Hydrogen pressure-swing purification using Aspen Adsorption 1/2
ฝัง
- เผยแพร่เมื่อ 22 มี.ค. 2020
- Simulation of a breakthrough experiment and basic PSA cycle (2 beds, 4 steps) to purify a H2/CH4/CO gas. Software by Aspen Tech Inc.
Other references:
Lee & al., 2008 (doi.org/10.1002/aic.11544)
Xiao & al., 2015 (doi.org/10.1016/j.ijhydene.20...)
DISCLAIMER:
these workshops are organized freely and made for free; they are not part of a regular course.
Hi Antonio, thanks a lot for your aspen tutorials, easy to follow and not boring at all! I have a question regarding the initial condition inside the column: why didn't you change the initial default values of parameter "Gas velocity within first element"? Besides, does it refer to velocity as v=Volumetric_flow_rate/cross-section area or to the interstitial velocity, considering the bed porosity too.?
Hallo Giuseppe.
It refers to the overall volumetric velocity. Transfer inside pores is treated (if the relative option is chosen) via a dedicate flux term in a term of the continuity equation for the particle, but it's not the case of this workshop because I used the "lumped" mass transport approximation.
Actually, I tried several times to change the "first element velocity" in my aspen adsorption works, but it never seems to have any effect. If the simulation is meant to crash, that number won't fix it, and viceversa. Anyway, I'm not a real expert of AA (I'm more into AP), so there may be some glitch that I'm not aware of.
Hi Antonio, thanks for your step-by-step tutorials. There is two questions. Firstly, the changing of the parameter of methane, k(3,"methane"), at 43:28 is reasonable? Many papers have displayed the initial value but as the same of you the initial phenomenon is not followed with the result of paper. Secondly, with dynamic model the simulation is very slow and even discontinue. The report shows that the variables, generally the concentrations, usually beyound the bound of variables(e.g. 0 for the lower bound of a concentration variable). Have you ever seen this phenominon, and how to figure it? Thanks for your anwsers.
hi antonio, at 28:15 you used 0.0015m as the adsorbent particle radius, while the paper used 0.00115m. Is there a reason for the difference or is it a typo? Thank you!
Could you please tell me how the concentration graph changed from linear to nonlinear
Hi Antonio, thanks a lot ! According to the experiments the breakthrough occurs around 600 s , while the simulation shows breakthrough around 0 sec.
Why is this happening ?
Hallo. I supposed that the delay is due to the gas line, because the paper's author don't mention its length (but surely there must have been one), plus the void volumes (that are not mentioned, either).
So I choose to focus on the curves' shapes, that depends primarly on the absorption isotherms, the mass-transfer coefficients and also on the diffusion coefficients (all these things are documented, and so the comparison can be made).
Unfortunately this explanation in the video is in Italian.
@@Antonio_Tripodi Thank you very much !
Hi Antonio @Antonio Tripodi. Thanks for the video. Are you sure the RHOs ( @28:29 ) is 850 kg/m3 and not 482 kg/m3? Also the Bed dia should be 0.0742 m and not 0.0371 (which is the value of internal radius)
Hallo.
Honestly, I did not check every single parameter. If you deem there are errors, proceed correcting them in your own work as you deem fit.
The main purpose of these videos is to show a correct method, not to replicate someone else's simulation exactly.
Thanks for the hint anyway.
Buongiorno. Nel diagramma è l'adsorbimento in upflow quando si simula anche questo upflow o non è importante in Aspen.
Hi Antonio, thanks for the tutorial, it is very helpful. I'm having problem with the product flowrate, it doesn't change. Also, my concentration graph is linear. Any idea how can i solve my problem?
Hallo.
Start with a single bed. Set correctly the valves (preferably in mode 2).
Then go on to do the full layout.
Hi Antonio , in hysys is possible thanks for this help , Grettings of Colombia , help me please
Great job, I just have a question, do you know how to choose the value of tank voids for column dimensions (1 m diameter and 5 m high) and given flow? Here the flow is on a lab-scale
Hallo.
In general, you choose the diameter to keep pressure drop low, then the height is 3 to 5 times the diameter.
But it depends also on the bed void fraction.
@@Antonio_Tripodi Thank you very much for your feedback! I understand because I am doing a simulation and through the bed itself the simulation is going, but after adding the tank voids I get the error "your simulation is badly posed structurally because a subset of equations are not independent" every time. Have you ever had a comment like this and maybe know what it might be related to?
@@alekasndraazniowska3091 are there more than one bed?
@@Antonio_Tripodi Yes, I wanted to perform a
simulation with those layers in one bed: silica gel & activated carbon
Hi Antonio! Great video. I am running into a problem though: whenever I try to run the program, the column does not seem to actually adsorb anything. Within about 5 seconds of the simulation, the outlet flow stabilizes to equal the inlet flow, and the outlet composition begins to match the inlet composition in about 10 seconds.
