GEOLOG 6.6.

1
Multimin Tutorial
Copyright © 1998-2006, Paradigm Geotechnology B.V. All Rights Reserved.

The information in this document is subject to change without notice and should not be construed as a commitment by Paradigm
Geotechnology B.V. or any of its affiliates (collectively, "Paradigm"). Paradigm assumes no responsibility for any errors that may
appear in this document.

The Copyright Act of the United States, Title 17 of the United States Code, Section 501 prohibits the reproduction or transmission
of Paradigm’s copyrighted material in any form or by any means, electronic or mechanical, including photocopying and recording,
or by any information storage and retrieval system without permission in writing from Paradigm. Violators of this statute will be
subject to civil and possible criminal liability. The infringing activity will be enjoined and the infringing articles will be impounded.
Violators will be personally liable for Paradigm’s actual damages and any additional profits of the infringer, or statutory damages
in the amount of up to $150,000 per infringement. Paradigm will also seek all costs and attorney fees. In addition, any person
who infringes this copyright willfully and for the purpose of commercial advantage or private financial gain, or by the reproduction
or distribution of one or more copies of a copyrighted work with a total retail value of over $1,000 shall be punished under the
criminal laws of the United States of America, including fines and possible imprisonment.

The following are trademarks or registered trademarks of Paradigm Geotechnology B.V. or any of its affiliates (collectively,
"Paradigm") in the United States or in other countries: Paradigm™, Paradigm logo, Artist™, Cement-IT™, EarthStudy 2D™, e-
Geoscience™, Epos™, Explorer™, EZdip™, Focus™, GeoDepth™, GeoLAN™, Geolog®, GeoScene™, GeoSec™, iMap™,
IntegralPlus™, LogLAN™, MapLAN™, NexModel™, OpenGeo Tools™, PlaNet™, Probe™, Pump-IT™, Reservoir Navigator™,
ResSim™, Rock & Fluid Canvas™, SeisEarth™, SeisFacies™, SeisX™, Slider™, SolidGeo™, Stratimagic™, Surveyor™,
Sysdrill™, Sysdrill Director™, Sysdrill DirectorGeo™, Sysdrill IDEAS™, Sysdrill SaG™, Vanguard™, Velocity Navigator™, VG
Tools™, VolLAN™, VoxelGeo™, VXplot™, and/or other Paradigm products referenced herein..

All rights not expressly granted are reserved.

The following are trademarks or registered trademarks of their respective companies or organizations:
Adobe Reader - Adobe Systems, Inc.
corEVAL is a trademark of ACS Laboratories Pty Ltd.
FLEXlm - Macrovision Corporation
Exceed and Exceed 3D - Hummingbird Communications
GOCAD - Earth Decision Sciences
GeoFrame, GeoQuest, GeoNet, GeoShare - Schlumberger Ltd.
HP, HPGL, HPGL/2 and PaintJet - Hewlett-Packard Company
IBM AIX - IBM Corporation
IRIS, IRIX, Onyx, Origin - Silicon Graphics, Inc.
JustCGM and Justplot - Justcroft International Ltd.
Linux - Linus Torvalds
Microsoft, MS, MS-DOS, Windows, Windows NT, Visual C++-Microsoft Corporation
Motif and OSF/Motif - Open Software Foundation, Inc.
Netscape - Netscape Communications Corp.
NuTCRACKER - Mortice Kern Systems, Inc.and MKS Software Inc.
NVIDIA, Quadro, GeForce - NVIDIA Corporation
OpenWindows - Sun Microsystems, Inc.
OpenWorks - Landmark Graphics Corporation
Operton, Athlon - AMD
Pentium, Xeon - Intel Corporation
Red Hat - Red Hat, Inc.
RP66/A - Schlumberger Ltd.
SDI and PrintMaster - System Development, Inc.
SunOS, Solaris - Sun Microsystems, Inc.
SPARCstation, UltraSPARC - SPARC International, Inc.
UNIX - The Open Group
WOW - Landmark Graphics Corporation
XVision Eclipse - The Santa Cruz Operation, Inc.
X Window System - Massachusetts Institute of Technology

All other trademarks are owned by their respective owners.

 Contents

Introduction to Geolog's Multimin Tutorial. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Document Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Tutorial Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Multimin Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
First Example - Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
First Example - Data Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Required Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Loading the Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Wireline Depth Matching and Curve Editing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Running the precalc Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Environmental Corrections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
First Example - Multimin Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Rock Descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Determine Minerals and Fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Fluid Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Analysis Intervals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Build a Multimin Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Run an Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Review the Multimin Analysis Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Completion Decision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
First Example - Alternative Interpretations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
A Further Review of the Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
An Alternative Way of Modelling Fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Multimin Analysis Review Revisited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Alternative Lithology Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Prepare a Printer Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Prepare a Plotter Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Review and Discussion of Model Maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Complex Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Review the Environmental Corrections and Pre-processing . . . . . . . . . . . . . . . . . . 45
Create a Simple Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Build a More Complex Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Handling Bad Hole. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Using Alternate Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

Geolog 6.6.1 - Multimin Tutorial 05 / 2007 Contents 1

 Verification Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Creation of Artificial Hole and Wireline Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Creation of Lithologies and Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Multimin Analysis of the Artificial Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
A Non-Linear Example. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Silt Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Review of Available Well Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Handling Coal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Independent Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
NIMBLE Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Modelling a "Pseudo-Mineral" . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Volumes Well Known But Some Parameters Poorly Known . . . . . . . . . . . . . . . . . . 75
Carbonate Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Review of Available Well Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Appendix A - Multimin Aliases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Alias Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Multimin Alias Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Appendix B - THOR Minerals from Rock Element Catalogue. . . . . . . . . . . . . . 80

Geolog 6.6.1 - Multimin Tutorial 05 / 2007 Contents 2

 Introduction to Geolog's Multimin Tutorial

Welcome to the Geolog Multimin tutorial.

This tutorial is designed, with a hands-on, practical approach, to assist you in learning how to
perform sophisticated formation evaluation using Multimin, the advanced petrophysics
application of Paradigm’s Geolog system.

The tutorial is taught by professional trainers to small groups of interested people from the
petroleum industry. The material is presented by following worked examples on a workstation.

You are guided through:

• a first example—a documented, worked example of a simple analysis, and then

further interpretation of the example.
• a complex example.
• a verification example.
• a silt example.
• a NIMBLE example.
• a Carbonate example.

Answers are provided for all problem examples.

This tutorial is intended to be used with Geolog Version 6.6.1 and later. The Multimin version
number is identical to the Geolog version number. If you are using an older version of Geolog,
there will be differences in the appearance of some of the menus and the GUI.

Prerequisites
To gain the most from this tutorial, the following prerequisites are advised to avoid
unnecessary frustration:

• A working knowledge of the Geolog system and its conventions, the methods to
import data into the software, and the way data is stored.
• A working knowledge of the methods used in Geolog to perform basic
preparation tasks for petrophysical data including:
— depth-matching, editing and despiking for petrophysical logs;
— environmental corrections;
— data displays;
— crossplotting.

Geolog 6.6.1 - Multimin Tutorial Introduction 1

 • A general understanding of petrophysics, including the concept of response
equations for tools, and the methods used to calculate shale volumes, porosities
and water saturations. An appreciation of cation-exchange based methods for
water saturation determination would be beneficial, as would an understanding
of mineralogy effects on tool responses.
• A thorough grounding is given in Paradigm's theoretical tutorial, "Probabalistic
Petrophysics - Optimizing Methods in Formation Evaluation". This tutorial
(multimin_theoretical.ppt) is based on the Multimin Technical Reference
documentation. An understanding of the detailed mathematics is helpful, but
not essential.

Document Conventions
In this document, all INPUT to the computer is in Bold Courier New, while all OUTPUT from
the computer is in Courier New, but not bold.

Tutorial Data
The following additional files (files not supplied with software) are used in this tutorial:

DATA: multimin_master.unl REPORTS:

LAYOUTS: (Copy from layouts_units) SECTIONS:
mm_check_lin.layout
mm_check_nonlin.layout
LOGLAN: mm_check.info SPECS: See "Answer Files in This Tutorial" on
mm_check.lls Page 3.
PLOTS: WELLS: halfway_test, heavy, secphi, silt

Powerpoint presentation — multimin_theoretical.ppt — download from the intranet.

Geolog 6.6.1 - Multimin Tutorial Introduction 2

 Answer Files in This Tutorial
As you perform the steps described in this tutorial, there is a possibility that you may
accidentally type in an incorrect value, or choose to do something different. This can cause
problems if later steps build upon an assumed correct starting point. To help avoid these
problems, answers are provided in the models (found in the /specs directory) as listed below:

First Example Complex Example Verification Example

mm_halfway_simp mm_heavy_simp mm_check_lin

mm_halfway_alt mm_heavy_pota mm_check_nonlin
mm_heavy_zircon
mm_heavy_unc
mm_heavy_phit

Silt Example Nimble Example Carbonate Example

mm_silt_simp mm_halfway_nim mm_carb_simp

mm_silt_indoa mm_carb_sec
mm_silt_indob
mm_silt_indo
mm_silt_juhasz
mm_silt_coal

Geolog 6.6.1 - Multimin Tutorial Introduction 3

 Multimin Overview

Multimin is a software application that provides advanced formation evaluation answers for
petroleum geologists, engineers, and petrophysicists. The application is based in the
probabilistic, or optimizing, approach to modelling wireline and rock data. The name
"Multimin" is an abbreviation of "Multiple Mineral", revealing the application’s strong
geological heritage and nature. However, current petrophysical theory, rapid numerical
programming, and an intuitive user interface underpin this geological nature. Multimin users
are able to rapidly produce detailed petrophysical analyses of formations using a wide range
of wireline and rock data.

Geolog 6.6.1 - Multimin Tutorial Multimin Overview 4

 First Example - Introduction

The following steps will guide you through your first Multimin analysis, and then explore,
review, challenge and change this analysis. After completing these steps, you will have a
broad understanding of how to accomplish a Multimin analysis.

The example you will follow is from a well that was drilled in the Alberta Basin of Canada during
the late eighties. The data you will analyze are the original core and wireline logging
measurements that were available at the time the completion decision was made on this well.
The well was drilled to test the Upper Triassic Halfway Formation, and the zone to be analyzed
is the only possible candidate for a gas or oil completion. Based on your petrophysical analysis
and recommendation, this expensive exploration well would either be abandoned, or
additional expense equal to half the original well cost will be incurred to complete the zone you
have predicted to be hydrocarbon productive.

