7 replies to this topic

[erwinfs]

Dear All,

 

In an article, I read that there are maximum velocity criteria in compressor. That article states that compressor suction velocity criteria is 20 m/s and for discharge is 30 m/s. But the article does not give the reference for this criteria.

 

Can anyone tell which standard or code does this article refer?

 

Thank you.

 

Erwin

[gegio1960]

constraints on compressor line sizing are usually given as rule of thumbs and each company has its own limits, often based on the compressor service. one can easily understand that could be some differences in compressing air, or hydrogen, or refrigerants....

anyway, the codes limit the gas velocities as a result of vibrations and noise.

the piping around any compressor is subject to detailed vibration (analog) study that can change the preliminary size of some pipe segments.

the 20-30 m/s reported by the article you mention are, for my experience, on the high side of the industry practice.

good luck!

[v7788999]

For gas velocity in the pipe line 20~ 30 m/s is not so fast. For gas line sizing normally will check the kinetic energy density x velocity ^ 2 shall be no larger than 10,000 kg/ms2

[breizh]

Hi ,

You may consider the standard attached .

 

https://www.cheresou...for-compressor/

 

Good luck

Breizh

Edited by breizh, 04 August 2018 - 10:16 PM.

[Fixerted]

Hi,

A basic question I don't seem to find the answer for easily.  When pipe velocities are quoted in m/s here and in the references given in this thread, that is Actual velocity, right?  I understand that the starting velocities or "rules of thumb" are just for initial sizing and we do the actual pressure drop calculations as we size the system but for general reference, would the guidelines of, for example, 20-30m/s be at Standard Conditions or Actual pressure, temperature, elevation?

 

For gases obviously the density and velocity in actual terms can be quite different than at STP.

 

For example 2, the IPS lists a table in 14.8 for noise from pipe with velocities corresponding to estimated noise listed in m/s.  To me there is no verification that this is standard m/s or actual. I would expect noise to be proportional to mass flow (Standard m/s) not volume flow (actual m/s) but I am not sure.

 

Thanks. 

Edited by Fixerted, 07 June 2019 - 04:21 PM.

[oilpmp]

The OP is quite old, but I will attempt to give an answer (for centrifugals) so that hopefully the thread can be resolved.

 

The guideline of 30 m/s max (often 35 m/s is acceptable too) velocity at compressor suction is because the vendor shall guarantee performance. Generally, but not always, the guarantee point is located in the best efficiency area which itself corresponds quite well to the area where the predictability of performance is "good" (which boils down to a statistical concept). This has to do with the smoothness/regularity of the flow as it flows across the inlet plenum towards the first stage, which is also the condition on which the machine is tested. Keeping in mind that a too low velocity (<10~15 m/s) is also not desirable for that purpose. This reasoning can be also made on discharge side (losses across discharge scroll).

 

In other words, if you have an alternative case operating on the choking limit line, or in the vicinity, it is clear that the velocity of the gas at the suction flange for that case will be well above 35 m/s (50 m/s is not unusual) ; yet, it is not forbidden to run on such operating condition but as far as predictability is concerned, the vendor wont guarantee such point performance, at least on a standard approach. If it is required to guarantee performance for such extreme cases (deal breaker), a risk assessment will be conducted, which typically implies extra contingencies and/or increased and hidden tolerance (determined by the probability to fail performance test vs. built in tolerance).

[Fixerted]

Oil PMP,

 

Thanks for the response. In your posting, are the velocities in actual m/s at the suction/discharge pressure & temp or standard velocity, m/s at STP?

 

Regards,

Fixerted

[oilpmp]

Velocity m/s are always based upon actual flow. Using any other reference (of pressure,temperature), does not make a lot of sense, physically speaking, as far as the analysis of the flow path vs. impact on performance is concerned (losses, velocity triangle representation). At the end of the day, standard or normal flow quantities depend on which convention has been agreed upon.