Simulating a PFR with multiple reactor tubes

[andrewsoton]

Hi,

I would like to know whether there is a solution to modelling a PFR, or more specifically a packed bed reactor that consists of multiple reactor tubes, like a steam methane reformer.

I can only think of modelling a single reactor tube, and then multiplying the product stream by the number of tubes, but then it would be to a different scale with the rest of the process in the flowsheet. Or modelling each tube as an individual PFR unit, but there are over 100 tubes. Is there a function to model the PFR as multiple tubes or to multiply the product stream by the number of tubes?

Many thanks

Andrew

[jasper]

Does COUSCOUS' PFR reactor not meet your needs?

[link]https://www.cocosimulator.org/index_help.php?page=COUS/pfr.htm[/link]

[andrewsoton]

Hi Jasper, thanks for replying.

I may have the packed tube and the heating/cooling tubes mixed up. It was my understanding that the internal tubes in white represent the heating/cooling tubes within the larger packed reactor tube in yellow, as in this diagram. But I am trying to model multiple packed tubes within a furnace. Doesn't this arrangement make it a single reactor tube?

tube 2.PNG (17.79 KiB) Viewed 17967 times

Thanks you helping.

Andrew

[jasper]

Is this not the same as multiple times one packed tube (without internals)? And a given wall temperature?

[andrewsoton]

Yes, agreed. Suppose I need a reactor with 170 tubes that feeds into the next reactor, how would I modify the single packed tube? Thanks.

[jasper]

I suppose what you are looking for is a unit operation that divides the flow rate by 170. Then run the resulting stream through your reactor. Then multiply the flow rate again by 170. Such unit operations (which would not be in mass balance of course) could easily be done as custom unit operations (Excel, Matlab, Scilab, ...)

Another option would be a splitter - split 1/170 off and run it through a reactor. Then to simulate the other 169 tubes, use a stream duplicator that duplicates the product, with a flow controller on it that multiplies the flow rate by 169. Then mix.

[andrewsoton]

This worked well, thanks.

I have a couple more queries on the PFR please. I'm also trying to find the thermal conductivity of the gas flowing through the PFR. Can the reactor profile of thermal conductivity be viewed?

A value can be entered for the wall heat transfer coefficient, but what if the heat transfer coefficient varies along the reactor? Or is an average used?

[jasper]

There is currently no interface to display arbitrary physical properties along the length of the reactor (although they can easily back calculated from the profile data that is printed in the calculation report). There is also no interface for a length dependent heat transfer coefficient (what is the application for this?).

Such untypical customization of a reactor would call for a custom reactor implementation - essentially any program that will allow you to solve a system of ODEs (Matlab, Scilab, ...) could be used to model a custom reactor like that:

https://www.amsterchem.com/matlabunitop.html

https://www.amsterchem.com/scilabunitop.html

[andrewsoton]

Thanks for the links. I'm modelling SMR at the moment, and calculating U from empirical correlations. I don't require this customisation right now but asking just in case.

[andrewsoton]

Another query about the PFR please. As stated, the Ergun Equation is used to calculate pressure drop but I would like the equation to incorporate 'shape factor' aka sphericity, φ. Is there any way to factor this in?