API6A-PR2 vs PR2F 2

[Thomas0111]

For the PR2 qualified API-10K valves, I have not to further request valve vendor to carry out the design validation tests unless there is design change on the new purchased valves, in which the design validation is compulsory in place in accordance with PR2F, meant product validated to PR2 level according to this annex. Do I understand correctly?

Secondly, do I need to request valve supplier to carry out any engineering analysis as part of verification for the API validated valves as supplied?

 

[jmec87]

I don't fully understand your question, but I'll try to answer.

If you specify PR2, the manufacturer has to validate that the valve is capable of performing the number of cycles required for PR2. That's all. PR2 does not mean PR2F. If you want PR2F validation, you have to specify PR2F validation, not just PR2.

The valve manufacturer is already required to complete the engineering analysis per API 6A, so I don't see any need request this again.

 

[SPDL310]

Thomas,

PR2 and PR2F are two separate rating levels. The requirements for PR2 validation of vales is that the manufacturer conducts a 200 cycle validation test. That test is generally conducted at ambient temperature with water. PR2F requires an extensive min/max temperature and pressure range cycling as well as a 200 cycle test. A majority of the PR2F cycle test can be conducted at ambient temperature with a given number of cycles performed at min/max temperature. PR2F also limits soft seal goods to compounds that have been tested for fluid compatibility in the specified temperature range. So yes if a valve has passed a PR2F validation it is able to be labeled as a PR2 product as well because it passed a 200 cycle validation. If you are currently buying PR2 valves with no issues you may not want to switch to PR2F unless there is a engineering need to do so, as purchasing a custom valve to PR2F may significantly increase the valve cost.

As jmec87 stated engineering design of the valve is required by 6A regardless of PR level.

 

[Thomas0111]

jmec87 and SPDL310,

Thanks for your clear explanation. However, I still have few other questions:
1. If that valve manufacturer had supplied similar API-10K valves with PR2 qualification before, the new ordered valves will require to have different connection (flange to hub or new hub manufacturer design to suit) or material to apply. How can the valve vendor to qualify these valves with PR2 compliance? Will PR2F to be the approach to validate the new design?

2. Understood that scaling method is listed in Annex F, could valve manufacturer to follow that to validate the new design valve like stated in item 1? If this is the approach, will these valves considered as PR2F validated valves?

3. What will be design criteria to select PR2 over PR2F?

4. When will be the engineering analysis as part of new design validation such as finite element analysis or calculation required to carry out? Does this implemented for new designed valves (different connection) or including all valves even with design validated valves to procure?

 

[SPDL310]

Thomas,

1.) If you are purchasing a monogramed valve from a reputable manufacturer they will perform all required validation prior to shipping your vale.

2.)Scaling methods can be used for PR2 validations as well as PR2F. If a valve is tested PR2F then the scaled valve can be marked PR2F as well.

3.) PR2 is the general standard for production valves. I cant think of the document number off the top of my head but, there is an API 6A purchasers guidelines document that will go into more detailed recommendations on what application would imply a requirement for a PR2F validation. Generally speaking (not to API purchaser guidelines) you would want it when safety is more critical, and the possibility of leaks becoming detected is unlikely to occur or poses significant to health and safety. I would also recommend it for anything is TC: X or Y service, or service temperatures below -20F, and designs with single point of failure elastomeric seals that do not have a long service history.

4.) FEA analysis and calculations are part of the design process, and should be completed prior to validation. Once again the valve vendor will know when they are required to perform a new validation, and when they can scale to a previous validation. Moving a size up or down on connections is allowed by scaling i.e. 3" to 4". Swapping a welded 1502 hub from male to female would probably not require a new validation either. What type of end connection change is your vendor making that you are concerned about needing to re validate the valve?

 

[Thomas0111]

SPDL310,

Vey much appreciated your advice. In my case here, we have to re-select new hub/flange supplier and expected some dimensional changes which required to further consider into valve supplier end connection, and size changing where it can be further managed through the scaling method as per Annex F. I wonder whether these valves required to re-validate with new FEA analysis supported as proof for valve integrity?

