Positioners needed or not needed

Hi All
Customer had put me in a some what difficult scenario. I am working at a Gas plant which has lot of fishercontrol valves (D2 or L2 etc). Most of these valves are bench set at 14 to 30 psi (14 to start open and 30 to fully open). Their respective I/P are set at 6 to 30 psi....so with this difference there is a delay of opening these valves...around 40% plc output. But when they are open then they tend to control fine.....Customer has already made his mind that they need positioners.....and he wants me to suggest him to install positioners on these valves.
According to me, they do not need positioners...i know positioners will help to make the opening and closing of the valve faster....but if the process is not stable and we see huge swing in temperature of reboiler from 140 to 170 degree F even the valve is Snap shut.......then even positioner under these circumstances will do nothing.....
What you guys suggest.....do we really need positioners to install.....even though valves are controling just fine (except the initial start)? When the valves are in the range of 40 to 70 % they throttle great.
Your input on this will highly be appreciated and will able to guide my customer in a better way.

plcnovel said:

...Customer has already made his mind that they need positioners...

Click to expand...

That settles it for me. Don't argue with the customer. Install positioners. Optimize operation. PID control will help with the wild swings.
The positioners may not solve all problems, but some things will improve, and the customer will be happier.
I'm not that familiar with your valves - this is more of a customer relations answer.

The present setup provides for output control over a range of 33% - 100% (14-30 psi range in a total span of 6-30 psi)

Have you considered re-ranging the output of one of the I/P's for 14-30 psi instead of 6-30?

That would mean the valve would swing 0-100% in tandem with the control output of 0-100%; as compared to the present 33%-100% output control.

It would somewhat mimic the response of a positioner.

Dan

Add the Fisher DVC positioners. You won't regret it.

Positioners

I would install positioners on just about every control valve. I have been told that positioners make the response of the valve sluggish. But in my experience even on fast loops such as vacuum control the valves responded just fine.

I worked on a chemical plant startup that, to save money, did not install positioners. These valves were fisher control valves and a couple of samson valves.

We had a bunch of issues with control. We always came back to checking the bench set which is a major issue with installed valves. With a positioner, you can be fairly confident that a 50% controller output opens the valve 50%.

I cannot understand how a loop would control when the first 40% of its output has no effect on the process. I have encountered issues with face & bypass dampers on air handling units feeding process equipment. Often the first and last 20% of damper opening has no effect on the temperature control (no heating until about 20% and full heating at 80%). I have scaled the PID so that 0% output corresponds to 20% damper opening and 100% output equals 80% damper opening. This type of setup provides very good response.

Hi,

Can someone post a link to a typical "positioner" so I can get a mental picture of what y'all are referring to? Thanks
BD

bikerdude said:

Hi,

Can someone post a link to a typical "positioner" so I can get a mental picture of what y'all are referring to? Thanks
BD

Click to expand...

http://www.lesliecontrols.com/Products/actuators_posit.htm

Basically, for pneumatic control valves, you can either pump air into the hat, and try to control the valve CV, or you can use a positioner on the valve, and give it a command to open 10, 20, 35, 75% etc.

Positioners are a huge advantage in dealing with control valves, as the valve actuation is decoupled from the back pressure on the valve body. That, and you can precisely position the valve. Without a positioner, it's pretty random how the valve trim will move with applied air, and it will change from day to day depending on pressure across the valve body, wear of the valve packing, rust, humidity, and just about every other factor you can name.

Put a positioner on, you'll be glad of it.

Positioners provide 3 huge advantages over using just an I/P and the valve.


1: Speed of response - the positioner shifts alot more air alot quicker than an I/P can.
2: Accuracy - the positioner KNOWS where the valve is due to the feedback arm and will modulte its output to maintain that regardless of valve stem stiction, wear etc.
3:Scalability - No matter what the bench set of the valve or the input signal the positioner can be set up to give full stroke and positive shutoff of the valve.

To sum up, positioners are a must on any valve used for process control.

danw said:

The present setup provides for output control over a range of 33% - 100% (14-30 psi range in a total span of 6-30 psi)

Have you considered re-ranging the output of one of the I/P's for 14-30 psi instead of 6-30?

That would mean the valve would swing 0-100% in tandem with the control output of 0-100%; as compared to the present 33%-100% output control.

It would somewhat mimic the response of a positioner.

Dan

Click to expand...

Hi All
Thank you very much for your responses.....The problem with the I/P i am using is that it can only set to two ranges 6 to 30 psi and 3 to 15 psi......what i can do is to represent my output from the plc something like that right now 0% is 4mA and 100% is 20 mA, so i can shift my 0% to 9mA or what ever it takes the valve to open. One disadvantage of doing this is that even though my output from plc is 0%(9mA) but the I/P will be producing the output of lets say 12 psi which would be on the valve all the time...what do you guys suggest doing it......do u think even though valve needs 14 psi to open, by providing 12psi all the time will harm or damage the valve or even possible leakage......

Thanks in advance,,,

plcnovel said:

do u think even though valve needs 14 psi to open, by providing 12psi all the time will harm or damage the valve or even possible leakage.

Click to expand...

Answer, No, it will not harm the valve. But . . .

But, the question is why the valve actuator needs 14 psi to open? Someone spec'd it that way and had it set up that way.

