We have a client who is asking for us to review all pipe stress calculations for relief valves on a very large offshore facility.
The relief valves are Tyco Pilot Operated.
Tyco publishes a "maximum discharge force" for their relief valves.
The published force is for a failed pilot condition with an open relief system.

We do not have an open relief system, we have a closed system.
api 521 sect 4.4.2 "Determining Reaction Forces in a Closed Discharge System" states
"Closed discharge system, however, do not lend themselves to simplified analytical techniques."

So, if we used the max published, open system, discharge force we end up needing to recommend seriously beefing up the supports.  

Can anyone recommend a method of determining what sort of force we would really see?

It seems like a combination of overcoming momentum in the static volume of gas in the header, plus the rapid deterioration of pressure, therefore velocity, would produce a spike of acceleration and an over all force, to over come structurally, of less than that published maximum force.

Time history using Caesar would perhaps be the most technically correct assuming you can get and use the correct input data.  (re-read the last few words a couple more times).

Don't rely on process data sheets because these don't contain the right data for force calcs.  Listen more to the valve supplier but don't believe him 100%.  The suppliers are experts on valves and will on balance give you the best info. However, I've had problems listening blindly because the supplier has not seen all of the big picture.  You have got to be your own expert because you are sitting between others who are looking at the situation from different view points.

However, a word of dire and serious warning - no joke whatsoever - been there and got shafted. Many self proclaimed hot shot pipe stress engineers can't be bothered to understand and read the manual.  By doing "fancy and unnecessary" calcs they haven't the brains to understand is not going to win any friends whatsoever and will make enemies.  If your going to do calcs that might make the project manager happy on this job, make sure you don't get well and truly stuffed for the next.  Tread carefully.

MORE NEWS