Using FCI-70-2 or API-598 1

[srhash]

I am working at a petrochemical complex. We have a lot of control valves that need to be tested during Shutdown for leak tests.

I am a little confused as to which standard should be used for valve seat leakage tests and hydrotests (shell test). As i understand, FCI-70-2 is specifically dedicated to control valves testing for seat leakages. On the other hand, API-598 details both seat leakage tests and hydrotests along with other tests. API-598 is not specific to control valves and there is no mention of class-wise segregation of valves and the corresponding allowable leakage rates.

Also, at the same time i have valves wiht classes defined (e.g. Class-V) which are only performing isolation (emergency shut-off) and no other controlling function similar to that of a normal control valve; should i test their leakage rates based on FCI or as per API-598?

Would really appreciate some feedback and comments on practices used elsewhere and which of the 02 standards to refer to.

 

[Danlap]

This is a broad topic with no real precise answer.
It really depends on the end user agreed standard or Recommended Practice or how the SIL analysis view it. Country or governing NOBO which regulates the end user standard.

In general there are 4 known Valve testing criteria (Manual and or Control) aside from PSV which are: API 598, API 6D, FCI-70-2 and ISO 5208.

Control valve general interpretation is Manual valve with actuator (pneumatic, hydraulic, electric) which regulate the valve movement (Fully Open to Fully Close).
Therefore, shall the actuator already tuned (according to Manufacturer standard as new built), it should comply with FCI-70-2.
This is closely related to seating force (fully closed position). Small difference in the seating force value (which driven by the actuator setting) have a big difference on seat leakage criteria.

However shall this valve being tested as a manual valve (shall possible) or without actuator mount on it, I would recommend ISO 5208 as a standard.
API 598 is quite general (it states general material and general minimum testing pressure) and you should combine it with ASME B16.34. And ISO 5208 already taken into account API598, even though you still need to see ASME B16.34 in accordance with MAWP (further Shell test is 1.5xMAWP, Seat 1,1xMAWP, and so on)

Just some metaphor: A really strong operator could over tightened 2" Class 300# Metal to metal Globe Valve with hands and reach Tight Shut Off (Class A ISO 5208), but not necessarily reach the same result (Class IV of FCI-70-2 or higher) with fine-tuned actuator.

Others probably have different view.
Regards,
MR

Greenfield and Brownfield have one thing in common; Valve(s) is deemed to "run to fail" earlier shall compared to other equipments

 

[mbelly84]

Muktiadi, i think the answer is much simpler.

He should use FCI 70-2 for seat leakage tests, for the simple reason that API 598 / API 6D allowable leakages are meant for shut-off valves, therefore much more restrictive than FCI 70-2. Control valves are meant for control, not shut-off.

As for the shell test, it's ok to use API 598 / API 6D, since FCI 70-2 doesn't cover such tests.

 

[srhash]

@muktiadi and @19081984 >>> Thanks for the feedback.
Would welcome more view points on the discussion from other members.

Also, i need an answer to the question as in my original post; that if i have a Class-V valve (defined as control valve in datasheet) but only performing the funtion of a shut-off valve (EIV); do i need to carry out it's seat leakage test based on FCI-70-2 or simply follow API-598?

 

[Danlap]

Hi 19081984,
Thanks, Agree with you. The answer could be much simpler. Just want to broaden the view for others.

Hi srhash,
Is this a control valve (Open at certain position to choke the flow) which also has function as an EIV / SDV?
Not sure about the type, size and connection (Butt weld or flange) of it. So I am taking a guess of its complexity
Upon shutdown, does this valve refurbished? If yes, then it should be considered as new and respect the datasheet (Class V according to FCI-70-2). Shall not, contact the Process Engineer and re-evaluate the Safety aspect/SIL analysis whether Class IV or lower is also acceptable. Use the same standard as stipulated on the datasheet.

In respect with its function (EIV / ESDV / etc.), shall something goes wrong in the Plant then the valve must be isolated to avoid exothermic chain / catastrophic in the downstream side.
Use FCI-70-2 for Seat test. What leakage rate threshold is acceptable to avoid catastrophic until the other ignition problem (fire, etc.) are solved?
if you have some tools such as Valve diagnostic or even a good Control Room system, Use OIM and datasheet as threshold to measure other parameter (Closing time, Hysteresis, Seating force, etc.).

Even though no emergency ever occurred (valve never being self-operated), however on service and after the warranty years is due there are still risk of wear and erosion on the internals which leads to lower Seat tightness Class.

Regards,
MR


Greenfield and Brownfield have one thing in common; Valve(s) is deemed to "run to fail" earlier shall compared to other equipments

 

[srhash]

Hi Muktiadi,

The question is not specific to a particular valve but to many such valves.

These valves remain in open position (100%) and function as EIV when the upstream supply is cut or some abnormality downstream of the valves (fail safe close).

In one such case this is a 8", 900# RF flange connection type valve. The plan is to leak test the valve; soft parts will be replaced; however decision on ball/ plug/ seat replacement will be based on as-found conditions. Additionally, this is a SIL rated valve.

From Process/ Production perspective they always desire "perfect" scenarios which don't necessarily exist; so for them ideal leakage rate is ZERO (0)

I don't understand what u meant by using other diagnostic tools/ measuring other parameters. How can they help in valve leakage testing decision making??

Also, i agree to your last point that years of service even without operation have their due impact on the internals and decrease in valve seat tightness.

 

[Danlap]

Hi srhash,

Agree with you, zero leakage can only be achieved with a very expensive valve on a very clean service (let say pure oxygen, etc.) however within some period of time only.
That is why there are leakage class standard which in favor for manufacturers, since most of process requirement are not necessarily requires Tight Shut Off.

Measuring control valve parameters are not directly related to leakage testing itself, however it is assurance of how it will comply with the SIL parameter.
For instance:
--> Emergency occurs (time travel from fully open to fully close) under x seconds?; A lag on closing time can be catastrophic.
--> What voltage / current required for its to be seat tight? This can be converted so you will gate Seating force.

​

Excessive seating force may damage the seat and or ball/plug/disc, even though it may perform as tight shut off for the first time.
--> Hysteresis. Lag time between manually override (from control room) input to be received by positioner / actuator in order for it to start moving. Etc.
This activity is normally done by the valve field engineer or manufacturer representative.

Apology for expanding this to broader horizon. Complex refinery requires assurance not just commissioning report after start up but also reliability of the valve until the next shut down.

Greenfield and Brownfield have one thing in common; Valve(s) is deemed to "run to fail" earlier shall compared to other equipments

 

[srhash]

@Muktiadi >>> Thanks for the info and detailed clarity.