Control valve maintenance strategy

Hello there,

seeking your experience in the field of control valve best practices in term of preventive maintenance for control valve while the valve in operation mode
 
Check these notes from the web:

Periodically inspect and perform maintenance service to the control valve for the following: Check that there is no air leak from between the diaphragm case and diaphragm. Check that there is no process fluid leak from the bonnet and flanges. Check that the yoke clamp bolt is not loose.


How do you test a control valve?

Low pressure closure test
  1. The test flange must be installed at one side of the valve, one side will be injected with pressure air and the other side will be filled with test liquid.
  2. The valve must be closed fully.
  3. Air pressure must be injected into the valve after that snoop at the inlet point of the valve.
https://automationforum.co/how-to-d...-what-are-the-types-of-control-valve-testing/

https://instrumentationtools.com/installation-and-maintenance-of-control-valves/


You can search on the web there are many good informations on the subject!

Also one has to check the OEM or O&M manuals for that purpose....

Do you have a OEM name ?
 
If you have smart positioner such as fisher DVC6200 , you can do online diagnostic testing. This will give you a good diagnostics information about the healthiness of the Valve. There should be S/W to be used such as Valvelink in Emerson/Fisher positioners that can be installed on Asset Management System ,if you have at your site.
 
Hello there,

seeking your experience in the field of control valve best practices in term of preventive maintenance for control valve while the valve in operation mode
Preventive Maintenance means routine checks. Cleaning, air leakage checking, gland leakage checking etc. are some normal stuffs which are generally done for a control valve. Apart from that depending on your plant's risk taking for a particular process you may devise other stuffs too like opening/closing a valve to some particular position and checking the time, stiction etc. but then again it all depends on a plan and close coordination between the process and the maintenance team, documenting it and following it.
 
The whole HART diagnostics thing is supposed to be about using the data in HART diagnostic counters that run continuously in the valve positioner and then analyzing that data to tell you which valves appear to need work.

The concept is predicated on having (relatively expensive, licensed) Asset Management Software with the functions to fetch and display the diagnostic data from the positioners along with the associated HART pass-through analog hardware needed to talk to and communicate with the positioner. I suspect that transmitters equipped with Foundation Fieldbus have the same functionality, but would also need the Asset Management software to deal with the diagnostics data.

My one run-in with a process engineer at a toilet paper manufacturer (where the machines run 24-7 except for a 14 day turn-around scheduled long in advance) revealed that the 200 odd hours (on an annual basis) he spent analyzing the data was well worth it because of the relatively limited resources available during the turn-around. The diagnostics enabled him to determine which valves needed an overhaul and which could wait, so that the resources could be targeted properly.

It was also evident from that conversation, maybe 12 years ago, that there is no cheat-sheet telling you how to interpret raw diagnostic information. It takes an experienced process engineer who understood the process, the media and valves to make judgment calls from the raw data about valve condition. No red light, yellow light, green light assessments from the software.

Maybe all that fluff about artificial intelligence has worked AI into industrial diagnostics, but I doubt it.
 
The whole HART diagnostics thing is supposed to be about using the data in HART diagnostic counters that run continuously in the valve positioner and then analyzing that data to tell you which valves appear to need work.

The concept is predicated on having (relatively expensive, licensed) Asset Management Software with the functions to fetch and display the diagnostic data from the positioners along with the associated HART pass-through analog hardware needed to talk to and communicate with the positioner. I suspect that transmitters equipped with Foundation Fieldbus have the same functionality, but would also need the Asset Management software to deal with the diagnostics data.

My one run-in with a process engineer at a toilet paper manufacturer (where the machines run 24-7 except for a 14 day turn-around scheduled long in advance) revealed that the 200 odd hours (on an annual basis) he spent analyzing the data was well worth it because of the relatively limited resources available during the turn-around. The diagnostics enabled him to determine which valves needed an overhaul and which could wait, so that the resources could be targeted properly.

It was also evident from that conversation, maybe 12 years ago, that there is no cheat-sheet telling you how to interpret raw diagnostic information. It takes an experienced process engineer who understood the process, the media and valves to make judgment calls from the raw data about valve condition. No red light, yellow light, green light assessments from the software.

Maybe all that fluff about artificial intelligence has worked AI into industrial diagnostics, but I doubt it.
Thanks for your reply David... Kindly I have a question. What about ESDVs/BDVs/XVs maintenance strategy during turnaround??? How you select the valves to overhaul during turnaround?? As you know these valves not having smart positioners like control valves..
 
The move towards partial stroke testing is to exercise a test that assures operation and avoids the wait-until-shutdown for testing and by that point, emergency repair.

There's a relatively reasonable overview of the partial stroke testing (PST/PSVT) and its 3 implementation methods here:
https://automationforum.co/partial-valve-stroke-testing/

One of the methods, mechanical jamming is discussed in a report on PST implementation at a chemical plant, here:
https://www.sofisglobal.com/mechanical-partial-stroke-test-reveals-hidden-esd-valve-failures/

Without testing, there's assurance or data on whether the valve works or not. Like all things, there's no free lunch.
 
Top