GRIDSTAT defines faults in two different ways:
Faulted 3D models in GRIDSTAT can be handled in three different ways:
There are two ways to define faults by lines in GRIDSTAT: import fault locations from a file, or add the faults interactively on Basemap.
Fault lines can be imported into GRIDSTAT from file as polylines.
The format for a Fault File is:
Example: Fault File in GRIDSTAT Format
67892.3 12367 1
66721.1 11118 1
69048.4 11039 1
51834.7 9209 2
53943.6 9733 2
..............................................................
where the first column is the EASTING (X) coordinate of a point on the fault, the second column is the NORTHING (Y) coordinate, and the third column is the fault segment identification.
If no fault line file is available, faults can be entered interactively from the DataQC panel Basemap. Use the Add button and select Fault by Polyline. Click the fault polyline on basemap. Use the right mouse button or click the same location to end a fault. All points in a polyline have same FAULTid where id is different for different fault.
To remove fault points, use Remove then Fault by Polygon and click a polygon to enclose the points to remove.
To edit fault points, use Edit then Edit Fault and then Move or Insert or Delete Vertex. For easier identification of the fault points, change the PolyFaults to Line&Text. From Options then GraphPar.
Fault lines are part of the data file and will show up on basemap and cross sections graphs.
There are two ways to define faults by surfaces in GRIDSTAT: mapping a 2D grid from fault at wells as markers, or start from seismic interpretation.
The most common use of faulting in griding is to serve as separator. Control data from the other side of the fault will not be used. This also provides a means of limiting the area of the model: use a polygon as a fault line to circle out the area of interest, then remove all the data outside that area. The resulted grid will contain nulls outside that area.
When horizon grid is built within GRIDSTAT, data on the other side of a fault are not used.
If a fault block has no data, the gridding with fault approach will leave the fault block blank. To avoid this problem, map the horizon without the faults, then Sharpen the horizon with the faults.
In gridding attribute grid, data on the other side of a fault are also used, assuming the attribute is continuous across fault after restoring to pre-fault position. This assumption may not be correct for such attributes as saturation. To separate across fault in attribute gridding, from the Tools pull down menu go to Preference and turn on Fault Separate AttrGrid.
If a horizon has been mapped, discontinuity can be calculated from the grid (Calcul. then Discontinuity on the left of the GridOutput panel). A Search Radius and a Fault Zone (for excluding cells too close to the center) are used to estimate the directional discontinuity on the surfaces. Faults would show up as line features and the throw may be approximated by the discontinuity.
A faster and simpler way of mapping discontinuity is to calculate the first derivative, also from the Calcul. button.
If a fault has been identified but the horizon surface is not properly structured near the fault, as is often the case with software limitations or seismic interference near faults, a Sharpen Fault function is available (Calcul. then Sharpen Fault on the left of the GridOutput panel). This function also requires a Search Radius and a Fault Zone (for excluding cells too close to the fault) to extrapolate the surface to the fault.
The fault picks from seismic may be simplified for faster calculation and easier export for reservoir simulation. If there are hundreds of fault points, the plotting of contours on cross sections and 3D view will be slow.
One way to simplify the faults is to import the fault locations as XYZ Data. Add faults on basemap to follow the faults points. Remove By Name to remove the data. Save the faults. Then add the saved fault file to the desired project.
To map slopping faults, there are two different approaches in data handling. One is to load the fault cut points at wells as markers, and then to map the fault surfaces as horizons. Select the active marker as the fault to be mapped, define the grid dimension in the GridSize panel, then click the Map button and select Horizon Depth.
The other approach is to import the fault data as XYZ data with X, Y, and elevation as Data. Gridding of fault surface is similar to horizon gridding. There are two key points in fault surface gridding: turning on Areal Trend option from Options pulldown menu Grid Select, and turning off Match Data Distribution in GridSize. These options are automatically taken care off when mapping as horizon from markers.
When fault data can not completely describe the faults and horizons with fault cut are available, additional fault data can be derived from the horizon data. If horizon grids are not generated yet, use fault polygons as vertical faults and map the horizons first. For each horizon, graph Top View and show each cell. From the Edit pull down menu and select Polygon Boundary. Pick the polygon to enclose the grid cells around the fault to be mapped. Use Calcul. Func. Transform then A+B*X with A=-99999 and B=0 to remove cells outside the polygon. Use Range Edit Range then Sample Data to convert the remaining cells to data. This data can then be used for fault mapping when combined with data from other horizons and fault data. Because a horizon may have multiple fault cuts, this extraction must be done for each fault separately.