Steam Stripping Process 1

[thewiseguy]

Hi, I have a process where I react an acid with an alcohol to make an ester, and water as waste. I need to use a 5 mol% excess of alcohol to complete the reaction in less than 12 hours. During the reaction I use a Dean Stark setup to take the water of reaction off and reflux the alcohol back to the reactor (the alcohol is not water-soluble). After the reaction the alcohol content is 1.5 wt% (some comes off during the reaction) and I need to reduce it to 200 ppm because the odor threshold is low and our customers are adverse to the odor. I currently steam strip the product under vacuum to do this. Unfortunately the steam stripping is a lengthy process and the process is low-yielding as a result. I typically get 80-85% yield and I want to increase that and, like everyone else, cut down on cycle time.

I have tried reducing the temperature of the strip from 90-95C to 70-75C, as well as raising the pressure from < 150 mmHg to ~300 mmHg. Both approaches were intended to strip off a higher fraction of the low boiler (alcohol) but added 2 days to the steam strip and neither improved the yield by much. We have a 110 deg. C steam jacket on our stack and I have also tried shutting that off to improve the yield. That worked a little better than changing the temperature and pressure but the strip time was still long. My reactor is used for many other things and it doesn't have a packed column. I am thinking I either need that packed column or I need to reflux the organic phase during the strip to have any chance of solving the yield and cycle time problems.

Does anyone have any ideas about how to improve my separation process? The normal boiling points of the alcohol and the ester are, respectively, 185C and 375C. The chemist at my plant has looked into some common solvents for LLE but so far he hasn't found anything that works.

Any and all suggestions are welcome. Thanks!

 

[ash9144]

I'm a bit confused by your actions to "reduce" cycle time. Normally higher temperature and lower pressure are used to accelerate stripping.

If you have some means of agitating (agitator or circulating pump) that would help. Another possibility is steam sparging directly into reactor.

 

[Compositepro]

Why are your yields lower due to stipping? Are you getting hydrolysis of the ester? That would explain the long cycle times. You may be generating alcohol while trying to strip it. Stripping should not take a long time. You should probably be using a stripping solvent that does not react with your materials. You could use nitrogen but that is not condensible and would make recovery of alcohol dificult.

 

[25362]

IMHO Compositepro is right and merits a star. Nitrogen stripping is possible and needs a well-designed condensing system.

 

[thewiseguy]

ash9144 - I took those actions to try to improve the yield, and I wasn't sure how they would affect the cycle time. Also I didn't mention that I am using a turbine agitator, a recirculation loop, and the steam is sparged through a gassing dip tube directly into the product.

Compositepro and 25362 - The yields are low because I am stripping off product. I have done alcohol balances on the reactor and receiver and I did not have more alcohol than when I started, suggesting no hydrolysis occurred. Both nitrogen and IPA are readily available and I have a vapor recovery unit so maybe that is the way to go. I tried nitrogen once but the alcohol content didn't go down at all, maybe because I just didn't have enough CFM's. (Batch size is 3,500 gallons and I was using 24 SCFM)

Thanks for your suggestions!