Well Integrity Technical Section

 View Only
  • 1.  Xmas Tree Valves Field Leak Rate Test

    Posted 08-05-2016 03:50 AM

    Hi everyone,

    I work in a field in which we are implementing a new WIMS in which valve leak rate test theory is explained. The problem is that this theory is not straight forward specially when it comes to apply it in the field. I'd like to know if you can provide any guidance on how to test xmas tree valves leak rates in the field; some rule-of-thumb to get similar results to those of commonly used theory that our technicians can easily use to determine if a valve is leaking and if it needs reparation/substitution.

    Best regards 

    ------------------------------
    C. Lopez
    Production Engineer
    ------------------------------


  • 2.  RE: Xmas Tree Valves Field Leak Rate Test

    Posted 08-06-2016 09:52 PM

    Lopez,

    Recently, I worked on this issue. We have two standard to follow: API RP 14C and ISO 16530-1. The first one is flexible than the later. You can pick either one best fitted to your location if you don't have any country regulations.

    ------------------------------
    Nasir Ahmed
    Well Reliability & Optimization Advisor
    Chevron Bangladesh Ltd
    Dhaka



  • 3.  RE: Xmas Tree Valves Field Leak Rate Test

    Posted 08-08-2016 10:57 AM

    Carlos,

    You can calculate generic pressure change limits for valves in a group of similar wells (or well types). Your field technicians can then use those as test limits for that well type.

    Setting a field useable test using API RP14 for tree valves is difficult, primarily because of small tree cavity volumes.

    ISO/TS16530-2:2014 helpfully refers to 2cc/min per inch of valve size as an allowable leak rate (see Table 1 on page 22). This is more conservative. An external leak at this rate would be easier to deal with and unlikely to erode the leaking valve. This criteria gives you pressure change limits over the test duration that are easy to manage.

    As a valve leaks the pressure differential across it changes so the leak rate changes also. You should account for this when setting your pressure change limits.

    I have previously worked on this with your colleagues in Madrid. You can contact Paul Villar about this or e-mail me at stuart.girling@girlingmcintish.com

    Regards,

    Stuart

    ------------------------------
    Stuart Girling
    GirlingMcIntosh
    London



  • 4.  RE: Xmas Tree Valves Field Leak Rate Test

    Posted 08-09-2016 06:00 AM

    Dear All,

    just would like to highlight that API6A - ISO 10423 also provides leakage acceptance criteria in its annex F.

    Regards

    PR

    ------------------------------
    Philippe Rabis
    Well Integrity TA - Completions Lead
    Maersk Oil CTP
    Kobenhavn K



  • 5.  RE: Xmas Tree Valves Field Leak Rate Test

    Posted 08-10-2016 03:16 AM

    As I understand just calculating the void volume behind the valve to be tested and using a maximum acceptable leak rate value I could calculate an acceptable pressure variance over time for each valve, right? But other questions arise, for example: wich sequence is preferred to test the valves? If both kwv and sv are leaking how can I get a proper measure of the leak rate? Wich volume I use for the pwv, lease pipe volume?

    ------------------------------
    Carlos Lopez
    Production Engineer



  • 6.  RE: Xmas Tree Valves Field Leak Rate Test

    Posted 08-11-2016 05:19 AM
    Youre spot on it is therefore not very relevant to Know the exact rate As the test is only a representation if the leak if it is bad or worse.

    Hence the 100 psi rule Will give you that Information, exact calculaties are always subject to discussions. For a regular Maintenance crew a simple 100 psi / 10 minuts is easier to work with, and if exceeded you investigate with Some more greasing, testing .

    Verstuurd vanaf mijn iPhone

    ------Original Message------

    As I understand just calculating the void volume behind the valve to be tested and using a maximum acceptable leak rate value I could calculate an acceptable pressure variance over time for each valve, right? But other questions arise, for example: wich sequence is preferred to test the valves? If both kwv and sv are leaking how can I get a proper measure of the leak rate? Wich volume I use for the pwv, lease pipe volume?

    ------------------------------
    Carlos Lopez
    Production Engineer
    ------------------------------


  • 7.  RE: Xmas Tree Valves Field Leak Rate Test

    Posted 08-09-2016 12:54 PM
    Valves have no specific defined leak rate apart from ESD valves they are defined in API 6 AV and refer to API 14 B that quotes 15 SCF or 400 cc of of hydrocarbon liquid. These Rates are based on the risk of a leak on an offshore platform and would generate a flame of 1.5 meters that able extinguished by a person wearing proper PPE and a portable fire exthinghuiser.

    If you test valves, the valves are positioned in X mass tree or side out lets of wellhead, they are single valves or multiple valves, when you test by inflow test, the volume used to observe the build up rates depends on how many valves or spool pieces are involved. I.e.

    A lower master tested with flow wing closed and kill wing closed and observed at tree cap would involve, half a volume of lower master excluding gate cavity , full volume of upper master plus gate cavity, Cross volume, half a kill wing plus gate cavity, half a flowing valve plus valve cavity, full swab valve volume plus gate cavity and swab valve tree cap.

    The gate valve cavity is normally 1/2 the volume of the full valve bore, however cavity can be full of grease, oil or still under pressure, this makes the whole debate on what is acceptable and what not a bit of a endless discussion . Note a full valve volume with cavity would be 3 X 1/2 valve bore volume.

    When we drafted ISO 16530-1 we took the API 598 for flow line check valves and that specifies 3 cc or 0,025 scf per inch diameter per minute as an acceptance criteria for leak testing check valves which has been used for some years in North Sea by some Operators as the standard for testing tree valves.

    If you want to go the rule of thumb approach then you could use 100 psi per 10 minutes as. A default , if you exceed the rate investigate / grease , if still exceed build up rate investigate further.

    This means that if a lower master is tested and builds up 100 psi of and you calculate per volume as explained above using q= [4 x ( delta p ) x V] / t ( API 14 B for low pressures ) that would provide you leak rate per minute.

    Please note when going the exact calculation route you need to consider pressure class as valve lengths from flange to flange change with pressure class.

    If you use a default 100 psi / 10 min for each test irrespective the position of the valve in sequence, then you test the inner valve at a higher acceptable leak rate then you do the test the outer valves because of the change of inflow test volume, which I assume appropriate as the outer valves are more critical I.e. you're last line of defense.

    Hope this helps and confuse you to much.





    Verstuurd vanaf mijn iPad

    ------Original Message------

    Carlos,

    You can calculate generic pressure change limits for valves in a group of similar wells (or well types). Your field technicians can then use those as test limits for that well type.

    Setting a field useable test using API RP14 for tree valves is difficult, primarily because of small tree cavity volumes.

    ISO/TS16530-2:2014 helpfully refers to 2cc/min per inch of valve size as an allowable leak rate (see Table 1 on page 22). This is more conservative. An external leak at this rate would be easier to deal with and unlikely to erode the leaking valve. This criteria gives you pressure change limits over the test duration that are easy to manage.

    As a valve leaks the pressure differential across it changes so the leak rate changes also. You should account for this when setting your pressure change limits.

    I have previously worked on this with your colleagues in Madrid. You can contact Paul Villar about this or e-mail me at stuart.girling@girlingmcintish.com

    Regards,

    Stuart

    ------------------------------
    Stuart Girling
    GirlingMcIntosh
    London
    ------------------------------