PFR with axial dispersion - CSTR in series conversion?

1 reply · 3K views · last reply
Andrew O Regan
Messages
3
Reaction score
0

Homework Statement


Here is the problem description:

Develop an Excel file that given a set of data from an RTD pulse injection will determine the model parameters of the following schematics, and then predict the conversion in a CONTINUOUS reactor with a n-order reaction (where n is not equal to 1)

Block diagram of two reactors in series: a plug flow reactor with axial dispersion labelled alpha V and Dax, followed by a CSTR labelled (1-alpha)V.


Homework Equations


Four equations for residence time distribution: E(t) = C(t) divided by the integral of C(t) dt from 0 to infinity; F(t) = integral of E(t) dt from 0 to t; mean time tm = integral of t times E(t) dt; variance sigma squared = integral of (t minus tm) squared times E(t) dt; and an infinite series expression for E(t) with Pe and alpha_i terms.

Equation: tg(alpha_i) = 4.Pe.alpha_i / (4.alpha_i - Pe^2), with a trailing comma.

Equation: sigma squared over t_m squared equals 2/Pe minus (2/Pe^2) times (1 minus e to the power of negative Pe).

Equation for conversion X in a PFR with axial dispersion: X = 1 minus [4q e^(Pe/2)] divided by [(1+q)^2 e^(Pe q/2) minus (1-q)^2 e^(-Pe q/2)], with q = sqrt(1 + 4k tau / Pe) defined below.

The Attempt at a Solution


The above formulae were used in this assignment to take the user inputs and populate the excel spreadsheet.

Equation for the exit age distribution E(t) for a dispersion model: E(t) equals e^(Pe/2) over tau, times an infinite sum over i from 1 of (-1)^(i+1) times 8 alpha_i squared times e^(-(Pe squared + 4 alpha_i squared) t / (4 tau Pe)), divided by (4 alpha_i squared + 4 Pe + Pe squared).

The above formula is simplified to form the equation for the model which is calculated using varying values of alpha.

The flow model needs the roots of:
Equation: tg(alpha_i) = 4 Pe alpha_i / (4 alpha_i - Pe^2), with Pe in the numerator and Pe squared in the denominator.


Pe is calculated from the known result:

Equation for the variance of an RTD curve in a dispersion model: sigma squared over t_m squared equals 2/Pe minus (2/Pe^2)(1 minus e^(-Pe)).


This then generates the flow curve for Pe in excel.

Solving the equation above for this trial gives Pe=7.676. This was done in Excel using the Goal Seek function to find when the first term of the equation above was equal to the second.

As can be seen from the figure below, the term chosen for Pe gives a difference of 0.0152 between the terms.

Excel Goal Seek table titled "Goal Seek to Find Pe" listing 1st term 0.211162316, 2nd term 0.226370373, and error -0.015208056.


To calculate the conversion, a further two equations are needed:

Equation for conversion X = 1 minus a fraction with numerator 4q.e^(Pe/2) over (1+q)^2.e^(Pe.q/2) minus (1-q)^2.e^(-Pe.q/2).


Where q was calculated using formula mentioned above.

The eventual conversion calculated for this run was 68%.

This is for the conversion exiting the PFR.

The difficulty I am having is with the CSTR. I don't know how to relate the two reactors.

I am treating this setup as closed-closed.

I do not know how to translate the information I have into setting up the CSTR conversion section, to find my overall conversion?

Any information/help in guiding me to a solution would be very much appreciated.

Thank you.
 
I don't understand what they are asking of you. You show a plug flow reactor followed by a CSTR, and the plug flow reactor has axial dispersion. Somehow, you are supposed to find the parameters for this reactor system (the dispersion coefficient, the total volume, and the volume split) based on an RTD they give you. Is that correct? Let's see the RTD.