PID Output and Valves

plcnovel

Member
Join Date
Feb 2006
Location
LHR
Posts
215
Hi There
I am having some problems in stroking the valves from the pid controller in plc. I am using Controllogix platform. None of the valves are having any positioners on them (simple I/P, and they are fairly big size pressure/temperature control valves). The problem i am experiencing is , none of these valves will actually start to move untill PID output reaches around 35% to 45%. Once the valves are moved from the fully closed position, then they seem to follow the pid output without any problems. But just at the starting they are late to react. Is there anyway i can achieve the solution to this problem through my logic? I will appreciate your help and time. Thanks in advance

Regards
 
Check the pressure range of the actuators, and the output pressure from your regulated I/P transducer. I would bet the supply pressure is set to low. Another thing might be not enough volume to the actuator.
 
Certainly that valve has some kind of positioner on it?

Ideally I would think that you would want to calibrate it, rather than try to adjust for it in the program. What happens when the calibration drifts further? Then you have to go back into the program and adjust again. At least with proper calibration, you wouldn't have to mess with PLC logic to get your loop working properly.

Maybe I just don't understand the situation.

edit: I guess I type slow. Good suggestion nswu1
 
Sounds like your I/P and control valves aren't using the same control range.
Is you I/P 3-15 psig, what about the valves are they set up for 3-15? Sounds like they may be 9-15.
 
plcnovel said:
Hi There
I am having some problems in stroking the valves from the pid controller in plc. I am using Controllogix platform. None of the valves are having any positioners on them (simple I/P, and they are fairly big size pressure/temperature control valves). The problem i am experiencing is , none of these valves will actually start to move untill PID output reaches around 35% to 45%. Once the valves are moved from the fully closed position, then they seem to follow the pid output without any problems. But just at the starting they are late to react. Is there anyway i can achieve the solution to this problem through my logic? I will appreciate your help and time. Thanks in advance

Regards

check the basics first:

-Confirm the pressure output of the I/P with a pressure gauge in series with the air signal to the valve

-Confirm the ma readings from your analog output to the actuator

If necessary,rescale your analog output instead of 0-20ma ( 4 -20ma ) to start at 10ma or whatever output it takes to get the actuator to move.

Ian
 
I would highly recommend adding positioners to the valve actuators. If you are having problems when the valves are new, just wait until they have some age on them. Tuning the PID loops will be a pain without positioners.
 
Before you go changing *ANYTHING* about the valves...

Are the valves under operating conditions? That is do they have the normal pressure drop across the valve? Control valves have what is called a "bench set". This is the amount of extra spring force required to hold the valve closed against the specified shutoff pressure. If there is no pressure on the valve (like during pre-startup checkout) you will see the behavior you describe. It is normal. When you have pressure present, it will balance out the spring, and the valve will come off its seat with a much lower signal. It probably won't be at 4.1 ma, but it should a lot closer than 35%.

Back in the old days, Fisher used to recommend not using a positioner on fast flow or pressure loops. Positioners were generally recommended on temperature loops. Now, it's very common to see positioners on all control valves.
 
plcnovel said:
Hi There
I am having some problems in stroking the valves from the pid controller in plc. I am using Controllogix platform. None of the valves are having any positioners on them (simple I/P, and they are fairly big size pressure/temperature control valves). The problem i am experiencing is , none of these valves will actually start to move untill PID output reaches around 35% to 45%. Once the valves are moved from the fully closed position, then they seem to follow the pid output without any problems. But just at the starting they are late to react. Is there anyway i can achieve the solution to this problem through my logic? I will appreciate your help and time. Thanks in advance

Regards

More basics stuff:

can you post:
-what analog output card you are using
-what I/P
-what Control valve
-the application

Is this one valve or multiple valves that have the same issue? I have seen a pinhole or a kink in an airline to the control valve cause the exact same problem you desribe.

Also from past experience I have found that most 3-15PSI control valves are actually 0 - 17PSI, meaning at 4ma ( 3psi ) there is a slight leaking through the valve. We solved that by using a solenoid to supply air to the I/P. Whenever the I/P is not in use, solenoid OFF, no air. Then at 20ma (15PSI) the valve was not fully open. In the old days most I/Ps had a span adjustment that allowed the 20ma signal span to be raised. Now most AB 4-20ma outputs will actually give you 21ma if you supply the analog output channel the max value .

