Introduction
This documentation goes through the working steps and menus in Geocap for depth conversion of time surfaces applying a time-velocity cube. There are other methods for depth conversion than those described here, but this method is favourable when stacking velocities and well data are present. This depth conversion documentation deals with quality control and analysis of input data. Then gridding into a time-velocity cube and finally performing depth conversion of time grids in a Layer Model into depth grids.
The velocity cube reflects the information in the velocity data in an organized way. Working with velocity cubes will make depth conversion of time data to depth data more easy.
On this page:
Steps in depth conversion
Steps in depth conversion of time data using time-velocity cube
- Reading stacking velocities into a folder as: x y time velocities
- Inspecting the velocity data
- Gridding into a velocity cube
- Displaying the velocity cube
- Checkshot wells can be used to trim the velocity cube through sophisticated updating techniques
- The velocity cube is then used for depth conversion of time grids into depth grids
- Calculation of cubes for RMS and Interval velocities using the generated velocity cube. (Option)
The folders used to save data in this project are of type Generic and renamed to indicated their contents. The various datasets will also be set to their proper schema type so that only relevant command objects are shown.
The documentation follows a specific dataset and is thus close to a case study. This implies that other dataset may apply slightly different display options due to size and attribute saving in scalar or field data. The general principles for depth conversion should however be the same in all cases when using the method of a velocity cube.
Example of velocity cube holding stacking velocity data
Reading stacking velocities
Geocap provides import command objects (CO) for reading stacking velocities organized in ASCII column data.
If the file is large, keep the input as low as possible. To generate a velocity cube only x y time velocity are needed.
Example input: stacking velocities in ASCII columns, size ~ 400MB.
Reading stacking velocities into a project
- Create a Generic folder, give it a proper name
- Right click on the folder and select *Import->Stacking Velocities.
- Browse in all stacking velocity datasets
- The cell structure is kept by specifying 'Value change' and set up 'cell separator' for the actual column
- The schema type is set to 'Stacking velocities' to get the correct commands visible on the dataset
- In this case the time values are in the z coordinates and the velocity values are scalar values
Geocap project for stacking velocity data and velocity cubes
The actual project was generated the standard way by opening a new project and gradually populate it with folders when needed. The stacking velocities was imported into a generic folder given the name All points in line cells, just to indicate that the stacking velocities are organized as vertical lines, called line cells in Geocap syntax. The schema setting of the dataset is Stacking Velocities.
Inspecting the velocity data
The stacking velocity file has nearly 7 million points and there are dedicated COs to make a fast display if necessary like Utilities->Map vertical cells in fast mode. This example applied Map Data and Display survey lines (will only display the top points as lines) to get a visual impression of the different surveys in this dataset.
A closer look at the display shows that it is a collection of many different surveys. The transition between the different surveys may show inconsistencies in velocity data. It is possible and advisable in some cases to adjust surveys to get a better match by standard data manipulation. Gridding stacking velocities into a cube will also smooth out errors and tends to make transitions acceptable.
To display a smaller part of the surveys one can use the CO under Utilities->Display stacking velocities in line mode. Set always the correct attribute (scalar or velocity if field data) and select for example Eliminate outside cursor: A small part of the dataset around the cursor is selected for display and also saved in workspace as elioutcursor. Similar for an area zoomed into when in 2d mode and part of dataset can be saved as workdata in workspace.
Gridding a cube from stacking velocities
The stacking velocities are gridded into a time-velocity cube that will be used for depth conversion. Read more about this theme in Gridding of time-velocity cube.
Depth conversion using a dedicated command object
The velocity cube is useful for depth conversion of a set of time grids into depth grids. This formula is used: depth_grid = time_grid x velocity_of_time_grid.
The size of the cube in z direction should be at least a little bit greater than the extent of the top and bottom surface. Also the x y extent should completely include the grid window. Then all the surfaces will fit into the velocity cube.
Requirement
All the time grids should be located in a folder which have the schema Layer Model.
Active the menu command by Layer Model->Depth conversion of time grids in folder using velocity cube.
Read the description on the depth conversion panel thoroughly. Here is a summary.
Fast and convenient method for depth converting all time grids in a folder using a velocity cube
- Set schema of the folder with the time grids to Layer model
- Copy the entire time grid folder to a new folder
- Rename the copied folder to Depth grids. It is still time grids, but will soon be converted to depth grids
- On the folder with depth grids, activate the commands *Layer utility->Depth conversion of time grids in folder using velocity cube
- Browse in the velocity cube and hit Execute
- As a result all grids in the folder that were time grids are now converted to depth grid
The central command in depth conversion is the pro command which does the probing of the time grid into the velocity cube. The time grid must be converted to polydata in order have a scalar part: x y time scalar. After probing this dataset into the velocity cube, the scalars are assigned values from the cube at their positions and the whole dataset is now transferred into x y time velocity.
The rest of the procedure is to work out the formula: depth_grid = time_grid x velocity_of_time_grid and saving the converted dataset back into its place.
If by some reason one wants to redo this operation, one has to bring a new fresh set of time grids into a folder and perform the procedure over again.
It is also possible to convert just a single grid from time to depth using the velocity cube. Apply the commands for grids Grid operations->Depth conversion using velocity cube.
Example of depth conversion from time grid to depth grid
The above picture shows an example of depth conversion a time grid into depth grid using a velocity cube and the same procedure as described above. All grids in the layer model were depth converted simultaneously.
Creating a RMS velocity cube and interval velocity cube
Having a stacking velocity cube, Geocap can do calculation on the cube and convert it into other cube types. Activate the commands Utilities->Perform calculation on the velocity cube.
The RMS velocity cube is calculated according to the formula:
VRMSn = sqrt((Sum_i(Vi^2*Ti)/Sum_i(Ti)))
The Interval Velocity Cube is calculated according to Dix formula:
Vn = sqrt((VRMSn^2*Tn - VRMSn-1^2*Tn-1)/(Tn-Tn-1))
These calculated cubes will help the professional to understand the geology of the subsurface based on velocity changes. The below picture displays the RMS and Interval velocity cube in the lower viewports.
Four velocity cubes displayed in a viewport presentation
Upper left Raw velocity cube. Upper right Smoothed grid updated cube. Lower left Root mean square (RMS) cube. Lower right Interval velocity cube.
The above viewport presentation applies the viewport icon on the toolbar to make 2x2 connected viewports with different background color. Then each viewport is selected for displaying its appropriate cube. A cursor point is set within the cube using the keyboard letter p. Then the command Cursor view>Map X Y Z pales at cursor position* is executed for each cube. The shell command tx2 lle col bla txt "text" was used for displaying a bottom text.