5 replies to this topic

[aju_1807]

We have few relief valves in supercritical zone. I would like to know what Kd factor we should consider for RV sizing. Our RV vendor has offered Kd as 0.6516 which is normally for liquid service. the same vendor has considered Kd of 0.6516 for liquid relief cases.

I am not convinced why the Kd of liquid is considered for RV in supercritical zone. Could somebody share their experience of RV in supercritical zone. WHat Kd factor we should consider? One of the technical paper for sizing of RV in supercritical zone shows kd of .975.

appreciale valuable feedback on this subject.

Rgds,

AJ

[rxnarang]

Richard Doane's article in HP Jan 2010 mentions Kd=0.85 as in API 521, to be used with consistent API equations, when vendor information is not available. I would think ASME Kd =0.65 would be conservative. This is to used with ASME formulae. I hope you have considered the difference in Kd for ASME and API.

Would choosing a higher Kd lower the valve size and save you money? If so, then it is worth pursuing, else why bother.

The problem is the Red Book does not cover supercritical flow certication Kd. All the vendors I know will not size supercritical or two phase PSVs and certify the valve. They tend to take your sizing as final.

Regards
Rajiv

Edited by rxnarang, 20 February 2010 - 11:02 PM.

[aju_1807]

Richard Doane's article in HP Jan 2010 mentions Kd=0.85 as in API 521, to be used with consistent API equations, when vendor information is not available. I would think ASME Kd =0.65 would be conservative. This is to used with ASME formulae. I hope you have considered the difference in Kd for ASME and API.

Would choosing a higher Kd lower the valve size and save you money? If so, then it is worth pursuing, else why bother.

The problem is the Red Book does not cover supercritical flow certication Kd. All the vendors I know will not size supercritical or two phase PSVs and certify the valve. They tend to take your sizing as final.

Regards
Rajiv

Hi Rajiv,

Appreciate your response on this query.

Actally, as you mentioned normally vendor will not size relief valev for supercritical zone as they are not familiar with teh correct methodology for sizing. Anyway, in our case we are providing mass flux and mass flow rate to vendor. vendor earlier has considered Kd of liquid for sizing (i.e. 0.65) as a conservative apporach. However, subsequently, as a corporate decision we decieded to indicate relief valve size reqd directly to vendor so now ball is in our court and we have to tell RV size.

As mentioned in your reply, the kd factro of 0.85 is as per API but dont know where it has been mentioned. Appreciate if you can tell me the clause for ready reference. Further, appreciate if you can send me HP artical also whch you are refering in your reply.

Once again thanks for your reply.

Regaards,

AJ

[aju_1807]

Richard Doane's article in HP Jan 2010 mentions Kd=0.85 as in API 521, to be used with consistent API equations, when vendor information is not available. I would think ASME Kd =0.65 would be conservative. This is to used with ASME formulae. I hope you have considered the difference in Kd for ASME and API.

Would choosing a higher Kd lower the valve size and save you money? If so, then it is worth pursuing, else why bother.

The problem is the Red Book does not cover supercritical flow certication Kd. All the vendors I know will not size supercritical or two phase PSVs and certify the valve. They tend to take your sizing as final.

Regards
Rajiv

Hi Rajiv,

Appreciate your response on this query.

Actally, as you mentioned normally vendor will not size relief valev for supercritical zone as they are not familiar with teh correct methodology for sizing. Anyway, in our case we are providing mass flux and mass flow rate to vendor. vendor earlier has considered Kd of liquid for sizing (i.e. 0.65) as a conservative apporach. However, subsequently, as a corporate decision we decieded to indicate relief valve size reqd directly to vendor so now ball is in our court and we have to tell RV size.

As mentioned in your reply, the kd factro of 0.85 is as per API but dont know where it has been mentioned. Appreciate if you can tell me the clause for ready reference. Further, appreciate if you can send me HP artical also whch you are refering in your reply.

Once again thanks for your reply.

Regaards,

AJ

Hi Rajiv,

PLease send it to me on ajay.sharma@petrofac.com or ajaysharma_2001@rediffmail.com

[JoeWong]

Safety related device, take conservative approach to minimize and/or mitigate the risk if you are not sure.

Supercritical fluid is dense phase fluid carrying vapor and liquid like properties. It is not easy to clearly define good properties of a dense phase fluid. In the regards, vendor has taken a conservative approach by "assuming" it is "liquid like" properties and adopting Kd = 0.65. This approach may not be correct mathematically, however it carry conservatism in it and worth for moving forward without a proper approach right in front of you.

You have mentioned that "One of the technical paper for sizing of RV in supercritical zone shows kd of .975.". Can you please advise the article ?
Author may have provided the argument of using Kd of 0.975 e.g. author may have analysed the properties of supercritical fluid and understood that it is "vapor like " and decided to go for 0.975.

For two phase flow relieve, there is no one commonly accepted derivation method by all expert as far as i aware. Infact, there are many methods in determining the Kd factor for two phase flow. Unlike vapor and liquid where proper testing method available to obtain Kd method e.g. Kd = o.6515 used by vendor, there is no "defined" method in determining Kd factor for two phase flow. Therefore, vendor has taken a conservative approach. For your information, Dr. Leung has recommended to use Kd of 0.85 in Omega method. Read more in API omega method.

I have done some literature works in the past, many methods were proposed for their particular system and testing range. Suitability of each method for each system and condition subject to understanding of the methodology and fluid behavior. This is where the expertise is called for.

[rxnarang]

API 520 5th Edition, Appendix C recommends 0.85 for two phase flow, which is akin to supercritical flow. See Appendix B for an example of supercitical flow of ethylene.

Quote from Appendix C

Kd is the discharge coefficient. For a preliminary sizing estimation, a discharge coefficient of 0.85 can be used
for a two-phase mixture or saturated liquid entering the PRV inlet. For the case of a liquid entering the PRV
inlet, a discharge coefficient equal to 0.65 is consistent with the single-phase method in Equation (28) and
Equation (29). Note that a value of 0.65 may result in a conservative valve size for liquids that are only
slightly sub-cooled. The user may select other methods for determining a discharge coefficient [13], [14],

Unquote

Also CCPS handbook mentions that if the fluid chokes in the valve use Gas Trim, else use Liquid Trim.

Does this help?

Regards

Edited by rxnarang, 02 March 2010 - 03:51 AM.