Control valve positioner

R

Ronald Deepak

What George said earlier is true. The positioner is in fact working in a cascade (inner) loop while the DCS is controlling the process variable (flow/pressure/level,etc.) From what I know, the positioners in fact have integral gains, so the error in the main process variable is not accumulated. However, like I mentioned earlier, there is an integral dead-zone for the valve travel, so the integral action is disabled when travel varies within this zone outside the set-point. So practically, the integral gain is almost non-existent in most positioners because the dead-zone is set high enough so that the valve position does not oscillate.

However, I do not quite agree to the comment about the usage of positioners in fast loops. At present, positioners are used in all kinds of loops and I don't see any problem with using a correctly sized valve and optimally tuned positioner in any kind of flow loop.

To answer your question about pressure drop :

The flowing pressure drop is used for Cv calculation. You might be aware that

Cv (approx.) = Q * Sqrt(SG/dP)

Where, Q is the vol. flow rate, SG is fluid specific gravity and dP is the pressure drop. In case of gases/vapors, the equation is different -dP is replaced by (dP/P1).

The shut-off pressure drop is used in actuator sizing. Like you said, the shut-off pressure drop occurs when the valve is fully closed. So this value is used in calculating the total actuator thrust required to keep the valve closed, because the actuator has to keep the valve closed opposing this pressure drop, in addition to providing the desired leakage class and overcoming friction.

In short, flowing pressure drop is used for Cv calculation and subsequently for sizing the valve. Shut-off pressure drop is used for sizing the actuator.

Hope this helps. If you still have questions, please let me know.

Ronald Deepak
jronalddeepak [at] gmail.com
 
R

Ronald Deepak

Harold,

Thanks a lot for pointing it out. I used the term %opening because it is frequently used these days by valve vendors and end-users to refer to % travel. In fact, the more appropriate term is % travel (this is the term that is used in ISA specification sheets for control valve).

However, I didn't refer to the actual flowing area or the effective opening inside the valve that is available for flow. Only the total effective flow area depends on the valve characteristic.

Ronald Deepak
[email protected]
 
The problem I have is in understanding what I call, the resolution of the unit. How much of a change in current does it take to create actual movement for the positioner, and what is the specification that defines this? Every vendor defines their product by what appears to be whatever spec puts it in the best light. Yes, they do, for the most part, use common terminology, and maybe I'm just stupid, but none of the spec's clearly identifies the 'number of step changes' available for a given 4-20mA signal.

mshopkins [at] dow.com
 
R

Ronald Deepak

Mr.Hopkins,

I would like to answer your questions. First of all, when it comes to the issue of resolution, it is more appropriate to talk about the resolution of the entire control valve assembly (valve,actuator & positioner)than that of just the positioner alone. With the advent of smart digital positioners, the valve&actuator assembly is the one that limits the resolution of the entire unit, not the positioner.

Like you mentioned, the actual movement of a valve in response to a control signal will not be a smooth ramp, but in 'steps'. The question is what is the minimum step change a valve exhibits (this defines the resolution)? Friction inside the valve tends to 'hold' the valve or make the valve stem "stick". Stem packing, seal rings, Actuator O-rings - all of these account for friction and subsequently contribute to dead-band. When there is a significant difference in static & dynamic friction values, the valve tends to exhibit "stick-slip" motion and this affects resolution. In the case of large size valves, the valve & actuator internal components like valve plug/stem, actuator piston (in the case of piston actuators)themselves are very heavy and the sheer forces involved in moving them, in addition to the friction components already mentioned play a significant role in determining the response of the valve.

Having said all this, today's positioners are tuned to handle such large valve constructions to meet typical resolution requirements. Yes, there is no universal specification that specifies an acceptable resolution for a control valve assembly. But several End-Users and Engineering contractors specify the resolution for valve response. Some applications like compressor anti-surge control require precise response to control signals and accordingly the resolution is specified. Most valve assemblies offered these days have the ability to respond to control signal changes as small as 1%.

When highly precise control is required, for ex., in applications like pulp flow control to the paper machine in the pulp & paper industry, basis weight control valves with Electric actuators are used. These valve/actuator assemblies have very high resolutions (somewhere in the range of 10,000 steps for a 0-100% control signal span i.e, a resolution of 0.01%).

However, the resolutions offered by standard globe valves with pneumatic actuators are good enough to handle most other applications encountered across process industries.

Hope this helps.

Ronald Deepak
jronalddeepak [at] gmail.com
 
How do one calculate the actual air consumption of the positioner?
I know that the amount of air used in general for sizing is approx 0.6 scfm to 1 scfm.

But the actual consumption varies for different manufacturer. What data do one need in order to calculate the actual air consumption of a valve positioner? and how to calculate it.

Thanks in advance
 
Yes, there is a big difference in air consumption between manufacturers because there are a number of designs, most of which I think got installed in our power plant. We've got 'em all.