Do you know ways to fix this? I copied all of the column internals that you had specified in this video so maybe it is something with the software?
Thanks for the helpful tutorial!
Are you talking about one column, or did you set up the whole tandem?
@@Antonio_Tripodi Just the one column: Nothing is being adsorbed.
Actually, with just one column, to have the same inlet and outlet is just what you expect, as time goes on. To check the model behaviour, you could run this test starting from "clean" bed (i.e. filled with hydrogen). If you deem that the equilibrium between the streams is reached too soon, you may modify the contact time (lerger bed or lower flowrate).
How can I find the document he used!!!
Hi Antonio! I simulated exactly what you did, but the concentrations of all components remained unchanged. What should I change?
Hallo.
Please, be more specific. To start out, check te isotherms' parameters.
@@Antonio_Tripodi If the process adsorbs a certain substance (eg hydrogen), the product stream may have a lower mole fraction of hydrogen because the hydrogen was adsorbed in the bed. However, the final concentration of the effluent is the same as the feed.
What do i have to change in the procedure to make a steady state analysis for this process? Could you make a tutorial about that?
Hi. Do you mean turning the dynamic analysis into a Gas-Steady-State Analysis, or to represent it into an Aspen Plus steady-state simulation?
@@Antonio_Tripodi I am making a process design in ASPEN PLUS V8.8 in which i produce a gas mixture of CH4,CO,H2 and i want to separate the hydrogen to use it as a pure reactant in another reactor afterwards.
Please, check out the channel, I've just added something that may be useful: let me know.
@@Antonio_Tripodi Hi. First of all thank you so much for making these videos. I have a stream that is comprised of 20% H2 and 40% CO2 and 40% H2O. I want to separate H2 by using ASPEN PLUS. Would you please make a short comment here to explain little bit.
Sorry for another comment.
You had made reply to previous comment, saying that ''You have added something related to this on your channel''. I tried to find that video but unable to get that. Would you please share its link here.
How did you determine Cv for VP valve? I am confused about specifying this value.
First step: set the process pressure and the product stream pressure, i.e. Pfeed and Prod
Second step: figure out a reasonble product flow, Flow_prod < Flow_feed
Third step: adjust Cv = (Pfeed - Pprod) / Flow_prod with the proper units.
@@Antonio_Tripodi Great! Thank you for such fast answer! Just, I believe that equation should be Cv = Flow_prod/(Pfeed - Pprod)
sure. sorry for mistyping
I enlarged the adsorption tower about 10 times, but the single tower penetration has been failing. What is the reason for this? What parameters need to be changed，thanks！
Respect to the original paper, I modified the flowrate acting on the valves opening. I did not change the bed dimensions.
Can you provide me a hard copy of simulation Of Pressure swing Adsorption
can i know what reference paper you used fir this simulation pls
Lee & al., American Institute of Chemical Engineers, 2008
Xiao & al., International Journal of Hydrogen Chemistry, 2015
@@Antonio_Tripodi thanks alot 👍🏻
At 41:34 what did yοu change in order to come closer to the figure of the paper
I forgot to reinitialise after the change in valve f
I changed the cv ov the bottom valve to decrease the overall gas flow in the system.
Then I surmised there could be a typing error in the paper in the adsorption parameters section, and I modified one of them.
Then I increased again the overall flowrate modyfing the Cv of the bottom valve.
If you forget to reinitialize after such changes, try to rollback to a time-zero snapshot.
Hi Antonio is there anyway you can send me this model to validate against mine?? Thanks!
you can validate your model against literature data
Thank you so much for your detailed video
6.8/60/1000*9.8*101300/8.314/298/9.7=0.045mol/s
0.000045kmol/s
0.000045/9.7=0.000004639 kmol/s/bar
Where Δp =9.7bar is obtained?
I thank you so much for your precious time.
Hallo.
in the paper I used as a reference, the beds were charged to about 10 bars, but discharged to the atmosphere, so the product control valve must bear the greatest part of the pressure drop, assuming that the pressure drop in the bed is small.
In hysys, is possible, this simulation?
Not that I know
🎉 sir can provide provide me a hard copy of simulation Of Pressure swing Adsorption
No, I cannot hand over these files, they were meant for my students only and I have not distribution rights.
Is it possible to at least add english subtitle?
I will work on it
Still waiting on the translation please
@@ituaogbewele8230 you can set to auto translate. It was good enough for me.
Hi , Do you have an email where I can contact you?
antonio.tripodi83@gmail.com
can you do this in English please
As this video /lecture available in English Language
no, only the last sessions about styrene process