3 replies to this topic

[joker1]

Can someone help me design a vacuum system based on liquid ejector ? This process involves the condensation of solvent (eg. hexane) at 20 Torr. The solvent must be water-free (because water is detrimental to the catalyst used in the process) which eliminates the use of steam jet. The solvent is reused.
Vapor in: 3000 kg/h hexane vapor + 30 kg/h air at 150 C, 20 Torr
Utilities available: refrigerant (down to -40 C)
I am thinking of using refrigerated solvent as the motive fluid for liquid ejectors as well as to quench the vapor by direct contact (similar to a barometric condenser).

A mechanical compression/ liquid ring pump system is currently used. I have had little assistance from companies such as Croll Reynolds.

Many thanks

[Art Montemayor]

Joker:

Have you consulted with Graham at: www.graham-mfg.com ?

They allow you to download their VacWorks II software design program to assist you in such vacuum solvent applications.

They may not help you with the liquid ejector scope you have fixed, but they may have some other help for you. I would try them - if you haven't done so yet.

You seem to be tolerating a lot of air infiltration (30 kg/hr = 14.4 Scfm = 802.2 Acfm @ 150 oC & 20 Torr. Is this your estimate, or is this a real figure?

[joker1]

Joker:

Have you consulted with Graham at: www.graham-mfg.com ?

They allow you to download their VacWorks II software design program to assist you in such vacuum solvent applications.

They may not help you with the liquid ejector scope you have fixed, but they may have some other help for you. I would try them - if you haven't done so yet.

You seem to be tolerating a lot of air infiltration (30 kg/hr = 14.4 Scfm = 802.2 Acfm @ 150 oC & 20 Torr. Is this your estimate, or is this a real figure?

Hi Art,

Yes I have contacted Graham although I have not tried their VacWorks software. Graham as well as the majority of companies do not offer large or multiple parallel liquid jet systems. There seems to be very little literature available especially when the motive fluid also acts as the quench medium. Bird, Stewart & Lightfoot's Transport Phenomena has only a short example about the liquid-liquid ejector in Chapter 7. In this case I believe the major phenomenon is condensation which shrinks the vapor volume and contributes most to creating vacuum.

You are right about the air ingress. The estimate of 30 kg/h is extremely conservative.

Thanks for your input.

[Ali66]

Can someone help me design a vacuum system based on liquid ejector ? This process involves the condensation of solvent (eg. hexane) at 20 Torr. The solvent must be water-free (because water is detrimental to the catalyst used in the process) which eliminates the use of steam jet. The solvent is reused.
Vapor in: 3000 kg/h hexane vapor + 30 kg/h air at 150 C, 20 Torr
Utilities available: refrigerant (down to -40 C)
I am thinking of using refrigerated solvent as the motive fluid for liquid ejectors as well as to quench the vapor by direct contact (similar to a barometric condenser).

A mechanical compression/ liquid ring pump system is currently used. I have had little assistance from companies such as Croll Reynolds.

Many thanks

Hi Joker1,

I don't have specific numbers but I think using liquid ejectors would require a very high flow rate and/or multiple parallel units to pull the vapor load you stated. There may be other, easier options.

I have three questions:

1) What is the challenge/issue with your current setup?

2) What is the first compression stage you are currently using, i.e. from 20 torr to what pressure? If the 1st stage is reliable maybe it can be disconnected from your current setup and become part of a new, hopefully easier, arrangement.

3) Why can't you limit air ingress? 20 Torr is not too-deep a vacuum and air ingress is readily preventable (substantially). Unless you are using very large and old equipment.

Ali