Seat Leakage Rate for Check Valves

[uaepiping]

As per BS 1868 the leakage rate given is Rate A for resilient seated valves and Rate C (As per BS EN 12266-1) for metallic seated valves.
API 594 states leakage rates as per API 598, which (in table 6) states all resilient seated valves to have ZERO leakage and different rates are listed for metal seated valves.
In ISO 14313, leakage rate for soft-seated valves is rate A and for metal seated check valves, the rate D is given.
All above standards for check valves require to have certain seat rates for soft and metallic seated check valves.
But one of our engineer is insisting, we should not include the seat leakage rate requirement in our check valve specification. As per him, the check valves leak all the times in backflow. So it is not needed. I am of the opinion, we should give these rates.
What piping engineers think in this forum?

 

[MickMc]

Typically in my experience we do not stipulate leakage rate for check valves in data sheet for the referenced reason but would just stipulate design code. We do for all other valves with maybe more stringent for severe service.

 

[Danlap]

Check ISO 5208, Quote "Rate A as applied to DN <= 50mm, Rate AA-Gas and Rate CC-Liquid for other than metal seated check valves and for check valves Rate EE-Gas and Rate G-Liquid" unquote.
I would always assume worst case condition regarding Check valve:
- somebody design it solely to protect compressor/pump from back flow or surge
- volume between protected comp/pump/etc. and (solely) check valve is small enough and cannot bare pressure built-up within short period (it might be initially only bubbles/minute but after days/weeks the pressure built up might be sufficient influence upstream component).
- check valve will slam/chattering frequently thus wear is unavoidable thus lead to higher leakage rate (shall compare to as new condition).
- the information of why check valve selected is not clear enough / even the person not available any-longer in the future to tell the design philosophy. So ones might guessing what would be the effect shall leakage rate lead to pressure built-up
- etc.

So yes, initial leakage rate (from factory/after refurbishment) need to be defined. Agree with MickMc for severe service, it might be wise to push the manufacturer a bit harder, even up to Rate B shall possible.
To pass the QA check list, manufacturer need to state something on their pressure test sheet anyhow. referring to some standard.

When its in Operation, you may consider API 521 for whether its going to be single or double check valve.

Regards,
MR



All valves will last for years, except the ones that were poorly manufactured; are still wrongly operated and or were wrongly selected

 

[uaepiping]

Danlap,
What do you mean by pressure buildup. When the pressure on downstream side of the check valve is more, the valve has to shut down. for pressure build-up release other devices should be there in system.
Question is should we give the rates in the check valve spec or leave it to manufacturer/engineering contractor to decide.

 

[Danlap]

uaepiping,
"the valve has to shud down"? what is the definition of shut down / close (in terms of leakage rate as per ISO 5208/ API 598 / API 6D)?? It will passing for sure to the upstream side.
I understand some (relatively old) Plants design do not have anti surge and or pressure build-up release device, thus Yes I believe there should be some guarantee over "Class I Check valve(s)" installed in the system in terms of leakage rate / reliability. Class I refer to Check valve that are periodically check based on criticality.

In the nowadays world where Engineering department are segregated e.g. Pump/comp belong to Rotating Engineer, Valve belong to Valve/Mech Engineering, Process belong to Process Eng and so on. It is hard to defined each other expectations, Rotating eng don't want excessive leakage rate however they are not the one who order the check valve, same also with Process Engineer. Mech/Valve Eng knows this but they have to trust their contractor design "method" which quite often only see the segmented P&ID and never be on the field. Contractor engage with a better knowledgable manufacturer, and so on.
IMHO, this should be define by the End user and cascaded clearly to either contractor / manufacturer by means of standard.

Regards,
MR


All valves will last for years, except the ones that were poorly manufactured; are still wrongly operated and or were wrongly selected

 

[gerhardl]

Interesting discussion!

The real question could be rephrased as: 'Whom has the answer for the total mechanical safety of the process design?' This could of course be delegated down to the single engineering groups, but not without specific guidelines.

The discussion here arises from unclear or doubt about necessary component quality and specification, eg. unclear guidelines given. A too low quality will have negative impact on safety, a high quality will give a higher degree of safety and a higher initial cost, but perhaps lower and decrease the cost if damages.

If this could not be decided as a clear decision within the framework, it should be kicked upwards. The correct procedure from an engineers side could be to in brief consider alternatives and cost/impact for the project, write a summary and recommend a solution.

The reason is that this often not only is a technical question, but in addition juridical, budget-wise and process and factory maintenance related.

Good luck to all!