 

[SPDL310]

Thomas,

I suspect that the dimensional changes of flanges/hubs from vendor to vendor for same sized hub or flange would be well within the acceptable range for scaling. If you are looking at a valve the rules state 1 nominal size larger or smaller based on the bore of the end connection. If you have a valve with a 3" hub you can scale that validation to the same valve with either 2" or 4" hubs for end connections. Some suppliers use the same weld preparation and make a transition to the new flange/hub. This is especially common for chokes where nominal size is determined by orifice size rather than end connection size. A choke validation can be scaled to any end connection size as long as the orifice is within 1" up or down from the original size and stress levels are similar.

Are you the manufacturer of this valve? If you are uncertain about the validation status or applicability of previous FEA to the new geometry, then you should test it.

 

[Thomas0111]

SPDL310,

I am the contractor who purchase the valves for our project client. And, I would like to understand more on what will be the practical approach in industry to set the criteria for selecting suitable valve supplier which will be served the good reference for me.
Few more questions below and would to have your sharing with me if you had encountered any before:

1. Review on API purchaser guideline, there are not specific standard for selection between PR2 and PR2F. PR2F is more demanding in term of temp./pressure testing cycles as specified on Figure F.1 (Test Procedure). What was your experience when choosing the type of validation over one another? Follow with pressure rating, material types or criticality of the services, etc...

2. Does supplier require to perform FEA analysis once again for new purchased API valves? It is regardless of previous validated valves, or new valves selected based on Annex F Scaling method...

3. When will be scenario supplier has to perform operating cycles testing? Or I can avoid this if all the ordered valve types which had been validated PR2 or PR2F to previous client before.

4. Scaling can be implemented when there are 2 design requirements to meet. One of them is referred to design stress levels. Can I understand that this is referred to mechanical properties of valve material? If existing valves are PR2 validated, I can assume that new valves with same configuration and higher strength will be considered validated based on Scaling method.

 

[jmec87]

1. Choosing one type of validation over the other should depend on valve design service history and criticality/severity of the application. If you are using a proven valve design in normal service (not extreme temperatures, pressures, or fluids), then you don't need to specify a high level of validation testing - actual usage of the valve has demonstrated that it is fine. However, if you have a novel valve design and/or you are using the valve in extreme conditions where the valve has not been used before, you may want to specify a more stringent validation level (depending on the consequences of failure if the valve doesn't function).

2. There is no requirement in API 6A for the valve supplier to perform FEA on any valve designs. The supplier is required to design the valve body and bonnet per API 6A 5.1.3, which includes an option to perform design calculations per API 6X, and also allows the manufacturer to use "other methods" (such as FEA) to justify the design. The design requirements are not dependent on the validation requirements.

3. The manufacturer has to validate the design (normally by testing including several operating cycles) for a new design. If you order a valve that has already been validated to your desired validation requirements, then the manufacturer does not have to do the validation testing again (unless there have been substantive changes to the design after validation).

4. The meaning of the current "design stress levels" requirement (F.1.14.2b) is simply that the stress criteria is the same, i.e. general primary membrane stress at operating condition does not exceed 2/3 of yield strength. In general, this should imply that validation of a equipment manufactured from low strength material also validates equipment with the same geometry and pressure rating that is manufactured of higher strength material.

 

[SPDL310]

Thomas,

I think jmec87 has provided an excellent response, which is more concise than I could attempt. I will expand on a few of his point from my experience.

1. For chokes where most of my experience is focused, PR2F is rarely requested. If the criticality is deemed that high many users are ordering their chokes 16C monogramed. For valves I have seen more requests for PR2F especially for metal-seated ball valves, which can vary widely in reliability for vendor to vendor. Additional testing requirements (such as fire safe and and slurry) are often applied to ball valves.

2. I believe that you would be hard pressed to find many valves with designs based solely on FEA. Most are likely based on 6X calculation with critical or complex geometries being evaluated by FEA.

3. nothing to add here.

4. I agree with jmec87's statement in general, but there are a few instances where I believe that may not always be the case. For example when the coefficients of thermal expansion are significantly different seal extrusion gaps may change depending on a material change. I.E. a low yield stainless body vs a high yield alloy steel body. I think best practice would to have a separate validation for the alloy vs stainless, but there is nothing written in the standard which would require it.