Finer control of pressure on upstream of valve 2

[jeyaselvan]

In a compressor test rig, we intend to use an automated throttle valve to load the compressor at finer resolutions of say 0.05 bar or better at working pressures of 5 to 15 bar.The pressure downstream of the valve is near atmospheric. Currently we are using a 1" ball valve with pneumatic actuator and positioner with i/P converter. We were unable to achieve / maintain / adjust to the required upstream pressure due to dead band, hysteresis, etc. The operating pressures range from 5bar to 15bar,flow from 15 to 350scfm, temperature around 60 deg.C. I am on the lookout of alternate products / techniques to sort out this. Any suggestions?

 

[JimCasey]

If you just grabbed a ball valve off the shelf and slapped an actuator and a positioner on it, you're lucky it works as well as it does.

A >>CONTROL<< ball valve has:
(1) Zero lash connection between the ball and stem
(2) Characterized port carefully sized to match the capacity demands of the application.
(3) zero-lash connection (Clamped) between the actuator and the valve stem, and between the positioner and the actuator.
(4)Valve, actuator, and positioner all rotate about the same centerline.
(5) Actuator sized with generous safety factor to minimize the effects of friction, and sized with the ability to overcome valve breakout torque throughout its travel, not just at the end points.

(6) Premium positioner with 2-stage relay or inner-loop feedback to make it more responsive to small incremental changes.

(7) seat material selected with consideration of throttling service, not just resilient for shutoff as used in an isolation valve

An off-the-shelf isolation-type ball valve has 2-3 degrees of lost motion in the stem-to-ball connection, and more lost motion in a standard milled coupling. An "economy" positioner is a single-stage, force-balance unit that may not even detect an incremental change of 1% due to internal friction.

If you want excellent control, you need an excellent control valve.

Also, and just as important, you need to tune the control loop to respond appropriately to the system dynamics. It is entirely possible to have the system tuned so poorly that you won't get accurate control with the best valve in the world. A control valve just goes where the control loop tells it to go. A good control valve goes exactly where the control loop tells it to go.

 

[jeyaselvan]

Thank you Mr.JimCasey for your wonderful suggestion.I totally agree with you that the control valve is as good as the control loop.
I am little more interested in knowing more about the premium positioner, which I have never come across. As put out by you, the off-the-shelf isolation-type ball valve has 2-3 degrees of lost motion which I could really visualise in my application, wherein the valve could not respond to signals of upto 0.5 - 0.7 mA in a typical 4 - 20 mA electropneumatic actuator cum positioner, and at 0.8mA it overshoots, which is unsuitable for my application.

As put out in my problem definition, my application demands for precise control of valve upstream pressure.I am not sure as to the controllability of electric actuator. Any views on this? Of time, I am also looking for stepper motor controlled actuators, wherein much better positional accuracies are claimed. Your views please. I fully agree that the overall controllability is not only dictated by the actuator but also on the backlashes on the various stages of the valve assembly. Of sure I will going for characterised port for the application.

 

[JimCasey]

I assumed you were using an air-operated valve. There are premium electric actuator positioners, too.
Worcester, for one, offers a digital electric actuator positioner that is used in their type 75 electric actuator.

I have also always thought that a stepper motor would be a good idea, but I don't know of anybody using a stepper motor to drive a valve actuator. Lash in the gear reduction would present an engineering challenge to make the stepper as precise as hoped.

 

[jeyaselvan]

Yes, we are currently using pneumatically actuated ball valve with positioner and i/P converter. I am getting few alternate suggetsions to try out with globe valve instead of ballvalve citing that manufacturing/assembly tolerances in globe valves are little closer and backlash associated with them being lesser. Any suggestions?

Meanwhile I also come across Bernard electric actuators of modulation class I, wherein they claim precision positioning of 0.5% and better. Any experiences?

 

[JimCasey]

Globe valves are a proven technology that has been around for 200 years. Compared to ball valves, globe valves have lower capacity per size, are larger, heavier, more expensive. Since globe valves do not need internal elastomers, they have a higher temperature limit than ball valves. But without internal polymers, they don't shut off as tightly. Rising stem valves such as globe valves have greater packing leakage than rotary valves, because rotary packing does not have contaminants dragged in and out of it as the valve strokes.

I don't think the tolerances in globe valves are inherently any better, and there is much more machining involved in making a globe valve. Concentricity of the body/seat/plug/bonnet is an issue in globe valves.

The latest generation positioners used on globe valves have rotary input shafts originally compatible with rotary valves. To use the positioner on a globe valve there needs to be a linkage to convert the straight-line motion of the rising valves stem to the rotary motion of the positioner. If this is not carefully done, there is the possibility of nonlinearity in the linksge affecting the valve's response to a signal.

 

[hazel]

does anyone think to adjust the positioner is possible?
Why the signal need more than 0.8mA as the actuator start to move?
How do we reduce the damping?

 

[BSQUAREDBUCKEYE]

The gentlemen is right about a floating ball type valve having significant backlash in the ball to stem engagement. You may want to look for a characterized segmented ball valve that has a splinded and pinned connection eliminating the backlash. The Paper Industry uses a valve like this for an application called "Basis Weight". This controls the thickness of the sheet of paper on the paper machine. Some machines travel at thousands of feet per second so an error can add up very quickly. They typically use a "stepper motor" attached to a zero backlash ballscrew train. This produces a control valve assembly with zero backlash, zero hysterisis and thousands of position increments. The input signal is typically pulse since one can get fine enough resolution with 4-20mA. Oh, and they are expensive... $10K

 

[sterl]

Plenty of Globe Style valves out there with high resolution steppers. Long stroke devices usually can be characterized linearly and very accurately in terms of flow, which can yield large advantage when devising your control loop...

Their limitation is essentially Speed, at least a couple can take more than 2-minutes to stroke 0 to 100% and of course the breakaway condition from Open to Close requires a fairly high motor torque so it implies a "lag" from signal to winding response in angular terms.

Some have Full Balanced Port arrangements; many operate without gears but will have lash in their rotary screw (which I suppose is just a worm gear, after all).

The stepper motors can be set up to offer:

Linear, slow opening or high initial flow characterizations

Tight Seating at Full Closure

inherent positional feedback

Longevity from the seat materials

Resistance to Flashing or erosion

Very high pressure difference capabilities

Fail safe (open or closed as required) control arrangements

torque limiting / force limiting

a Balanced Port Globe will take care of some of the seated-force inclemencies.


and so on:


But I am not sure why you need any form of external actuator...Unless you require a very precise pressure control, a very high turndown, or are working with a pretty ugly fluid: there are pilot operated, no-external-moving part devices out there, some of which even offer positional feedback...Kunkle, Metrex, Danfoss, Egelhoff, Hansen and a couple of others all offer such devices, that utilize the pressure difference across the valve to effect the required positioning.