Watch videos of Presentations from SEG 2017
Using Attributes to Interpret the Environment of Deposition - A Video Course. Taught by Kurt Marfurt, Rocky Roden, and ChingWen Chen
Dr. Kurt Marfurt and Dr. Tom Smith featured in the July edition of AOGR on Machine Learning and Multi-Attribute Analysis
Rocky Roden and Ching Wen Chen in May edition of First Break - Interpretation of DHI Characteristics using Machine Learning
Seismic interpretation and machine learning by Rocky Roden and Deborah Sacrey, GeoExPro, December 2016

LOGIN / MY PROFILE | LOG OUT

Show geologic features and anomalies in 2D and 3D data

SEG 2015 | Patricia Santogrossi

An excerpt:

Upper EF can be readily distinguished and multiple character variations are observed. Ash beds (top EF Sh) may be mapped and separated from potential reservoir facies. Basal clay rich facies can be mapped and separated from Buda and from Eagle Ford shale target. Eagle Ford reservoir interval, previously characterized as “Low Res” or “High Res”.

...read more >

Full Article Text...

Sub-Seismic Resolution Below Tuning

Slide 3

What Paradise Can Do That Other Software Cannot

  • Combines multiple attributes on 2D as well as 3D data
    • Quickly –
      • a 2D line in a twinkling
      • 10 sq. miles w/ 40 attributes in minutes
      • 50 squares with 60-70 attributes in under an hour.
  • Software automatically documents analysis details, results. Reports planned.
  • Normalized by standard deviation, results eliminate tuning, velocity concerns.
  • Characterizations can be delimited by horizons; guided by key line(s)

Slide 4

What Paradise Can Do That Other Software Cannot cont'd

We use:

  • Principal Component Analysis (PCA) Identify and quantify the key attributes in the suite/seismic volumes
  • Self Organizing Maps (SOMs) - Classifydata in clusters that
    • show geologic features/anomalies
    • enable specific calibration
  • Interactive2D ColorMapsthat can use transparency to focus

Slide 5

Eagle Ford Project Basis

  • Nearly 216 sqmi of merged 3D seismic data
    • PSTM volume from 0-5.0s at 2ms sample interval
    • Enhanced volume (SPEN)
  • 4 Mapped Horizons -Austin Chalk, Upper EF, TopEF Shale, Buda
  • Over 300 wells
    • 10+ pilot holes with tops, logs, deviations, T/ D corrections
    • 274 horizontal wells with “attributes”*
    • Type logs for 23 wells

Slide 6

Conclusions to Date

  • Upper EF can be readily distinguished and multiple character variations are observed
  • Ash beds (top EF Sh) may be mapped and separated from potential reservoir facies
  • Basal clay rich facies can be mapped and separated from Buda and from Eagle Ford shale target
  • Eagle Ford reservoir interval, previously characterized as “Low Resistivity”,“High Resistivity”or “Mixed”
    • can be separated into four or more non-layer cake zones
    • includes an unrecognized, underdeveloped (?) sweet spot
    • No problem seeing faults
  • Spectral Decompworks best in the Austin Chalk and Buda….

Slide 7 - Paradise Preliminary Stratigraphic Results

Slide 8

Instantaneous Attributes SOM Recipe from 3 Eigenvectors

Attributes for PCA (16)

  • Acceleration of Phase
  • Band Width
  • Dominant Frequency
  • EnvelopeModulated Phase
  • Envelope Second Derivative
  • Envelope Time Derivative
  • Imaginary Part
  • Instantaneous Frequency
  • Inst. Freq. Envelope Weighted
  • Instantaneous Phase
  • Instantaneous Q
  • Normalized Amplitude
  • Real part
  • Relative Acoustic Impedance
  • Thin Bed Indicator
  • Trace Envelope

Key Attributes for SOMs (10)

  • _3D_merge_pstm_180 (Amplitude)
  • _3D_ps_Envelope Second Derivative
  • _3D_ps_Imaginary Part
  • _3D_ps_Instantaneous Frequency
  • _3D_ps_Normalized Amplitude
  • _3D_ps_Real Part
  • _3D_ps_Relative Acoustic Impedance
  • _3D_ps_Thin Bed Indicator
  • _3D_ps_Trace Envelope

