release/2025.04/final

Release Highlights

• Support for well on/off cycling (WCYCLE keyword)
• Fix for initial in-place fluid volumes reported to the .PRT file, both in base runs and in restarted runs
• Initial support for reservoir coupling
• Initial support for calculating and reporting per-cell quantities in compositional runs
• Support mnemonic value WELLS=2 in the RPTSCHED keyword
• Performance improvements in property evaluation
• Initial support for deferring property evaluation dedicated hardware accelerators/GPUs
• Support for the AmgX library of linear solvers
• Support version 11 of the libfmt library
• New and experimental consistency checks for saturation functions, especially for end-point scaling. Off by default.

What's Changed

• Enable New Saturation Function Consistency Checks by @bska in #5596
• FIX gas-water restart by @totto82 in #5668
• In partitioning, only account for wells that will be active at some point in time by @vkip in #5609
• in flowexp_blackoil, invalidate output data after solution is updated by @GitPaean in #5664
• Separate non-DUNE methods in LevelCartesianIndexMapper by @aritorto in #5649
• Fix issue with stopped injectors in thermal runs by @totto82 in #5605
• Bugfixes #5609 by @vkip in #5686
• enableThresholdPressure_ initialized to be false by @GitPaean in #5691
• Add Logging for Python Actions by @bska in #5699
• ecl style output for compositional simulation by @GitPaean in #5687
• fixed: set active indices in fieldprops by @akva2 in #5701
• Print gpus used on simulator startup by @multitalentloes in #5611
• fixed: division by zero if no gpu device was found by @akva2 in #5705
• Update flow manpage with help2man. by @blattms in #5708
• fixed: test-suites needs a name by @akva2 in #5710
• dont switch from GRUP if check_group_constrains_well_iterations=false by @totto82 in #5695
• Fix gaslift 1 by @totto82 in #5694
• fixed: building pypi release packages by @akva2 in #5719
• Add LINRAD (hybrid linear-radial) filter cake model by @vkip in #5717
• Added the missing plus in gplv3+ for the Python packages. by @blattms in #5722
• Removed GPL copyright header from licensed under Boost license. by @blattms in #5721
• Fix index bug in gpubridge separate wells by @razvnane in #5720
• Correctly mark data files with ODBL and DBCL license by @blattms in #5723
• Gpu support co2gaspvt by @multitalentloes in #5636
• fixed: RESTART_SCHED parameter never took effect by @akva2 in #5725
• fixed: fmt::join is in ranges.h using {{fmt11}} by @akva2 in #5726
• Avoid potential segfault due to invalid vector.end() after erase() by @vkip in #5740
• Activate New Regression Tests for ACTIONX Name Matching by @bska in #5750
• Do not try to compute initial solution for inactive multi-segment wells split across processors by @vkip in #5751
• Make sure all errors from parsing show up in the log files. by @blattms in #5707
• Feature/ms wells - part 1: Initial assembly of B C D and the residual by @lisajulia in #5680
• Cmake: Use upper case for COMPONENTS of FindSuiteSparse by @blattms in #5745
• dont shut well with GRUP control that struggles to converge by @totto82 in #5735
• Well testing using gas lift facilities by @svenn-t in #5757
• Subtract reduction rate for current group when checking higher group constraints by @steink in #5729
• Only update target of group-controlled wells during updateAndCommunicate by @steink in #5748
• Support GCONSUMP on multiple levels in the group hierarchy by @vkip in #5739
• Use formattedErrors() and simplify. by @atgeirr in #5765
• Update second stage preconditioner for CPR instead of recreate by @jakobtorben in #5758
• Restore threshold pressures from restart by @vkip in #5743
• Network sub iter by @totto82 in #5767
• added: ability to pass command line parameters from python by @akva2 in #5760
• Changes in how the simulator handles NUPCOL by @totto82 in #5724
• Write correct THRESHPR values also from restart by @vkip in #5770
• fixed: well model could deadlock by @akva2 in #5773
• bump minimum dune version to 2.9 by @akva2 in #5500
• changed: do not use boost.test in test_tasklets_failure by @akva2 in #5782
