All,
This topic has been discussed many times before, by I WANT to ask questions by which your answers would summarize all the previous answers:

Every company has its own spread sheets for evaluating liquid, and gas PSV's forces(for both open and closed systems).
The question is:
1- Where to apply this force (at valve point or at pipe elbow)?
2- What is the direction of that force to be applied (vertical or horizontal?)
3- If u chose one force, why do you say that the other is balanced?

Please ignore the dynamic (time history) from your answers, we are here talking about equivalent static approach (which most off us use), experienced stress engineers are invited to give junior engineers their accumulated experience approach shortly and to the point.

no easy answer here EL NEGM,
like all things in life, on must search for the truth yourself, as the path to learning is as important as the lesson itself.

search the forum,
you will find umpteen diferent answers, do the research and come up with your own solution, which you can have others (your company) validate.


Tim

I hope one day a Company – why not one in the Middle East?- would say "I WANT to spend resources to answer such questions"!…. It would be interesting to publish the experimental results, giving a "shortly and to the point" gift to the community.

As individual approach, you can start with API520 part II and B31.1 figures, because are parts of the Codes. There are some articles on the subject, and after carefully evaluating the assumptions you may agree with them or not.

Best regards,

EL NEGM,

If one accepts a code standard as the summary and agreement of a number of users and authorities, then you should consider the approach as detailed in the ASME B31.1 Power piping code, Appendix II Rules for the Design of Safety Valve installations. It does state that it is non-mandatory, so you could choose to ignore it. I think that it presents some good tools and typical examples for safety valve installations.

1.) There are many shapes and geometries for safety valves, but the most common would be vertical bottom inlet and horizontal side outlet with a 90 elbow directly on the outlet, pointed up. Many have a larger diameter vertical pipe stack for the 90 elbow to discharge into without a connecting fitting. The 90 elbow outlet fitting is the location for primary vertical load. The horizontal loads are basically cancelled by the horizontal components of pressure inside the valve body versus horizontal component of pressure at the 90 elbow. Only at the 90 elbow is the vertical downward pressure component not balanced- the open end of 90 elbow does not provide balance.

2.) The Appendix II presents a method to calculate the outlet pressure of the safety valve. This outlet pressure is used for calculation of the force on the 90 elbow on the discharge. The net force is downward.

3.) If you wanted to follow the pressure wave from the safety valve orifice, thru the body, to the outlet, then there would be milliseconds of time intervals when there would be unbalanced forces. After flow has been established through the safety valve discharge elbow, then there is only the net vertical downward force at the discharge 90 elbow. That is why a support - from the elbow back down to the header pipe or vessel - is a very effective method to deal with the discharge force. When the flow stops then there might be a momentary negative pressure to result in a net vertical upward force of a few millseconds duration.

It is more complicated for a safety valve discharging into a closed system - flare header for instance. A safety valve with a horizontal discharge pipe would have a net horizontal force on the safety valve outlet. These scenarios are discussed in the B31.1 Appendix II. If you have a liquid service safety valve then you have density differences, but the physical principles are the same.

Richard,
Let's summarize your reply in few points-according to my understanding- and correct me if I am wrong:
1- For an open system (vent to air by a 90 degree elbow): The pop force is vertical downwards and to be applied at the elbow.

2- For closed systems (reliefing to a flare header for instance):The pop force is horizontal in the opposit direction of the flow and to be applied at the valve

EL NEGM,

1.) Yes, agreed the 90 degree elbow up to open air has vertical downward force. We have made progress in understanding.

2.) No. The safety valve with a open horizontal pipe -no discharge elbow-piped to outside thru a wall penetration, to the side of a platform past the hand rail, etc. would have a horizontal force. The open end of pipe is in the horizontal direction (no balance force) so that the net force at the safety valve outlet connection is horizontal.

The forces in a closed system may have some intervals of unbalanced forces as the pressure wave travels downstream. Assume a pressure wave is traveling at speed of 300 meters per second in a pipe run of 600 meters between two 90 elbows. There is a force at the first elbow. There is a net force on the pipe until the pressure wave reaches the second elbow. After the pressure wave has passed the second elbow the net axial force for that run of pipe is zero. The net force acts for the 2 seconds time during the time interval while the pressurre wave travels between the pair of elbows. The concept of elbow pairs is basic to understanding the force unbalance which could cause the pipe to move.

Closed systems would have pipe runs of different lengths in several directions possibly. The first elbow pair could be the discharge elbow on the safety valve outlet. The short run from safety valve body to the elbow would result in very short duration of milliseconds force unbalance. The runs of straight pipe between the next elbow pairs could be as much as the 600 meters of earlier example to have a resultant force with duration of maybe 2 seconds in the direction of each straight run of pipe. So you could imagine the effect of a force that travels through the closed system pointing in different directions for different time durations from the geometry of the piping. Is a force of less than 2 seconds duration a significant load? Are there systems of lengths over 600 meters? Are there pressure wave velocities much slower or much faster? Each case of discharge pressure, directions, and run length will need to be evaluated.

Richard,

It’s interesting your example. Just a remark: the article "Simplified analysis of steamhammer pipe supports loads", by E.C. Goodling includes a procedure to calculate the unbalanced force on each leg, based on the leg length, the speed of the sonic wave and the closing/ opening time.
In my opinion it’s a realistic calculation for PSV case.
The only remark I have is about the real growing-up wave time- to know it we need the PSV opening profile rather than PSV opening time. In fact, Goodling has made a similar correction in the article, based on the closing profile. Unfortunately, it’s hard to obtain such information from Vendors.

Caesar II has a Relief Valve Thrust tool based on the article "Steam Flow through Safety Valve Vent Pipes" by H. E. Brandmaier and M. E. Knebel. This article assumes that there is flow acceleration after PSV orifice, in supersonic domain (more exactly a supersonic underexpanded free supersonic jet). In this case, a momentum calculation must take into account this acceleration, and this is not transient scenario.
I have a lot of comments on the article assumptions, but this is not a Fluid Mechanics forum….

I'd be intrested in looking at how Gooling calculates unbalanced forces. Mariog.

Do you have a link or a source of that procedure?

Much appreciated

Tjn

Sure.
In Power, this article is well known. By CraigB generosity you can download it, please see:
http://www.coade.com/ubbthreads/ubbthreads.php?ubb=showflat&Number=14930

The Goodling’s paper is not referring to PSVs, but he has presented his paper as valid for both traveling waves of increasing and decreasing pressure. In my opinion, it would be engineering satisfactorily to consider such approach for either expansion or compression waves.

I’ve tried to make some comments on subject at:
http://www.coade.com/ubbthreads/ubbthreads.php?ubb=showflat&Number=21000

Best regards,