Is Choked Flow the Key to Designing an Effective Relief Valve Tailpipe?

[0507476h]

Hi,

I'm a final year master's student doing a study on relief valve tailpipe design for my final year project. The system which the relief valve is fitted to is a high pressure breathing air gas cylinder (300 bar) on board a large ship and it vents to atmosphere via piping which can be between 30 and 100 metres long and include bends and changes in section. I am looking for any advice on how to go about this study... From what I have read etc so far I think that when overpressure occurs and the relief valve opens the gas will flow through the tailpipe and increase in velocity until the flow is choked. This choked flow will create large resultant forces against the inner walls of the piping and it is the magnitude of this force that I should be trying to work out in order to determine pipe diameters/mountings requirements etc. Does this sound like I am along the correct tracks? Also I am working under the assumption that the gas involved is compressible and unsteady and therefore I will have to use isothermal gas laws rather than bernoullis? Any suggestions on literature/resources etc would be gratefully appreciated since I am a little lost with this project so far!

Cheers.

 

[Q_Goest]

Hi. Very briefly, ASME code (and virtually all RV manufacturers) prohibit pressure in excess of 10% of set pressure at the outlet of a relief device. So if the relief valve is set at 100 psi, outlet pressure during flow should not exceed 10 psi.

To accomplish this, the flow through the pipe should be calculated using standard methods. Generally, Crane Technical Paper #410 is used in industry as a guide to flow. They use the Darcy-Weisbach equation with Bernoulli's to determine pressure drop through a piping network. So basically, you start with the known flow rate which is determined by calculating the flow through the relief device. Use ASME or API code equations for this.

Next is to determine the flow restriction and pressure drop. For air, the pressure drop is generally sufficient that density changes must be taken into consideration, so you will probably need to break your line up into smaller sections and calculate each section individually, taking into account the change in density and temperature. To determine fluid states, you will also need to apply the first law of thermo and possibly heat transfer with the walls of the pipe. Or you can assume adiabatic or isothermal conditions, but that's often a judgment call and for a student that's doing a master's program, probably isn't advisable. I'd expect a grad student to be able to create a numerical analysis by creating a computer program that can handle changes in state as the air travels along the pipe. Start at the outlet of your RV header, you will need to determine if a shock wave exists. Sometimes, the flow through a vent header is too high to exit at atmospheric pressure. Regardless, you can work backwards from the outlet, determining the pressure drop from each individual flow restriction.

That's a very short description of what's done. Start with Crane TP #410. Also attached is a copy of a general discussion. Also go to your library and pull out ASME Section VIII, Div 1 and go through paragraphs UG-125 through UG-136. I also have a lot of old papers written on the topic which provide a cook-book methodology if you'd like.

 

[0507476h]

Q_Goest, thank you very much for your reply. I will start by trying to obtain a copy of Crane's Technical Paper 410 through our library at University and also the ASME document you mentioned (I already have ASME b31.3). Any old papers on the topic would be very much appreciated, thank-you very much.

 

[Q_Goest]

The attached "Calculation of Flow Losses in Inlet and Discharge Headers Associated with Safety Relief Valves" has been used quite a bit throughout industry.

(Note: Pg 7 is blank and has no information on it.)

 

[FredGarvin]

What a great reference Q. Thanks for posting that.

 

[0507476h]

Again thanks, much appreciated...

I am now busying myself with pressure transient surges of compressible gas caused by the choked flow, much like a water hammer effect apparently

 

[CFDFEAGURU]

Q_Goest,

Excellent paper on relief valves. I have a waste heat boiler project at work right now and I put that paper to good use. The head of my department (by default) let our subscription to the journals run out years ago and now we aren't allowed to renew them because of the cost.

Anyways,

Thanks
Matt