Geolog 6.6.1 - Multimin Tutorial First Example - Introduction 5

 First Example - Data Preparation

Procedure
Before performing an analysis, you need to gather the required information (provided below
for this tutorial) and then you will:

• Start the Connect application and load the data.

• Start the Well application, and review wireline depth matching and curve
editing options.
• Run the precalc module.

Required Information
Before analyzing the example well, you need to gather all the information and data
that will be required. This includes loading the "hard" data such as wireline logs
and core analyses. It also includes reviewing all fluid analyses that have been
performed for the formation, including formation water, oil and gas analyses. The
"soft" data would include knowledge of the geology of the zone for analysis.
Initially, all geological and engineering information from adjacent wells should be
reviewed before a Multimin analysis. Once a few wells in the area have been
analyzed, you will develop particular models that can be applied to analysis
intervals. In this context, model means an appropriate combination of:
• Choice of minerals for available measurements.
• Choice of fluids for available measurements.
• Response equation parameters for all minerals and fluids on each measurement.
• Fluid analysis parameters.
• Electrical properties such as cementation factor.
• Mineral properties such as cation-exchange capacity.

Wireline Logging Data

A fairly comprehensive set of logging data is available:
• Phasor Induction/Sonic/GR
• Lithodensity/Neutron/GR

Geolog 6.6.1 - Multimin Tutorial First Example - Data Preparation 6

 Routine Core Analysis Data
For this first example, routine core analysis will be used for comparative purposes
only. It will not be used as part of the "hard" measurements, such as wireline logs,
from which Multimin calculates volumes of the chosen minerals and fluids. Rather,
it will be used as the "ground truth", against which the Multimin analysis will be
compared. In particular, we will want to match porosity values from both Multimin
and core analyses.
• In the laboratory, fresh surfaces were cut through the core. Visual examination
confirmed that whole mud invasion of the reservoir rock had taken place
variably into permeable intervals.
• Sampling was based on lithological and petrophysical description, and the
samples subjected to careful cleaning and oven drying.
• After cleaning, a routine core analysis was performed, and values for helium
porosity, air permeability, grain density and retort water saturation were measured.

Exercise 1

Loading the Data

All the data required for this Multimin tutorial is stored in the Geolog ASCII unload
file, multimin_master.unl, located in the stars/data directory.

1. Start Geolog and open the STARS project.

2. Start the Connect application.

3. Select and load the multimin_master.unl file.

4. Accept all sets and all default names.

5. After loading is complete, close the Connect application.

Geolog 6.6.1 - Multimin Tutorial First Example - Data Preparation 7

 ) The core data has been depth shifted to match the log data.

Note also the "standard" names for the regular core analysis results—
PHIT_CR, KA_CR, RHOG_CR and SWT_CR have been applied on
loading.
These are the names used in the Paradigm-supplied Multimin layouts,
crossplots, and models. You could choose to use alternative names, but you
would then need to edit the layouts, crossplots and models, or include the
alternative names in the alias.alias file. See "Appendix A - Multimin
Aliases’ on Page 78 for a review of the implementation of aliases in relation
to Multimin.

Exercise 2

Wireline Depth Matching and Curve Editing

1. Start the Well application.

2. Open the HALFWAY_TEST well.

Depth matching and curve editing are processes that are common in
petrophysical analyses. They are a normal part of a structured approach to
any petrophysical analysis, including one that includes a Multimin analysis.
The curve editing process may include despiking, splicing, and even
wholesale graphical and text replacement of clearly spurious data.
Various Paradigm-supplied layouts (e.g., composite) can be used to view the
data to verify its quality, or you can build your own layout. The zone of interest
is from 4820-5085 feet (1470-1550 meters).
For the sake of this tutorial, let us consider this data set to have no need for
either depth matching or editing. If you disagree, please feel free to perform
these tasks now using the normal Geolog processing options.

Geolog 6.6.1 - Multimin Tutorial First Example - Data Preparation 8

 Exercise 3

Running the precalc Module

The precalc module is a compulsory precursor to any petrophysical analysis for
several reasons:
• Various fluid properties that are used by Multimin are dependent on both the
mud pressure in, and the temperature of, the borehole and the immediately
surrounding rock. Some environmental corrections also use these parameters.
The precalc module places curves of these calculated properties into the
database.
• Some environmental corrections are dependent on mud cake thickness.
Commonly, the mud cake thickness measured on the density/neutron porosity
curves and the resistivity curves will be different. The pre-processing option
creates mud cake thickness curves for both of these tools.
• Mud, mud cake, and mud filtrate resistivities from samples need to be translated
into resistivity curves that vary with depth and temperature.
• Conductivities for the flushed zone (CXO) and unflushed zone (CT) will be
calculated from the deep and shallow resistivity logs as they are listed in the
alias.alias file. These will be overwritten if environmental corrections are
subsequently applied to the resistivity logs.
• The photoelectric absorption cross-section (U), will be calculated as the product
of the photoelectric index and electronic density. This will be overwritten if
environmental corrections are subsequently applied to the density log.

Geolog 6.6.1 - Multimin Tutorial First Example - Data Preparation 9

 1. Select Petrophysics > Precalc to open the precalc module.

Figure 1: Precalc Module

Geolog 6.6.1 - Multimin Tutorial First Example - Data Preparation 10

 2. Ensure the parameters match those listed in Table 1 (taken from the log
headers):
Table 1: precalc Module Parameters

Name Imperial Metric Mixed

TLT 41 (degF) 5 (degC) 41 (degF)
BLT 114.8 (degF) 46 (degC) 114.8 (degF)
BLI 5315 (feet) 1620 (m) 1620 (m)
DFD 10.85 (lb/g) 1300 (K/m3) 10.85 (lb/g)
RMS 0.53 (ohmm) 0.53 (ohmm) 0.53 (ohmm)
MST 78.8 (degF) 26 (degC) 78.8 (degF)
RMFS 0.37 (ohmm) 0.37 (ohmm) 0.37 (ohmm)
MFST 78.8 (degF) 26 (degC) 78.8 (degF)
RMCS 0.4 (ohmm) 0.4 (ohmm) 0.4 (ohmm)
MCST 53.6 (degF) 12 (degC) 53.6 (degF)
BS 7.875 (in) 200 (mm) 7.875 (in)

Note the value of DFD is high due to the presence of barites in the mud.

3. Click Start to run the module.

The precalc module has a layout, precomp, associated with it. It is worthwhile
to check over the curves that are presented on the display.
Verify reasonable values for all curves as a mistake here will make the
Multimin analysis faulty.

Environmental Corrections
The next phase of the structured processing chain would normally be the application
of appropriate environmental corrections. These would be applied to the edited and
shifted curves to produce a new group of curves that had the suffix _COR.
For this example, and at this point, we are not going to perform any environmental
corrections. This is a matter of expedient choice, and could be argued on the basis
of the data to hand. The hole appears to be mainly in gauge, has an optimal
calibration diameter (200mm), and shows only moderate invasion. Even the heavy
barites mud would not provide a significant neutron correction.

) Should you wish to perform the environmental corrections, you must

remember to use the appropriately named environmentally corrected curves
in all subsequent Multimin processing.
The environmental corrections loglans each have associated layouts, and you
should check the results to verify correct application of the corrections.

Geolog 6.6.1 - Multimin Tutorial First Example - Data Preparation 11

 First Example - Multimin Analysis

Procedure
In this step, the Multimin analysis is performed. All ancillary data which is, in this example, rock
descriptions, fluid analyses, and analysis intervals should be reviewed. You then:

• Determine the minerals and fluids to be used, and take into consideration what
unusual effects whole mud invasion will have on log responses.
• Build a Multimin model for the HALFWAY_TEST well.
• Run an analysis.
• Review the results of the analysis.
• Make a completion decision based on the review.

Rock Descriptions
Multimin is a petrophysical package that calculates values for porosity, and the
relative volumes of water and hydrocarbons contained in the porosity. It is also a
geological package, calculating the volumes of each of the minerals that you
choose to include in the analysis.
The appropriate choice of which minerals to model is often very easy. However,
sometimes the accuracy and time to perform an analysis is critically dependent on
the choice of the right minerals, and it is worthwhile to use all sources of
knowledge to make this choice:
• The most definitive data is obtained from core descriptions and analyses—X-Ray
Diffraction (XRD), Scanning Electron Microscopy (SEM), particularly for clay
types, and thin section petrography.
• Extremely useful insights are also gained from reviewing binocular microscope
descriptions of drill cuttings.
• In difficult zones, SEM viewing of cuttings can provide vital clues. In using the
SEM, Energy Dispersive X-Ray Spectra (EDX) can be measured on particular
mineral grains. This not only identifies a mineral, but also gives the composition,
enabling a better choice of response parameters for wireline logging equations.
In new exploration areas, or in areas of restricted information, there will be times
when the only data available to determine the correct mineralogy and fluids will be
the wireline logs themselves. In these cases, your Multimin analysis will be very
much from a log analyst's perspective!

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 12

 However, for the HALFWAY_TEST well, there is an abundance of good
geological data:
• During the drilling of the well, an experienced well site geologist has prepared
accurate descriptions of cuttings.
• The data includes lithology descriptions and estimates of visible porosity and
permeability.
• All samples were subjected to fluorescence examination. This was performed
under an ultraviolet light using solvent, and all fluorescence has been attributed
to hydrocarbon only, and not to mineral fluorescence.
No zones of interest were observed until the penetration of the Halfway Formation.
The drill cutting descriptions that were received from the well site are recorded in
Table 2. Mud logging gas response was strongly indicative of producible gas in the
Halfway Formation.

