Nested Hybrid Grids

🧪 Experimental Feature

This module is currently experimental. It may undergo breaking changes in future versions without notice.

Overview

The nestedhybridgrid module creates nested hybrid grids where a selected region of a coarse grid is replaced by a refined (subdivided) sub-grid. The two grids are merged into a single grid and connected through Non-Neighbour Connections (NNCs).

The typical workflow is:

1. Define a coarse grid and a region property that marks cells to refine with value 1.

2. Use the NestedHybridGrid to produce the merged grid and an NNC table. You may need to export the NNC file to csv at this stage.

3. Do rescaling from the original gridmodel (e.g. a finer geogrid) to the merged grid, e.g. by using software like RMS.

4. Make another script, that computes transmissibilities with xtgeo.Grid.get_transmissibilities() passing the NNC table.

5. Export the NNC transmissibilities for the flow simulator.

Quick-start example

The example here runs within RMS, but similar workflows can be created for file i/o.

from fmu.tools.nestedhybridgrid import NestedHybridGrid

# Create nested hybrid grid (refine region 1 by 2×2×1)
nhg = NestedHybridGrid.from_rms(
    project,
    grid_name="Simgrid",
    region_name="Refinement_region",
    refinement=(2, 2, 1),
    properties=["Zone"],  # Optional list of properties to transfer to the output grid
)

# store nested grid with properties in RMS
nhg.to_rms(project, "NestedHybrid")

# write the NNC table to disk; this will be applied for computing NNC's in the next script
nhg.write_nnc_table("path_to_some_csv_file.csv")

The next step is to do a rescaling from the original geogrid to the merged grid using e.g. the RMS tool.

Further, we need to create NNC transmissibilities and generate file for flow simulator:

import pandas as pd
import xtgeo
from fmu.tools.nestedhybridgrid import (
    nnc_to_flowsimulator_input,
    nnc_to_gridproperty,
)

GNAME = "NestedHybrid"

# Load grid and region property which may be stored in RMS
nested = xtgeo.grid_from_roxar(project, GNAME)

# load the NNC table
nnc_table = pd.read_csv("path_to_some_nnc_file.csv")


# Load rescaled property input for transmissibilities and compute
permx = xtgeo.gridproperty_from_roxar(project, GNAME,"PERMX")
permy = xtgeo.gridproperty_from_roxar(project, GNAME,"PERMY")
permz = xtgeo.gridproperty_from_roxar(project, GNAME,"PERMZ")
ntg   = xtgeo.gridproperty_from_roxar(project, GNAME,"NTG")  # defaults to 1 if no NTG

# compute transmissibilities. Note that flow simulators do this for the normal cells/faults
# so strictly speaking, only nnc_hybrid is needed here.
tranx, trany, tranz, nnc_fault, nnc_hybrid, rbnd = nested.get_transmissibilities(
    permx, permy, permz, ntg, nnc_table=nnc_table
)

# Export NNC keyword for Eclipse / OPM Flow
nnc_to_flowsimulator_input(nnc_hybrid, "some_path/NNC_HYBRID.INC")

# Or map NNCs onto grid properties for visualisation
tx_nnc, ty_nnc, tz_nnc = nnc_to_gridproperty(nested, nnc_hybrid)
tx_nnc.to_roxar(project, GNAME, "TRANX_NNC_QC")  # etc

Concepts

NNC table

The NNC table captures which coarse (mother) cells connect to which refined cells — information that xtgeo needs to compute transmissibilities across the refinement boundary. It is accessed via the property nnc_table on the NestedHybridGrid instance and is of type DataFrame with columns:

Column	Description
I1, J1, K1	Mother cell indices (1-based)
I2, J2, K2	Refined cell indices (1-based)
DIRECTION	Face direction from the mother cell toward the refined cell (I+, I-, J+, J-, K+, K-)

This table is passed to xtgeo.Grid.get_transmissibilities() via the nnc_table parameter. The transmissibility computation uses geometric face-overlap calculations (Sutherland–Hodgman algorithm) and two-point flux approximation (TPFA).

Eclipse / OPM Flow export

nnc_to_flowsimulator_input() writes the NNC keyword in Eclipse format. The output file can be included in the simulator deck:

INCLUDE
  'NNC_HYBRID.INC' /