PSV Model with sonic flow 1

[anubis512]

I've been giving a model of PSV, it's connection sizes, orifice size, flow rate, and set pressure. I'm attempting to model the discharge piping in AFT Arrow.

This particular PSV has a 3" outlet and is rated for 18,780 lb/hr of steam with a set pressure of 205psig. When it's run with just 3" piping discharged to atmosphere, there's sonic choking at the exit and the pressure drop is greater than 10% of the set pressure, so I tried to increase the piping to 4".

From what I've read online, this is called expansion choking, and just seems to make the situation worse in the model (increases pressure drop and seems to exacerbate sonic choking).

Any recommendations on how to eliminate sonic choking or reduce the pressure drop across the discharge piping so it's < 10%?


Thanks in advance.

 

[apetri]

to model discharging piping I proceed as follows
a) assume a pressure at PSV outlet
b) model the PSV to calculate the conditions (temperature, density, speed of sound) at PSV outlet (I use PRODE PROPERTIES which includes a specific procedure to model a PSV)
c) calculate a line diameter for a mach number about 0.5
d) verify pressure drop and density in different sections of piping to avoid mach numbers > 0.7 , if required (mach number > 0.7) increase line size.

 

[casflo]

Use the Liao´s paper "Analysis of Power Plant Safety and Relief Valve Vent Stacks" published in Transactions of the ASME, October 1975 or use the Crane Technical Paper No. 410 taking for K (resistance coefficient that corresponding to the discharge pipe plus the equivalent of the valve).
It doesn´t matter to have choking velocity at the end of the discharge pipe if there is enough capacity to discharge the required flow.
casflo

 

[cosh]

if it's for a single valve and have some pressure allowance you may accept sonic flow (mach number 1) in some section of discharge piping, providing that resulting pressure drop (note that higher pressure in some section = higher density = lower velocity = higher mass flow acceptable for sonic flow limit) doesn't influence PSV operating,
differently you may follow a procedure as that described by apetri

 

[anubis512]

Thanks for the response gentlemen,

I think I got my model sorted out. I didn't stop to think the placement of fittings could have such a large impact, but putting two elbows on the 4" section of piping after the area change made things much better.

 

[Longinthetooth]

We do a lot of verification analysis of existing instalations. You pretty well always have a choke at the exit, and it's not uncommon within the tailpipe somewhere. Not the way you would design it (as per Apetri's method), but as long as you satisfy the 10% rule, it works, although it may be noisy and you have to make sure it is well supported

 

[don1980]

You may not be able to avoid choke. A bellows PSV will solve the problem.