Friedel's Model (two-phase) and Noise Attenuation Questions

[TiCl4]

I have a vessel with MAWP of 50 psig, with relief case required of 920 lb/hr of steam. I've selected a Kunkle 910-series, with rated capacity of 1,062 lb/hr steam. The PSV will be directly mounted on the tank nozzle, and will discharge through a tailpipe. The tailpipe exact routing is TBD, as we are still in design phase, but total equivalent length will be <50'. Steam quality after expansion through the PSV will be approximately 0.92. I know this isn't a huge amount of water, but I wanted to evaluate the tailpipe pressure drop through a two-phase model to ensure we stayed under the 10% built-up backpressure limit of the conventional PSV. I've built a spreadsheet to evaluate pressure losses using Friedel's model, and had a few questions:

First, based on two-phase flow evaluations from research papers, "good" two-phase flow models seem to have a standard variance of 30-40% from reality. Have you found this to be the general case for 2-phase flow models? In my case, if I discharge through 100' equivalent of 2" schedule 10S pipe, I'll have a calculated BBP of ~2.2 psig. My allowable BBP is 5 psig, so even with a 40% factor added to the 2.2 psig, I am still under the 10% limit. Is this approach sound?

If the calculated BBP were instead 4 psig, what would you do, given the inherent uncertainty with two-phase flow calcs? Would you limit the BBP to 5psig/1.4 = 3.6 psig, or would you take another approach?

Secondly, using API 521 approach to estimate sound energy, I get 120 decibels at a distance of 2 meters. Given this PSV is discharging through a tailpipe instead of straight to atmosphere, is this noise calcuation an estimation at the actual relief point at the end of the tailpipe, which is away from operators, or at the PSV itself, which is near operators? I assume the PSV discharging into a pipe would attentuate the noise for nearby operators, but don't have any experience to suggest what an actual attenuation value would be.

 

[pierreick]

Hi,
More correlations available to support your work.
You won't see many differences.

Distance Attenuation Calculator

The distance attenuation calculator finds how the sound level in dB decreases with distance from the sound source.

www.omnicalculator.com

If I remember well, the exit (after silencer) should be below 90 dB at 5 meters distance to protect workers.
Pierre

 

[GregLocock]

" I assume the PSV discharging into a pipe would attentuate the noise for nearby operators, but don't have any experience to suggest what an actual attenuation value would be."

A dangerous assumption. That depends on a host of factors, but you have basically created a steam powered organ pipe. if the end of the pipe is a long way from the operators you may be OK.

 

[TiCl4]

Greg,

That was a main question. The end of the tailpipe will vent outside the building (above the roof), and will be well away from any workers. I guess my questions was this: Is the noise calculation taken as the sound level at the discharge point of the tailpipe, or is there inherent noise at the PSV itself, regardless if it is discharging to atmosphere or not? I know the flow at the PSV will induce some noise, but not sure if the calculation only applies at the discharge point to atmosphere.

 

[pierreick]

I believe your calculation is incorrect, the delta P should be smaller based on data submitted.

Pierre

 

[GregLocock]

I'm /guessing/ that the 120 dB at 2m is for a PSV venting to atmosphere. Compared with say an IC engine that is pretty small beer, so you should be OK with truck type exhaust pipe. To get around the organ pipe effect you could add a Helmholtz resonator, or an expansion/contraction muffler, which can be retrofitted fairly easily if you get a pure tone which is too loud.

 

[The Obturator]

API 521 Calculates Noise based on an open valve outlet discharging to atmosphere. I will mention that there are manufacturers who can supply dedicated safety valve silencers that are fitted to the outlet stream.