Reducer/Expander effect on PSV? 1

[ipocoyo]

My current D orifice PSV has an inlet size of 20mm. If I were to attach a 8mm reducer to the inlet, what effects does it have on the PSV in terms of its set pressure, flow capacity etc?

Thanks in advance!

 

[RVAmeche]

PSV inlet piping cannot be any smaller than the inlet connection of the device. So if you have a 20 mm inlet connection, all inlet piping needs to be at least 20 mm. Furthermore, the total inlet losses cannot exceed 3% of the set pressure at rated capacity.

 

[ipocoyo]

Is there a reason for this? Are there any references for it?

 

[Latexman]

You still haven't read API 520 and 521 have you?

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[RVAmeche]

The PSV has a defined flow rate. Reducing the area of the piping will either result in a large pressure drop between the vessel/system and device at the flow capacity, or the device will not be able to reach it's flow capacity. API 520 and ASME both spell this requirement out.

3% losses is the general standard to prevent chatter. Basically the when closed the system pressure rises; once the valve begins to open the pressure seen at the valve drops below its set pressure due to the high losses of the piping and the valve re-closes. Since the system pressure is still high, the valve keeps opening and closing and destroys itself.

 

[ipocoyo]

Thank you all. I did not realise there as a second part to API 520.

 

[RVAmeche]

I'd highly recommend you have a seasoned engineer review your relief valve calculations and design. It's not something to be rubber stamped

 

[ipocoyo]

Just going through api 520 clause 7.3.2 and I have another question.

Since the excessive pressure loss will cause chattering and hence, lower the capacity of the PSV and quite possible cause damage to the seating surface, will the PSV still be able to relieve excess pressure in the event of an over pressurisation?

 

[e43u8]

PSV might still be, more or less, able to relieve excessive pressure in an over pressure case; but its seating surface damage may cause having continuous leakage through the PSV even in normal case...

 

[RVAmeche]

The PSV may still operate but could leak due to seal damage like e43 mentioned. However if we're still talking about an 8mm pipe for a 20mm inlet PSV, it's very likely you won't get the flow capacity. The system pressure would likely need to go much higher than your allowable overpressure to result in the PSV reaching it's flow capacity since your 8mm pipe will have significant losses.

 

[don1980]

Seat damage is a trivial matter, and a risk that's generally tolerable. Note that every metal seated PSV has a high risk of leakage after activation, due to seat damage. Chatter is a concern because it risks: (1) insufficient relieving capacity, and (2) loss-of-containment due to mechanical failure of the installation. The flowrate through a chattering valve is only a fraction of the valve's rated capacity, and this will cause the system pressure to rise. In most cases an equilibrium will be reached, at which point the PSV operates in a stable manner, albeit at an pressure that's probably greater than the allowable accumulation. The risk of loss of containment, due to mechanical failure, is much higher for PSVs in liquid service. That's because the force of impact, which occurs during each open-close cycle, is much greater in a incomprssible fluid service. Loss of containment can occur with a chattering PSV in incompressible fluid service, but the risk isn't as great.

To minimize the risk of chatter, most design codes require the inlet pipe (x-sec area) to be at least as large as the PSV inlet connection. In this particular case (8mm pipe leading to a 20mm PSV inlet), the restriction is severe. That's not a good design. It's a code noncompliance, and most likely it's an significant safety risk.