Commit Graph

3504 Commits

Author SHA1 Message Date
Tor Harald Sandve
dc8f811cbe Remove WellStateFullyImplicitBlackoilDense
After the restructuring of of the well model, keeping an extra class for
the "Dense" model is not needed. The only thing still left in
WellStateFullyImplicitBlackoilDense was some solvent related stuff, this
PR moves this to WellStateFullyImplicitBlackoil and removes
WellStateFullyImplicitBlackoilDense.

In addition to a cleaning code this PR fixes missing solvent well output.
2017-10-11 10:54:59 +02:00
Atgeirr Flø Rasmussen
de1be6b5ef Merge pull request #1273 from GitPaean/different_oscillation
avoiding the oscillation treatment applied to well solution
2017-10-10 14:29:04 +02:00
Atgeirr Flø Rasmussen
63b38cfe20 Merge pull request #1281 from totto82/fixRESV
Use pore volume weighted averaged hydrocarbon state in rateConverted.
2017-10-09 13:42:43 +02:00
dr-robertk
682424a7bf Merge pull request #1286 from blattms/communicate-ebos-solution-in-run
Communicate the ebos solution at the beginning of the run method.
2017-10-09 13:13:58 +02:00
Markus Blatt
9adc4d5d3c Communicate the ebos solution at the beginning of the run method.
Currently, the ReservoirState passed to the method and converted
to the initial solution of the Ebos simulator might have wrong
values in the ghost layer (this is the case for SPE9, Norne, Model 2).
This commit uses the syncOverlap method to update these values once they
are converted to the Ebos solution. The  ReservoirState remains untouched.
This is enough to have consistent values throughout the simulation and only
introduces one additional communication step.
2017-10-05 14:33:48 +02:00
Arne Morten Kvarving
35fbe03ca7 remove unused variables
quells warnings
2017-10-05 12:02:10 +02:00
Arne Morten Kvarving
167147acc4 use std::remove_const instead of Dune::remove_const 2017-10-05 12:02:10 +02:00
Tor Harald Sandve
ebc2f46967 Use pore volume weighted averaged hydrocarbon state in rateConverted.
- pressure, rs and rv is averaged using hydrocarbon pore volume weights.
- pvtRegions is used as input in the conversion factor calculations.
- the pvt cell of the first well cell is used as the pvt index.
(Completing a well in two different PVT regions sounds like a very bad
idea anyway)
- FIP region support is added to the rate converter also for the ebos
interface.
2017-10-03 10:25:56 +02:00
Kai Bao
28a7167d61 avoiding the oscillation treatment applied to well solution
it is a design to avoid some performance degeneration for parallel
running.
2017-09-29 17:15:39 +02:00
Atgeirr Flø Rasmussen
f7529cf33e Merge pull request #1269 from GitPaean/fixing_well_efficiency
correcting the way to use efficiency factor in assembleWelEq
2017-09-25 10:26:55 +02:00
Atgeirr Flø Rasmussen
83ca6d41e3 Merge pull request #1263 from totto82/gasoil
Make 2p gas oil compile and run
2017-09-21 14:03:33 +02:00
Tor Harald Sandve
5839962e31 Some trivial cleaning 2017-09-21 12:16:46 +02:00
Atgeirr Flø Rasmussen
c1024ce2d1 Merge pull request #1268 from blattms/log-wells-debug-messages-only-on-master
Log debug messages about wells only on master.
2017-09-21 10:33:04 +02:00
Tor Harald Sandve
992ab84435 Remove support and tests for pseudo 2p model in flow ebos 2017-09-20 08:11:44 +02:00
Tor Harald Sandve
b452d16f46 Fix parallell run 2017-09-19 11:16:34 +02:00
Tor Harald Sandve
69c608829f Include RESV in the scaling factor
- solvent + RESV is not correct. Make the simulator throw until this is
sorted out.

