Dew Point of Tower OH

[ngb23]

I am working in a refinery in the NE and am having trouble remembering how to go about the thermodynamic method of determining the dew point of water in a multi-component stream. I am considering multiple distillation columns that run at a variety of different temperatures and pressures, but the idea is the same. The first droplet of water that forms as the temperature of the OH drops will be very corrosive; therefore, it is useful to keep the OH line above the dew point temperature prior to a water wash of the OH stream. So I need to calculate the dew point temperature of water at a given pressure and an assumed mol% water in the OH stream (will be a very small mol%, as the towers I am considering are run "dry"). I am not sure what other information I need; I thought that was it, but I am having trouble. Any nudges in the right direction would be helpful!

Thanks for your help,

NGB

 

[25362]

For a variable and unknown moisture content, would it be too expemsive (or impossible) to keep top temperatures above the corresponding water boiling points at the prevailing operating pressures?

 

[ngb23]

A few things:
1) Overestimating moisture content allows me to have an acceptable cushion in regards to this calculation
2) It is my understanding that boiling point is different than dew point
3) Knowing this temperature is useful not only for control, but for identifying where reliability issues may occur, whether in the tower or in the OH system

Thanks,

NGB

 

[ngb23]

After considering this and consulting some of my old textbooks, I think I know how to proceed.

I can get the tower OH pressure from our plant monitoring software, trend it back for a few years, and estimate an average tower pressure. Using anything to give me a hint (lab data, design data, modeling data, etc.) will let me estimate the mol fraction of water in the OH. Assuming that the first droplet that forms is 100% water, I can then use Raoult's law to calculate the saturation vapor pressure using:
P*(@Tdp) = *(P)
where P*(@Tdp) is the water saturation pressure at Tdp
Tdp is the dew point temperature
y is the mol fraction of water in the vapor phase
P is the OH pressure

I can then look up the dew point temperature here:

http://www.engineeringtoolbox.com/water-vapor-saturation-pressure-d_599.html

by iterating between values on the chart.

Let's say I find that my P*(@Tdp) is 5 psia. I would then have to linearly interpolate using:
(P(@Tdp) - P(@160))/(P(@180) - P(@160)) = (Tdp - 160)/(180 -160)

I thought I would just include the interpolation formula for fun.

In regards to the assumption of constant mol% H20, I know in reality this is not true. However, by overestimating this number, I will overestimate the Tdp, introducing an element of safety into the calculation. Also, company standard is to add 25% safety cushion above this calculation, so overestimation is desirable anyway. Any comments on this method welcome!

Also, I know using Raoult's law is not as accurate as I could be. Also, if the water dew point is lower than the HC dewpoint, I cannot assume the condensing fluid is 100% water at that point, but I'll cross that bridge when I get to it....

Any suggestions?

Thanks,

NGB

 

[25362]

The envisaged process hasn't been described in detail. Therefore, and as aside, have a look also at thread124-184008, concerning possible corrosion factors.