Ian
 
Thank you very much for your input guys. I am using 1756-OF8 control logix analog output card, and the application is the temperature loop for the reboiler. The thing is, as you guys suggested earlier, i can scale my card output from 4ma to something higher than 4ma to lift the valve from its seat. But on it way back, when the output is 0%, then valve would still be open adding heat to the reboiler which we dont want. So how can i solve this problem. To me the valves are not linear at all.



Regards







curlyandshemp said:
More basics stuff:

can you post:
-what analog output card you are using
-what I/P
-what Control valve
-the application

Is this one valve or multiple valves that have the same issue? I have seen a pinhole or a kink in an airline to the control valve cause the exact same problem you desribe.

Also from past experience I have found that most 3-15PSI control valves are actually 0 - 17PSI, meaning at 4ma ( 3psi ) there is a slight leaking through the valve. We solved that by using a solenoid to supply air to the I/P. Whenever the I/P is not in use, solenoid OFF, no air. Then at 20ma (15PSI) the valve was not fully open. In the old days most I/Ps had a span adjustment that allowed the 20ma signal span to be raised. Now most AB 4-20ma outputs will actually give you 21ma if you supply the analog output channel the max value .

Ian
 
Peter Nachtwey said:
These kinds of problems go away and the machine runs better for the rest of its lifetime.

Thanks Peter, i do agree with you , but now i am in this situation where none of the valves are lifting from its seat until abt 35 to 40% output. Any solution to make these nonlinear valves to work properly?
 
Use a soloneoid that is controlled by a PLC output to supply air to the I/P. Whenever the heating loop is not required, AIR OFF, and also use logic to test if the temperature is getting too hot. Normally I take the required SETPOINT add 5c or 10c and compare TEST SETPOINT against PROCESS VARIABLE. If the PROCESS VARIABLE is greater than TEST SETPOINT for XX seconds, shut off the air to the I/P

Are you sure the valve actuator is not sticking ?

Ian
 
Ian's last line is a big thing to look at. Without a positioner on the valve, your plc output is simply creating a force balance condition between the pneumatic actuator and any internal forces wanting to keep the valve closed. In a perfect world this force balance would linearly map pressure to position. In your case it sounds like there is a valve force offset so the force balance relationship is being skewwed.

Another possibility is the valve is designed to be a positive seal valve. This will apply extra closing force to make sure flow is stopped. You stated a couple of posts ago that if the PLC output doesn't go to zero on the way back to closed that you will still get media flow. Have you confirmed this? It is possible you simply have a spring force bias and that you need some minimum actuation pressure to get the valve to move. If that is the case your solution is easy.

If you do have binding you need to get rid of it. It's not worth messing around for a month coming up with a software fix for a bad component.

This shouldn't happen because of valve design, but it may be that the pressure of your media is helping to seal the valve closed. If this is the case, you would expect the valve to require some minimum pressure to open. However, when it opens it will open in a step, probably pretty close to the position expected from the command you give. As you decrease your output from this initial open point toward zero, the valve will act correctly until some lower limit, at which point it will snap shut.

Keith
 
I am sorry guys i got back to this thread very late. Actually i was in the process of getting some data about the valaves Bench settings and Transducers pressure settings. Here are the numbers:


Operating Range (I/P output) 6-30 Psi
Bench setting for the valve 14 - 30 psi
PLC card output (4-20 mA to I/P)

Which means that valve will not start to open until I/P has a pressure around 14 psi(about 9mA output from Output card, 33%). If the I/P operating range is not equal to Valve Bench Settings, experts say we need to add a positioner, What are your comments on it? Also if the positioner is needed, can we eliminate the need of it and work around in our plc logic or output card span to make these I/P/Valves working properly. Thanks a lot for your input
 

Similar Topics

Hello Friends, I am looking for a way to take my PID Output( it is the result of my PI Algorithm in Real format) and generate a PWM Signal in...
Replies
16
Views
2,320
Hello, I have a flow control PID that keeps locking up. It seems to control fine but after a while the output no longer moves. For instance...
Replies
4
Views
946
I have an electric actuator with a 75% duty cycle. I am sending an analog output from a PID to a controller board on the actuator. My concern is...
Replies
6
Views
2,386
Hello. I am trying to figure out how to setup a PID in Control Logix 1769 Controller in a way that it would control a drive speed by sending...
Replies
3
Views
1,598
In Studio 5000 Logix Designer, I've noticed some odd behavior of the ladder PID instruction. Using the .MINO tag, you can set a minimum output...
Replies
7
Views
3,921
Back
Top Bottom