When a valve is in a dynamic flowing pipe, the applied pressures on both sides of the valve trim might tend to open the valve if only 3 psi is applied at 0% output. So it could be that a preset bench set of 14 psi was specifcally done to keep the valve closed at 0% output. If you apply 12 psi at 0 % output, you could be helping the vavle move off its closed position.

.....The problem with the I/P i am using is that it can only set to two ranges 6 to 30 psi and 3 to 15 psi......

Click to expand...

Considering everything, including the probability that valve needs to be set at 14 psi to open, why not look at replacing the I/P converters with ones with an adjustable range of 14 to 30. I/P converters are cheaper than positioners. Then re-evaluate the situation to see if positioners are also needed.

I would bet that the discontinuity was deliberate to prevent wild swings in the valves due to wide swing in temperature. In which case a temperature-averaging routine in the PLC might be all that is needed.

Firstly, where there is a mismatch between output ranges leaving a dead zone in the output it is easily solved by having an output LOW limit that does not allow the output to go below that limit. I always configure output HIGH and LOW limits in case they are required.

Regarding positioners, I always use positioners even though I know that with everything tuned properly they could in some cases be elliminated.

I am a great believer in cascaded control. For example in many cases of temperature control I cascade my temperature control to fuel control to fuel valve position control. I have always found that this strategy provides better more stable control, and when any problem arises it is easy to see exactly where the problem lies.

Hi,

The picture below (thanks for the link!) is similar to the valve actuators I am famaliar with.



That upper part is what rdast refers to as the "hat" I assume. I use these all the time bench set for 3-30psi. I use the I/P that have "zero" & "span" adustment screws. If the OP has these adjustments on his I/P he can tune the I/P output to match the actuator. Otherwise he will need to adjust the PLC analog output to eliminate the dead zone from 4ma up to whatever current is required to start the actuator movement & then back it off 'bout 3 psi so it will not remain "cracked" open when no flow condition is desired. I guess I have never seen a positioner in real life. I can't speak as to their value in this particular application. My thoughts are: if your system has direct feedback of the process such as a flow meter down stream of a valve/actuator, a positioner is a redundant element. I mean who cares what the actual valve position is? You just need to keep driving it one way or the other to acheive the goal. If your system doesn't have direct feedback, then a positioner will provide a set valve position for a corresponding PLC output. Not sure what applications would benefit from that setup. I have no clue what a "Gas Plant" does or what a "Reboiler" is. Sounds like a temperature control application though. If if is critical to your application to control the flow of whatever medium it is flowing through the valves in question, then you need to put some kind of feed back device in the process pipe to let the PLC know how it's doing. Whether that is a positioner or not I can't say from where I'm sitting.
Just had to throw in my $0.02 on the subject
BD

danw said:

Answer, No, it will not harm the valve. But . . .

But, the question is why the valve actuator needs 14 psi to open? Someone spec'd it that way and had it set up that way.

When a valve is in a dynamic flowing pipe, the applied pressures on both sides of the valve trim might tend to open the valve if only 3 psi is applied at 0% output. So it could be that a preset bench set of 14 psi was specifcally done to keep the valve closed at 0% output. If you apply 12 psi at 0 % output, you could be helping the vavle move off its closed position.

Click to expand...

Ok Thanks once again.....the reason i do not feel a need of a positioner right now is as follows: The application is like that, the liquids come from the top of a tower and go into a reboiler tube from the bottom and get out from the top after being heated. The heat medium in the reboiler is the hot oil circulation. My temperature trasnmitter is mounted on the top of the reboiler tube vertically. If there is no enough flow in the reboiler tube then all this element is measuring is the vapour not the temperature of steady state flow of the liquid. Thats why the temperature is over shooting and udershooting a lot......Once this flow is stable i am sure the control will be stable as well. Also all these valves are throtlling type valves which means they work best within a specific range, not fully open not fully closed......

Lancie1: thats a good idea of replacing I/P instead of positioners.....i will be looking into it if i think even after getting a stable flow, fluctuation is there. Also you mentioned something temperature averaging routine in PLC, does that mean take a average over a time...is there any specific instruction i can use or i have to right my own....also how would i determine the time span for a average if the overshooting and undershooting is wild........Thanks in advance

We have a large installed base of Fisher control valves many of which have bench sets on them. The requirements for a bench set is determined by the vendor given process conditions such as pressure drop etc., the valve size, the valve type and actutator combination, and shutoff leakage class required.

What surprises me is that positioners were not specified originally for the valves as you have a gross mis-match between the valve operating pressure of 14-30 psi and the I/P output range.

Instruments usually work with standard signal ranges of 3-15 psi, 20-100Kpa, 4-20mA and should be calibrated accordingly to ensure that correct operation is achieved over the entire range. Valve positioners also work with standard inputs as above but can give outputs which are able to match the required bench set for correct operation of the valve because they are essentially high gain proportional controllers who's ouputs can vary between 0 and 100% of the valve air supply pressure.

Valve positioners are helpfull in any kind of control loop except flow or liquid pressure control which are self regulating (Quote from Greg Shinskey who I reckon is one of the Guru's of Control).

Under stable conditions your loops will work correctly as will virtually any loop. What you need is the ability to control your loops under upset conditions as well (or at least be given a fighting chance to contyrol). With the current mismatch in your system you have very little chance of controlling anything under upset conditions as you have already stated. Until you address the mismatch problem you will not achieve anything. So I would suggest that you fit positioners and then you will need to re-tune the loops (They must have some strange settings at the moment anyway).