Direct and indirect CDE equlibrium

[jourak]

Hi,

I'm working with modeling the phosphorous transport in commercial filter materials. I have done two experiments for the model.

A batch experiment was done in order to find the sorbtion isotherm and a column experiment with a solution of phosphorus. The phosphorus solution was taken from the column in different time intervals to find out the effluent P concentration. I used the break through curves to find out the V and D value with given R value using CDE system with flux averaged concentration. The curve was fitted quite good with regression of 0.99 and values for D and V were found. Then I tried to use the direct mode of CDE to find the same curve but couldn't able to produce it. Here are my data, and it would be great if I can send my CXT file to you to check it.

1- dimensional time and length has been used
2- Time [day], length[m], concentration [mg/L]
3- Length of the column 7 cm
4- Kd value is 0.0185 m^3/kg
5- R value is 42,47
6-thetha is 0.446
7-inlet concentration is 11 mg/L

RSquare for regression of observed vs predicted = .99492190
(Coefficeint of determination)

Mean square for error (MSE) = .1129E+00


Non-linear least squares analysis, final results
================================================

95% Confidence limits
Name Value S.E.Coeff. T-Value Lower Upper
V.... .6172E-01 .3327E-03 .1855E+03 .6102E-01 .6242E-01
D.... .2937E-04 .3804E-05 .7721E+01 .2135E-04 .3740E-04



I want to check if I can get the same curve with direct mode. What can cause it not to work? I tried to use the dimensional direct CDE mode.

I know this was discussed a lot but when I'm collecting the effluent and measure the concentration shall I use flux average?

Your help is so appreciated.

Amir.

[rvang]

Jourak:

Something simple must have gone wrong. Send me the effluent data data and I will have a look at it (email to rvangenuchten@yahoo.com).

Did you use CFITIM within STANMOD the analyze the data (easiest)? Yes, you should use flux-averaged concentration data for the break-through curve.

Regards. --Rien van G.

[ntoride]

Thanks, Rien,
Jourak, If it were a problem of CXTFIT input such as "output structure", I will be able to fix your problem if you send me a project folder with a cxt file. Nobuo