Help needed on PID tunning

[jcp]

Hi Experts;
I attached an image of pid parameters setting.
When i increase the speed of gas turbine the speed goes 1% up then come back 1% down then stable at setpoint.
The parameters is as under
PV=0 to 11520= 0 to 100%
SP=7143 to 11520= 62% to 100%
When i start the turbine the intial setpoint is 7143 that is 62%. when engine reaches (PV) at 62% the speed goes to up 63% and then back to 60% after 2 sec it stable at 62%.
Also when i increase the SP suppose 74% the PV goes to 75% first then 73% and after 2 sec it stable at 74%.
What parameters i increase or decrease?
Is deadband effect on it or slew rate?
Please help me in this issu.
Regards

 

[jcp]

Plz help me.
Regards

 

[jcp]

Experts;
What more required to answer this or help me?
Regards

 

[MasterBlaster]

I am totally unfamiliar with your PLC, and you don't say how often the PID is triggered or scanned, but I'm certainly curious as to how the "Sample Time" can be zero and function properly.

Just a thought.

Bill

 

[cjd1965]

Hi Slew rate is the time it takes to move from 0-100% output... think of a big motorised valve or a overhead crane that may take a while to traverse. In your screenshot it is 0 so it is effetively ignored.

Interesting the enable is OFF so the PID is not being scanned

 

[Mickey]

The picture shown is an offline snap shot so I don't think we are seeing the true picture.

If it were me I would decrease the deadband.

 

[SD_Scott]

I don't see what the problem is. When you introduce a step change it will overshoot slightly and then stabilize. Unless its causing a problem, I wouldn't mess with it. I was a design engineer for Solar Turbines and they do have quite a bit of enertia.

 

[SD_Scott]

Is this a generator or a compressor....etc. What is the turbine driving?

 

[shooter]

You can set the deadband to zero as it is a proportional output.
And yes it is pretty normal it goes up a little as the I is slow. The process is rather slow, and will not react fast it has a looptime, so be happy with your result it is fine.

 

[jcp]

Hi SD_Scott;
Gas turbine drives the Dynomometer. It is two shaft gas turbine Saturn 1100.
All the sequences doing well upto 100% but i want that this becomes more accurate when increase the speed. I carefully examine the centaur turbine as well which controll design by solar turbines by using controll logix. I noted that it also 0.5% overshoot if SP chnage to achive that SP.
You see the images if i change the proportional gain 11.18 to around 7 or 8 then what happen?
Regards

 

[SD_Scott]

First of all a two shafter is going to have some slop between the gas producer speed(NGP) and the power turbine speed(NPT). Are you measuring the PV(oops I Meant PV) from the power turbine? If you have a special case where you don't want as much overshoot, you can open loop a ramp function and slowly ramp up or down. when you are within x% then switch over to a PID that has zero error and go from there.

I said SP, I meant PV.

 

[SD_Scott]

You can also ramp the SP of the PID to avoid so much reset windup. If you have looked through Solar's code you can see where they do that.

 

[jcp]

Hi SD_Scott;
Good day. Yes i agree but i want to discuss in detail here.
When start the engine after purge time complete which is 10 Sec for that turbine model After purge time the Ramp Generator function and give 4 to 20mA (Converted via module 0 to 50mA) in 8 Second if ignition occur mean T7 reaches 350 Deg F then control shift on PID. Now turbine required a 50mA to ramp up becuase i used electro hydralic servo actuator with old model fuel control valve (Loder 1,2,3 if you know act on PCD).
The ramp function after ignition can not be used in this case. Becuase no Delta P Transmitter for PCD AND Fuel. Its mechnical Fuel control Valve i think you know that, Actuator attract and extract the throttle butterfly valve.
Do you have any suggestion?
Thanking you i will remain

 

[SD_Scott]

Hi SD_Scott;
Good day. Yes i agree but i want to discuss in detail here.
When start the engine after purge time complete which is 10 Sec for that turbine model After purge time the Ramp Generator function and give 4 to 20mA (Converted via module 0 to 50mA) in 8 Second if ignition occur mean T7 reaches 350 Deg F then control shift on PID.

This part I understand.

Now turbine required a 50mA to ramp up becuase i used electro hydralic servo actuator with old model fuel control valve (Loder 1,2,3 if you know act on PCD).
The ramp function after ignition can not be used in this case. Becuase no Delta P Transmitter for PCD AND Fuel. Its mechnical Fuel control Valve i think you know that, Actuator attract and extract the throttle butterfly valve.
Do you have any suggestion?
Thanking you i will remain

Are you using a new actuator with the old valve body? Are you measuring PCD? I assume you are. It sounds like you are retrofitting some new stuff on an old engine. Where you have old systems that were directly connected via air signals, black boxes etc. They have to be built into your software. If an old fuel valve used direct air signals or a black box, you will have to develop an engine map in the PLC to replace it. Without more info that is all I can offer.

 

[jcp]

Hi;
Yes, The old system consist of relay logic and black boxces like Main fuel actuator, NGP, NPt ramp Generator temp. Control and Speed control boxces.
More Air signal direct to enter in fuel control valve and i measure the PCD only for reading.
In Short i convert the relay logic system on to PLC.
Every steps are ok except speed overshoot to 1%.

Are you using a new actuator with the old valve body? Are you measuring PCD?

no old electro hydhralic actuator with old fuel valve. Yes i measure the PCD only for display.

you will have to develop an engine map in the PLC to replace it

I am very kindful if you give me an example how to develop?
My email is [email protected]