Velocity Calibration

Description Corrects a velocity model so that well markers tie their corresponding layer-model horizons.
Module(s) QI

Corrects avelocitymodel so that well markers tie their corresponding layer-model horizons. The depth errors (misties between well markers and layer-model horizons) measured by aDepth Errorsprocess are interpolated across thevelocitymodel extents using universal kriging, producing a correctedvelocityvolume.

Create a Velocity Calibration process

  1. In theControl Panel, open theProcesstab.
  2. Click the blue+icon, and selectNew Process.
  3. Scroll down and selectVelocityCalibrationAdd.
  4. Type a name for the process, and clickOK.

Configuring Velocity Calibration

  1. Input depth errors: Select theDepth Errorsprocess that supplies the misties. The wells, layer model, and inputvelocityvolume are all taken from this process.
  2. Datum Horizon: Select the water bottom, topography, or other reference horizon below whichvelocitycalibrationbegins. No correction is applied above this horizon; the correction ramps smoothly from zero at the horizon down to the full kriged value over theRamp length. If no horizon is supplied, the datum is assumed to be at time zero. Note that any layer-model horizons that fall within the ramp window will not be fully tied in depth.
  3. Ramp length (ms): Set a ramp length value that represents the two-way-time distance from the Datum Horizon over which the depth correction ramps from zero to its full value. For offshore data, the datum horizon is typically the water bottom. If the water column in the inputvelocitymodel is already reliable, use a ramp to keep it unchanged while allowing the correction to build gradually below the datum horizon. When a constant trend term is enabled, a ramp is required to avoid non-physical velocities near the datum horizon.
  4. Horizontal kriging range: Specify a lateral search range used by the kriging (in survey distance units). This controls how far horizontally a well’s misties influence the output.
  5. Vertical kriging range (ms): Specify a vertical search range used by the kriging in two-way-time milliseconds.
  6. Vertical kriging taper (ms): Set a vertical kriging taper value that represents the two-way-time distance below the deepest tie point in each well over which the correction tapers back to the background trend. This prevents the kriged correction from oscillating in untied regions below the wells.

Trend terms

One or more trend terms form the background (deterministic) component of the kriging. Adding appropriate trend terms improves extrapolation away from wells.

  • Constant trend: Select to include a spatially constant offset as a trend term. Recommended in most cases.
  • Depth-based trend: Select the appropriate option to add a term that varies with two-way time (or depth), capturing systematic depth-dependentvelocityerrors. The following are available options:
    • None
    • Depth
    • Datum depth
    • Depth below datum.
  • Geobody horizons: Select to include pairs of horizons that define geobody regions. A separate drift volume is constructed for each geobody and included as a trend term, allowing layer-specificvelocitycorrections.
  • Drift volumes: Select to include externalvelocity-domain volumes used as additional trend terms. Useful when a prior model (e.g. from tomography) can guide the spatial variation of the depth correction.
  • Geobody taper length (ms): Specify a length value that represents the taper applied at the top and base of each geobody drift volume, to avoid sharp discontinuities at the horizon boundaries.

Outputs

  • Open table viewer: Click to open a dynamic table of the residual misties after the trend correction is applied — i.e. the misties evaluated against the trend-correctedvelocity. These are the values that the kriging then corrects. Table views can be used to display bubble plots of these residuals (seeCreate mistie bubble plots).
  • Open theVolumetab to view the following outputs:
    • Mistie trend: The trend component of the kriged depth error (in metres). This represents the large-scale, spatially smooth depth error predicted by the selected trend terms.
    • Mistie kriged: The full kriged depth error (trend plus residual, in metres). Both mistie volumes describe the depth errors of theinputvelocitymodel, not the output.
    • Velocitytrend: The input intervalvelocityvolume corrected using only the depth-error trend. Useful for understanding the large-scalevelocityadjustment.
    • Velocitykriged: The input intervalvelocityvolume corrected using the full kriged depth error. This is the primary output: avelocitymodel whose depth-converted horizons match the supplied well markers.