Hi all,
I thought I'd re-post this here, in addition to the original post in the Completions TIG. We plan to validate a new subsurface safety valve and I'm reaching out to see if we're missing anything.
We're considering a passive check valve as an SSV for water injection and gas injection wells. Its literally a ball check where a spring holds the ball on seat when the well is static, and flow pumps the ball off seat when injecting. Its brutally simple and does away with an SCSSV's hydraulics. The closure mechanism is insensitive to A annulus pressure, tubing pressure, and many of the other challenges we can face w/ SCSSVs. I won't go into details but erosion, thermal cycling and similar issues have been, or will be, addressed thru CFD and physical testing.
We plan to bring the valve to a test facility to validate for our well conditions, to API 14A subsurface injection safety valve (SSISV) criteria. We'll test with high-rate water and high-rate gas. Ensuring that the as-tested water rate is a good analogue (ie, an applicable validation) for the water injectors is easy. Getting a good analogue for the gas injectors is far more challenging: being compressible, the gas volume under well conditions during injection (cold and high pressure) is far less than than a similar volume will be at test facility conditions (hotter and far lower pressure). If I push 50 MMSCFD across the valve under downhole conditions, the gas velocity is pretty modest. That same volume would be screaming and likely damaging at the lower pressures and higher temperatures that'll exist in the test facility.
So how do I test the valve to ensure its suitable for gas injection service? I looked at the following:
1) Mass flow rate thru the valve: the planned water injection rate yields about twice the mass flow rate as the gas. The water portion of the test should suffice.
2) Viscosity and hydraulic drag: the gas is far less viscous and, even at its higher velocities, imposes less hydraulic drag than the water under downhole conditions. Again, the water test should suffice.
3) Kinetic energy: I don't think this is relevant. Unlike an SCSSV in a producing well, this valve cannot be slammed closed. Short of a piece of debris suddenly plugging it (which is not part of the API test regimen), the injection fluids cannot be stopped by the valve and the fluid's kinetic energy simply passes thru the valve. That said, the kinetic energy of the water is about the same as the gas.
4) Velocity: the gas velocity is quite a bit higher than the water velocity. I plan to flow gas at >125% of the maximum expected velocity when in service (a worst-case transient start-up condition). I think this is a robust validation but when measured in MMSCFD, the gas rate required to achieve this velocity is only 20% of the intended injection rate. The intended injection MMSCFD, if flowed thru the valve under test facility conditions, would have a velocity of almost 11x what we'd expect to see in service, and almost 8x the highest rate we could possibly see under transient conditions. We cannot physically test to these conditions, nor (IMO) would it make sense to do so.
My plan is to validate to 1.1x our maximum water injection rate. This exceeds the highest mass thru-put, hydraulic drag, and (unconverted) fluid kinetic energy that the valve will ever see in service, in either well type. I'll also flow gas thru the valve at 125% of the maximum expected velocity under worst-case transient conditions. I think this is a good validation, but unfortunately the API monogram will list a gas volume that's 20% of the intended MMSCFD. IMO, API would be wise to have included velocity in that monogram, but that's for a future committee to consider.
My question to you: should I be considering any other criteria to ensure the valve is suitable for our intended gas rate?
Many thanks in advance,
R
------------------------------
Ron Nelson
Subsea Completion Consultant
ron@deep-blue.ca------------------------------