Table 2: HALFWAY_TEST Drill Cuttings Description

Depth (Feet) Depth (meters) Lithology Description

4822.83 - 4831.03 1470.0 - 1472.5 Tight siltstone with major clay, 40% dolomite, no
fluorescence.
4831.03 - 4839.24 1472.5 - 1475.0 Very fine grained marl with 10% dolomite, minor clay,
6% intergranular porosity and fair permeability.
Moderately well sorted quartz grains. Poor
fluorescence. Coring recommended.
4839.24 - 4884.19 1475.0 - 1488.7 Cored interval. 100% recovery. Separate core
description. Mud was oil flecked during coring
operation.
4884.19 - 4896.65 1488.7 - 1492.5 Fine grained marl with well sorted quartz grains.
10 - 20% dolomite. Porosity from 6 - 15% decreasing
with degree of dolomitization. Good permeability.
Poor to moderate fluorescence.
4896.65 - 4954.07 1492.5 - 1510.0 Fine grained marl. Very well sorted quartz grains.
10 - 20% dolomite. Porosity from 6 - 9% decreasing
with degree of dolomitization. Fair permeability and
poor fluorescence.
4954.07 - 5019.69 1510.0 - 1530.0 Similar to 1492.5 - 1510.0 meters with decreasing
fluorescence with depth.
5019.69 - 5085.30 1530.0 - 1550.0 Similar to 1492.5 -1510.0 meters with no
fluorescence.
5085.30 - 5150.92 1550.0 - 1570.0 Tight siltstone with major clay, 40% dolomite, no
fluorescence.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 13

 A core was taken through part of the zone of interest, and the core description
made at the well site was augmented by a detailed description at the core
laboratory. Three major zones were described, and are presented in Table 3.

Table 3: Halfway_Test Core Description

Depth (Feet) Depth (meters) Lithology Description

4839.24 - 4840.88 1475.0 - 1475.5 Upper very fine grained, moderately well sorted
quartzose sandstone variably (5-15%) dolomitized.
Minor illitic clay. Fair permeability with 5%
inter-granular porosity. Poor oil stain
4840.88 - 4854.00 1475.5 - 1479.5 Tight marl with abundant unidentified clay. Matrix is
composed of roughly equal quartz and dolomite. No
oil stain.
4854.00 - 4884.19 1479.5 - 1488.7 Upper fine grained quartzose sandstone. Minor illitic
clay. Very well sorted. Dolomitization varies from
5 - 25%. Dolomitization reduces porosity, which
varies from 9 - 15%. Good permeability with whole
barites mud invasion seen in open pores. Poor oil
stain.

Exercise 1

Determine Minerals and Fluids

The basic question to be asked at the beginning of a Multimin analysis is—which
minerals and fluids should be used?

1. Review carefully the cuttings descriptions in Table 2, and the core

descriptions in Table 3.

2. Consider the minerals and fluids you would choose.

Based on these cuttings and core descriptions, it is appropriate to choose
both quartz and dolomite as matrix minerals, together with illite as the major
clay mineral. (In adjacent wells, SEM analysis has shown that no clay other
than illite is present). The oil stains suggest that we are definitely dealing with
a hydrocarbon zone. It is appropriate to use free water (mud filtrate) and gas
in the flushed zone, and free water (formation brine) and gas in the unflushed
zone. Clay bound water will automatically be included in both zones as we
have chosen a clay (illite).

3. The other information that piques our interest is the invasion of the pore
system by the whole mud, presumably carrying barites into the formation. In
setting up the model, the following question should be considered:

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 14

 The core shows whole mud invasion, and we can assume that barites has
invaded the formation. What unusual effects will whole mud invasion have on
log responses?

Fluid Analyses
We have already chosen gas as the likely hydrocarbon present in this well, and it
is desirable to know the gas specific gravity so that appropriate logging responses
to gas can be estimated.
If the zone was subjected to a production test, a drill stem test, or a wireline
formation fluid test, a sample of the hydrocarbon might have been recovered.
However, for the purposes of this example, no such recovery was available at the
time of performing the Multimin analysis. There are 3 possible sources of
information:
• From the mud logging data.
• Using data from nearby exploration wells that tested gas, or from wells within
the same gas field if this were a development well.
• By inference from reviewing data from elsewhere in the basin.
Nearby producing wells have gas analyses with a specific gravity of 0.65.
No formation water was recovered by testing. Fortunately, data from offset wells is
present, and a comprehensive formation water catalog of the formation suggests a
water resistivity of 0.06 ohmm at 77 degF (25 degC).
Resistivities of mud, mud filtrate, and mud cake were measured by the wireline
logging contractor.

Analysis Intervals
Multimin analyses are generally performed on an interval basis, although a simple
analysis could be run over a defined range. For the HALFWAY_TEST well, only
one interval will be analyzed, but it is common practice to have many analysis
intervals in a single well. Interval subdivision could be a result of lithological or fluid
changes, or possibly as a result of hole conditions, etc. Interval subdivision would
be performed by reviewing the wireline logs, mud logs, core descriptions, test
results, and supported by crossplot analysis.
The HALFWAY interval has been preset in the INTERVAL.INTERVAL log which
was loaded from the Geolog ASCII file.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 15

 Exercise 2

Build a Multimin Model

This is where we need to integrate all available knowledge—wireline logs, rock
and fluid data from the mud log, rock data from cuttings, core descriptions and
analysis, and XRD/IR/SEM, fluid information from samples (RFT/MDT or DST/
production tests) and any chemical analysis.

1. Select Petrophysics > Multimin > Model Maintenance from the Well menu
to display the Multimin Model Editor window (see Figure 2).

Figure 2: Multimin Model Editor screen

The Model Maintenance option can be used to build new models or change
existing models. Models are well-independent, and so can be used in any well
where that particular lithology/fluid model is applicable.

2. Select Model > New (Old Defaults) to open the default model. The initial
display is the Unknowns+ > Responses screen (see Figure 3).
To start, you can choose either of 2 default models. The difference between
these 2 default models lies in the wireline log responses of the different
minerals (see the "Note" on Page 20 for further information).

) When you select any option from the Unknowns+ menu, the Unknowns (top)
section is always displayed and you can therefore select/change the
Unknowns options from any of these combination screens.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 16

 The initial review work we did on the cuttings and core descriptions in
Exercise 1 on Page 14 suggested the minerals quartz, dolomite and
illite should be selected, as well as gas, in both the flushed and unflushed
zones. These components are toggled on/off by clicking the appropriate
checkboxes.
A clay bound water component is implicitly included in both the flushed and
unflushed zones if any clay is selected. This is because the default saturation
equation for the default model is the Dual Water equation. If different
saturation equations are subsequently chosen, bound waters and clays or
shales will be selected/deselected, as appropriate for the chosen equation.
3. Update the Unknowns panel as shown in Figure 3.

Figure 3: Unknown and Responses screen

Note user defined special minerals and fluids are possible. More on this in
"Complex Example" on Page 45 and " Verification Example" on Page 59.

The Responses section allows you to assign response values to each of the
selected Unknowns (volumes) for each of the chosen logs. We will do this in
the next section.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 17

 We should now select the wireline (and any other, e.g., core) logs we want to use
for this model.

4. Select Equations > Wireline.

Figure 4: Wireline Equations screen

5. Logs are turned on or off under the "Selected" column in the Wireline
Equations section. Ensure the following equations are on (set to YES):
— Formation density
— Neutron
— Sonic transit time
— Photoelectric absorption
— Total gamma
— Unflushed conductivity
— Flushed conductivity

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 18

 The name of the database curve that contains the data to be applied to the
equation is displayed in the Log column. All of the defaults are present in the
current well database, so no action is needed and generally will not be unless
non-standard names have been used when loading the data.

) The use of non-standard names may cause problems, particularly when

trying to apply models to other wells that do contain standard log names.
Note the "Responses" section from the "Unknowns+ > Responses" screen
has been reproduced under the "Wireline Equations" section.
As the logs are chosen, the default parameters for the previously chosen
volumes are displayed in the Responses section. Default parameters and
parameter selection are discussed at the end of this section.

6. We will accept the default parameters for our chosen volumes and equations.
The green cells in the "Method" column can be toggled to allow the choice of
different response equations. There are options for Neutron, Sonic transit
time, Unflushed conductivity, and Flushed conductivity.

7. At this stage, leave the Neutron at LINEAR, the Sonic transit time at WYLLIE
LINEAR and select the DUAL-WATER LINEAR for both conductivity equations.
Each of the available options will be explained by your instructor.

8. In the Unc. Method (Uncertainty Method) column, leave the defaults as shown
(Value for all logs except CT and CXO, which are Interval).
The uncertainty method can be toggled between value and log for all wireline
equations. There is a third (Interval) option for CT and CXO. Uncertainty
values are presented for each response equation. The role of uncertainties is
to provide a relative weighting for the effect of each of the equations in the
Multimin analysis. A higher uncertainty will cause the equation to have a
lesser effect on the solution. See "Multimin Uncertainties" in the Multimin
online help documentation for further information.

9. The seven equations chosen represent a typical suite of wireline logging data.
However, other tools are represented by equations here, and you should
scroll through the complete list of equations. Below the partial list shown are
many more types, including user defined response equations.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 19

 Note
The Responses section contains the responses of the chosen tools or measurements, as
represented by their response equations, to the minerals and fluids previously chosen for
this model.

The default values you see are dependent upon the initial default model chosen. The "Old
Defaults" (as we have chosen for this model) are based on various published sources. Three
references are suggested to represent (but not to limit to) the myriad of data sources:

"Element Mineral Rock Catalog", O. Serra. Schlumberger Manual SMP- 7040, 1990.

"Mineral Logging Parameters: Nuclear and Acoustic", A. Ellis et al. The Technical Review,
Schlumberger, v. 36, no. 1, p. 38-52, 1988.

"Applying Mineral Knowledge to Standard Log Interpretation", D. Cannon and G. Coates.

Paper V in Transactions, 31st SPWLA Annual Logging Symposium, Lafayette,
June 24-27, 1990.

The "Herron Defaults" are based on more recent work described in the following paper.

"Elemental Composition and Nuclear Parameters of Some Common Sedimentary

Minerals", Michael M Herron and Abigail Matteson. Nuclear Geophysics, Vol 7, No 3
pp 383-406, 1993.

Which defaults you use will depend on local experience. After choosing the correct minerals
and fluids, setting these responses is the most important factor in obtaining a good Multimin
analysis. As a Multimin user you will probably build up a small library of references containing
tool responses to various fluids and minerals. Multimin is a modeling program where a
modeled or predicted set of logs is compared to a measured set of logs (and we include core
measurements as part of these available logs). The closer the match, the better the analysis.
The match can be improved by varying the responses shown here. However, only reasonable
changes, based on real evidence, should be made. The change in responses should be within
the range of known values for a mineral or fluid.

