On/Off Butterfly for LNG Service - Zero Leakage Testing

[juliusdarius117]

Good day, I am in the process of designing a flow metering skid for LNG with bidirectional flow. We will have a Z configuration and parallel lines, each with a butterfly valve at the inlet and outlet of each meter run. My background is chemical process design, but for oil and gas. I was wondering if anyone had recommendations on what to look for in a butterfly valve to prevent leakage when closed. I have seen triple-offsets quoted to me, and a 5-offset from V-AXX, but what standards are applicable to cryogenic temperatures for leakage? Especially as this skid may not always be operating, and I am concerned with leakage associated with temperature differentials on S/U and S/D. Additionally, for meter proving, I don't want any bypass of the block valve into the standby flowmeter line in parallel.

For the 5-offset I have been guaranteed:

Class A EN 12266
E Class 1 DIN 3230 BA/BO/BN
E Per API 598 Resilient / API-6D
E Per BS 6364

Would you seek out more proof or documentation from a vender on either side?

 

[bcd]

Most common standards for cold temperature testing are MSS SP-134 and ISO 28921-1.

MSS SP-134 is referenced by ASME B16.34 and some API standards, including API 609 which is butterfly valves for the oil & gas industry.

ISO 28921-1 was developed to replace BS 6364 as the BS standard is withdrawn. I think the last revision to BS 6364 was 1984 or 1991. A very long time ago. People still refer to it, but it was replaced by ISO 28921-1.

My suggestion is to look for a cryogenic valve that complies with API 609.

 

[juliusdarius117]

bcd,

I will look at API 609, MSS SP-134, and ISO 28921-1 for sure!

Do you happen to have any experience with butterfly valves in LNG service? Right now, the pipe size is 10" for the LNG and 16" for the boil-off gas, so our valves will be relatively large. I'm not well versed in the nitty gritty details from a mechanical perspective, but is it feasible to find a triple offset with zero leakage at LNG temps?

 

[GBTorpenhow]

Yes, triple offset butterfly valves with a resilient seat configuration can provide zero leakage capability in LNG service.

A good summary:

https://globalsupplyline.com.au/wp-content/uploads/2014/10/Valve_Leakage_Rates_Test_Std.pdf

 

[bcd]

API 609 refers to APi 598 for pressure testing butterfly valves. Per API 598, all resilient seated butterfly valves are "no seat leakage" during the closure test. API 598 also states that triple-offset butterfly valves with laminate seats are to be considered resilient seated. So triple offset with laminate seat is no leak during the API 598 closure test.

 

[HitokiriAoshi]

In a cryogenic application, it's often the problem that ice or somethings forms on the seat surface, preventing you from achieving perfect sealing. While you could feasibly specify in "zero" leakage (every valve leaks, it's how much and against what size media vs how long)

That being said, when specifying tests for FAT.

Pressure, Temperature, Test Media, Duration, Direction, Allowable Leakage.

API 598 is not a cryogenic test, and fails to appreciate the deformation of valves due to large temperature change
6364 test have a production and a prototype test, the prototype test is what gets low.

For ultra critical application, you should specify production cryogenic test, to your desired allowable leakage rate. both very low and very high pressure are hard to seal. If not possible, you want to cryo treat your metals and conduct at least a high pressure penumatic seat test if feasible (may not due to safety concerns)

The 4th/5th offset butterfly valves are all attempts at creating a body/seating area that will expand and contract evenly that helps with the valve's sealing performance, i know a certain vendor's 4 offset butterfly valve is used at all SPACE X and Blue Origin's lunch pad. In most oil and gas application, 3 offset from a competent manufacture may be adequate.

Luke | Valve Hax |

https://valvehax.com/