7 replies to this topic

[chemengrocks]

Hi !

 

I would like to know how to find the formula which relates temperature to the heat capacity of compounds in Hysys? (something that goes like Cp= a +bT +CT^2 +DT^3+...). 

 

Another question is whether I can model a cracking reaction which occurs in a furnace as a non-isothermal reactor with constant heat flux supplied by the furnace using MATLAB? Then for the value of the overall heat transfer coefficient (U) in the energy balance differential equation, how can I calculate it? (i.e. what correlations are available?)

 

Last question is in HYSYS, when I model an isothermal reactor (by setting the inlet and outlet temperature the same), the duty calculated by HYSYS in the reactor(shown in performance tab) is changing with the reactor length (in other words, variable amount of heat is supplied along the reactor length). So, if I want to assume a constant heat flux supplied to the reactor, can I instead specify a value in the energy stream that is attached to the plug flow reactor such that the inlet and outlet temperature of the reactor becomes the same?

 

Thank you.

[thorium90]

Hysys calculates the heat capacities differently, so the constants (if they are there) they use are different, but they can also be found in the properties for each component on the page where you first add components. But it wont be helpful if you want to do it in matlab. Best to get the constants from nist or books

 

This thread could be helpful

http://www.cheresour...gn-with-matlab/

 

If you supply the inlet and outlet temperatures, hysys calculates the heat required. ie: hysys fills up the remaining degrees of freedom. You can also put in your own value of heat, but you would need to remove the outlet temperature to keep the DOF the same.

Attached Files

•  111.jpg   77.81KB   26 downloads
•  222.jpg   26.32KB   24 downloads
•  1111.pdf   178.71KB   84 downloads

Edited by thorium90, 09 March 2013 - 01:23 PM.

[chemengrocks]

Hi !

 

Thanks for your quick reply to my queries.

 

I have another related question on HYSYS simulation.

 

I tried to simulate a reactor that heats up from 250C to 320C. Since I am assuming a constant surface heat flux to the reactor, I specify a value for the energy stream that is attached to the reactor such that it can heat up to 320C. By doing in this way, I see a constant heat duty profile shown in the performance tab of the reactor. However, when I look at the temperature of the tube surface, it decreases with reactor length. As for the temperature of the fluid, it increases with reactor length. This is an unreasonable trend because since I want constant surface heat flux (which is equal to the heat transfer coefficient multiplied by the temperature difference of the tube surface and fluid temperature), so as the temperature of the fluid increases, the temperature of the tube surface should also increase accordingly to give constant surface heat flux!!!

 

May I know how can I resolve this problem in HYSYS or do I have no choice but to simulate in MATLAB?

 

Thank you very much =)

[thorium90]

Perhaps you can post a screenshot of the following:

1) Under "Performance" tab, "Conditions"

2) Under "Performance" tab, "Flow"

3) Under "Worksheet" tab, "Conditions"

 

And any other pages you think might be relevant like your tube surface temperature?(Havent seen that page before)

 

Btw, the heat duty you entered, is the heat supplied to the reactor. Heat flux is the rate of heat flow. There is a difference between the two...

[chemengrocks]

Hi ! Thanks for the reply. I have decided to model my reactions in a series of adiabatic reactors with interstage heating with heat coming from the furnace.
 
However, I have some problems in reactor optimization. Firstly, for a specified conversion, how many adiabatic reactors should I have in series? Can I pass the reactor the reactor outlet stream for reheat when the reaction rate in the reactor starts to remain constant and there is no point to keep increasing the reactor length to increase the conversion? Secondly, how can I justify the conversion which I am aiming for? For your information, my reaction is irreversible and endothermic so it will not be limited by thermodynamic constraints.  Is it through calculating the costs associated with increased conversion and the separation costs incurred with decreased conversion (i.e. the trade off between reactor size and recycle) and then use the conversion that give me the minimum costs? Or it there other ways to go about justifying the conversion that I have chosen.

Edited by chemengrocks, 17 March 2013 - 12:33 AM.

[thorium90]

Oh, not limited by thermodynamic constraints? Perhaps what you mean is that since the reactions are irreversible, there is no equilibrium composition per se, but taking the sample in http://www.cheresour...gn-with-matlab/ , everything would become D after a sufficiently long time.

 

Perhaps the following will jolt your memory on chemical kinetics. Taken from "Chemical Reaction Engineering - Levenspiel O"

Quote from the same book, "The optimum temperature progression.... will always be at the temperature where the rate is maximum"

Finding the balance between increased conversion and increased separation could be a possible optimization objective function. But if your reactor has multiple reactions with undesired side reactions, it would not be so straightforward. The conversion to more products may also mean more conversion to undesirable side products which also needs separation. Plots of concentration of product vs reactor length at various inlet temperatures for isothermal reactors would make it easy to see an optimum. You would then see a peak for the product concentration for that reactor length at each temperature.

If your reactions are endothermic, non catalytic and irreversible, would being heated in a shell and tube exchanger do the job?

Attached Files

•  rctr size.jpg   80.96KB   6 downloads

Edited by thorium90, 17 March 2013 - 03:00 AM.

[chemengrocks]

Thanks for your fast reply. I would like to know if there is any heuristics saying how many tubular adiabatic reactors in series will be the optimum/best arrangement for a specified conversion? In other words, how do I decide how many adiabatic reactors should I have in series to achieve a specified conversion?

[thorium90]

Im afraid there are no heuristics on the optimum number of reactors. All reactions are different. As I have suggested previously, "Plots of concentration of product vs reactor length at various inlet temperatures for isothermal reactors would make it easy to see an optimum. You would then see a peak for the product concentration for that reactor length at each temperature."

If you are willing to follow through with the suggestion, you can post your plots here when you are done for the benefit of others

 

Might I suggest reading Chemical Reactor Design and Control by Luyben. It would be very useful in understanding more of your problem

Edited by thorium90, 20 March 2013 - 02:49 AM.