It is suggested that the Herron defaults would be more applicable with authigenic clays while the
old defaults might in fact be composite values (possibly including small percentages of heavy
radioactive minerals and other impurities). For these reasons, you might find that the Herron
defaults work well in carbonates while the old defaults work well in shaley sands. There is no
hard and fast rule. You will need to see which ones work best in different areas or even fields,
and it might also depend on the number of logs you have available—with more logs you can
work with more minerals and so use more pure minerals rather than pseudos or composites.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 20

The unflushed zone gas, UGAS, creates no response in any of the chosen tools, and the
unflushed zone conductivity, CT, only shows a response to the free water present in the
unflushed zone and bound water.

Clay bound water responses are set equal to the responses of the flushed zone free water
(except for conductivity), and the responses for clays (illite, in this example) are those for dry
clays. When the Dual Water saturation equation is used, Multimin internally computes the
volume of clay bound water associated with the clay as a function of CEC, temperature and
alphax/alphau. Response parameters for wet clays are then calculated. See the "Multimin
Technical Reference" documentation in Geolog's online help for details. When the analysis is
executed, Multimin will initially solve for volumes of wet clays. In storing answers to the well
database, it saves the volumes of each dry clay, and the combined volume of clay bound
water. Whether bound water volumes, clay or shale volumes are computed depends on the
saturation equation chosen.

10. We will not use any core data in the analysis at this stage, so we will not
select any equations from the Core, XRD-IR, or Petrography screens.
However, you should look at these screens (under the Equations menu) to
see what other types of data can be used to help in a Multimin analysis.

11. Select Method > Parameters to display the Dual Water Saturation
Parameters screen.

Figure 5: Saturation Parameters Screen showing Dual-Water Options and Parameters

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 21

 The Saturation Parameters screen shows the options and parameters that
relate to the application of the saturation model that you have chosen (e.g., in
Step 7 above, if you had changed the saturation model to "Archie Linear", the
Archie Linear Saturation Parameters screen would have been invoked).

) White fields can be edited; gray fields cannot be changed.

12. In the Method row in the Dual Water Saturation Parameters section, ensure
the Saturation factor (W) is set to Value.
Notice the values for Cementation exponent (M) and Saturation exponent (N)
are grayed out—WHY?

Setting Saturation factor (W) to Log would give you the option to account for a
variable cementation exponent, but leave as Value for now.

At this stage, ensure "Allow expansion of Clay Bound Water" and "Allow
application of Beta-dilute" are toggled ON—WHY?

13. Select Unknowns+ > Properties.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 22

 Figure 6: Properties screen

Note these are mineral properties, and not wireline equation responses. The
grain densities listed here have different values from the formation density
wireline responses—WHY?

The Cation Exchange Capacity (CEC) values listed here for various clays
should be treated as starting points only. Clays are chemically extremely
variable, and their CECs are also extremely variable. It is a good strategy to
vary these values to assist in the matching of measured to predicted
conductivity curves. For instance, if the predicted unflushed zone and flushed
zone conductivity curves were consistently less than the measured data,
consider increasing the CEC of the dominant clay. This will put in place more
clay bound water, increasing the predicted conductivities.

If "Cation Exchange Capacity" has been selected as the "Clay Porosity

Source" in the "Saturation > Parameters" screen, the user can enter a
constant value for CEC (and a corresponding value of Wet Clay Porosity will
be calculated), or nominate an external log source. If the log option is chosen,
a corresponding PHIT_<clay_type> log will be created.
If "Wet Clay Porosity" has been selected as the "Clay Porosity Source" in the
"Saturation > Parameters" screen, the user can enter a constant value for the
Wet Clay Porosity (and a corresponding value of CEC will be calculated), or
nominate an external log source. If the log option is chosen, a corresponding
CEC_<clay_type> log will be created.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 23

 14. Accept the default values for now.

15. Select Unknowns+ > Constraints to display the Constraints screen.

Figure 7: Constraints Screen

The program constraints (which are not user editable) are displayed.

Note the "Prog Water Mud" would be replaced by "Prog Oil Mud" if Mud Type
under Verify > Edit Parameters had been set to Oil instead of Water.

Look at the Prog X Bndwat and Prog U Bndwat constraints. What do the
numbers mean?

User constraints are defined by the user, and are shown on the menu to be
available for user editing. They must be toggled to "Yes" to be made active.
When this is done, the entire row on the screen is highlighted in green. See
"Constraints" in the Multimin online help documentation for information on the
application of User Constraints.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 24

 16. Select Unknowns+ > Bounds to display the Bounds screen.

Figure 8: Bounds screen

17. Bounds are placed on the upper and lower values the chosen volumes can
have. Leave the default values.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 25

 18. Select Verify > Edit Parameters.

Figure 9: Updated Edit Parameters screen

19. Leave "Calculate fluid responses" set to YES. Note the color of the
corresponding Units field is gray.

20. Change KCI concentration of mud to 0.

21. In Well, select Well > View > Text.

22. Select the WIRE set and click the Log Headers tab.

23. Locate the mid-zone values for Formation temperature (FTEMP) and
Formation pressure (FPRESS), and enter these values.

24. Copy both logs to the Interval set, and from there enter the values into the
Model FTEMP and FPRESS parameters.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 26

 25. Enter the parameter values for the interval of interest for the
HALFWAY_TEST well. Remember that (from "Running the precalc Module"
on Page 9 and "Fluid Analyses" on Page 15):
— Rw: 0.06 ohmm at 77 degF (25 degC)
— RMFS: 0.37 ohmm at 78.8 degF (26 degC)
— gas specific gravity: 0.65

26. Select Equations > Wireline again.

27. Check the Responses. Note that some of the fluid fields are colored gray.

28. Select Verify > Verify Model.

Figure 10: Model verified successfully

The gray Response fields and the uncertainties (Unc. Value/Log column) for
CT and CXO (since the method is set to "Interval") change value. They have
been calculated from the parameters entered in the Edit Parameters screen
using the fluid properties package within Multimin. When "Calculate fluid
responses" is set to YES, none of the gray Response fields are user-editable.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 27

 The interaction of the Calculate fluid responses YES/NO and the editability of
fluid properties fields will be explained by your instructor, and is also detailed in
"Multimin Uncertainties" in the Multimin online help documentation.

29. Select Verify > Display Statistics.

Figure 11: Statistics Screen

Do not be alarmed by this mass of numbers! They are explained fully in the
"Model Quality" section of the Multimin Technical Reference and Multimin
online help documentation.

) The major feature is the interval condition number, reported as 3.25033. The
condition number is a method of describing how well the measured curves
are capable of discriminating the unknown components. One can have the
best quality measurements, but if they are not suited to discriminating the
chosen unknowns, then this number will tell you. The smaller the number
the better.
Multimin describes this model as having a good interval condition number so
we can feel confident in running this analysis. However, we should still run
some models with bad condition numbers, even up to values approaching 10,
the limit at which Multimin will cut out for fully linear models (you can set this to
a higher value in the Verify > Edit Parameters screen—experienced users
only).

) Please take careful note of the last two lines of the report. Should you not
like the statistics at this point, you should edit the parameters, either
removing or changing components, or adding measurements.

30. Select Model > Save As.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 28

 31. Enter a name in the Selection field. You can enter whatever name suits; enter
a description, if required, then select OK to save to the default specs
directory.
It is probable you will want to alter your model several times to optimize your
analysis so, each time you alter your model, it is better to save and then leave the
model editor running in the background.
You have now completed building your first Multimin model.

Exercise 3

Run an Analysis

1. Select Well > Default Datum... from the Well menu.

2. Set the interval processing to use the HALFWAY interval.

3. Open the multimin_recon layout.

Paradigm-supplied Multimin layouts are listed in the table below.

Table 4: Multimin Layouts

Layout Name Description

multimin_analysis.layout Shows lithology, porosity, fluid and quality information.

multimin_recon.layout Combines the previous layout with displays of raw
and predicted data for a "basic" logging set.
multimin_ngt.layout Same as multimin_recon layout but with NGT data
added.
multimin_ngt_mril.layout Same as multimin_recon layout but with NMR data
added.

4. Select Petrophysics > Multimin > Run Analysis to display the tp_multimin
Launcher screen.

Using Figure 12 on Page 31 as a guide, enter the following settings:

5. Leave WIRE as the default data Set names.

6. Select DEPTH for OPT_FLUID.

See "Multimin Analysis" in the Multimin online help documentation for an in
depth discussion of this option. We will use the other option (MODEL) in
" Verification Example" on Page 59.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 29

 7. Accept the gas specific gravity (GASSG) default value of 0.65.

8. Accept the default value of 30 for the Oil API (OILAPI).

9. Enter 0.06 for RWS; the formation water resistivity is known to be 0.06 ohmm
at 77 degF (25 degC).

10. Enter 77 (25) for RWT.

11. Select WATER for MUDBASE.

The MUDBASE field has the options of either WATER or OIL. The mud system
in the example is water based. This parameter controls the relative volumes of
free water in the flushed and unflushed zones. This is accomplished by
mathematical constraints in the formulation of Multimin, and is explained in
detail in the Multimin Technical Reference online help documentation.

12. Accept the default value of 0 for the KCl content in the mud (KCL_MUD).

13. Double click in the MM_MODEL field, click on the Dropdown List button
and select your model from the displayed list (in this example, halfway_1
is selected).

14. Verify that FTEMP, FPRESS and RMF logs have been selected at the bottom
of the screen.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 30

 Figure 12: Multimin (run analysis) Module

We do not have any alternate models to run so we are ready to run the
analysis.

15. Click Start to run the module and the numerical optimization process.
The Message area provides you with information about the depths and
intervals that Multimin processed at run time. When the processing is
complete, the results are displayed in the multimin_recon layout (see
Figure 13).

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 31

 Figure 13: Multimin analysis results

Exercise 4

Review the Multimin Analysis Results

A close inspection of the resulting display should be performed. Lets look at a few
features of the analysis:
1. Consider the amounts of the major minerals quartz and dolomite. Do they
appear to realistically reflect the volumes described from cuttings and core?

