Sizing Vacuum Pump for Batch Distillation

[BenjaminM]

We've got several processes which we perform vacuum distillation on a batch basis to remove solvents. In some cases we use Nash liquid ring, they are real work horses. In other cases it is a smaller Kinney vacuum pump.

I'd like to develop reasonable justification for what vacuum pump we use, and have a better understanding of what is actually required.

Does anyone have any know of any good resources they can point me towards for sizing a pump for this type of application? Google / search has been less than productive, and I'd prefer something more concrete that my personal gut feeling of bigger than a bread box, and smaller than a sedan.

 

[RVAmeche]

Definitely not a one size fits all use, but this vendor has this guide with a decent table of flows/pressures:

Link

 

[BenjaminM]

Thank you for the link RVAmeche.

Unfortunately it isn't a selection guide, but a sizing guide for which I am looking.

An example might be, I have a 10,000 gallon vessel from which I want to distill off 2,000 gallons of MEK, over a 12 hour time period at some temperature, say 60*C. What would be the minimum size pump to complete this?

I've found some online calculators for determining time required, but they don't provide a sufficient explanation of how they function.

https://www.chemengonline.com/partner-content/vacuum-pump-sizing-estimating-flow-rate/

https://vacuu-lan.com/vacuum-drying-process-time-calculator/

 

[georgeverghese]

Batch mode distillation is much more complicated than continuous steady state; suggest you run a process simulation model on Simsci or Aspectech Hysis Batch mode distillation unit operation, and tune the model to suit your needs. Most critical is the selection and correct inputs for the thermodynamics model for this, since you have very non ideal MEK to boil off. Also carefully study the Application Guidelines and Example Simulations CaseBook in Pro II, which has some simulations for non ideal components such as MEK also. Get help from the simulations company also for this challenging task - they have considerable expertise in the USA for Simsci-Pro II.

 

[Latexman]

What is the vacuum target at top of column?

Good Luck,
Latexman

 

[pierreick]

Hi,
Consider this resource to support your work.
note : I found your post confusing , the purpose of the vacuum pump is to pump non condensable material, not the solvent!
Between the distillation column and the vacuum pump you should have a total condenser to recover organic material . Solvent is not supposed to be carried over .
Good luck
Pierre

 

[BenjaminM]

pierreick,
We have a vessel with a solution containing a solvent. That vessel is connected to an overhead condenser which drains to a receiver tank. Vacuum is applied to the receiver tank, the solvent evaporates in the vessel, is condensed, and caught in the receiver. Ideally the outlet of the condenser is cold enough that there is little lose to the vacuum system. Thank you for the link. I actually may already have this document. Let me start digging through it again.

Latexman,
There is no column. Hopefully my response to pierreick bettwer explains the configuraiton. Minimum pressure (vacuum) may vary depending on the process. In the example case, we may limit vacuum to 4 PSIA to limit evaporation of the solvent collected in the receiver.

Georgeverghese,
Unfortunately I have been less than successful in convincing the powers that be of the value of Aspen. I believe there would be significant value in performing the simulations as you've described.

My interest is in what the minimum capacity of a vacuum pump is necessary at a set pressure for the process at hand. If our pump gives us 300 CFM at 4 PSIA (21 inHg)and it works, that's great, BUT could we get by with a vacuum pump that is 100 CFM or 10 CFM?

Intuitively I think the volume of the head space is important (is it a 10 gallon tank or a 10,000 gallon tank?), as is vapor pressure, and rate of evaporation, etc.

Thank you all again for your time.

 

[georgeverghese]

If there is no column and no reflux at column top, you can expect high "solute" loss into the overheads.
At 4psia, condensing temp is about 42degC for pure MEK, seems reasonable for cooling water as coolant in condensor

 

[JoeChem]

The most important piece of information needed when sizing/rating vacuum systems is the actual air leakage rate. Without this data you will be wasting your time as this item cannot be determined with any reasonable accuracy using correlations, nomographs, etc.

Plan for an opportunity when the equipment will be clean and dry, draw down the pressure, valve off the system and record the pressure rise over time. Use well documented equations to calculate the air leakage rate. Everything else you need can then be calculated with decent accuracy.

 

[jari001]

Do you have the scale up data/process chemistry info for the groups of products you work with? If you can't simulate the solvent evaporation process, you'll have to lean on some data to understand what capacity you need.

 

[georgeverghese]

To minimise the size of this vac pump, this MEK condensor would need to enable a subcooled liquid. So if you operate at 4psia at condensor top, the exit MEK may need to be at say 38-40degC. This overheads system design can get a little tricky, a controlled hot gas bypass is often installed in these cases where total condensation is required. You may get by with condensing and subcooling in a single HX.
What is the solute in this case ?