The Siemens and ABB digital positioners move the control valve using fast acting valves, similar to fuel injectors. You can hear them rapidly fire their valves when they're maintaining a control valve in a position if the control signal varies or if the valve's pneumatic actuator is leaking.

The advantage of this design is that air consumption goes to zero in an air-tight system, unlike the old nozzle and flapper types that use far more air, particularly when they're applying zero pressure to the actuator. It doesn't shut off the air supply to the actuator--it vents it off.

We've gotten years of reliable service out of the ABB's in indoor applications on small valves that aren't subjected to vibration or water hammer. The Siemens PS2's weren't so fortunate and we've had to replace them in harsh applications. They do offer an upgrade for harsh duty, but we I&C types tend to cull out the weak in favor of the reliable.

We've gotten good, reliable performance out of Masoneilan SVII's on single acting actuators. They use pneumatic relays which use more air and are less precise than the ABB's or Siemens but they make a lot of mounting brackets for a variety of control valves, and that means a lot for the techs who have to install and calibrate them.

One of our plants had problems with double acting SVIIs on drum level control which were never resolved. They were replaced with Flowserves' Logix positioners that are easy to configure, reliable and reasonably priced. They're a definite favorite.

We're impressed with all the gee-wiz firmware and software available that gives you advanced diagnostics and a zillion configurable parameters but for all of its 50 plus parameters on board the PSII's, we prefer the simplicity of the Logix and we haven't got time to play with all the bells and whistles. If a positioner is causing process problems and can't quickly be corrected, it gets replaced, not studied.

PMV's are primitve by todays standards but the only thing that's gone out in 10 years of reliable service on several of them is their position feedback to the DCS which isn't critical.

Fisher DVC 5000's are one of the few that are accepted in nukes because they're rock solid, too, but overpriced and are designed to fit their valves. When applied to non-Fisher products, they tend to require a lot of levers to translate stem travel to rotary motion which creates a lot of linkage slop. We've replaced those with Logix.

Unfortunately, we've got other manufacturer's positioners such as STI on CCI DRAG valves (very robust but difficult to callibrate) to Siemens nozzel/flapper whatever. They all have their pros and cons, but do us maintenance guys a favor: standardize when possible and don't let the gee-wiz factor sway you unless you have lots of play time.
 
Hi there,

Thanks for the effort, very interesting read.

When you mentioned Siemens PS2 I all off a sudden felt sick. Biggest crap the world have ever seen.

Made me think about how some people went on in another thread about Siemens and their very advanced technological knowledge and better than most PLC's(S7).Might be true but then they should stick to PLC's and stay away from trying to design SMART positioners.

In contrast the Fisher DVC 6000 series from Fisher is the best I ever worked on. We mounted them on Valtech valves as well without any problems.
 
I have used the PS1 and PS2 positioners. I recently went through a process of standardizing basic service positioners across a site. Had 6 different types of positioners due to various sections of plant being package supplied. Siemens won as e PS2 was comparable to the DVC5000 on cost and ease of set up - doesn't require a 375. Diagnostics is still new to us but we looking to start utilizing existing diagnostic packages to aid in determining maintenance strategy for valves.
 
Sir,
I have a doubt. Suppose a control valve with a pneumatic positioner receives instrument signal from an IP. If the I to P is disturbed.. Then which is the better way

(1) To Calibrate the I/P with a master gauge separately and then calibrate the positioner

Or (2) calibrate the positioner first and then calibrate the I to P with respect to the positioner..

Please give me a reply..
 
N

North Sea Tiffy

Consider this, please.....the more error factors that exist within the system components, the greater the overall error will be within the complete system.

I see much very informed advice in this thread, but if you will indulge me a moment, I would like to take you back to some more basic elements of setting up a valve positioner.

First of all, set up the basic v/v and positioner alignment (mechanical alignment) and stroke length for the given v/v.
The safest way to do this is to disable the positioner's ability to control the valve, by removing its supply air and output tubing.
Connect a regulated air supply directly to the diaphragm of the valve. Slowly increase the air feed until the v/v is at 50% of its travel, and maintain it at this position.(Do not exceed the stated max working pressure for the diaphragm!)

Next , check the data plate on the valve for its stroke length. This figure will be used when setting up the Stem coupling drive pin on the positioner's feedback arm graduated scale(for example, it is typically given on a Kent P3300 positioner as being X mm, in multiples of 10mm).

With the v/v still in mid-position, carefully align the positioner feedback arm as horizontal as you can judge it to be, and engage the drive pin in the feeback arm slot , setting it at the required figure derived from the data plate.

Tighten the stem coupling arm, (use a suitably sized spanner to counter the movement of the stem coupling arm as you finally tighten up).

Finally, check the following:
a) that the v/v stem is still at mid-position,

b) that the positioner feedback arm is horizontal

c) that the drive pin is still located in the feedback arm drive slot at the required length as above.