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 32

 2. Do the core porosities confirm what you have calculated from the Multimin
analysis?

3. What about the quality indicators? The quality curve indicates a reasonable
result, except for short intervals where it exceeds one.

4. Do the measured and predicted curves match each other sufficiently closely?

Which ones are different?

What would be the effect on the analysis if the different predicted curves were
forced to match their measurement curves?

Which mineral and fluid volumes would increase or decrease?

This part of the analysis is critical. We have performed an intensely numeric

analysis, based on reasonable petrophysical and geological assumptions, and
now is the time to review the results. Some of our analyses will be fine first run, but
most of the time we will want to improve them. This is achieved by first, observing
what is wrong, and second, by altering the analysis in several possible ways:
• Reviewing the measurements, environmental corrections and pre-processing.
• Changing the minerals and fluids in the model. Perhaps including some exotic
mineral (or fluid!), with unusual response parameters into the analysis.

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 33

 • Altering response parameters and petrophysical properties.
• Incorporating bad hole logic.
• Removing anomalous measurements (e.g., uranium) from the analysis.
• Increasing or decreasing the uncertainties on measurements we wish to have a
lesser or greater effect on the analysis.

Exercise 5

Completion Decision
Your Exploration Manager has just walked in the door; he announces that the
project geologist, reservoir engineer and drilling engineer cannot agree on
anything about this well. He is looking forward to your completion decision, and
would appreciate having it within the next fifteen minutes. He may order a drill
stem test to verify your advice.
What is your completion decision?

_______________________________________________________________
_______________________________________________________________

Geolog 6.6.1 - Multimin Tutorial First Example - Multimin Analysis 34

 First Example - Alternative Interpretations

Procedure
This step will explore, review, challenge, and change the Multimin analysis results of the first
example completed in "First Example - Introduction" on Page 5.

) The exercises in "First Example - Introduction" presented material in a very detailed,

stepwise fashion. There was an emphasis on learning the conventions of Multimin, and
on practicing the detailed procedures of performing a Multimin analysis at your
workstation.

This step will assume you are aware of conventions and procedures, and will start to
explore the geological and petrophysical "what-ifs".
You will:

• Learn an alternative method of modelling fluids.

• Learn to use alternative lithology models.
• Prepare a printer report.
• Prepare a plotter report.
• Review and discuss Model Maintenance.

A Further Review of the Analysis

The analysis performed in "First Example - Introduction" contained several flaws.
A further review of the analysis display reveals the following:
• The volumes of quartz and dolomite do not reflect the amounts described from
either the cuttings or core. There is far more dolomite in our analysis than seen
in the core. You should ask yourself the question: is there a measurement or
measurements, not well modeled, that would benefit from more dolomite and/or
less quartz?
• Multimin overestimates porosity when compared with the core data.
• The quality indicator indicates a reasonable result over most of the interval,
except in short intervals where it exceeds one, which are indicated by red
shading. You should use the quality indicator as a flag to check which logs are
not being properly reconstructed.

Geolog 6.6.1 - Multimin Tutorial First Example - Alternative Interpretations 35

 • In and around the zones of poor quality there are several curves that are not
modelled satisfactorily. In particular, the unflushed zone conductivity has a much
lower predicted value than the measured value. We have no justification for
increasing conductivity through either increasing water conductivity, or the
volume of clay, and hence clay bound water. We must conclude that there is not
enough connate water in the answer.
• The photoelectric absorption curve is predicted low, possibly causing Multimin to
prefer dolomite over quartz.
The most likely resolution to these problems is that we have chosen a mineral or
fluid incorrectly. The evidence has been presented for whole mud invasion of the
formation, probably carrying some volume of barites into the formation.
The presence of barites would explain the high photoelectric absorption
measurements compared to the predicted or modelled values. Barites has an
extremely high photoelectric absorption cross-section value at 1065 barns/c3, and
only a small volume would be required for the observed effect. This would explain
Multimin's preference for dolomite (U=9.0) over quartz (U=4.78). This small
volume would have relatively little effect on the density curve, even though the
density of barites is 4.084 g/cm3 (4084 k/m3). The proposition is borne out by the
close match of measured and predicted density curves. The extra density
predicted by preferring dolomite to quartz would be counterbalanced by the
inclusion of larger volumes of gas.
It is this kind of thinking and rationalization that is a key to obtaining good Multimin
analyses. Never lose sight of the fact that Multimin is a modelling program, and we
must always be exploring realistic, alternative scenarios.

Exercise 1

An Alternative Way of Modelling Fluids

Having decided that including barites into our model would help to resolve many of
the problems, we need to change the analysis. Fortunately, we have already noticed
that it is possible to include a special, or user defined fluid, in both the flushed and
unflushed zones. Barites entering the flushed zone is best modelled as a fluid,
because it is part of the contents of the porosity, and displaces the initial fluids.
Follow these steps to include barites as a flushed zone special fluid:

1. Go back to the model editor and, if necessary, open the model you built in
"First Example - Introduction".

2. Select Unknowns+ > Responses.

Geolog 6.6.1 - Multimin Tutorial First Example - Alternative Interpretations 36

 3. Add the X Zone Special Fluid to the volumes already chosen—WHY only the
X Zone?

4. Take a look at the parameters for X Zone Special Fluid. Note that the default
values are for barites. The sonic travel time is set equal to the default for
water, reflecting that barites is part of the fluid filled porosity, and not firmly
attached as part of the rock matrix.

5. Select Equations > Wireline.

We have discussed that barites may have a significant effect on the
photoelectric absorption measurement, and it is necessary to increase the
uncertainty on this curve.

6. Double the uncertainty from 0.32 to 0.64 in the Unc Value/Log field for the
Photoelectric absorption. As we are using a least squares format, this will
mean this curve will now only have a quarter of its previous effect on the
analysis.

7. Select Verify > Display Statistics. The message indicates the model has
been modified, so you must then select Verify > Verify Model.
A new set of descriptive statistics is displayed. The values are more adverse
than those from the previous analysis, but are still quite acceptable for
attempting an analysis.

The description of a POOR interval condition number should not be a cause

for concern—let the statistics of the solution, particularly the quality curve, be
your guide, rather than the statistics of the model.

The reason that most of the minerals have been tagged is the very low
predicted error for X-ZONE SPECIAL FLUID, the barites component. Any
mineral that has a predicted error greater than twenty times the minimum
predicted error is tagged. In reality, the absolute values of the predicted errors
may be acceptable, and we do not know until we run the optimization. See the
Multimin Technical Reference online help documentation for a discussion of
these statistics.

) We are now ready to try this new model. You might want to save as a new
name so you can easily go back to your first model, if necessary.

8. Select Model > Save As, to save the model with a new name.

Geolog 6.6.1 - Multimin Tutorial First Example - Alternative Interpretations 37

 9. In Well, rerun the Multimin analysis using the revised or new model.

) This is an example of how easy it is to change models and rerun analyses

to explore various "what-ifs".

Multimin Analysis Review Revisited

The analysis is remarkably different from that produced in the "First Example -
Introduction". A quick comparison of the two displays and the lithology descriptions
reveals:
• The lithology calculated has the right mix of mainly quartz, with lesser dolomite.
• The match of core and Multimin porosities are extremely close, improving on the
poor previous match.
• The quality indicator is excellent, with a value less than one throughout.
• All measured and predicted curves match each other extremely well.
This supports the conclusion that we have gained a superior analysis by including
just one extra component. This is very typical of a large proportion of Multimin
analyses—success is dependent on choosing the correct components, and the
review of both petrographic descriptions and fluid analyses are important steps.
The down side of this improved analysis is that we have a much more pessimistic
analysis. The gas saturation is much less, and does not show a zone at irreducible
water saturation, but rather part of a transition zone. Local experience suggests
that production is likely to be mainly water, possibly with gas cut.
Note the very small volume of X Zone Special Fluid that is plotted on the left side
of the porosity track. This is an estimate of the amount of barites that has been
carried into the formation during mud filtrate invasion. It is a very sensitive
permeability indicator, and can be used to show fractured shales.
Should you wish to load an answer file for this well, open the mm_halfway_alt
model. The answer file for the original model is mm_halfway_simp.

Geolog 6.6.1 - Multimin Tutorial First Example - Alternative Interpretations 38

 Exercise 2

Alternative Lithology Models

This exercise is designed to show how quick it is to examine various lithology
models with Multimin. There are three possible minerals that are found elsewhere
in this formation: anhydrite and rarely calcite; destroying the sand porosity; and,
rarely kaolinite as a diagenetic clay.
Perform the following analyses:

1. Run your second model with "Allow expansion of claybound water" turned off
(select Method > Parameters).

2. Run your original model (without the special fluid) without using the
photoelectric absorption (select Equations > Wireline).

3. Attempt to add the three minor minerals at the same time to the current
analysis. You will not succeed, but this will illustrate a safety feature of
Multimin. WHY were you not allowed this many volumes?

4. Add anhydrite by itself to the existing quartz-dolomite-illite model, using the

same fluid components. Run the analysis and review the display.

5. Repeat Step 4. with calcite instead of anhydrite.

6. Repeat Step 4. with kaolinite instead of calcite.

By comparing the three different analyses (Steps 4., 5. and 6.) you can determine
the next most useful mineral to add to a Multimin analysis.
• Anhydrite would seem to be the probable choice.

• The absence of kaolinite agrees with the petrography descriptions.

• Calcite volumes calculated are inconsistent with the petrography descriptions

and this illustrates the importance of analyst control in model composition.

• The barites effect on photoelectric absorption tries to boost the calcite.

Geolog 6.6.1 - Multimin Tutorial First Example - Alternative Interpretations 39

 • Another approach could be to include the barites as a special fluid but drop the
photoelectric absorption as a log. This would require us to return to the simpler
model (no anhydrite, calcite or kaolinite) but would support the previous results.
Trying various alternatives in the model and "fine-tuning" is a common technique
in building Multimin models.
Once you are happy with the results of your analysis, you can generate report files
and plots to include in your well report.

Exercise 3

Prepare a Printer Report

Printer reports of the parameters you have used in the Multimin analysis of the
well can be created, when required.
To create a printer report
1. In Well, select Petrophysics > Multimin > Report.

