compositional equilibration with COMPVD - #7373
Draft
GitPaean wants to merge 16 commits into
Draft
Conversation
GitPaean
added
the
manual:irrelevant
GitPaean
force-pushed
the
compvd_equilibration
branch
2 times, most recently
from
August 31, 2026 11:53
b1bee28 to
ba75c0e
Compare
The linear solver leaves the rows of ghost cells untouched, so their primary variables must be fetched from the owning processes before the intensive quantities are recomputed. A virtual postSolutionUpdate hook between the two steps of the base updateSolution carries the sync, so the sequence itself stays in one place.
The scatter resized the distributed data to hold every value of a multi-valued keyword such as ZMF, but left the keyword information at one value per cell, so numValuePerCell(), numCells() and compress() read the distributed field as a scalar one.
Type 1 integrates the whole column with the single EOS root chosen from the datum side of the contact. With the contact inside the region that root can be the wrong one of a three-root cubic on the far side, but the warning only covered a depth-invariant ZMFVD; a varying one was initialized in silence. A genuinely two-phase column needs item 10 = 3, which splits the integration at the contact.
Cover the supported EQUIL envelope on the verified 1D column: a continuous liquid with the composition following ZMFVD (item 10 = 1), a gas cap over a liquid leg with the contact pressure at the saturation pressure of the contact liquid (item 10 = 3), two independently equilibrated EQLNUM regions, and the error on a missing ZMFVD. The pressures are anchored to the reference-simulator values the implementation was verified against.
EQUIL_1D_ZMFVD initializes a continuous liquid column (EQUIL item 10 defaulted) and EQUIL_1D_ZMFVD_OIL a gas cap over a liquid leg with the contact pressure at the saturation pressure (item 10 = 3), covering the supported envelope of the compositional equilibration end to end.
COMPVD was refused outright, and a water-oil contact inside the region with it. Read the composition from either COMPVD or ZMFVD, take the phase COMPVD states for the EOS root of a single-phase region, and give the water its own hydrostatic column: connate saturation above the contact, full saturation below, and a pressure integrated from the contact rather than derived from the hydrocarbon one. A datum below the water-oil contact states the water pressure, not the hydrocarbon pressure, so the integration starts there and hands the hydrocarbon its pressure at the contact.
A COMPVD table that names the vapour phase for some rows and the liquid phase for others describes a gas zone over a liquid one, meeting at the gas-oil contact. Build a composition from the rows of each phase and give the gas zone its own hydrostatic column, continuous with the liquid one at the contact, instead of forcing the whole region onto the EOS root of the datum's side.
A composition-versus-depth table states the composition where it was measured; continuing its slope past the shallowest or deepest row invents a fluid. A customer deck whose vapour rows span 2550 to 2573.5 m was extrapolated up to 2454 m, where the methane fraction reached 0.796 against the 0.7495 of the shallowest row, and the density error that followed left the top of the column 0.13 bar off the reference. Clamping brings it to 0.05 bar, the same as the deck without the shallower row.
Below the water-oil contact the pore space holds water alone, but the cell pressure came from the oil pressure function, which carried the lighter hydrocarbon gradient into the water zone. On a customer deck the deepest cell came out 0.65 bar below the reference run, and the error grew with depth. The water pressure is already integrated to place the contact. Use it for the cells below the contact as well. The same deck now matches the reference to 0.00003 bar there.
Successive substitution on its own does not always converge: two of six compositional test decks abort during initialization at the hundred iteration limit, and the user is left with no result and no hint that another method would work. Newton switches back to successive substitution when it fails, so the combined method reaches everything the plain one reaches. The two decks now initialize and match their reference runs to 0.0005 bar, and the four that already ran are unchanged.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.