Think about this. All (almost) prime movers can have both primary and secondary control. At any time the governor should respond to only one type of control, either primary or secondary. Not both. To make thing worse both controls take system frequency as one of the inputs.

Assuming your prime mover has both type of controls. You want your prime mover to operate under AGC- Constant frequency. Obviously you don't want your prime mover to operate under droop mode. How you are going to "tell" the controller?

About 15 years ago, it was achieved by making separation based on frequency dead band. Within this dead band, say 49.85 -50.15 Hz (Constant frequency regulation band), the droop response was disabled. The prime mover was under AGC-Constant frequency's command. In case the actual frequency felt outside this range, AGC suspended. Droop response took over.

Honestly, I don't know of much if any advancement in this area today.

 

Namatimangan08,

You make a lot of assumptions (and I edited my posting in anticipation of being edited or even rejected).

Every governor is not NOT configured or programmed to operate in the same way or to achieve the same purpose.

Every prime mover does not have secondary frequency response (in the usage you refer to).

Most of the turbines I have started up around the world did not have, nor were required to have, secondary frequency response--whether it be by AGC or some other means.

Thank you for the clarification, Namatimangan08, but can you provide a direct answer to EC&I Engineer's question about Mark VI control systems?

In response to EC&I Engineer's question, you should be referring to the application code running in the Mark VI at your site or ask the provider of the Mark VI at your site.

 

It is possible to have a control system that can handle primary and secondary response simultaneously. In this case separation of commands can be done via internal control logic.

During my days when I worked as a shift engineer for a system of power plants with the total capacity of the 540MW, AGC was an add on function. It did not come with the plants that I used to operate. That was 15 years ago.

 

Namatimangan08,
Indeed, it is possible to have a control system that capably handles both primary and secondary frequency response "simultaneously".

The question was whether or not a GE Speedtronic Mark VI had a 0-150 mHz deadband.

 

>Namatimangan08, You make a lot of assumptions<

Ok. Let us find out where are those assumptions.

>Every governor is not NOT configured or programmed to operate in the same way or to achieve the same purpose.<

Agreed. No argument.

>Every prime mover does not have secondary frequency response (in the
usage you refer to).<

Every prime movers CAN. This doesn't mean they have special designed systems to execute secondary response. The fact that, AGC-constant frequency (typical secondary response) does not need many prime movers to operate under this mode. Definitely not all the prime movers. To make it simple, secondary response is a supplementary function. The grid needs it badly. Not the prime movers. We can operate our prime movers for 25 years without this function. In fact we can operate them longer than as if we participate in the secondary response.

>Most of the turbines I have started up around the world did not have, nor were required to have, secondary frequency response--whether it be by AGC or some other means.<

1) The turbine does not require secondary response. No argument. The grid does.

We have more than 60 prime movers in our system, ranging from 150kW-700MW per unit. I personally have visited about 75% of them.

Not more than 30% of our prime movers have secondary response function. The rest are none. At any time we don't need all those 30% to operate under secondary response such as under AGC-constant frequency mode or AGC-tie-line load flow. We can select which prime movers to be put under this mode.

AGC has many modes. I know a few of them.
1) Constant frequency
2) Tie line net interchange
3) Constant frequency and tie line load flow -Bias on tie line load flow
4) 10 minutes future
5) 45 minutes future.....

Which AGCs you are referring to? Secondary response is associated with 1), 2) and 3) above. 4) and 5) can be AGC too, but there are tertiary response. They are equivalent to operators' interventions but the operators' function is replaced by AGC system.

>Every prime mover does not have secondary frequency response (in the usage you refer to).<

The basic secondary response does not need anything other than raise and lower button to control the load. The mean is always there as long as a plant operator can raise and lower the load. Unless there was no operator or the prime mover takes load set point from the scheduled ahead load.

That was how we did 20 years ago! 5 years later we were introduced to the so called AGC-Constant frequency and a few other AGC modes. Now, many years since then. Thing have changed I suppose.

>Thank you for the clarification, Namatimangan08, but can you provide a direct answer to EC&I Engineer's question about Mark VI control systems?<

No I don't. Last year, I was called to solve one specific problem at one of our plants (GT) which uses Mark V or Mark VI (I can't remember which version). Unfortunately I didn't go through all the logic control since I had already traced the problem without having to explore everything that that Mark V or VI is having. That was my first encounter with Mark V or VI. It won't be the last...

You have to know I'm not a control guy. But I know what the grid should expect and can expect from EACH prime mover.

I have to admit that you are still the best with Mark V or Mark VI or whatever Mark that you have. No doubt about it. I'm not challenging your authority. We share what we know.

