A Method of Creating Pseudo Acoustic Compressional and Shear and Density Logs Involving Fluid Substi

1
Title
A Method of Creating Pseudo Acoustic
(Compressional and Shear) and Density Logs
Involving Fluid Substitution
Prepared for the Canadian Well Logging
Society by Michael Holmes Digital Formation,
Inc. October 8, 2003
2
Outline

• Introduction
• Data Analysis
• Examples
• Conclusions

3
Introduction Part 1

• Gassmanns model showing the effect of gas
  saturation on the slowing of rock velocities has
  been adapted to petrophysical analysis.
• Components of an analogous model of Krief have
  been integrated into the Gassmann model, to
  include
• The entire range of shaley formations
• Analysis of shear as well as compressional
  velocities
• Using standard wireline log suites, it is
  possible to calculate compressional and shear
  pseudo travel time curves for any input of gas
  vs. water saturation. Additionally, it is
  possible to calculate pseudo density curves,
  again for any input of gas vs. water saturation.

4
Introduction Part 2

• Source data required are
• Porosity as curves
• Shale as curves
• Water and Gas Saturation as curves
• Matrix as component input
• Shale as component input
• Fluid as component input
• An extension of the analysis involves estimating
  water saturation for each porosity log,
  independent of resistivity data. The technique
  consists of solving the porosity log equation for
  fluid content assuming an accurate knowledge of
• Porosity
• Matrix
• Shale
• From the saturations determined from porosity
  logs, it is possible to create pseudo resistivity
  logs, using reasonable values of m, n, and RW.
  These yield a resistivity invasion profile, which
  can be compared with actual resistivities.

5
Data Analyzed

• Four wells from Western Canada, plus one US well
• Belly River Alberta
• Shell et at Rosevear
• Amoco et al Hanlan
• BP Boundary Lake South
• Wamsutter, Wyoming
• All wells have the following log suites
• Compressional DT
• Density
• Neutron
• Gamma Ray
• Resistivity
• Two wells (Amoco BP) also have a Shear DT

6
Data Analysis

• Corrected porosity logs for bad hole.
• Performed standard shaley formation analysis
  using density/neutron combination for porosity.
  This minimizes any gas and matrix effects on
  porosity calculations.
• Ran fluid substitution to derive pseudo logs
  assuming
• Wet formation
• Residual Gas Saturation
• Uninvaded Zone Gas Saturation
• Pseudo compressional and shear acoustic and
  pseudo density curves were created.
• Ran saturation calculation from porosity logs,
  and created pseudo-resistivity logs for each
  porosity log.
• Compared pseudo logs and pseudo saturation curves
  with raw data and calculated saturations.

7
Example 1

• Belly River

8
Belly River Raw Data
Minor correction required
9
Belly River Pseudo Logs
Acoustic log sees gas
Pseudo log heavily affected by very small amounts
of gas
10
Example 2

• Shell Rosevear 714

11
Shell Rosevear 714 Raw Data
Large correction due to hole washout
12
Shell Rosevear 714 Pseudo Logs
Acoustic log does not see gas
Acoustic log sees gas
Generally good correlation between actual and
pseudo acoustic compressional logs
13
Shell Rosevear 714 Pseudo Logs
Acoustic sees little or no gas
Good correlation between actual and pseudo
density curves
14
Shell Rosevear 714 Composite Log
Fair correlation between petrophysics and core
data
Probable bad hole
15
Example 3

• Amoco et al Hanlan

16
Amoco et al Hanlan Raw Data
Significant corrections were required due to bad
hole
17
Amoco et al Hanlan Pseudo Logs
Cycle skips
Acoustic compressional log mostly tracks the wet
pseudo log except where gas occurs
Gas
Good correlation with measured shear curve
18
Amoco et al Hanlan Composite Log
Fair correlation between petrophysics and core
19
Amoco et al Hanlan Pseudo Logs
Compressional acoustic log does not see gas
Good correlation between pseudo and actual shear
acoustic
20
Example 4

• BP Boundary Lake South

21
BP Boundary Lake South Raw Data
Extensive curve editing required due to bad hole
22
BP Boundary Lake South Pseudo Logs
Acoustic log sees gas
Acoustic log does not see gas
Anomalous response, possibly due to bad hole
Generally good correspondence between pseudo and
measured shear acoustic
23
BP Boundary Lake South SW Resistivity from
Porosity Logs
Apparent cycle skipping due to gas in wellbore
Regular invasion profiles in high porosity rocks
24
BP Boundary Lake South Composite Log
Good comparison between petrophysics and core
25
BP Boundary Lake South Pseudo Logs
Fair to good comparison between actual and pseudo
acoustic logs
26
Example 5

• Wamsutter, Wyoming

27
Wamsutter, Wyoming SW Resistivity from
Porosity Logs
Shows variable invasion profile, acoustic log
mostly SW1.0
28
Conclusions Part 1

• Pseudo compressional acoustic logs compare well
  with actual data for the full range of shale
  content. However, it can be shown that recorded
  compressional acoustic data sees gas saturation
  in an irregular and unpredictable fashion. This
  is a consequence of erratic and discontinuous mud
  invasion.
• Small amounts of gas have large effects on the
  acoustic log. Unless fluid (particularly gas)
  saturation is known from other sources, the
  acoustic log cannot be used reliably to determine
  porosity.
• For the two wells with shear acoustic there is
  good to excellent correlation with pseudo logs.
  This gives confidence in the ability to estimate
  shear acoustic curves when no measurements have
  been made.

29
Conclusions Part 2

• Saturation calculations from porosity logs show a
  varied invasion profile. In the general case,
  the following pattern is observed
• Lowest SW
• Resistivity
• Neutron reflecting the changing depth of
  investigation
• Density - reflecting the changing depth of
  investigation
• Sonic reflecting the changing depth of
  investigation
• Highest SW
• Pseudo resistivities logs frequently show better
  bed definition than resistivity logs (especially
  induction) due to better vertical resolution.
• The analytic techniques, in combination, show
  whether or not the raw measurements are
  consistent with the deterministic
  interpretational model. Any inconsistenency in
  the model, and/or bad/miscalibrated raw data will
  be reflected in impossible output such as
  porosity log saturations less than zero, or
  unrealistic values of calculated fluid properties.

Increasing
30
The endHave a nice day!