I don't tell others to do things that I will hesitate to do if I'm the one that need to solve the problem.

 

Think about the possibility that the grid management instructs plant owners to set their droop at the wrong set points. What will you do? It can happen. And it did happen.

 

One more question on the droop governor settings of GE Speedtronic Mark VI. i try to find which is the control constant where GE sets the dead band for droop control, the are aof setting is 0-150 mHz. Can you advise?

 

What Control Constant are you referring to?

 

Let us get clear about the sources and the nature of the imbalance. Such imbalance could be dedicated to the droop response or poor frequency regulation or the combination of both.

How you define the load imbalance? Is it based on real time error or time integration of load error over the specific period wrt time?

 

> One more question on the droop governor settings of GE Speedtronic Mark VI. i
> try to find which is the control constant where GE sets the dead band for
> droop control, the are aof setting is 0-150 mHz. Can you advise?

If I get you right the dead band is for a purpose to disable the droop. Normally the dead band is set within +/- 0.15Hz around the nominal frequency. This is done to ensure clear separation between primary & secondary frequency response. Within this dead band primary response via the droop is not greatly required.

 

Dear CSA,

i refer to a control constant that sets the dead band of droop control. For example, eny droop controller has a dead band of 0-150 mHz where it does not take any action. Can you advise?

 

I have no idea what either of you are trying to say.

Absolutely none.