I guess that you want to describe the course of the process in the packed catalyst bed. The form of the kinetic equation is known, where the rate is defined per mass unit of the bed. I do not understand, however, the reason for changing the porosity of the bed. This situation can occur eg. in the case of a gas-solid reaction, where the solid is a substrate. Nothing is also known about the assumptions made. It is also worth remembering that, according to the definition, the reaction rate is non-negative (it should be r = K ....., not r = - K .....). It is not mass balance.

But really, what the porosity concerns: a monolith, a bed or a particle of a bed? And whether it is about changes in time or along the bed? Regards,

Most Sulfur containing compounds consume more than 1mole of H2 per mole of S removed. Mercaptans are the exception:

R-SH + H2 => R-H + H2S

Sulfides require 2 moles of H2 per mole of S removed

Thiophenes and benzothiophenes generally require 3 or moles of H2 per mole of S removed.

There is usually also H2 consumed in side reactions like aromatics saturation, Hydrocacracking olefin saturation and other heteroatom removal.

What reactions are you attempting to model?

Stoichiometric equation and initial substrate concentrations are important in stoichiometric analysis. Here, it seems to me that the question is whether the variable porosity of the bed can affect the form of the kinetic equation (given above). In my opinion, it has no effect, and the variable porosity of the bed should be included in the mass balance equation. However, what is the reason for the change in porosity, I do not know. Regards,

The more rigorous way to mode HDS kinetics is with a separate 1 st order rate constant for each S compound. There nay also be langmuir type inhibition terms for H2,H2S and each individual sulfur compound. Typcal side reactions are cracking and Isomerization which become more important at higher temperatures. Some of the higher boiling S containing compounds like substitured Dibenzo thiophenenes have multiple raction paths/ Usually the path that saturates t leas one aromatic ring before removing the S and converting it to H2S is faster that direct Hydrodesuphirization. If you gove me some more details about what your feed looks ike (boiling range, S,N and H content etc) I may be able to help you a bit more.

Continuing the thought, I do not know what numerical method you are writing about. You probably want to solve a partial differential equation written against two or maybe more (3D) variables? Regards,

Dear Dr. Mirosław Grzesik ,

I will address below your points in order to obtain further help from you on this problem:

1. The variation of porosity is a common feature of the packed bed. A variable porosity with respect to axial position is provided in the article that I am trying to validate by a given function. But however, that was just to define the problem. I do not have an issue implementing a variation in porosity. My situation is a gas-liquid reaction and by variation of porosity, the liquid distribution is non-uniform and therefore, the gas-liquid contact is non-uniform which leads to a change in the reaction.

2. The following reaction rate equation has been taken from the article:

rHDS = −KHDS CArS1.6 CH20.56 / (1+kad CH2S) [kmol/(kgs • s)] .

So are you suggesting me not to consider the negative sign while running the simulations?

3. My system is basically a simple cylinder with spherical particles arranged in a random order which is why there is a porosity variation. I am running transient state simulations so changes are occurring both along the length as well as flow time.

4. I have actually described the system that I am modeling which basically has a single reactant gas and a reactant liquid which produces one product gas and a product liquid. Both my reactants are pure components and are set to an mass fractions of 1 (since my model is a two phase system). I am not concerned with the change in concentration with porosity. The problem that I am facing is that after running the simulation there is no formation of product and the reactant leaves the bed with the same concentration as at the inlet.

I am modeling the spherical particles as a porous media. For the two phases I am using the Eulerian 2-phase model by a relative permeability concept which is the standard method. I am new to modeling reactions in the software ANSYS Fluent 15.0 and need help with the steps to model a two phase reaction using the rate equation.

Warm Regards

What compound is represented by ArS1.6 CH20.56 . Is Ar reallt Arsenic or is it an abreviation for "aromoatic"

Dear Rick Manner,

The rate expression actually represents [ArS]^1.56 and [H2(hydrogen)]^0.56. Its basically the concentrations of Aromatic sulfide and Hydrogen raised to powers of 1.56 and 0.56, respectively.

What compound (s) do you consider to be ArS? The naturally occuring S compounds I am familiar with are Mercaptans (RSH) sulfides (R-S), Thiopenes (C4H4S) and Benzotiohenes (C8H6) Dibenzothiophene etc. I do not know what you mean by "Aromatic sulfide" do you mean thiophenes and dibentothiophenes? What compounds do you consider to be "Aromatic Sulfide. If you are actually performing pilot plant HDS reaction what compounds are you using as feedstock? 1,56 and .56 are also curious choices for exponents.

The typical "aromatic" S compounds you find in crude oil are thiophenes, benzothiophenes and dibenzothiophenes. I have attached a picture of what they look like. These compounds cannot be accurately characterized as ARS. What compounds are you actually using and what reactions are you modeling?

📷

Now it seems to me that I know what is going on and I will try to say something about it. I believe that the porosity of the bed is a global quantity and it should to be treated as a local one. I can not imagine how you can clearly define porosity as a local quantity. Of course, one can imagine the division of the bed into parts. The problem, however, is that such a division can be made in various ways, and there are limitations due to the size of such a part. Besides, in my opinion it would lead to flow discontinuity. A question arises: why introduce a change in porosity in the function of location (if of course it would be possible)? Will we get any additional benefit thanks to this? ........ I do not think so.

As I wrote above, the porosity of the bed can of course change as a function of time. But it is a completely different problem.

Another problem. I wrote that the rate of the process or phenomenon is always positive. The rate of a process or phenomenon expressed by a kinetic equation is something other than a mass balance of which the kinetic equation is an integral part. Here, we must remember that for the substrate there is a negative sign and for the product a positive sign. Regards,

I will ignore the fact the the reaction doesn't make sense to me and refocus on the fact that the reaction is not occurring in your simulation. Apparently the rate is at or very near 0 at the conditions you specified. This may be because the rate really is near 0 at the conditions specified or because the reaction hasn't been programmed properly. I suggest increasing the reaction rate constant by an order of magitude. If the reaction still doesn't occur increase it by several orders of magnitude. If you can't enter a reaction rate constant high enough to make the reaction occur at all you have probably modeled the reaction incorrectly. It could be an undefined variable or a variable always assignd to 0 or an error in the way you coded the reaction. something as simple as a . where a , should be could cause the problem

If it is not possible to carry out intermediate printouts, then the answer to the question of the type: "why zero at the output?" is difficult. It is also worth remembering that all variables can be set to zero before starting calculations. Just forget about one variable and the result will be incorrect.

I suggest starting with the assumption of a plug flow. Here, Excel and the Euler method are enough. Regards,