Governor modes of operation

P

Thread Starter

Process Value

GOvernor modes of operation

Speed control and governor modes of operation has been a pet peeve in control.com countless discussions, arguments and fist fights had taken place trying to

explain the topic. well this is my own jab at explaining the concepts relating to governor modes of operation. the paper is divided into the following

a. Droop mode
b. Load control mode
c. Isochronous mode

i have tried to provide the basic details and i have also explained in detail the implementation in GE controllers.
Note - I have not still worked with turbines with constant settable droop, so the paper will explain the normal droop only.

http://www.2shared.com/document/adSmDXpx/Governor_Modes_of_Operation.html

comments , reviews , criticisms are always welcome :). and i hope this will find its way into controlwiki :)
 
Nice piece of work Process_Value, will come in handy as a reference.

Can I add another governor mode for consideration.

d. Isochronous mode load sharing mode

How do you run multiple units in isoch with equal loading on an islanded site?

This something we have looked at doing but havn't received much OEM support. It's common in the reciprocating engine world. Cheers
 
Process Value,

This is basically a good work, and I have several comments.

You have used examples of droop in the beginning of the document that talk about reducing speed reference as load is increased. That's <b>one</b> method of implementing droop. Then in the second section of the document where you describe GE's implementation of "straight" droop control you correctly describe GE's philosophy of increasing turbine speed reference to increase load.

This would seem to be very confusing to many people--decreasing the speed reference in the beginning of the document to increase load, and then increasing speed reference at the end of the document to increase load. True, it is all about the error between the speed reference and the actual speed. Why not use a consistent approach to the error throughout the document?

And, how many prime movers operating in droop control change their actual speed as the load increases? I think this is another very confusing point for many operators and technicians. In a properly monitored and controlled grid of any size the frequency should be fairly stable, which means the speed of the prime mover (presuming it's directly connected to the alternator) will also be stable and will not change as the load of the alternator changes.

And if the load on the grid changes the grid operator must take appropriate action through the means available to them to keep the frequency constant, which will keep the speeds of all the prime movers (presuming they are directly connected to the alternators they are providing torque to).

So, my comments are basically that there is an inconsistent description of droop control in the two sections of the document, and that the description of prime mover speed changing when the load of the prime mover driving the alternator is being changed are both very confusing.
 
P

Process Value

isochronous load sharing mode

309Eguy thanks for the support. well about iso load sharing mode, er it is something i do not have a good experience with. actually i have a very bad experience with iso load sharing scheme. i have seen a implementation go real bad. 5 unit tripping before it was abandoned. prime mover was 3 x 20 MW steam turbine units.

i think i have the original control philosophy somewhere in office i will check it out. The OEM's do not support it perhaps they know it do not work well. from what i remember of the incident, the iso load sharing unit was not the OEM's it was a external vendors. it took a single frequency data and in case of a change in frequency sent setpoints to the controller. the controller was ABB's Procontrol and the iso load sharing unit was from some electronica ltd. i will check up and write about iso load sharing scheme.
 
P

Process Value

Droop explanation

thanks CSA for the comments. well i made a small mistake in the initial "Turbine droop control mode" section. i hope you were referring to this

"When the Turbine is put in droop control mode, the speed reference is decreased with increase in the load of the machine. The droop in effect refers to the change (reduction) in frequency when the machine is loaded from no load to full load. The droop is expressed as a percentage of the change in frequency from no load to full load to the turbine rated frequency."

i meant only "speed" not speed reference. i was trying to explain speed reduction in the turbine with increase in power output when "speed reference" is held constant. in the accompanying figure the speed reference is fixed at 50hz. the figure shows how the "speed" will reduce in droop mode when speed reference is held constant at 50hz and load is increased. i will make the correction and upload the doc once again.

Grid control

grid control is slightly different. in grid level control is a three layered control these days.

primary response - 3-10 sec
secondary response - 1-2 min
scheduling - 10-15min

the droop comes into play in the primary response period. here with a rise in the system frequency the load reduces and vice versa.
the secondary response is the AGC which is used to maintain a desired frequency level. this activates only if the frequency deviation is higher than the predetermined setpoint.

scheduling is for optimal operation, planning for maintenance etc. here the whole days's scheduled power is given one day in advance and it is required to be followed by the generating station. it is typically a 96 slot one (15 min schedule).

hope i have explained to your satisfaction :).
 
Process Value,

The more I read the first section of the document the more it reads like explanations from some Woodward manuals.... including the charts and figures.

And the more I read the second part the more I fear even more questions and problems when people start willy-nilly trying to change the droop setpoint of their GE-design heavy duty gas turbine for whatever purpose they think it will serve. We already get enough questions about droop as it is, and people still don't understand that <b>droop is primarily only a method for stably producing power when operating in parallel with other prime movers</b>.

Process Value, you now own the droop discussion here on control.com.

The baton is passed.
 
P

Process Value

"The more I read the first section of the document the more it reads like explanations from some Woodward manuals.... including the charts and figures."

ha ha ha , well not by woodward manuals but by CANDU manuals. i was mostly inspired by them. they had the droop explained beautifully , but i did not like the explanation of multimachine droop control. the figures of the multimachine droop settings were also inspired by their manual. but i expanded a bit by providing points of operation and how the machine reaches the points. rest of it are mostly my own (excluding the GE application code).

"And the more I read the second part the more I fear even more questions and problems when people start willy-nilly trying to change the droop setpoint of their GE-design heavy duty gas turbine for whatever purpose they think it will serve. We already get enough questions about droop as it is, and people still don't understand that droop is primarily only a method for stably producing power when operating in parallel with other prime movers."

OK that worries me, i definitely DO NOT WANT anyone to try adjust the droop. It is not only here we get governor droop questions , i get a lot of them in real life too. i wrote it primarily for new commissioning engineers. in droop adjustment is done a lot more in smaller machines running in independent micro girds like a refinery or a heavy industry. this is because GE application comes with the ISO droop set at 4%. the refinery already will have a machine running at 5% droop , so even though it is not strictly necessary commissioning engineers are told to change it to 5%. and it is mostly new personnel who are sent to site (without any training). this was intended to help them. idea behind it was to get the concepts right :) for all personnel involved in power generation.
 
GT 501F is already commissioned and was running normal and now need to implement Frequency response as per requirement of Grid. I looked into the logic and they have 4 % Speed regulation. Any Suggestions? I know it is not easy to show logic here but not an expert and do not want to destroy nor change without knowing what I have been doing.

Please advise.
 
One thing that never ceases to amaze me is how complicated and ignorant usually smart people make droop.

How bout this definition: Droop is a specific formula that is used to set the gain of the purely proportional control on a governor.

Now to me the above seems over complicated...but I guess to many other engineers that prefer to explain stuff with formula and math it may help guide them. Some don't like qualitative or conceptual answers I rekn'.
 
P

Process Value

Primary frequency response

Susan , what is needed out of a primary frequency response is decided by the grid operator. they will provide you with

a. the droop to be set (4-5%)

b. Min spinning reserve to be kept at all times (as this will be a loss of production. there will be a clause to pay for the maintaining the spinning reserve). if the grid is running at a deficit (like india) this will be replaced by a frequency based power purchase agreement.

c. Frequency response bandwidth (the grid frequency band in which frequency response is expected form the utility)

based on the above data you will have to configure the controls. how to configure depends on the type of the controller. your OEM will have a standard frequency response package to suit your needs. it will not be a good proposition to do it yourself, it is always better to run a tested and proven code from a turbine manufacturer for something important as this.
 
Top