>In response to EC&I Engineer's question, you should be referring to the application code running in the Mark VI at your site or ask the provider of the Mark VI at your site.<

He asked about the dead band. I was telling the purpose of the dead band. Probably not all of us here know it. It may or may not have anything to do with Mark VI.

Additional info:

"Effect of Prime-Mover Speed Control Characteristics on Electric Power System Performance- IEEE"

http://ieeexplore.ieee.org/Xplore/l...073891.pdf?arnumber=4073891&authDecision=-203


Abstract
"During normal operation the prime-mover speed governors function principally as stabilizers, automatic power-frequency control taking over the job of maintaining tie-line power schedules, frequency, and time. The prime importance of dead band is reemphasized. During disturbances governors affect both frequency drop and relatively long-term power swings, but generation schedules are of equal importance in certain cases."

My notes:

1) governors function principally as stabilizers

- Primary response (3-15 seconds or 3-30 seconds according to EU grid codes)

2) automatic power-frequency control taking over the job of maintaining tie-line power schedules

- Secondary response -Constant frequency and constant tie line net interchange mode. (*4s- 10 minutes)

3) but generation schedules are of equal importance in certain cases.

- Tertiary response. Could be AGC or manual loading up and loading down.

* Without the introduction of frequency dead band there will be an overlap region between primary & secondary responses. In general it is not a big problem. Unless the same prime mover does both functions!

 

>Indeed, it is possible to have a control system that capably handles both primary and secondary frequency response "simultaneously".
>
>The question was whether or not a GE Speedtronic Mark VI had a 0-150 mHz deadband. <

In 2006 there was one poster asked markvguy the same question in control.com. The question was "Does MarkV has any settable dead band?" Then he rationalized why such dead band is important. This poster was in electrical supply industry. He said that.

Before markvguy provided his answer the poster had answered his own question. He said there was. Sorry, I can't remember the figure. He said the dead band was already built in and cannot be easily changed. That was for Mark V. Not Mark VI.

I will try to find back the discussion between markvguy and the said poster.

 

http://www.control.com/thread/1026227827#1026228299

I was trying to get the original poster to be a little more specific.

Fortunately, you were able to read his mind.

 

Thank you CSA.

 

As I know, the AGC, secondary control /frequency control etc. exist in a control center. The used communication protocol between control centers is the ICCP (TASE 2), at least in Europe. Could you please specify the used communication protocol between the control center and the unit for Droop control, frequency, active and reactive power, etc.? I think that MODBUS and other proprietary protocols are in use for both Europe and USA.

THANKS

 

I don't know about the communication protocol but I do know what to communicate.

The control center want to get your plant operating parameters since your plant participates in frequency control. When your plant is doing frequency control, normally governor speed droop will be disable since your plant cannot be allowed to take commands from two diff sources.

Note that governor speed droop and frequency control do not exist in the same domain. As you know the droop is proportional control. AGC frequency control is typical integral and derivative control. <pre>
Here is general equation for AGC -Constant frequency mode:
Area Control Error (ACE) = k*(f_act-f_sch)

Where

k = stiffness constant (MW/(0.1Hz.T)
f_act = act freq of generator (Hz)
f_sch = scheduled frequency (Hz)
T = AGC regulation time interval (s)</pre>
ACE solves for "step function" power to reset system frequency to the scheduled frequency over a period of Ts in future. It is updated every T seconds to ensure system actual frequency approaching the scheduled in such a manner that the scheduled frequency will become a tangent line to the actual frequency -time plot. This is in theoretical sense. There are reasons for this. In reality, this objective is difficult to achieve.

ACE belongs to control center. After the ACE is known, it will be shared "economically" between all the prime movers that are participating in AGC- constant frequency mode. Then the ACE will be sent to a module that is called economic load dispatch module (ELD). ELD module perform calculation in order to allocate ACE that minimizes incremental cost. ELD or Dynamic Economic Load Dispatch (DELD) is one of the most complicated and interesting subject in power generation & transmission. It doesn't really matter how complex it might be, however, it still has to be done since it is all about money!

It is important to note secondary response via AGC does NOT mean to match supply and demand. One might want to believe that if grid frequency is lower that the scheduled frequency, this indicate supply is lower than demand. Or vice versa. This absolutely not true. The right indicator to measure supply-demand mismatch is by measuring frequency rate of change (dF/dt). If dF/dt is negative this indicates supply is less than demand. Otherwise supply is greater than demand. If dF/dt =0, we call it as a steady state condition, where supply and demand are equal. it can happen at whatever frequency that the grid can exist.

Supply and demand balancing task is done by primary response, i.e. governor speed droop. The control center will instruct each prime mover within its area to set the droop in according to frequency control & regulation requirements. Normally the droop percentage set point is revised once every one year or just after a new and bigger that the existing per unit capacity turbine generator is added.