Geolog 6.6.1 - Multimin Tutorial First Example - Alternative Interpretations 40

 Figure 14: Module for Creating Multimin Reports

Geolog 6.6.1 - Multimin Tutorial First Example - Alternative Interpretations 41

 2. Enter a file name for the report in the REPFILE field and, if required, change
the values in the PAGE WIDTH and LENGTH fields.

3. Click Start to create the report.

The report is automatically displayed as a PDF and can be printed using the
procedures normally followed for printing at your site. You will have to obtain
these procedures from your Systems Administrator. The report file is also
placed in the project reports directory (in this tutorial example, it is placed in
the reports directory of the STARS project).

4. Close the report.

Exercise 4

Prepare a Plotter Report

Plot reports are similar to printer reports in content (see " Prepare a Printer
Report"), but are specifically designed for attaching to the foot of Multimin layouts.

To create a plotter report:

1. Select Petrophysics > Multimin > Plot to display the plotter module which is
partially reproduced in Figure 15.

Geolog 6.6.1 - Multimin Tutorial First Example - Alternative Interpretations 42

 Figure 15: Module for Creating Plots for Multimin Layout Footers

Geolog 6.6.1 - Multimin Tutorial First Example - Alternative Interpretations 43

 2. Enter a file name for the CGM plot in the PLOT_FILE field and, if required,
change the values in the other Parameter fields to suit your requirements.

3. Click Start to create the CGM file.

A CGM file is automatically displayed and also placed in the project plots
directory. This file can be attached as a footer to your Multimin layouts.

4. Close the file.

You may wish to create a hard copy of the display. The implementation of
particular plotters will depend entirely on your local installation; you should
consult your Geolog administrator for the correct procedures.

5. Close all open views.

Exercise 5

Review and Discussion of Model Maintenance

Before we start on the next example, it is a good time to review Model
Maintenance and look a little deeper into how the Model editor is structured and
how it can interact with tp_multimin.
The instructor will lead this review with references made to the Multimin online
help documentation.

Geolog 6.6.1 - Multimin Tutorial First Example - Alternative Interpretations 44

 Complex Example

Procedure
This section initially focuses on a review of the Model Editor and then introduces an example
that is designed to cover much more complex procedures. Variable uncertainties, model
switching and depth reference conventions are also introduced. You will:

• Review the environmental corrections and pre-processing.

• Create a simple model.
• Build a more complex model from the simple model.
• Handle a bad hole.
• Use an alternate model.

Exercise 1

Review the Environmental Corrections and Pre-processing

1. Open the HEAVY well.

) There are many subtleties to this analysis, so do not be surprised if you find
some apparently unusual parameters!

Fluid properties:
Oil 45.4 DAPI Nearby well DST
Gas 0.613 SG Nearby well DST
Rw 0.5 ohmm @ 77 degF (25 degC) DST on Heavy
Temperature of target zone = 120 degF (49 degC)
KCl = 3%

Table 5 on page 46 summarizes the petrophysicist's description of the core

immediately surrounding the points from which the plugs used in the
measurement of electrical properties were selected.

Geolog 6.6.1 - Multimin Tutorial Complex Example 45

 Table 5: Heavy: Plug Surrounds

Depth (feet) Depth (meters) Lithology Description

7428.81 2264.3 Rubble zone above (7424.54-7425.30ft; 2263-
plug 5 plug 5 2263.23m). Appears more shaley and is subdued on
fluorescence. Described as more argillaceous above
and below on core description. Has more "dark grains"
than sample below. Quite friable sample but cement
has well lithified rock below. Siderite? Poor sample -
about 25% of plug (not present) appears to have
argillaceous zone through it.
7441.27 2268.1 15.75" (40 cm) below this sample is a very argillaceous
plug 8 plug 8 thick band; also many argillaceous fragments 7.87"-
15.75" (20-40 cm) below. Very highly graded.
Somewhat argillaceous 11.81" (30 cm) above. Very
open pore system; many opaques; no obvious
argillaceous material.
7454.72 2272.2 Much more open pore system than sample 5 (i.e., low
plug 10a plug 10a m is more credible). Poor sample—partly more lithified
around rock fragments, but overall very friable. Much
more argillaceous streak above.
7498.69 2285.6 Very argillaceous 15.75" (40 cm) band about 19.69"
plug 13a plug 13a (50 cm) above; appears more subdued fluorescence
3.94" (10 cm) below. Very clear change in lithology
1.97" (5 cm) above to more argillaceous material
(shows up well in wetted core).
7531.17 2295.5 Only half of this sample (plughole) present in box
plug 17 plug 17 presented. However, was a very clean sample (far
cleaner than either 5 or 10a above). Rock below is
similarly clean until a very argillaceous band 19.69"
(50 cm) below. Photos show a similar band 19.69"
(50 cm) higher. A great deal of opaques (rather than
argillaceous rock fragments) may be giving an
inappropriately high clay content from logs; logs predict
highest CEC!

Table 6: Heavy: SCAL Electrical Properties Results

Depth KA_CR PHIT_CR RHOG_CR M N

m md % g/cc
2263.25 51 20.6 2.67 1.74 1.96
2267.05 92 20.1 2.6 1.73 2.07
2271.15 357 16 2.68 1.57 1.56
2284.35 102 21.8 2.7 1.84 1.83
2294.2 61 24.8 2.68 1.96 1.87

Geolog 6.6.1 - Multimin Tutorial Complex Example 46

 Exercise 2

Create a Simple Model

In this example, little information is available on the mineralogy of the formation
through the interval of interest. This is often the case, particularly in exploration
wells and early in the appraisal stages of a field. In order to perform the analysis,
we can use crossplots to help identify which minerals may be present, and which
parameters to use for special minerals.

Basic Mineral Assemblage

In order to make an educated guess of which minerals may be present within the
formation, we can:
— calculate the apparent matrix properties of the formation.
— use an RHOmatrix(app)-Umatrix(app) crossplot or a MID plot to deduce the
minerals present.

1. Select Well > Default Datum... and remove interval processing.

2. Select Petrophysics > Parameter Picking > Apparent Matrix Properties to

open the appmat module.
This module calculates RHO_MAA, DT_MAA, U_MAA, PHIT_DN, PHIT_NS,
NLITH and MLITH logs. These logs can be used in crossplots for mineral
identification.

3. Change the DT_Quartz value to 55 µs/ft (180 µs/m).

4. Click Start to process the module.

5. Select Petrophysics > Parameter Picking > RHOMA-UMA plot to open the
RHOMA-UMA crossplot (see Figure 16).

Geolog 6.6.1 - Multimin Tutorial Complex Example 47

 Figure 16: RHOMA-UMA crossplot for lithology identification

This plot shows the position of end member minerals. From this plot, what is
the simplest mineral assemblage that would adequately explain the data?

Geolog 6.6.1 - Multimin Tutorial Complex Example 48

 6. Look at the MID plot (Petrophysics > Parameter Picking > MID plot) and
consider the same question (see Figure 17).

Figure 17: MID crossplot for lithology identification.

Geolog 6.6.1 - Multimin Tutorial Complex Example 49

 7. Create a Multimin model using your newly decided mineral assemblage and
run the model to see the result.
At this stage, do not use any of the information from the spectral gamma data.

8. Iterate through your model a few times to obtain the best possible result.
9. Once you are happy with your result, open the multimin_recon layout that
shows the spectral gamma data. In particular, have a look at the interval
7513.12-7529.53ft (2290-2295m). What do you think about your prediction?

10. Add the POTA log from the spectral gamma ray to your analysis. What is the
result?

So what does this mean? . . . It probably means that we have either picked a
mineral incorrectly or that a mineral is missing from our model.

Exercise 3

Build a More Complex Model

In the above case, it is likely that a mineral is missing from the model.

1. Have a look at the Plug surrounds descriptions (Table 5 on page 46) to see if
any clues are given.

2. Also look at the following crossplots to see if there is anything unusual about
the data:
— thor_pota_uran.xplot
— rho_nphi_gr.xplot (try changing the Color Expression to URAN_COR,
THOR_COR and U, as well as GR)

) You may need to adjust the xplot (e.g. limits and color bar) and Datum in
order to pick up the subtleties (see Figure 18).

Geolog 6.6.1 - Multimin Tutorial Complex Example 50

 Figure 18: Crossplots for lithology identification

Geolog 6.6.1 - Multimin Tutorial Complex Example 51

 From these crossplots, it should be apparent what is required.

Having decided that an opaque or heavy mineral may be required, and that this
mineral needs to be high in thorium and uranium, and low in potassium, the first
step would be to go to the rock element catalogs and try to find a suitable
mineral (see "Appendix B - THOR Minerals from Rock Element Catalogue’ on
Page 80).

Zircon could be a suitable mineral and is a common accessory mineral.

However, the parameters associated with mineral (in particular, the THOR and
URAN values) are quite poorly defined. In fact, if we are going to incorporate
the spectral gamma data (i.e., use URAN/THOR/POTA instead of GR), then
the most critical parameter is the ratio of URAN to THOR in the heavy mineral
rather than its absolute values. Because the actual volume of the heavy
mineral is expected to be so small, it is not critical that we get the volume
exactly right . . . just that we give the model enough GR response so that it can
'soak' it up as it needs instead of putting in false amounts of clay and feldspar.
We can determine the required ratio from the spectral gamma crossplots.

3. Look at the spectral gamma crossplot (thor_pota_uran).

Geolog 6.6.1 - Multimin Tutorial Complex Example 52

 Hot Sand Response

Normal Sand Response

Figure 19: Spectral Gamma Response for Heavy Interval

The strong zircon influence in the crossplot is at THOR=14, POTA=0.005 and

URAN=6.5. From the crossplot, normal sands read THOR=2.5, POTA=0.005
and URAN=2. Therefore, the relative influence above the normal sand
baseline is 10.5 ppm for THOR and 4.5 ppm for URAN; this corresponds to a
ratio of 2.6 units of THOR for every URAN. In this case, if we have a URAN
parameter of 1000 ppm, then the corresponding THOR should be 2600 ppm.
If we want to work out what the corresponding GR_COR value should be,
then that can be easily estimated from the existing wireline data.

