Recommended % of Sonic Velocity for Relief Systems 1

[UW1981]

I am trying to find out if there are some industry guidelines regarding relief system sizing and what the maximum velocity (% of sonic) one should design for. I realize API recommends tip velocities of around 500 ft/sec for proper dispersion, and that we want to stay below sonic velocity (choked flow), and API uses a velocity of 0.5 sonic in one of their examples, but I am not aware of any other industry recommendations other than to keep the velocity in the piping low enough to preclude erosion corrosion, manage the delta P, and minimize the noise...but this is pretty vague.

Any other thoughts/recommendations would be appreciated.

 

[zdas04]

There are a lot of "recomendations", "suggestions", and "rules of thumb" out there.

The one I use is to ensure that the friction drop in the line results in no more than 80% of critical pressure for choked flow through the PSV. As long as the pressure on the exhust side of the PSV is below the critical pressure I am pretty confident that the valve will releive near its design rate and the system I'm protecting is protected. On the other hand if the downstream pressure gets above critical then the releiving rate is pretty much a crap shoot. The 20% hit I'm taking is to account for trash or condensation in the exhaust piping.

David

 

[4Pipes]

The reference to mach number is for a closed system. If it steam straight to atmosphere you will end up with mickey mouse piping with huge outlet pipe sizes hanging on an over stressed valve. This reduces the mach number or so you will be told by process engineers applying Crane. If you apply Fanno you could easily find the reducers are choked for higher set pressures and for bigger valves or both.

What I have never found and would be interested to hear is what happens to velocity in a choked reducer. Being a simple sole, a reducer (expander) looks a bit like the diveregent section of a nozzle and would accelerate the flow to M>1. As we all know M>1 cannot exist in a parallel conduit and thus at some point there will be a shock wave to bring the flow back to M<1. My recommendation would be to speak to the relief valve vendor. All the main names have good guide lines and will be able to offer advice. (Althought most don't want to take the responsibility for the tail pipe design and thus you might not be given a straight answer which you will have to interpret).

If I understand zdas04's post - which admitted I might not - I am not convinced. I would suggest speaking to Crosby, Dresser or however is supplying the valve PLUS investing in standard text book dealing with compressible flow.
(Crane 410 is a good publication but I wouldn't use it for this situation unless I fully understood what I was doing).
Best Regards

 

[UW1981]

zdso4/4pipes - thanks for the responses and insight. Given your advice, what we can glean from the PSV vendors, and our own internal experience, we will keep velcity <0.75M for this application and ensure we are nowhere near the critical pressure.

 

[zdas04]

I guess I wasn't clear so let me try again. The steps I take are:
1. Calculate releiving rate as though there were no backpressure (for compressible flow this will always be the mass flow rate associated with a velocity of 1.0 Mach).
2. Calculate critical downstream pressure
3. Calculate pressure drop in the pipe using atmospheric pressure downstream and flow rate from "1".
4. Verify that upstream pressure less than 80% of critical downstream pressure.

I would think that the velocity in the vent piping will be well under 0.75M if the line is any length at all (since pressure drop at high velocities tends to be really high).

David

 

[4Pipes]

Zdas,
Thanks for the explanation but I am still not convinced !

 

[zdas04]

4Pipes,
I'm not sure I was trying to convince anyone of anything. I was just describing a technique that I use and find defensible

In the PSV design course I took at the Mary Kay O'Conner facility several years ago, the instructor stressed that for relief headers you need to be exceptionally careful that there is no point in the header where the pressure is above the Critical Pressure for the lowest PSV set point in the header. As long as you ensure that, then your PSV calcs are valid. They are not valid if the rate through the valve falls below sonic because of back pressure in the header due to friction drops.

David

 

[gene2007]

If the relief system contains a relief valve or relief valves and discharges to a flare, I would consult with the flare tip manufacturer to find out maximum and minimum discharge velocities. If the discharge is to atmosphere, I suggest calculating the built-up backpressure in the discharge line to see if the pressure will adversely impact the operation of the relief valve or its capacity.

If the relief system contains no relief valves, there may very well be critical flow at the discharge location. For this case, as long as the system provides adequate flow and discharges to a safe location, critical flow is acceptable.