• Zoltan partitioner with all perforated cells of a well merged to one vertex by @michal-toth in #5749
• fixed: always calculate initial FIP by @akva2 in #5784
• Expose finesmoother pre/post smooth parameters in CPR setup by @jakobtorben in #5786
• Feature/ms wells - part 2: Solving, straightforward option by @lisajulia in #5746
• Feature/ms wells test by @lisajulia in #5747
• Alternative way of avoiding group control oscillations by @totto82 in #5753
• Ensure tracer flow for cross-flowing injector connections is included by @vkip in #5789
• Allow for the third group switch by @totto82 in #5791
• remove 16 days restriction for timestep in predicition mode by @totto82 in #2970
• Use regular OpenMP for loop, then iterate over chunks inside. by @atgeirr in #5777
• Add Hypre BoomerAMG Support by @jakobtorben in #5762
• Make OpenCL optional by @kjetilly in #5793
• Feature/ms wells fix restart file error by @lisajulia in #5802
• added: support WCYCLE by @akva2 in #5792
• Add test for distributed MSWells with two MSWells by @lisajulia in #5805
• Amgx integration by @jakobtorben in #5808
• Do not write initial FIP for CpGrid with LGRs for now by @aritorto in #5813
• Use initial pressure if ROCKOPTS item 2 is STORE by @svenn-t in #5785
• Allow query for efficiency scaling factor in off-process wells where required by @akva2 in #5816
• LevelCartesianIndexMapper constructor for AluGrid taking only a grid as an argument by @aritorto in #5715
• Allow group guide rates in case that a group is an auto choke group by @plgbrts in #5754
• Remove the mpi calls from the business logic by @lisajulia in #5819
• Move implementation of WellOperators out of BlackoilWellModel by @akva2 in #5823
• use reservoir rates not reduction rate in target calculator for VREP by @totto82 in #5783
• adjust output and parameters in gpu autotuner by @multitalentloes in #5779
• split AdaptiveTimeStepping to use an impl file by @akva2 in #5834
• changed: use standard header guard structure for _impl.hpp clangd helpers by @akva2 in #5835
• [Nit] Add EHYSTR(HYSTMOD) = 4 to list of supported values in message by @gdfldm in #5837
• Reduce size of enums to save 15 bytes per cell. by @atgeirr in #5841
• changed: move GasLift code in BlackoilWellModel to separate class by @akva2 in #5832
• add test for gpu brineco2pvt by @multitalentloes in #5839
• Fix clang complaints about incorrect constexpr usage. by @atgeirr in #5844
• Add test for inserting keyword WTMULT from a Pyaction by @lisajulia in #5830
• Add test for WLIST keyword in Pyaction and ACTIONX by @lisajulia in #5836
• Add gpudilu mixed precision by @multitalentloes in #5674
• Fix errors in documentation. by @lisajulia in #5853
• Use new linearizer for blackoil + thermal by @totto82 in #5843
• Fix RFT output for shut distributed wells by @vkip in #5824
• Add parallel versions of the PyAction tests by @lisajulia in #5860
• Add mixed precision double diagonal gpu ilu0 variants by @multitalentloes in #5688
• gas water for multisegment well by @GitPaean in #5818
• Switch default load balancer to Zoltan with vertex = all cells of a well by @blattms in #5800
• regression test for gas water and multisegment wells by @GitPaean in #5868
• BlackoilModel: move to has-a instead of is-a modelling of penalty cards by @akva2 in #5871
• Add tests for GEFAC for ACTIONX and PYACTON by @lisajulia in #5846
• Fixes related to zero guiderates from NETV by @totto82 in #5822
• Fix WCYCLE for multiple blocks by @akva2 in #5877
• Skip allocTrans when !(enableEclOutput || loadBalancerSet == 0 || partitionJacobiBlocks) by @daavid00 in #5817
• Transmissibility: add a cache for centroids and use this in update() by @akva2 in #5888
• Prevent spawning more threads if we limit OMP thread count by @michal-toth in #5856
• Feature/py action insert kw welpi wpimult alternative by @lisajulia in #5892
• added: option to only compare summary files in regression tests by @akva2 in #5897
• Compositional simulations with more cubic EOS formulations by @svenn-t in #5886
• Reservoir coupling: Implement time stepping by @hakonhagland in #5620
• Skip empty positional arguments by @vkip in #5679