4. Format the datum such that the well is being processed from the range
WELL_TOP to WELL_BOTTOM.

Geolog 6.6.1 - Multimin Tutorial Complex Example 53

 5. Select General > Statistics > Multiple Regression to open the
tp_regress_multiple module.

6. Fill in the fields as follows:

FILE_OUT heavy_spectral_analysis.rpt
MACRO_NAME gr_from_spectral
LOG_DEP GR_COR
LOG_INDEP POTA_COR THOR_COR URAN_COR

7. Run the module and look at the resulting report file.

Figure 20: Heavy Spectral Analysis Report Produced From

Multiple Regression Module

From the statistics calculated by the module, the relationship for GR from the
individual spectral components can be obtained and it is expected that this is
consistent, as it is a function of tool calibration.

8. Now build a model incorporating the spectral gamma data and a heavy mineral.

Geolog 6.6.1 - Multimin Tutorial Complex Example 54

 ) You can use the rock element catalog for the other properties for Zircon.

9. For spectral data, the following could be tried as an example:

POTA = 0
THOR = 2600
URAN = 1000
GR_COR = 2600*4.5+1000*7.5 = 19200

10. Also try running the model without one of the spectral gamma logs, using the
GR instead.

Exercise 4

Handling Bad Hole

One way to handle bad hole would be to generate a flag which was based on
normal indicators of bad hole (e.g., caliper or density correction excursions). This
could then be used to introduce an alternate model where the uncertainties on the
logs affected by the bad hole would have been increased. "Using Alternate
Models" on Page 56 discusses this in further detail.
An easier and more elegant way to handle bad hole would be to generate variable
uncertainty logs.

Variable Uncertainties
Uncertainties are values that reflect the precision of measurements and vary as a
result of tool physics, tool calibration, borehole conditions and other factors.
Uncertainty on each wireline log measurement is a function of several factors,
usually similar for most types of measurements. For example, the uncertainty on a
gamma ray log is a function of:
• Changes in hole size and mud weight from nominal conditions.
• Uncertainty in depth recording - does the measurement represent the value at
the exact depth, or part way to the frame of data immediately above or below?
• Precision of detector physics.
• Accuracy of tool calibration.

Geolog 6.6.1 - Multimin Tutorial Complex Example 55

 Variable uncertainties are calculated using loglans based on the Schlumberger
environmental corrections. Loglans are provided for:
• gamma ray (GR) / spectral gamma ray (POTA, THOR, URAN)
• thermal neutron (TNPH),
• formation density (FDC),
• litho-density (LDT),
• sonic (DT), and
• a range of resistivity and conductivity logs.
These loglans compute output uncertainty logs for each tool
(e.g., GR_COR_UNC) and logs to display the uncertainty range
(e.g., GR_COR_PLUS and GR_COR_MINUS). Note that, although the loglans
are based on the environmental corrections, they do not output environmentally
corrected logs. The relevant service company environmental corrections should
be applied to the logs (if desired/required) before calculating uncertainties.

Exercise 5

Using Alternate Models

Depth Reference Conventions

Input Sets
There are several different sources of measurements that can be used in Multimin.
We have loaded two kinds of data, with different data structures:
• Wireline data in a set called WIRE. This set is a regularly sampled point data
set. Geolog sampling concepts are explained in the Multimin online help
documentation.
• Core data in a set called CORE. This data is irregularly sampled point data. It
will not match depth for depth, measurement for measurement, the regularly
sampled wireline data.
Other data could have been loaded into two other kinds of sets called (by Multimin
convention):
• XRD_IR contains XRD/IR (X-Ray Diffraction / Infra Red spectroscopy)
measurements of mineral weight fractions.
• ROCK petrography descriptions of cuttings. Their sampling could be either
limited or extensive, and with variable regularity.

Geolog 6.6.1 - Multimin Tutorial Complex Example 56

 Sampling Control Logs
It is because of the great variability in sampling that Multimin imposes the following
convention: the base depth reference in all Multimin analyses will normally be the
DEPTH curve from the wireline set. This is because it is normally the highest
sampling rate curve of the complete data set. Multimin will analyze at each depth
frame present in the wireline set, and measurements used from other sets will be
interpolated to depths present in the wireline set.
This has implications for how the data in sets such as CORE, XRD_IR and ROCK
are stored. For example, if there are significant gaps in the core data and these
gaps are not identified by "missings", then the core data will be interpolated across
these gaps with resultant impact on the analysis.
The reason that DEPTH is chosen as the curve to base the sampling on is that it is
the first input log called by tp_multimin. This, of course, could be changed by
nominating a different "Sampling Control Log", or even nominating a different
default input set in the runtime window.
This convention also forces an apparently unusual data structure on some results.
All results—volumes, predicted curves, etc.—will be placed in the default output
set. So the predicted curves for the measurements used from the sets CORE,
XRD_IR and ROCK would normally be placed into the set WIRE. However, this is
appropriate, as these values would normally be calculated at the depth frames
defined in the wireline data set, rather than the depths present in the sets CORE,
XRD_IR and ROCK.

Output Sets
Just as we can nominate a default input set, we can also nominate a different
output set. This can be very handy for exploring "what-ifs" and comparing results
from different models. It is also an efficient way of knowing which volume came
from which pass—if some minerals or fluid volumes are included in some models
but not in others then, if all results are sent to the same default set, it is difficult to
know which result log version goes with which pass.
Multimin features logic for model switching between a primary model and up to
10 alternate models.

) The use of alternate models is optional.

The purpose of primary and alternate models is to allow the application of more
than one model in a particular interval. Reasons to have alternate models in the
same interval could include:
• The inter-stratification of a different reservoir type. This could be a change of
mineralogy or the need to switch fluid models (e.g., from gas to oil).
• Evaporitic or coal sequences.

Geolog 6.6.1 - Multimin Tutorial Complex Example 57

 • Shales that need different porosity/fluid modelling.
• A bad hole condition with the need to downgrade or even abandon poorly
measured logs and simplify to a smaller model.
Alternate models are chosen over a primary model by the application of a set of
user defined conditions. They are defined via the "Selection Expression" which is
similar in construction to "Evaluate" statements (Boolean expressions). The
primary model pass is run over all levels, then each level is flagged with the first
alternate model whose associated selection expression is passed (stored in the
MODEL log). It is therefore possible to use results from the primary model pass as
elements of the selection expression for alternate models. Multimin loops once for
each alternate model whose associated selection expression has been passed,
applying the alternate model previously stored in the MODEL log.
See "Multimin Analysis" in the Multimin online help documentation for a
description of the special alternate models "SKIP" and "NONE", and examples of
alternate model selection.
This is a very efficient computational process as only the primary and possibly one
alternate model are calculated for each depth frame. The fluid properties
associated with each model are calculated only during the loop that deals with that
model. Only one set of results is stored.
Over the zone of interest in this well, we have excellent core coverage. We could,
therefore, use the core data as part of the analysis solution.

1. Create a new model which includes PHIT_CORE and RHOG_CORE as input

logs. Remember about gaps in the core data and how they should be
handled.

2. Rerun Multimin with your new model, including the CORE logs selected by
the expression "PHIT_CORE <> missing".
3. What would we need to do to make the model with the CORE logs the
primary model and our original model the alternate?

Geolog 6.6.1 - Multimin Tutorial Complex Example 58

 Verification Example

Procedure
This section is designed to show that the technology behind the Multimin program really
works. To do this, you will:

• Create a controlled data set. Seven distinct lithology intervals of different

complexity will be created in an artificial well.
• Create, for the common wireline logs, artificial data for each lithology using the
Multimin response equations.
• Perform a Multimin analysis to calculate the volumes of minerals and fluids, and
the predicted response curves.
• Generate a display showing the component volumes, and observe the exact
match of predicted to measured wireline logs.

Overview
After studying the theory of probabilistic petrophysics, and tackling the
mathematics behind response equations and optimization schemes, we have
moved on to perform several practical examples. Complex and involved formulae
have been replaced by an easy to use program that hopefully makes probabilistic
petrophysics fun.
If Multimin has achieved this, there will appear to be little connection between the
theoretical and practical aspects. You may therefore be wondering—how can one
verify that Multimin is really doing what it claims to do?
This process will be performed for an example with only linear response
equations, and then repeated for an example with non-linear conductivity and
neutron equations.

Geolog 6.6.1 - Multimin Tutorial Verification Example 59

 Exercise 1

Creation of Artificial Hole and Wireline Set

This test can be performed under any units system.

To generate a new hole:

1. Close all open views—Well > View > Close All.

2. Select Well > New.

3. You will be prompted to provide a Well Identifier. Enter an appropriate name

(e.g., MULTIMIN_CHECK).

4. Save the well changes.

5. Select Well > View > Text.

6. In the Text view, insert a set named WIRE with:

— Top of 1
— Bottom of 700
— Sample Rate of 1

7. Click Create—the WIRE and a REFERENCE set will be created.

8. Select Well > Save.

Exercise 2

Creation of Lithologies and Measurements

The project includes a loglan called mm_check.

To review and prepare the loglan

1. Select General > Loglan.

2. Open the mm_check.lls loglan.

3. Select Tools > Compile to compile the program to run on the version of
Geolog currently being used for the course.

Geolog 6.6.1 - Multimin Tutorial Verification Example 60

 MM_CHECK creates volumes of the following components:
— quartz
— calcite
— dolomite
— illite
— clay bound water
— flushed and unflushed zone free water
— flushed and unflushed zone oil

We are going to create seven distinct lithologies of varying mineralogy and fluids.
Within each lithology, the relative proportion of each mineral remains the same, as
does the relative proportion of each fluid. However, the porosity is increased from
0 to 0.5 over 100 depth frames.
Each of the seven lithology and fluids models is progressively more complex. The
first six are reasonable models, with the volumes of minerals and flushed zone
fluids summing to 1. The seventh model is deliberately unrealistic, as these
components are contrived to sum to 1.1 rather than 1.
Artificial measurements are created using response equations as detailed in
"Response Equations" in the Multimin Technical Reference online help
documentation. The loglan also creates the following artificial log curves:
— RHOB formation density
— TNPH neutron linear
— DT sonic transit time
— U photoelectric absorption cross-section, rather than PEF
— GR total gamma
— CT unflushed zone conductivity, and the corresponding RT
— CXO flushed zone conductivity, and the corresponding RXO

For the TNPH, CT, and CXO linear responses, corresponding non-linear
responses called TNPH_NL, CT_NL and CXO_NL are created.

