3 replies to this topic

[directone]

I'd like to know what concentrations of propylene glycol I would need in a propylene glycol / water mixture to get a partial pressure of about 3 bar at a temp of 190C.

[MrShorty]

More questions than answers:

1) I can never remember if the common name propylene glycol refers to the 1,2 isomer or the 1,3 isomer. I'm guessing it refers to the 1,2 isomer. If you're not sure, you should probably verify which compound you're looking for.
2) Do you have any VLE data? I would almost expect that most literature data out there were measured at much lower temperatures/pressures, but it might give you an idea as to how non-ideal this system is.
3) Do you have access to a simulator? Not that I would blindly accept the simulator's answer without knowing where the interaction parameters came from or some sense of how reliable they are, but it would be a beginning.
4) How accurately do you need to know the answer? If you just need a very rough guesstimate, you could use Raoult's law. You could also use UNIFAC to estimate activity coefficients, but you'll need to recognize that, in addition to errors associated with estimating gamma, you may also run into issues using activity coefficients at that high of system pressure.
5) Unless you can find some good literature data, sometimes the only way to know the answer (to within reasonable uncertainty) to a problem like this is to take it into the lab and directly measure it.

[directone]

I'm after a rough answer really to start with, so I can choose some safety values for an experimental vessel wall thickness. My understanding of Raoult's law is that the molar fractions weight the pure partial pressure of each substance in the mixture.

If the mixture is "non-ideal" will this tend to increase or decrease the overall pressure of the system at > B.P. ?

UNIFAC?

[MrShorty]

I'm after a rough answer really to start with, so I can choose some safety values for an experimental vessel wall thickness. My understanding of Raoult's law is that the molar fractions weight the pure partial pressure of each substance in the mixture.

That's a basic statement of Raoult's law.

If you're trying to design a pressure vessel, I would think you would be more interested in the total system pressure rather than the partial pressure of the heavier component. Of course, on closer inspection, the boiling point of 1,2-propanediol is about 190 C, so it would be near impossible to get a partial pressure of 3 bar. Is that 3 bar your desired total system pressure then? If so, then an activity coefficient equation should be entirely adequate for modeling this system. In fact, I would fully expect that someone at some point has measured an isobar at 1 bar or 1 atm, which would be a good starting point for estimating activity coefficients at 3 bar. Do you have access to DeChema's database (university library maybe)?

Another thought in terms of safety. I don't know what you're going to be doing in your pressure vessel, but the vapor pressure of water at that temeprature is 12.5 bar. Depending on what I was doing, I might be inclined to design the vessel for this pressure, just in case someone added too much water or forgot to add the glycol or something.

If the mixture is "non-ideal" will this tend to increase or decrease the overall pressure of the system at > B.P. ?

Depends on whether or not the system exhibits positive or negative deviation from ideality (positive deviation means system pressures are higher than ideal, for negative, system pressures are lower). I would expect this system to exhibit some positive deviation, but you really need either some literature data to know that for sure, or you need to use something like UNIFAC to estimate the activity coefficients.

UNIFAC?

UNIFAC is a group contribution method based on the UNIQUAC equation that can be used to estimate activity coefficients. I'm not going to begin to try to explain the details. You might try putting UNIFAC into your favorite internet search engine. A few years ago, there were a few downloadable UNIFAC calculators out there, they still might be there. You might also find a good description of the method. I might also suggest referring to a good text (like Poling et al's The Properties of Liquids and Gases) for an explanation of how to use UNIFAC to do what you want to do.