Mass balance error in non-reacting multiple continuum model
36 views
Skip to first unread message
Jordi Sanglas Molist
Apr 7, 2025, 3:07:35 AMApr 7
to pflotran-dev
Hi all,
I think that I have found a mass balance bug in the multiple continuum capability (unrelated to the other recent post in this forum). I simplified the model that I was simulating so that it only has a non-reacting tracer and I reduced the number of cells to 2 primary cells, each one having 2 secondary continuum cells. In the simplified model, the mass balance error is still present.
I attach the input deck in this message. The model is 1D and has uniform flow. Initially, there is almost no tracer (concentration of 1e-20 M). Then, a short pulse of tracer is injected through the inlet. The tracer travels through the model and leaves through the outlet, so after a long time we should get that the same amount of tracer injected through the inlet has left through the outlet.
Instead, we get that more than twice as much tracer has left through the outlet compared to the mass that was injected. The mass balance file is attached, and we can see the difference between "outlet Tracer [mol]" and "inlet Tracer [mol]" for late times. We can confirm that the model is initially empty, because if I remove the pulse injection we get close to 0 mass leaving the outlet.
Do you have any idea what could be the issue?
Many thanks,
Jordi
I think that I have found a mass balance bug in the multiple continuum capability (unrelated to the other recent post in this forum). I simplified the model that I was simulating so that it only has a non-reacting tracer and I reduced the number of cells to 2 primary cells, each one having 2 secondary continuum cells. In the simplified model, the mass balance error is still present.
I attach the input deck in this message. The model is 1D and has uniform flow. Initially, there is almost no tracer (concentration of 1e-20 M). Then, a short pulse of tracer is injected through the inlet. The tracer travels through the model and leaves through the outlet, so after a long time we should get that the same amount of tracer injected through the inlet has left through the outlet.
Instead, we get that more than twice as much tracer has left through the outlet compared to the mass that was injected. The mass balance file is attached, and we can see the difference between "outlet Tracer [mol]" and "inlet Tracer [mol]" for late times. We can confirm that the model is initially empty, because if I remove the pulse injection we get close to 0 mass leaving the outlet.
Do you have any idea what could be the issue?
Many thanks,
Jordi
Leone, Rosie
Apr 9, 2025, 2:43:21 PMApr 9
to pflotr...@googlegroups.com
Hi Jordi,
I was able to replicate the issue on my computer. I think the problem might be with the reflective boundary condition at the last cell in the secondary continuum. When I increased the length of the secondary continuum to where the tracer was not reaching the
boundary the mass balance problem went away. I'm working on a fix now and will keep you updated.
Rosie
From: pflotr...@googlegroups.com <pflotr...@googlegroups.com> on behalf of Jordi Sanglas Molist <jordi.s...@gmail.com>
Sent: Monday, April 7, 2025 1:07 AM
To: pflotran-dev <pflotr...@googlegroups.com>
Subject: [EXTERNAL] [pflotran-dev: 6321] Mass balance error in non-reacting multiple continuum model
Sent: Monday, April 7, 2025 1:07 AM
To: pflotran-dev <pflotr...@googlegroups.com>
Subject: [EXTERNAL] [pflotran-dev: 6321] Mass balance error in non-reacting multiple continuum model
--
You received this message because you are subscribed to the Google Groups "pflotran-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pflotran-dev...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pflotran-dev/24363df6-75f1-4010-8551-f5805d8895a1n%40googlegroups.com.
You received this message because you are subscribed to the Google Groups "pflotran-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pflotran-dev...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pflotran-dev/24363df6-75f1-4010-8551-f5805d8895a1n%40googlegroups.com.
Jordi Sanglas Molist
Apr 10, 2025, 3:32:02 AMApr 10
to pflotran-dev
Hi Rosie,
Thanks for your help!
Regards,
Jordi
El dia dimecres, 9 d’abril del 2025 a les 20:43:21 UTC+2, Leone, Rosie va escriure:
Reply all
Reply to author
Forward
0 new messages