vacuum Distillation scale-up

[StoneCold]

We are attempting to scale up a batch vacuum distillation. The orriginal unit is very small. It has a still that holds about 25 gallons a has a 4" diameter column. The column is packed with a dumped expanded metal packing. The distillation is ran at a receiver pressure of about 1 torr.
We are trying to move the process over to a larger column with not much success. The new column is on a 400 gallon still and is 10" in diameter. The column is packed with a similar dumped packing. The small column has a flapper/timer type splitter and the big column utilizes a gravity fed three-way ballvalve. The problem that we are having is that the bigger column does not distill much faster than the small column. Does anyone else out there try to distill at such a low pressure? Has anyone else out there tried to use a three-way ball valve as a splitter? Have you had any success? Have you seen problems with subcooling of the reflux? The reflux flow is very slow. Maybe 100 ml/min. Any ideas?

Thanks

 

[25362]

When comparing the small unit to the prototype: to distill you must provide heat; to distill faster you need more heat. What about the pressure drop from the top receiver at 1mm Hg abs. pressure to the pot ? Is the packing of equal size to that with the smaller column ?
Wiped film evaporators and molecular distillations use this level of vacuum or better. I know of lube re-refiners using them.

 

[StoneCold]

Heating the still is not a problem. The problem is that when we add more heat the pressure drop increases in the column and the temperature rises in the still bottom. This causes the product to decompose. I believe that the pressure drop in the larger column is close to the same as the little column. However the larger column is much further from the vacuum pump than the little column. Ultimately it may be that we are not getting as low a vacuum at the large column. Use of the large columns still as a single stage flash unit yesterday seemed to indicate that the pressure is higher in the large still than in the small unit. Based on a comparison of the boiling point of the products.

 

[25362]

There is the answer to your query. You must solve the delta P problem, first. But beware ! the smallest leak of air may ruin the whole exercise.

 

[wbasinger]

"Three Questions"

1) Is the larger vacuum distillation column and take-off head well insulated (very important)?
2) Does the vacuum pump you are using have the "horses" to handle the desired vapor rate, offset system leaks, and to overcome the column dP (very important- check pot and head vac. readings and do a leakdown check with nitrogen to detect major leaks if any are present)?
3)Most important-if my memory serves me correctly, dump packing is typically not the ideal packing for pilot size or larger vacuum distillation units (too large of a dP for each HETP--would require you to pull a much lower vac. than with the smaller unit to increase throughput). Have you contacted Norton to see what they recommend for your system (ie. screens, etc.)?

Good Luck,

William Basinger
Ajay-SQM Group

 

[chemtiger]

i have some experience on the pilot, because i have scaled up my small glass column to 2.8 m diameter vacuum plant.
the key problem is vacuum, then is the boiler and insulation. If you just build a small pilot plant, just solve the problem of vacuum and insulation , then all is ok. If you want to build a big plant, you must choose the boiler type and packing from the begining.

 

[StoneCold]

Wbasinger
1. Yes the column is well insultated. I had thought that this might be a problem as well and we went back and increased the insulation from two inches to four inches of fiberglass.
2. Vacuum pump size is a problem. But it is mostly the way that the vacuum is ran to the column. It is routed though 1" hoses with camlok fittings. This causes additional leaks and excessive pressure drop. We are revamping the vacuum system to try to fix this problem.
3. You are right in saying that dumped packing is not the best for pressure drop. However it was about a third the cost of going with structured hastelloy packing, and I can get an HEPT of about 3" on this little column. (confirmed by testing)
4. We may have a water leak in the condenser that is complicating things. I have tried putting air pressure on the shell side and watching for a tube side pressure rise but I have not had much luck yet. Gettting the time to explore all the options is not easy.

Thanks for the input.