The regional velocity model VELMOD-4 is the successor of VELMOD-3.1 and is built in a Joint Industry project in the time period 2017-2019 (partners: Total and NAM). For the first time seismic stacking velocities have been used as input to the VELMOD velocity model in an effort to better capture the lateral velocity variation in between wells. The model integrates velocities measured in boreholes (sonic logs and check shot data) with the stacking velocities derived from seismic data.
The primary application of the VELMOD-4 model is time-depth conversion for large scale (regional) seismic interpretation and mapping.
The project was a cooperation between TNO and Estimages in which TNO provided the data and geological knowledge for the Dutch subsurface; Estimages has built the velocity model using their geostatistical expertise in regional velocity model building (https://estimages.com/).
A total of 179 3D-velocity datasets and 136 2D-velocity datasets were used for the project. Most of these data were available from the TNO DINO database, but several 2D-velocity datasets have been digitised by TNO in 2017. These datasets have been QC-ed resulting in discarding some sets. Spatial filtering of the individual datasets was followed by merging of seismic velocity datasets and a subsequent calibration by well velocities.
There are two end products available:
- VELMOD-4a: Velocity cubes and uncertainty cube (merged velocity cube, calibrated velocity cube and velocity uncertainty).
- VELMOD-4b: Vint and V0 grids for the main stratigraphic units as defined in DGM-deep.
The workflow for these products is described in the report and accompanying PowerPoint presentation. The coordinate system used for the products is ED 1950 UTM Zone 31N
Well control in VELMOD-4a is poorer than in VELMOD-4b since the number of wells with reliable velocity information along a significant depth interval to calibrate the velocity cube is more limited than the number of input wells for individual stratigraphic units in VELMOD-4b.
VELMOD-4b has been tested in a time-depth conversion on a set of blind wells and compared with VELMOD-3.1 resulting in smaller residuals on average in the blind well set.
When using the VELMOD-4 velocity models, it is recommended to also test the VELMOD-3.1 velocity model and take notice of the differences in the end result. An updated version of VELMOD-3 is expected to be released in the course of 2022 (VELMOD-3.2).