API 3% Inlet Loss for PSVs: concentric reducer directly at inlet safety valve included in loss calc?

[Gnoom]

Hi all,

With a colleague I'm having a discussion on how to interpretarte the API 520 Part II recommendation of not having more than 3% pressure loss at the inlet piping to the PSV. I'm already aware of the following: “The calculation should contain only NON-RECOVERABLE losses. For gases, you DO NOT include acceleration losses as these are recoverable losses. Frictional losses in piping (including entrance and exit losses), fittings and valves, equipment and rupture disks if there was any are the only losses you need to consider. This means you should only be using the incompressible flow equation (i.e. Darcy) in calculating the piping losses.” Which is clearly explained in threat

http://www.eng-tips.com/viewthread.cfm?qid=50179

Due to this 3% loss recommendation, you often have to "oversize" the piping to a larger diameter than the actual inlet connection diameter of the safety valve. In our case, we are having a 3" inlet pipeline to the safety valve, but the safety valves inlet connection is 1 1/2" (enough capacity for the relief case we have). Medium is natural gas (compressible fluid), set pressure 10barg. At the very end of the inlet piping we need a concentric reducer 3" x 1 1/2" to be able to install the valve.
In our flow calculations, the pressure drop of the inlet pipe is now mainly caused by the concentric reducer and just a little bit by the 3" pipe (which is a pretty long pipe in this case due to the vessel's double wall design). The concentric reducer is already causing approx. 3% of the pressure drop according our flow calculation.

In these pressure loss calculations, do you normally include the final reducing element to fit the safety valve as part of the inlet piping? Or can this be considered as "part of the safety valve"? I have to add to this that this case concerns a conventional safety valve, not pilot operated or balanced.

Thanks in advance for the feedback!

 

[EmmanuelTop]

3% refers to inlet piping, which encompasses everything up to the inlet flange of the PSV.
As for the 3% rule, you may find the following article useful:

http://www.eng-tips.com/viewthread.cfm?qid=383741

Dejan IVANOVIC
Process Engineer, MSChE

 

[Gnoom]

Thanks EmmanuelTop, I had a look at the pdf, interesting. It still seems awkward to me though, that just making the safety valve fit on the piping with this reducer, it is the reducer that's already leading to the 3%. Maybe I should doublecheck the flow calculation of the pressure drop over the reducer...

 

[Latexman]

Gnoom,

Is this a general discussion you are having with your colleague, or is it for a specific application? If it's about a specific project or installation, the scope of this discussion could be narrowed quite a lot with some details (PFD, P&ID, specs of the PSV, etc.).

Good luck,
Latexman

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

 

[Gnoom]

Hi Latexman, the discussion was originated because of a particular case, but we've had a similar discussion already for other cases, therefore for us it was a general discussion about how to work with these situations. Since in this particular case we are working with a threaded male NPT inlet of the safety valve (we can do with a pretty small orifice, not a lot of discharge mass flow required), we've found a good solution: going all the way up to the last flange with 1 1/2" piping and flanges and then install the safety valve in a threaded hole of a blind flange on top of the 1 1/2" flange (which is like working with a safety that has a 1 1/2" connecting flange, but a small orifice section). Furthermore, the type of safety we use will not close yet once reaching the set pressure, but will continue to blow off gas until the pressure is reduced a lot further. Since it is installed on a pressure vessel of several m3, pressure rise in the vessel will take time after a discharge of the safety.

 

[RVAmeche]

I'm having a similar discussion with a client. Our initial calculations for a liquid relief case showed the change in static pressure for the inlet piping exceeded the 3% rule, but when you only look at the frictional losses, it's less than 1%.

Is there any official interpretation or articles discussing using this topic? I don't understand from a practical standpoint how a decrease in static pressure due to a velocity change wouldn't effect a relief valve and potentially lead to chatter?

 

[Gnoom]

Hi RVAmeche,

from what I understood reading other threads, this is mainly due to the force the gas or liquid flow is applying due to its kinetic momentum. Try to picture it as a water canon shooting at you: the water has left the hose already (so static pressure equals ambient pressure), but still you fill a lot of force due to the momentum of the fluid. The disc you're lifting in a safety valve only receives static force due to pressure difference when it is still closed just about to open. But when it's fully open, there is a lot of mass flow crashing against the surface of the lifted disc.