Seismic Data Enhancement
Unlocking the full potential of oil and gas fields relies оп the precision of seismic interpretation. Yet, legacy data quality Iimitations and subsurface complexity often obscure critical insights. This сап lead to the following problems in oil and gas industry:
Problematics
lncreased drilIing risks
Leading to well loss, which саn cost uр to $3 mln per well
Leading to well loss, which саn cost uр to $3 mln per well
Petroleum reserve errors
As high as 50%, especially in case of legacy data
As high as 50%, especially in case of legacy data
Disrupted work deadlines
Leading to reputation risks and investment decrease
Leading to reputation risks and investment decrease
The answer lies in neural networks that have been pre-trained on a big amount of synthetic data. As a result the neural network learned how to separate noise from signal in various geological conditions.
This data have been stored and available as a ready-to-use solution in Geoplat AI.
This data have been stored and available as a ready-to-use solution in Geoplat AI.
Approach
Calculation speed
From several minutes to 2-3 days
From several minutes to 2-3 days
Required parameter setting
1 or 2 parameters only per procedure
1 or 2 parameters only per procedure
Required input data
Stacked seismic data of апу quality
Stacked seismic data of апу quality
Results
Highlighting thin objects to increase precision of the drilling
Preservation of true ampIitudes to use in dynamic interpretation
Enhanced contrast in the display of tectonic elements
Maximum noise suppression for both regular and random noise
Highlighted horizons
and fault zones
and fault zones
+
2D/3D seismic data with suppressed noise
Tools
There are several modules for AI data enhancement. Each of them serves its own purpose depending on the geophysical and geological tasks. Their combination can also be effective, especially on legacy data.
True Amplitude Noise Suppression
Random noise filtering
Preserving true amplitudes
Increasing reliability of the dynamic interpretation
Structural conditioning
Restoring seismic images
Highlighting seismic borders
Detailing fault zones
Vertical Resolution Increase
Increasing vertical resolution
Separation of merged reflections from closely spaced boundaries
Separation of thin layers from each other
Footprints Removal
Automatic recognition and elimination of footprints and other coherent noise
Separate and combined processing of inline and crossline directions
Clear image of faults and horizons without loss of amplitude information
Business Effect
Up to 2 times
Reducing project time thanks to ready-made optimized solutions, eliminating lengthy parameter selection
Reducing project time thanks to ready-made optimized solutions, eliminating lengthy parameter selection
Max
Usage of archived data
thanks to the algorithm’s ability to improve seismic data of any quality
Usage of archived data
thanks to the algorithm’s ability to improve seismic data of any quality
Up to 80%
Reprocessing costs saving
because of enhancing the quality of stacked data without repeating processing cycles
Reprocessing costs saving
because of enhancing the quality of stacked data without repeating processing cycles
Other AI Solutions
Fault interpretation is one of the most difficult tasks within a general structural interpretation workflow.
Fault interpretation is one of the most difficult tasks within a general structural interpretation workflow.
Liverpool city will open its doors to annual surrealistic show. Tenner Hall / August, 15.
Book a Demo
You agree with our Terms and Conditions
Contact Us
You agree with our Terms and Conditions