Basic alignment is now complete, and you may remove the regulator and tubing which you used, and re-instate the positioner air supply and output tubing to its original set-up

Next, check and adjust I to P calibration
Finally, calibrate the positioner to your requirements, (assuming full range linear application, a 4 to 20mA input producing 0 to 100% travel)

Check performance/linearity and repeatability in 25% increments upscale and downscale.

Bingo...job done,with all sources of error having been eliminated.(assuming correct cam profile etc are set)

The short of it is , that the more accurately you set up the individual components, then the more accurate will be the overall result

Hope this is of help to you !
 
N

Noth Sea Tiffy

Quite correct Dilip...as we say in the North Sea,
"There is no such thing as a daft question...there is only the daftie (fool), who doesn't ask it !"
 
Hi guys,

I'm doing a research project on 4 positioners.,two smart which are the Fisher DVC6200 and the Masoneilan SVI IIAP. I'm also doing Fisher 3610J and MAsoneilan 4700P for pneumatic positioners. I was wondering if anyone could give me any insight on what applications pneumatic positioners would be best for. Also i would like to know out of four positioners what one would you recommend for tight shut off applications and throttling.

Thanks,
chris
 
Both smart valve positioners are available with either HART or Foundation Fieldbus input and are the most adaptable choices. Pure pneumatic positioners with 3-15psi air pressure input can be used with DCS or electronic controllers if a I-to-P (I/P) converter is used to convert the electronic 4-20ma signal from the controller to the requisite 3-15psi air pressure. The only reason I can think of where I would use a pneumatic positioner is where there is an explosive gas mixture present at all times such as in chemical plants making ether or acetylene. Otherwise, smart positioners are always preferred unless their higher cost is a problem. The purpose of all control valves is to throttle the process flow over a narrow range of flow rates.

Tight shutoff cannot be expected with any control valve; the valve itself determines if tight shutoff can be achieved not the positioner; and globe valves used for control are not designed for tight shutoff. If you need tight shutoff, you should use solenoid operated quarter-turn ball valves that are unsuitable for control purposes.

Dick Caro
 
F

Fieldbus query

Hi Chris,

Tight shutoff for conventional pneumatic positioners with 4-20 mA is a functionality driven by the control system it is connected. For eg: if tight shutoff needs to be done the control system will deliver 3.6 mA or sightly less and for full open the current will be 21mA

if you take the case of foundation fieldbus positioners this functionality resides in the positioner and is tuned by appropriately making changes in the parameters of TB01 (tranducer01 block)
Not sure I thinks it can be done by changing

- final cutoff value_HI =110
- final cutoff value_LO =-10

OR
- Stop value_HI =110
- Stop value_LO =-10

Coming back to throttle / Stroke
you can throttle conventional positioner, if you feed mA from 4to 20(4mA=0%,8=25%,12=50%....)

Foundation fieldbus positioners - its done by digital communication and needs a host to execute the commands
 
Hey guys,

Thank you for the insight for what applications would best be suited for pneumatic positioners as it is helping me with my report. With this report i am basically trying to decide and convince my college faculty which positioner is for best for use out in the field and why. This is between the DVC6200 by Fisher and the Masoneilan SVI II Ap. If there is any opinion or features that will help defend my argument that the Masoneilan SVI II AP is better it would help greatly.

Also when looking at the calibration of the DVC6200 i noticed there are two modes of calibration which are travel calibration which consists of manual and auto and sensor calibration. The DVC6200 uses a HART communicator for these modes of calibration. The SVI II AP by Masoneilan uses push buttons and a LCD display for calibration and also can incorporate the HART communicator like the DVC6200. Does this feature give the positioner an advantage over the DVC6200

Also if there are any other advantage or comparisons between the SVI II AP and the DVC6200 that i should be aware of it would help out very much.

Thanks.
Chris
 
Hello to all!

First off: there has been some great discussion with some very qualified individuals in this thread.

I currently work in the nuclear field with the title Field Service Engineer on AOVs. The main advice I can offer to newbies is to tinker with all types of positioners on a bench. Have a qualified individual nearby to assist you if you need it and to also check your work.

In my line of work, we have realized that on a 3-15 positioner, setting the values to 3.2 and 14.8 Psi ends up being the optimal setpoints for an AOV.

3.2psi meaning the diaphragm is completely bled off at this point and valve travel has ceased.

14.8 means the the diaphragm is saturating with air and the valve travel has reached its peak and cannot move anymore.

Also, if you are working with Fisher positioners (3582) and they are straight out of the box, more than likely a beam alignment will need to be performed before calibrating the ZERO and SPAN setpoints. Also, Adjust your span before adjusting the zero. A 3582 positioner is different than most instrumentation and adjusting the span before the zero will eliminate excess time spent calibrating the positioner.

As to Chris's question: the nuclear field in the states uses mostly Fisher and ABB brand positioners. DVCs are making there way but only on the secondary side.

I presume the Fisher 3582 over just about any other positioner, except when I'm dealing with dual acting actuators (Piston actuators). Then an ABB Baileys is a great instrument for the configuration at hand.
 
Top