Scribd
Upload
41 views3 pages

Notes For The Depth - Conversion - in - Petrel - 2008 - 4661715 - 03

Uploaded by

thangdoan73
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
41 views3 pages

Notes For The Depth - Conversion - in - Petrel - 2008 - 4661715 - 03

Uploaded by

thangdoan73
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Notes for the “Depth Conversion in Petrel 2008” Lunch and Learn

Slide 5: The method of layer cake depth conversion is to treat each lithologic unit
separately and to define each unit by a different mathematical function

Slide 6: Some velocity settings explained:


• *Horizon from 3D Grid
• Well trace Md inc – the sampling interval for the TDR when calulating Vo and k
• Depth tolerance – minimum thickness of a zone (in project depth units) below
which it will not be used for calculation for Vo and k. This option ensures that
extreme values in thin parts of the zone do not affect the rest of the model.
• Clip Vo – if TDR is used then the extrapolation between points could lead to
extreme velocities away from wells. Enabling this option will truncate the surface
at the chosen level

Slide 7:
Velocity Model 1 uses the established Time/Depth relationship from wells to distribute
velocitiy trends throughout the area.
Velocity Model 2 is based on a loaded SEG-Y Velocity volume that is used to extract
velocities from.
Velocity Model 3 is based on Stacking Velocity points that are used to upscale and
distribute as a velocity property.

By selecting one of the velocity models, a wide range of objects can be domain converted
in accordance with that model.

Slide 8:
Datum=SRD - Seismic Reference Datum. In most cases SRD = MSL
Note that SRD can be different from MSL.

Slide 9:
If you have a segy cube with interval velocities you can convert it to average velocities
by using Volume attributes.

Slide 16:
• Calculate the error between the well tops times and the gridded interpreted time
surfaces,
• Determine whether the error is an offset or random?
• Correct time surfaces
It is strongly recommended to build Velocity Models with corrected TWT surface. It has
been observed that in some cases Velocity Modeling delivers wrong linear velocity laws
if the time surfaces deviate from the well top times.
Slide 18:
Sonic logs are densely sampled which allows to derive the velocity law more accurately

Slide 20:
This 3D grid is produced only for analyzing well velocity data. No interpolation of the
data is performed for this purpose, so the horizontal cell size is irrelevant.
Well tops defining the intervals should be gridded into surfaces and used to make
horizons in the analysis grid. As the only purpose of this is to separate the zones correctly
at the well locations, the interpolation method is irrelevant
Layering should be as coarse as well data permits. When checkshots interval velocities
are used, use the checkshot interval as layering thickness.

Slide 24:
Where well control is inadequate seismically derived velocities can help interpolate the
well velocity data and define velocity field.

Slide 25:
Seismic lateral (stacking velocities) are slower than the Vertical Checkshot velocities
CS velocities – closer to truth; Stacking – need to correct

Another recommended option is to use average seismic velocities as a trend when


modeling checkshot average velocity using Kriging. Or one can use interval seismic
velocities as a trend to model checkshot interval velocities and then convert the resultant
Vint property to Vave using a workflow.

Slide 28:
V orig and V final – velocities before and after well top correction

Slide 30:
Seismic Interpretation – cannot be edited in depth – i.e. time can be edited and depth will
updated automatically, but not possible to edit in depth.

Slide 35:
Assumption: at Vrms ~= Vstack. This assumption is only strictly valid for non-dipping
geology, hence the method must be used with caution, and always calibrated to well
velocity data before use in domain conversion.

Before running the equation Petrel will sort the columns of points by XY and then Z.
The Dix formula will then be applied to a column of points with the same XY before
moving onto the next column. So the demand for a surface is replaced by the demand for
the data points to occur at a regular grid interval

Vi = {(V2^2T2-V1^2T1)/(T2-T1)}^1/2, where V = RMS velocity to current sample

Slide 36:
Ordinary Kriging is recommended as it calculates the local mean rather than global mean
as in the case with Simple Kriging.

Demo description:
Cloudspin project was used to demo the Velocity Modeling using Well TDR showing the
QC of the input Checkshot data and the Velocity Model contents such as Velocity and k
Surfaces, Velocity well logs and well tops and Final depth converted surfaces.
Also the current Stacking Velocities workflow from loading the point data to Depth
Conversion using Average velocity property was demonstrated.

You might also like