• Bugfix: Upon restart, do not try to restore segment info for inactive wells by @vkip in #5861
• Implement gsatprod by @totto82 in #5894
• Add run SPE1CASE1_CARFIN.DATA by @aritorto in #5903
• fixed: use a vector in mpibuffer by @akva2 in #5907
• added more timing by @hnil in #5909
• Add multi-threading to transmissibility calculation by @akva2 in #5890
• Dont reset subtimestep from tuning/wcycle if time step is chopped due to convergence issues by @totto82 in #5902
• Allow empty leaf nodes that do not exist in GRUPTREE by @vkip in #5919
• Switch property evaluation to using ElementChunks. by @atgeirr in #5912
• Parameterize network sub-iterations and pressure update dampening by @vkip in #5918
• Improve well shutting logic by @totto82 in #5900
• hacking to register the parameter ForceDisableFluidInPlaceOutput by @GitPaean in #5926
• Expose CornerCells and OverlapLayers Load Balancing Settings by @bska in #5942
• Smart pointers for GPU memory by @kjetilly in #5946
• Added utility class to hold a GPU resource (stream, event, graph, etc). by @kjetilly in #5951
• Limit relative roughness in Haaland friction factor calculation by @vkip in #5945
• Added copy functions from GPU smart pointers by @kjetilly in #5956
• Separate keyword validation for fully supported keywords by @vkip in #5866
• Add Infrastructure for Geomechanical Analysis by @bska in #5801
• Fixes for NVCC older than 12.5.1 by @kjetilly in #5958
• Add VoigtArray and SymmTensor classes by @akva2 in #5959
• Add MechContainer by @akva2 in #5960
• Fixed: Data output should use move semantics by @akva2 in #5965
• Fix shared library linking issues by @akva2 in #5963
• clang-format now adds newline at the end of file. by @kjetilly in #5970
• Using the CriticalException in PreconditionerFactory and ISTLSolver by @kjetilly in #5969
• component indices are now int by @akva2 in #5950
• Improve gpuistl using cudaGraphs by @multitalentloes in #5852
• Add CompositionalContainer, a container for compositional data output by @akva2 in #5978
• Compositional simulator expanded to three phases by @svenn-t in #5851
• Consistent primary variables by @hnil in #5952
• Add TracerContainer, a container for tracer data output by @akva2 in #5987
• Simplify GenericOutputBlackoilModule::doAllocBuffers by @akva2 in #5989
• Support instantaneous flow rates in extended network (WEFAC and GEFAC item 3) by @vkip in #5410
• General third order time step controller by @erikhide in #5974
• Add RFTContainer, a container for RFT data output by @akva2 in #5996
• Add a container for FLOWS data output by @akva2 in #5954
• Fix to write residuals by @daavid00 in #6002
• FlowProblem(Blackoil|Comp): Return a reference to EclWriter by @akva2 in #6001
• Add RESIDUAL output to a regression test by @akva2 in #6003
• Allow empty non-group leaf nodes by @vkip in #6010
• Report Connection Level Fracturing Statistics to I/O Layer by @bska in #5811
• Fix Summary BFLOW* by @daavid00 in #6013
• Extract addLgrs, invoke it before loadBalance by @aritorto in #5981
• fix tracy support by @akva2 in #6024
• Add missing communication for parallel runs with MSWells by @lisajulia in #5937
• Added tests for error macros on GPU. by @kjetilly in #5962
• Adapt to ElementChunks now requiring partition argument. by @atgeirr in #6027
• add some profiling blocks to tracer by @akva2 in #6035
• Ensure MSUMLINS and MSUMNEWT is updated. by @atgeirr in #6039
• TracerModel: avoid setting up ILU if no solves will be performed by @akva2 in #6045
• Only check trivial_target for GRUP when checking in THP by @totto82 in #6018
• Make pre-solve network optional by @totto82 in #6048
• Tracer: add some structs for well rate output and avoid complex keys by @akva2 in #6044
• make nonlinear system possible to change by @hnil in #6046
• Fix parallel tracers by @akva2 in #6052
• Tracer: enumerate components by @akva2 in #6058
• Fix to write vtk (if enabled) for dry runs in cases without oilPhase by @daavid00 in #6057
• Improving terminal/dbg output by @daavid00 in #6059
• Fix gpmaint for region 0 by @totto82 in #5916
• modify stagnation and oscillation code for MSW by @totto82 in #6020