- remove unused parameter
2017-09-18 11:28:13 +02:00
Tor Harald Sandve
352dccd5e9 Make 2p gas oil compile and runs
-- use mapping from canonicalToActiveCompIdx from Ebos
-- add guards againts non-existing components
2017-09-18 11:28:13 +02:00
Alf Birger Rustad
7b1e034a90 Merge pull request #1270 from atgeirr/fix-output-timestep
Fix report step number associated with substeps.
2017-09-17 20:25:33 +02:00
Atgeirr Flø Rasmussen
60fe6b2687 Fix report step number associated with substeps.
It appears summary files by convention expect all substep data points to
be chunked together with the data point at the end of the report step.
2017-09-15 12:04:30 +02:00
Kai Bao
fa03db07f8 correcting the way to use efficiency factor in assembleWelEq 2017-09-11 13:01:55 +02:00
Markus Blatt
a1f960b467 Reallow logging fatal errors for well convergence on all ranks 2017-09-07 14:45:00 +00:00
Markus Blatt
5d1eff5107 Log debug messages about wells only on master.
Previously all ranks logged the message and did clutter the
output files quite a bit.
2017-09-07 14:17:56 +00:00
Kai Bao
cd4d17bfa7 removing a few unused nw variable in StandardWell_impl. 2017-09-07 15:04:09 +02:00
Atgeirr Flø Rasmussen
77f3bdaeda Merge pull request #1257 from totto82/add_invert4x4
Add invert4x4 specialization for invertMatrix()
2017-09-04 09:08:32 +02:00
Tor Harald Sandve
210f1edf25 Add test for invert 4x4 matrix specialization 2017-08-29 11:01:20 +02:00
Kai Bao
b4538af164 not throwing when encountering multisegment wells. 2017-08-29 10:01:31 +02:00
Kai Bao
9accb56c86 adding a test for well model.
It is just test for StandardWell.
2017-08-25 14:09:26 +02:00
Kai Bao
355be6c26c collecting the NaN and too large well residuals
make sure all the processes will throw if there is any of the processes
found abnormal residual values.
2017-08-25 14:09:26 +02:00
Kai Bao
8abe48a693 making destructor of WellInterface virtual
to avoid memory leaking.
2017-08-25 14:09:26 +02:00
Kai Bao
a60a969257 fixing one memory-leaking for flow_ebos running 2p case. 2017-08-25 14:09:26 +02:00
Kai Bao
96340a23ff computePerfRate and crossFlowAllowed protected in StandardWell 2017-08-25 14:09:26 +02:00
Kai Bao
bc78553686 renaming well_solutions_ and well_variables
to give slightly easier understanding.
2017-08-25 14:09:26 +02:00
Kai Bao
25869026e5 renaming a few functions
to reflect the functions better. No functional change.
2017-08-25 14:09:26 +02:00
Kai Bao
9a4a9a2bcc using well name and allow_cf form well_ecl. 2017-08-25 14:09:26 +02:00
Kai Bao
a908bd7cf1 makding createWellContainer a static function. 2017-08-25 14:09:26 +02:00
Kai Bao
f9b7094075 fixing of some comments
no functional change.
2017-08-25 14:09:26 +02:00
Kai Bao
52af70ad0e adding debugging output for solveWellEq
to tell if the solveWellEq is successful.
2017-08-25 14:09:26 +02:00
Kai Bao
a29be796e7 adding comments about function calculateBhpFromThp. 2017-08-25 14:09:26 +02:00
Kai Bao
277d26df8a cleaning up well residual related in BlackoilModelEbos
TODO: how to output the information for debugging well iteration
process.
2017-08-25 14:09:26 +02:00
Kai Bao
7a9fc2132e more cleaning up of StandardWellsDense.
not functional change.
2017-08-25 14:09:26 +02:00
Kai Bao
68edfd7235 removing several unused variables. 2017-08-25 14:09:26 +02:00
Kai Bao
fe3d2f91e0 adding functions to calcuate between thp and bhp
to reduce some code duplication in StandardWell
2017-08-25 14:09:26 +02:00
Kai Bao
8a12ec677f more cleaning up
not funtional change.
2017-08-25 14:09:26 +02:00
Kai Bao
66b5226039 removing some unused variables
to reduce warnings.
2017-08-25 14:09:26 +02:00
Kai Bao
f1c0e84878 cleaning up for StandardWellsDense. 2017-08-25 14:09:26 +02:00
Kai Bao
49df0c12f6 cleaing up and fixing the error from rebasing. 2017-08-25 14:09:26 +02:00
Kai Bao
af6155d15a updating the well group status before updating targets.
This recovers the running with group control.
2017-08-25 14:09:26 +02:00
Kai Bao
31ce880ba7 fixing the prallel running.
It needs WellSwitchingLogger for all the processes even there are not
wells in some of the processes.
2017-08-25 14:09:26 +02:00
Kai Bao
5cb0a5109e adding setWellSolutions to StandardWellsDense.
the function will calculate the well_solutions_ based on well_state.
It will be used when we recover the well_state from a failed iteration,
then we should also update well_solutions_ for consistence.
2017-08-25 14:09:26 +02:00
Kai Bao
5af15fa63f removing well_soltutions_ from WellStateFullyImplicitBlackoilDense
adding function setWellSolutions() to StandardWell.

The class WellStateFullyImplicitBlackoilDense is ready to be removed
now, while the only thing can go wrong compred with the original version
is the group control, which is not tested yet.
2017-08-25 14:09:26 +02:00