When I began this investigation of the capabilities of the model-based tracer methods, I wanted to see for myself how well they performed as the models became more complex. I expected success with simple models, but I have been surprised at the success with Model 4, a fairly complicated model. After all, we are using a single data curve to obtain several parameters simultaneously.
As I previously mentioned, we don't know how the noise due to the chaotic flow field will affect the accuracy of the estimation results. Thus, if someone can provide data from either (1) an actual tracer study on a fluidized bed, or (2) from a simulation of a tracer study using a CFD model of a fluidized bed, I would like to hear from you. Please use the contact form on this website if you wish to collaborate on this phase of the study.
Commercial tracer resources
Contrary to the belief that tracer studies can't be done on industrial size reactors, there are companies that specialize in just that technology. In addition, software can be obtained that allow you to build the complex flow models such as I have done and that then analyze the tracer curves. I won't mention these companies because I don't want to favor any in particular.
A very thorough tracer tutorial
I recently discovered a paper that provides a great tutorial and numerous applications of tracer studies. It is by the International Atomic Energy Agency...follow this link.
For another example of the model-based approach
There is nothing new with the model-based approach I have described, other than the fact that Mathcad greatly simplified the parameter estimation process. Shin et al. (2002) give an example for a simple model with a semi-dead zone. They too used the Levenberg-Marquardt optimization method, but they had to program that algorithm. Both Shin et al. and I minimized the errors between a model curve and the data curve. Shin et al. used the E(t) curve whereas I used the c(t) directly.
This ends the series on the tracer studies. The next post will discuss the FCC regenerator model, including the incorporation of the parameters from the tracer study.
Shin, Seungkwon, Jongbum Kim, Sunghee Jung, and Joonha Jin. “The RTD Measurement on a Submerged Bio-Reactor Using a Radioisotope Tracer and the RTD Analysis” 1, no. 2 (2003): 210–14.