There is an issue with the pressure along the wellbore of the horizontal well.

76 views
Skip to first unread message

Calvin Scott

unread,
Oct 12, 2023, 2:02:57 PM10/12/23
to MRST-users: The Matlab Reservoir Simulation Toolbox User Group
Hello,
I am using a multi-segment wellbore model to simulate the coupling of a horizontal wellbore and reservoir to calculate the frictional force of the horizontal wellbore. I have set a fixed bottomhole pressure production (600barsa) and the model runs successfully. However, the results show that the pressure of each node in the horizontal wellbore is constant and equal to the fixed pressure value I set, whereas in normal situations, the pressure should decrease gradually from the toe to the heel of the horizontal wellbore, with only the heel pressure being fixed at 600barsa. When I use the same data settings and run the model in Eclipse, the pressure decreases from the toe to the heel of the horizontal wellbore.
Could you please advise me on what might be causing this issue? Is there something wrong with my settings?
Below is the graph of the results of my calculations.

Thank you.

Best regards.
CalvinSnipaste_2023-10-13_01-57-41.jpg
GW_topo.m
topo_gaswater3phase.DATA

Stein Krogstad

unread,
Oct 13, 2023, 6:09:43 AM10/13/23
to MRST-users: The Matlab Reservoir Simulation Toolbox User Group
Hi Calvin,

Thanks for your justifiable question. You've stumbled across a situation where there is lacking some warning or error message. In essence, the MS-well model in MRST is sort of hard coded for 3-phase black-oil, and hence only works with the ThreePhaseBlackOilModel.

If you rather use model = ThreePhaseBlackOilModel(...), you will see results more along what you are expecting.

Note however that your well definition seems a bit problematic. Your well has 64 grid connections but only 8 of these are assigned to well-nodes. Either you could have a 1-1 correspondance (that is 64+1 nodes), or you could assign group of connections to individual nodes. For that latter, your cell2node could have ones in positions (2, 1:8), (3, 9:16), (4, 17:24) and so on.    

Best regards,
Stein

Calvin Scott

unread,
Oct 14, 2023, 12:46:13 PM10/14/23
to Stein Krogstad, MRST-users: The Matlab Reservoir Simulation Toolbox User Group
Dear Stein Krogstad,

Thank you very much for your guidance. You corrected a misconception I had, which was that the multi-segment well model was only applicable to oil, gas and water three-phase systems. Now I understand that it is not only limited to three-phase systems, but also only applicable to the ThreePhaseBlackOilModel.m function. Thank you very much for your help. However, it seems that there might be some issues with the results of my calculation. The theoretical pressure in the wellbore should decrease from the 9th well node to the 1st well node (600*barsa), instead of what is shown in the calculated graph below. Additionally, each well segment should have some frictional pressure drop, so it is not possible for the pressure at the 4th well node to be the same as the pressure at the 9th well node as shown in the graph. Did the frictional pressure drop disappear? Could you please help me with this? Thank you.
image.png

Also, I didn't fully understand your explanation of the node setting issue. My cell2node function seems to assign the first 64 cells to different well segments as you said. Could you please provide me with more guidance? Thank you.

Best regards,
Calvin

Stein Krogstad <stein.k...@gmail.com> 于2023年10月13日周五 18:09写道:
--
You received this message because you are subscribed to the Google Groups "MRST-users: The Matlab Reservoir Simulation Toolbox User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sintef-mrst...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sintef-mrst/96981f98-5a77-4f70-b5a6-5aa506542404n%40googlegroups.com.

Stein Krogstad

unread,
Oct 16, 2023, 4:19:05 PM10/16/23
to MRST-users: The Matlab Reservoir Simulation Toolbox User Group
Hi Calvin,

If the production from connection k shall end up in node n, then element (n, k) of the cell2node-matrix should be 1. That is, each column of the  cell2node-matrix should be all zeros except a single 1. In your cell2node there are only 8 non-zero entries, so 56 of your connections are not assigned to nodes. I think the results look reasonnable if I run your code with 
cell2node = sparse(rldecode((2:9)', 8*ones(8,1)), (1:64)', 1, 9, 64);

Note, however, that the rates entering nodes 4-9 are quite tiny, so the frictional pressure drops there are several orders of magnitude smaller than that of the first segment.

Best regards,
Stein 

Calvin Scott

unread,
Oct 17, 2023, 3:07:36 AM10/17/23
to Stein Krogstad, MRST-users: The Matlab Reservoir Simulation Toolbox User Group
Dear Stein ,

Thank you for your kind assistance. The problem has been perfectly resolved now.
I am very grateful and wish you good health and a pleasant work experience.

Best regards,
Calvin

Stein Krogstad <stein.k...@gmail.com> 于2023年10月17日周二 04:19写道:
Reply all
Reply to author
Forward
0 new messages