4 replies to this topic

[rosa]

Hi friends,

I want to calculate TEG vapor pressure from log(P)=A-B/(C+T).BY HYSIS I calculate A=18.462,
B= -8310,C=0.BUT by these value the calculated vapor pressure is very high and it is not reasonable.I think B VALUE is wrong.how can I found A,B,C FOR TEG(Triethylenglycol)??

Kind regards,
rosa

[Art Montemayor]

Rosa:

There is no need for HySys - or any other simulator engine. I keep telling engineering students: forget about Simulator engines and concentrate on gathering skills and abilities to handle and calculate values and data - EXCEPT when you are testing or simulating a process. Simulators were designed to simulate - not to design distilation towers, heat exchangers, regress or curve fit data, etc., etc.. Trying to use a simulator as a "Jack of all trades" results in erroneous and mis-applied answers. No single computer program can be found that will "design", compute, and define a total process. This particular problem of yours is a perfect example of that fact.

All you need to do is to gather an empirical listing of vapor pressures for TEG at differing temperatures. Take this data and regress it by applying the principle of Least Squares. This is the BASIC principle and method - the way we used to do it before we had computers. Today, the computer does the same, basic calculation we would have had to do in 1956, but it does it faster and more accurate. The practical and usual Numerical Method is to regress the values into a working equation by employing a reliable and accurate computer program. I have one that I use all the time for a variety of curve fits or regressions. It is called DataFit. You can obtain a cheaper program or down load a free version or trial version of other programs from the InterNet. By now, you should have been trained in Numerical Methods - a part of your Math education and a pre-requisite for engineering courses. There, you should have been taught the principles and the complexities of regressing data into a usable equation. The Antoine Equation is simply only one form of a variety of equations leading to the answer. Some forms are more accurate than others. I personally have found the Antoine form to be limited in its accuracy and prefer to select the most accurate model developed by my program, DataFit. DataFit, as most of these type of programs, generates over 100 model equations from the same data. Some models, including the Antoine variety, are some times less accurate. My program, DataFit, will generate a model of a custom User Format that I can specify.

You can find a lot of information and programs to download within the InterNet. I strongly recommend that you spend some time studying Data Regression and Numerical Methods so that you can understand the complexity of the process and appreciate the time-saving work that a good data regression program does for you.

[gvdlans]

Rosa,

I agree with Art that you can better use a curve fitting program than a process similator to estimate Antoine Parameters.

Apart from that, you fail to state the units that are used in 'your' Antoine Equation, and in which temperature range it is valid.
You can use the same equation but with different units for P (e.g. mmHg or psia) and T (e.g. °C, °F, R), but this will result in different values for A, B and C. Also sometimes the natural logaritm ln() is used instead of log10()

Following is from the NIST database (http://webbook.nist.gov)

log10(P) = A − (B / (T + C))
P = vapor pressure (bar)
T = temperature (K)

Temperature (K): 387. - 551.4
A: 6.75680
B: 3715.222
C: -1.299

So the vapor pressure at 400 K is 0.00274 bar and the vapor pressure at 500 K is 0.203 bar

[brindha]

how to calculate constants in Antoin equation................

[MrShorty]

brindha

Review post #2 above by Art Montemayor about regressing vapor pressure data. That is how you obtain Antoine equation constants (or constants for any other vapor pressure equation if you deem the Antoine equation unsuitable).

If you don't have measured vapor pressures to regress, then you are in the unhappy position of trying to estimate vapor pressures, which tends to be very difficult and full of uncertainties. If this is your situation, I would suggest you read chapter 7 of the text The Properties of Gases and Liquids by Poling et al. They have a section about estimating vapor pressures and the pitfalls associated with those methods.