Slide 9 - Original Result: Instantaneous SOM1 - Default colors

Slide 10 - Instantaneous SOM1 Original Result with "Tweaked" colors

Slide 11 - Instantaneous SOM1 Original Result Resolves Facies Bands

Slide 12 - Eagle Ford and Other Very Similar Objectives on Inline 2

Slide 13 - SOM 2 (2)

Slide 14 - Strat Complexity on Inline 3

Slide 15

Instantaneous SOM2 'High Res' Extracted Geobody

  • Interval shown in previous slide “extracted” by three neurons
  • Poorly sampled in key wells with logs. Missed sweet spot?

Slide 16 - Classification on Horizon (3)

Slide 17

Comparison of Detail In Classification SOM1 vs SOM2

  • Character changes in the Upper EF
  • Variations in thickness of Top EF Ash
  • Singles out basal clay shale

Slide 18 - SOM2 Classification on Eagle Ford Top

Slide 19 - SOM2 Classification Volume - 'High Res' Extraction

Slide 20 - New Topolgoy (2)

Slide 21 - Instantaneous SOM3 Result - Facies bands?

Slide 22 - Changes in Neuron Textures by Area

Slide 23 - Paradise Well Results (9)

Slide 24

Resolution on the Eagle Ford

  • EF Shale Objective 14msor 108’
  • EF Formation 28 msor 260’
  • EF “influence”42 msor ~350’

Slide 25 - Index to Wells Instantaneous SOM2

Slide 26 - Instantaneous SOM2 at Wells 2V and 1V - 'Red Zone'

Slide 27 - Instantaneous SOM2 - Wells 3V & 4V, 5V - 'Gold Overlay'

Slide 28 - Instantaneous SOM2 Wells 6V & 8V - Rust Core

Slide 29 - Instantaneous SOM2 Fault Resolution

Slide 30 - Instantaneous SOM2 Classification in Time Slice 2.578s

Slide 31 - Instantaneous SOM2 Well 11 V

Slide 32 - Situation of Additional Uncalibrated Facies Downdip

Slide 33 - Geometric Attributes SOM (1)

Slide 34

Geometric Classification in High Res Eagle Ford

  • This horizon has been shifted 16ms below the Top Eagle Ford
  • Well 3H’slateral borehole encountered sixfaults, the most of any in the well set
  • Well 2H’slateral borehole encountered 2 faults

Slide 35

Spectral Decomposition SOMs (6)

  • A useful tool for “below resolution” seismic interpretation, reservoir thickness estimation, depositional structure enhancement
  • Unravels the seismic signal into its constituent frequencies which allows the user to see phase and amplitude tuned to specific wavelengths
    • Amplitude component quantifies thickness variablility
    • Phase component detects lateral discontinuities

Best results in the Austin Chalk and the Buda!

Slide 36 - Instantaneous Spectral Decomp SOMs 24.7Hz vs 26.5Hz

Slide 37 - Instantaneous Spectral Decomp 24.7Hz SOM

Slide 38

Instantaneous Spectral Decomp 24.7Hz SOM in Time Slice 2.482s

  • Stacking variations apparent from NE to SW
  • Orange/Red & Yellow /Red is Top Buda
  • Purple zone is Basal Clay-rich shale

Slide 39 - Instantaneous Spectral Decomp 24.7Hz SOM

Slide 40 - Instantaneous Spectral Decomp 24.7Hz SOM in Time Slice 2.290s

Slide 41 - Instantaneous SOM Spectral Decomp 26.5Hz OCT Trace

Slide 42 - Instantaneous SOM Spectral Decomp 26.5Hz OCT Trace

Slide 43

Conclusions to Date

  • Upper EF can be readily distinguished and multiple character variations are observed
  • Ash beds (top EF Sh) may be mapped and separated from potential reservoir facies
  • Basal clay rich facies can be mapped and separated from Buda and from Eagle Ford shale target
  • Eagle Ford reservoir interval, previously characterized as “Low Res” or “High Res”
    • can be separated into four or more non-layer cake zones
    • includes an unrecognized, underdeveloped (?) sweet spot
    • No problem seeing faults
  • Spectral Decompworks best in the Austin Chalk and Buda….