PSV Reaction Force

[Forshei]

Dear Sirs,
As a structural engineer I've been assigned to design a support platform for 2 off Block Body PSVs on an offshore gas processing platform.

Downstream each PSV there is a short horizontal section, continuing via a 90 deg elbow downward to a connection to the flare header via a T-section.

The reaction force for the PSV (calculated by the piping department) is in the 250 kN region, pointing towards the PSV nozzle. Calculation of the reaction force is based on API RP 520, section 2.4.1, which is valid for an open discharge system.

I strongly feel that the calculated reaction forces are too high, and possibly in the wrong direction, as forces will cancel each other out as the velocity vector changes direction. However, the piping department claims that they have always used that formula and that the calculated reaction force is a transient force that will be present in the instant the valve opens.

I hope that there is a set of design rules somewhere which is more relevant for this system, and that someone could be kind enough to point me in the right direction.

Regards
Bjorn Forshei

 

[BobPE]

forshei:

I would imagine that your piping department got it right, you left out a lot of information for us to guess at it. I would however note that the forces at a PRV can be quite complicated. You have to take into account transient wave propogation for various scenarios that could place equal or magnified forces on the valve in BOTH directions, and not only upon opening of the valve. Again, depending on your system, physical restraint of the valves may be necessary. Overkill in factor of safety in this situation should be considered good thing.

BobPE

 

[keyardee]

The reaction force is maximum at the nozzle mouth because of the very high velocity of discharge.Subsequently in the downstream line the velocity comes down to a reasonable operating level subjected to the allowable back pressure acumulations.So go for the worst and support your valves against maximum anticipated forces.Perhaps you can think about double headed safety valves with twin outlet nozzles diametrically opposite,so that the reaction forces of each nozzle will cancel out the other's. Hope this explains your piping deparment's stand.
keyardee

 

[Forshei]

Any decrease in flow velocity downstream the PSV nozzle must give rise to reaction forces, F = d(mv)/dt. This will also apply for a bend, where the velocity changes direction.

So for a "black box" with a PSV arrangement inside, the net force on the system should be equal to the change in massflow x velocity in/out of the system.

Any transient forces, wave propagation, etc. are obviously not explained by this "global" model, but the force calculated by API RP 520 does not seem to do so either.

-b-

 

[hacksaw]

Too high!

Sounds like you've never seen a relief valve laying in the parking lot after it broke free and jetted off.

Even when the mounting nozzle does not break, the damage in the discharge piping is mind-boggling.

The ASME has a complete set of calculation guidelines for discharge reactions and even worked examples.

Good luck, but make sure the piping is well braced too.