Recip Compr Piping Vibration

[JohnSr]

In regard to acoustical analysis, API 618 App M.2 says "The piping system must be modeled to a point where piping changes will have insignificant effects on the parts of the system under study (usually a large vessel upstream and downstream of the units to be studied)."

My questions are:
1. How can I determine the length of pipe to be included? Is it based on the pressure fluctuations? Can it be determined before the analysis is done? Is it a function of the gas composition? Can it be determined from the compressor capacity, speed, and the pipe size? What is a "large" vessel? Is it relative to the volume contained in the piping or compressor? TIA

 

[sms]

It is really a bit subjective, but what you are looking for is either an acoustic volume that is large enough to be essentially infinte, or a pipe run that is long enough that it can be modeled as essientially infinite.

In a chemical plant or a refinery that usually means modeling upstream and downstrem to the first large vessel, usually a column or a tower or a reactor. Something on the the order of 10+ feet in diameter and 50+ feet tall. In my days of modeling natural gas compressor stations I would model upstream and dowstream to the main line and then terminate the main line with an infinite end, basically modeling it as an infinitely long pipe run.

-The future's so bright I gotta wear shades!

 

[PulseMaster]

Let the guy doing the study decide where to terminate the model. If you are doing a model, model more pipe than you think you need to, then reduce the scope by cutting out the piping with negiligible pulsation (i.e., long runs/after big volumes like sms said).

Don't pay too little for a pulsation study.