3 replies to this topic

[sreekanth]

dear friends,
how does a multicomponent vapor mixture condenses. will it condense between its bubble and dewpoint. if so how do i know it latent heat of condensation. my primary objective is to design a condenser that condenses product vapors from a reactor in dimethyl ether production by methano dehydration. the vapor mixture consists of 40%DME, 40% water and 20% methanol unconverted, coming out of a reactor at 15 bars pressure and
380 0c. i need to condense it to a 80%liquid mixture before i let it into distillation tower. so my problem is whether is it possible to condense 80% of total flow of vapor to liquid and if possible will the liquid and vapor phase composition be same.
pls help me.

[babucher]

Let's see how much I can help, and what I need to have my understanding corrected on by other more knowledgeable members.

1. The vapor mixture will always condense at it's dew point. That's what the dew point is. However, the dew point temperature (assuming constant pressure) will change as condensation occurs because the component mole fractions will change, since some components are condensed more rapidly than others.

2. Calculate the heat of condensation by summing the component heats of condensation.
Heat of condensation = enthalpy in the vapor phase - enthalpy in the liquid phase

2a. I'd like to see some comments on how non-idealities affect these calculations. Among any other things, how does non-zero heats-of-mixing affect things? Would a non-zero heat of mixing affect the condensation of the vapor, or would it have effects

3. "Can I condense 80%...?" Of course. Will the compositions be the same? No. If you want to determine the compositions of the phases I recommend looking up single-stage batch evaporation and reversing the technique to come up with a single-stage batch condensation model. Perry's (7th) has information on page 13-96.

Hope this helps,
Brian

[babucher]

Correction! You would not use a batch-like procedure. I don't know what I was thinking, but it makes no sense since batch procedures are used when the vapor is removed and so does not remain in contact with the liquid.

I believe you would use a flash-like procedure, solving the MESH equations. Exactly how, I'm not completely sure. Let me ponder it a bit more.

Brian

[babucher]

Ok, this is what I came up with. Input from others is highly appreciated.

Since it's an equilibrium state, it doesn't matter whether one condenses to it or vaporizes to it, so I'm conceptualizing it simply as a flash vaporization, but the equations don't actually require that. The H (enthalpy) equations are not used unless you want to determine the energy requirements of condensing the vapor. Here is a method that should work, if you assume Raoult's and Dalton's laws (ideal gas, ideal solution). I'm also assuming you use a constant pressure.

Write component mole balances.
Write component equilibrium balances xi = yi / Ki and plug into mole balances
Solve the equations for yi
Plug Ki = pi* / P into the new equations (pi* is the pure component vapor pressure)
Plug the aboved solved for yi's into the summation equation SUM(yi) - 1 = 0
Iterate on temperature (to get pi* values) until the equation equals zero.
Get the pi* values from the temperature and calculate the Ki values
Plug the pi* (or Ki) values into the previous yi = f(pi*) (or = f(Ki)) equations
Plug yi and Ki into these equations to solve for the xi values xi = yi / Ki
That should give you your yi and xi values.

If you have a system where you can't say Ki = pi* / P, then use software to do it like ChemSep.

Brian