A closed relief valve system with a discharge pipe of 4", discharges gas into a 12" header. The header is anchored to the ground near the discharge point to the atmosphere. Therefore, thrust at the header/atmosphere inteaface is not transferred to the 4" discharge pipe. In this case, is it reasonable to assume that there is no load on the discharge pipe due to relief valve operation (relief valve opens in 10 milliseconds runs fully open for 1 second and closes in 10 millseconds)?

It is my understanding that acoustic shock due to sudden change in fluid momentum and associated travelling pressure wave as result of relief valve operation in gas systems does not generate forces large enough to be concerned of the pipe stresses. This phenomenon is important only in liquid systems. Therefore, for gas systems these forces can be ignored. Can some please comment.

Concerning to the second paragraph, I think it is the other way around. Gas and steam systems can create large momentum during the release than the liquid systems in terms of safety valves.
Due to the short time opening transient dynamic forces occur with the direction changes until the released gas starts filling the discharge side and creating back pressure.

In your case I would try Caesar II dynamic option to solve your problem. I guess the dynamic transient forces for safety valve with 4" inlet are going to be large.
Safety valves are very sensitive to the thermal and mechanical loads and should be isolated from piping stresses through proper support, anchoring or flexibility of the discharge piping.

Regards,

Ibrahim Demir

I Demir,

In my second paragraph I am referring to acoustic shock waves that are generated as a result of sudden opening and closing of valves (water hammer in liquid lines) that would be seen in the upstream piping. I would think due to compressibility of gases, the loads caused by the moving acoustic shock waves would be small in gas systems unlike in liquid flows. However, you are right that relief valves in liquid systems, open partially or very slowly and as a result large shock waves are not generated in liquid lines.

Does this mean that we can always ignore loads due to acoustic shock waves in relief systems, whether gas or liquid and consider only the loads due to momentum change?

In my problem, inlet of the relief valve is 3", outlet discharge is 4" and the 4" discharge pipe runs into a 12" header before releasing to atmosphere. The header is anchored at the header/ atmosphere interface.

Therefore my questions are,

1. Should I ignore the effect of shock waves in the upstream piping (upstream of the relief valve)?.
2. Do I have to consider unbalance forces occur at the bends as the gas front moves from the relief valve towards the header. My understanding is that as the gas front moves at almost sonic velocity, the time of the unbalance force at bends are very short and hence these loads do not have time to develop to full magnitude and can be ignored. These unbalance forces exist only until the gas front moves between bend-bend pairs.
3. At the header, as the discharge pipe runs into a larger pipe at atmospheric pressure, should we assume that this situation is similar to header/atmosphere interface and apply the full thrust force at this point.


Regards
Vish

Vish,

"API520 Part II 4.4.2-Determining reaction forces in a closed discharge system" says:

"Pressure-relief devices that relieve under steady-state flow conditions into a closed system usually do not transfer large force and bending moments to the inlet system, since changes in pressure and velocity within closed system components are small.
Only at points of sudden expansion in the discharge piping will there be any significant inlet piping reaction forces to be calculated. Closed discharge systems, however, do not lend themselves to simplified analytical techniques. A complex time history analysis of the piping system may be required to obtain the reaction forces and associated moments that are transferred to the inlet piping system."

Unfortunately, it is difficult to give you direction other than this without seeing full details of the system, valve characteristics and possibly running a time history analysis.

Regards,

Ibrahim Demir

True.

It is worth to note a paragraph in API RP 521, Guide for Pressure-Relieving and Depressuring Systems. It’s under the chapter 3.20.3 Multiple Valves. But the validity of idea is outside of chapter limits, for me it’s a good argument against the extrapolation of the API formulas to a closed discharge system.
Sometimes a Vendor gives you a calculated value that may be based on a similar approach, even is not declared as API.

"The maximum probable mechanical stresses and forces resulting from the discharge flow through a pressure relief valve are given in API Recommended Practice 520, Part II. These discussions consider steady-state conditions and not the momentary, instantaneous forces that result when the valve first opens. Both impact forces and adjustment forces result as the inlet and outlet piping attempts to reach a force equilibrium for the relieving conditions of pressure and temperature".

If there is a 90 degree bend immediately downstream of the relief valve, that means the relief valve will not see any force due to gas release. If there is a considerable distance between two consecutive bends, then part of the full force will apply at one of the bend. Also it is suggested that when gas moves into a large header, the bend immediately upstream of the header will see a large force. So conservatively one can apply the full steady state force at this bend.

Any comments?
Regards

Vish