Fast Reaction in Equilibrium Reactor

[nrgeng]

A fast reaction recently ran after a CUP update. My last Forum submission stated that the update

...has made progress on another long-standing problem reported earlier (more later on this).

The reaction had a reported Keq of 5E+12. This from my viewpoint is progress on the previously-reported problem in Forum submission {Questions on CSTR & Equilibrium}.

Now in an attempt to extend the progress, I am submitting a new problem that may allow this work to expand. You previously stated

...because there is no unit operation available that mixes equilibrium reactions with reactions that are irreversible and instantaneous.

This new problem is similar to my last Forum submittal in that the flowsheet solves, but fails when a reactant flow rate is doubled. Please work your magic on this problem too. I would like to use this reaction with a reversible reaction in an Equilibrium reactor Unit.

I have e-mailed the flowsheet to you.

[jasper]

For the particular document you sent, there is only one reaction used in the reactor. This is the one with the equilibrium constant in the order of 1e12. If this is what you want to solve I suggest you use a fixed conversion reactor and let all oxygen convert.

If you want to mix this with another (the other) reaction, this will not really be feasible with the type of solver that is used. It wants to solve for a reaction extent such that the ratios of concentrations are so that the specified equilibrium constant is met. This requires a near machine precision solution (if not beyond machine precision) of the reaction extent. It is possible to solve such systems, but not with the formulation used by this reactor.

Solving for the particular case of your two reactions is rather straight forward (presuming you are not interested in how much oxygen remains exactly as it is near zero). You are looking for one degree of freedom, the extent of the reaction that is not taking place to full extent, such that its equilibrium constant is obtained; during the function evaluation you perform the other reaction to its maximum extent. Simple to do in for example a Matlab or Scilab unit operation.

In this particular case however the overlap in reactants appears to me such that there is an even simpler solution: first react all oxygen away in a fixed conversion reactor, then perform your equilibrium in a second reactor. I sent this setup back to you via e-mail.

I have put support for irreversible instantaneous reactions on the list of things to do.

[nrgeng]

A new flowsheet uses a reaction scheme for a fast reaction among several equilibrium reactions. Allowing a fast reaction to occur in an Equilibrium Reactor would be a useful development.

The process began using a Gibbs Reactor for all reactions. The Gibbs Reactor warned of at least one fast reaction so the fast reaction was isolated and placed in a Fixed-Conversion Reactor. Work on the equilibrium reactions has begun. Would like to include the fast reaction in an Equilibrium Reactor so as not to have to manage the energy effects externally.

A past Forum posting "Questions on CSTR & Equilibrium" and this posting thread suggested the possibility of combining the two reaction types, and this concept was placed on the list. Has this concept developed in the intervening time span? A current use exists so an update would be appreciated. Thank you.

[jasper]

A new flowsheet uses a reaction scheme for a fast reaction among several equilibrium reactions. Allowing a fast reaction to occur in an Equilibrium Reactor would be a useful development.

Agreed - it is on my list of things to do. This one will probably happen as soon as the new reactions spec is released by CO-LaN. There is is actually defined on a CAPE-OPEN level that you can have an equilibrium reaction with attribute "FullConversion"