• Make Zoltan's Hyperedge Size Threshold Runtime Controllable by @bska in #6036
• Fix to write vtk (if enabled) for dry runs in cases with drsdt by @daavid00 in #6060
• Add a first test case for reservoir coupling by @hakonhagland in #5893
• Prevent some nan creations during output generation. by @blattms in #6077
• Prevent zero division by @michal-toth in #6053
• Set SOIL to 0 when readExplicitInitialCondition and 1-SGAS-SWAT=0 by @daavid00 in #6079
• WellModel: make a copy of events at start of time step by @akva2 in #6080
• fixed: clear out events before processing actions by @akva2 in #6078
• PERMFACT for each saturation region by @daavid00 in #6074
• Fixes for NVCC by @kjetilly in #6054
• Fix network iterations by @totto82 in #6062
• Tracer: Code deduplication by @akva2 in #6083
• Added set -e to hipify_file.sh by @kjetilly in #6088
• Add GSATPROD to regression test by @totto82 in #6089
• Add ACTIONX_GCONINJE to regression test suite. by @blattms in #6087
• PRT: Use {fmt} to format injection log by @akva2 in #6090
• PRT: Use {fmt} to format production log by @akva2 in #6093
• PRT: Use {fmt} to format cumulative log by @akva2 in #6094
• Bug fix in linear solver by @daavid00 in #6091
• PRT: Use {fmt} to format fip logs by @akva2 in #6096
• Apply simulator update once by @vkip in #6070
• Added WELL as supported to GCONPROD message for unsupported options by @blattms in #6101
• Add Regression Tests for GCONPROD(7)=WELL by @bska in #6106
• removing undefined and unused function computeWellGroupTarget() from BlackoilWellModel by @GitPaean in #6102
• Support for RTPSCHED WELLS=2 by @akva2 in #6099
• Fix a recent bug indexing in the CPRW cell weights. by @atgeirr in #6109
• Support WCONHIST/WCONINJ in ACTIONX by @blattms in #6092
• Add support for MSW PRT report by @akva2 in #6110
• Added regression tests for WCONINJH and WCONHIST in ACTIONX. by @blattms in #6112
• Make wellModelMatrixAdapter serial only by @jakobtorben in #6113
• Prepare OPM for DUNE 2.10 by @blattms in #5585
• Add support for merging well cells for NLDD partitioning by @jakobtorben in #6103
• Support for RSW in the SOLUTION section by @daavid00 in #6111
• Set thp from network in well testing by @totto82 in #6116
• Calculate original FIP by reading restart data for step 0 by @akva2 in #6119
• Compositional well by @GitPaean in #6042
• Add network+wtest test by @totto82 in #6130
• Use well rates from well_state_nupcol instead of previous_well_state in updateWellRates by @totto82 in #6128
• Actually respect NUPCOL (not NUPCOL+1) and allow network to converge with GLO by @vkip in #6100
• BUGIFIX: Only respect item 7 in GCONPROD for individual constrains, by @totto82 in #6135
• using std::optional for function initialInplace() by @GitPaean in #6137
• For convective mixing, store the saturated dissolution factor. by @atgeirr in #5774
• Fix and improve nldd timing and output reporting by @jakobtorben in #6124
• Fix compositional field output names by @akva2 in #6143
• Trigger WELSPECS Report From Simulator Side by @bska in #6107
• Updating MICP by @daavid00 in #6105
• Pre-solve network only once by @vkip in #6138
• Dont resolve well equation when computing initial guess for thp in updateWellStateWithTarget by @totto82 in #6081
• Silence various warnings. by @atgeirr in #6148
• Changes in SOURCE structure by @svenn-t in #6140
• Extending the SUMMARY keywords for CO2STORE and H2STORE by @daavid00 in #6127
• Backport of pr 6104 by @multitalentloes in #6159
• Backport of PR 6153 by @multitalentloes in #6161
• Backport of PR 6156 by @multitalentloes in #6162
• Backport of pr 6157 by @multitalentloes in #6163
• Backport of pr 6164 by @multitalentloes in #6165
• Backport of pr 6151 by @multitalentloes in #6166
• backport of pr 6194 by @multitalentloes in #6201
• Backport of pr 6184 by @multitalentloes in #6202
• Backport of pr 6182 by @multitalentloes in #6203
• backport of pr 6198 by @multitalentloes in #6204
• Backport of pr 6210 by @multitalentloes in #6239

New Contributors

• @michal-toth made their first contribution in #5749

Full Changelog: release/2024.10/final...release/2025.04/final