) Note that the creation of CT, CXO, and U obviates the need to perform
preprocessing with Precalc. In fact, RT, RXO, and PEF are created, so
Precalc could still be used.

4. Open the layout mm_check_lin.layout.

Geolog 6.6.1 - Multimin Tutorial Verification Example 61

 5. Select General > Module...

6. Select the mm_check.info loglan.

7. The default values can be used initially (you might want to experiment with
different parameters and models later).

8. For later use in Multimin, make a note of:

— fluid responses: RHO _______________________________
_______________________________
TNPH _______________________________
_______________________________
U _______________________________
_______________________________
— temperature: _______________________________
_______________________________
— water resistivity: _______________________________
_______________________________
— filtrate resistivity: _______________________________
_______________________________
— saturation exponents: _______________________________
_______________________________

9. Run the module.

The names for the known volume and saturation curves are suffixed with
a _K. All of the volumes and responses have now been created and will be
displayed on the layout.

Exercise 3

Multimin Analysis of the Artificial Data

The next steps are identical to the procedures we have followed previously to
perform a Multimin analysis:

1. Go to the Model Maintenance window or select Petrophysics > Multimin >

Model Maintenance.

2. Build a model using the same minerals and volumes created by

mm_check.model.

Geolog 6.6.1 - Multimin Tutorial Verification Example 62

 3. Select the same equations which correspond to the artificial logs previously
created. The default logs names in the model will be OK.

4. Set Neutron, Unflushed and Flushed Conductivities to Linear.

5. Set Conductivities Unc Method to Interval and to Value for all other logs
(accept the default values).

6. Ensure all mineral responses are the same as were defined in

mm_check.model.

7. Select Verify > Edit Parameters and toggle "Calculate fluid responses" to NO.

8. Go back to a Responses screen and edit the orange cells to have the same
values as used in mm_check.model.

9. Select Saturation > Parameters.

— Ensure M, N, and W correspond to the values used in mm_check.model.
— Ensure that "Allow expansion of Clay Bound Water" is set to NO.

10. Save the model under an appropriate name (e.g., mm_check_lin_new).

11. Select Petrophysics > Multimin > Run Analysis.

12. Ensure that:

— OPT_FLUID is set to MODEL
— RWS and RWT are set as in mm_check.model.

13. Enter the values used in mm_check.model for the RMF and FTEMP logs.
FPRESS can take any value in this case—WHY?.

The analysis can be run over the complete range or an interval set could be
created to experiment with different lithologies.

14. Compare the Multimin computed volumes to the known volumes and the
predicted logs to the artificial logs.

Geolog 6.6.1 - Multimin Tutorial Verification Example 63

 You have just verified the technology behind Multimin. Note that the measured
data curves are almost exactly duplicated by the predicted curves in the first six
lithology examples. (Any differences are due to slightly different parameters used
in mm_check.model and Multimin; you may wish to edit these in Loglan or
the model).
However, the impossible seventh lithology example has not been exactly solved by
Multimin. It has correctly warned that it is not possible to model this zone, with the
chosen parameters applied to the (false) artificial measurement curves. The quality
curve is unacceptably large, and the measured and predicted curve fail to match.

15. You may wish to expand this kind of verification example by modifying the
loglan routine to incorporate any components, or response equations, that
interest you.

Exercise 4

A Non-Linear Example

1. Create a new model using non-linear equations for neutron, unflushed and
flushed conductivity.

2. Change the names of the corresponding logs to pick up the non-linear

artificial measurements.

3. Open the layout mm_check_nonlin.layout to view the results from running

Multimin.

4. Review the quality of the solution, as shown by the QUALITY curve.

The curve NFUN shows the number of iterations required during non-linear
optimization. Note that, whereas the linear method gives exact solutions for
component volumes, the non-linear gives close to exact solutions. This is because
the non-linear optimization scheme used minimizes the objective function to a
small, but acceptable, minimum error (see the Multimin Technical Reference
online help documentation).

Geolog 6.6.1 - Multimin Tutorial Verification Example 64

 Silt Example

Procedure
This section presents an example that is designed to examine choice of clay types and their
associated parameters, with particular emphasis on CEC. You will:

• Review the data.

• Create a coal model to evaluate coaly horizons.
• Create a model to evaluate the sands and silts, and perform an analysis using
model switching.

Exercise 1

Review of Available Well Information

1. Open the SILT well.

2. Review the data—this well has 3 predefined intervals:

A Zones A and B consist of siltstones at the
B top and shaley sands at the bottom.
Granite Is basement.

The log header advised that the PEF has been affected by barites
(144.8 lb/bbl of barite) and that the mud was contaminated with diesel. The
following parameter values were taken from the log header:

Value Value Value

Name
(Imperial) (metric) (mixed)
TLT 78.8 (degF) 26 (degC) 78.8 (degF)
BLT 236 (degF) 113.33 (degC) 236 (degF)
BLI 10459 (feet) 3188 (m) 3188 (m)
DFD 11.35 (lb/g) 1370 (K/m3) 11.35 (lb/g)
KCl 2.4 (%)
RMS 0.24 (ohmm) 0.24 (ohmm) 0.24 (ohmm)
MST 78.8 (degF) 26 (degC) 78.8 (degF)
RMFS 0.175 (ohmm) 0.175 (ohmm) 0.175 (ohmm)
MFST 78.8 (degF) 26 (degC) 78.8 (degF)

Geolog 6.6.1 - Multimin Tutorial Silt Example 65

 Value Value Value
Name
(Imperial) (metric) (mixed)
RMCS 0.803 (ohmm) 0.803 (ohmm) 0.803 (ohmm)
MCST 73.4 (degF) 23 (degC) 73.4 (degF)
BS 8.5 (ins) 215.9 (mm) 8.5 (ins)

Fluid Properties

Perforated interval 10,039.37-10,078.74(ft)

3060-3072(m)
Condensate flowrate 634(bpd)
Gas flowrate 11.9(mmscf/d)
Water flowrate 36 (bpd)
Choke size 32/64(In)
Oil 54.3(DAPI)
Gas 0.683(SG)
Water 5693(ppm NaCl equiv.)
Rw = 1 ohmm @ 77 degF (25 degC)

Exercise 2

Handling Coal
At approximately 9908.14 ft (3020 m), a washout zone is present and the logs
would tend to indicate that this may be due to the presence of a coaly horizon. As
this represents a significant change in lithology, we should attempt to use a
different model over this interval.

1. Select General > Evaluate to evaluate a log.

2. Call the log "coal" and set the values of one (1) in areas of coal and zero (0)
elsewhere by choosing an appropriate DT cutoff.

3. Create a coal model which will only consist of those components that we
should expect to find in a coaly horizon.

Geolog 6.6.1 - Multimin Tutorial Silt Example 66

 Exercise 3

Independent Analysis

1. Build a simple model using minerals quartz, illite and kaolinite with gas and
free water as fluids.

) As a TNPH log is available, you should select a non-linear response for

this measurement.

2. Run the analysis—run it first without the Coal model and then including the
coal model.

3. Include smectite in the model and rerun.

4. Which is more believable?

5. What about Rw? How well are the conductivity logs reconstructed?

Some Additional Information

Following the initial analysis of this well, some additional information has come
to light:
— Results from a regional DST of the target interval showed a water salinity
of approximately 20,000 ppm. Rw = 0.3 @ 77 degF (25 degC).
— A study of Oil-Water inclusions indicated that water salinity at the time of
gas migration could have been in the range of 50 - 80 kppm (see
Figure 21).

6. Run through these two extra scenarios.

7. Which one do you think is the most likely?

Geolog 6.6.1 - Multimin Tutorial Silt Example 67

 Having completed a Multimin analysis using the Dual Water model, it would now
be useful to try using some of the other conductivity equations.

8. Save the well.

9. Duplicate the WIRE set and rename the duplicated set to WIRE_SHALE.

10. Set the Default Set to WIRE_SHALE.

11. Open the multimin_recon layout and save it as multimin_recon_shale.

12. In the first wireline (Lithology) track, replace the following logs:
VOL_ILLITE with VOL_WCS1
VOL_KAOLINITE with VOL_WCS2

13. Create a model using the Indonesia conductivity model.

Remember that with this model, wet shales rather than clay minerals are
used. This will mean that you will have to define the best properties of the
shales to use. You can do this in a similar fashion to what you would for a
deterministic methodology. Note that it may be useful to further subdivide
the reservoir.

14. In this reservoir, there are probably two shale types—one in the upper section
and one in the lower section. First, run with a separate model in each, and
then run with one model containing two wet shales, and compare the results.
Which one do you think is better?

15. Now try changing the conductivity equation to Juhasz . . . what is the
difference?

See Figure 22 to Figure 24 for example plots showing possible shale

parameter picks.

Geolog 6.6.1 - Multimin Tutorial Silt Example 68

 Figure 21: Resistivity of NaCl solutions

Geolog 6.6.1 - Multimin Tutorial Silt Example 69

 φ N = 0.35
ρ b = 2.55

DT = 92
GR = 105

Figure 22: Parameter picks for shales in Interval A in Silt well

Geolog 6.6.1 - Multimin Tutorial Silt Example 70

 RT = 3

φ N = 0.32
ρ b = 2.605

Figure 23: Parameter picks for shales in intervals A and B in Silt well

Geolog 6.6.1 - Multimin Tutorial Silt Example 71

 DT = 86
GR = 115

RT = 4.5

Figure 24: Parameter Picks for Shales in Interval B in Silt Well

Geolog 6.6.1 - Multimin Tutorial Silt Example 72

 NIMBLE Example

Procedure
This section introduces the Numerical Inversion Modelling By Linear Equations (NIMBLE)
module (tp-multimin_inverse) which solves the inverse problem for a single model over an
extended depth range. The inverse, or calibration, problem is when response parameters for
single or multiple mineral and/or fluid volumes on single or multiple tools are calculated over
an interval. This technique might prove useful when, for example, volumes are well known
from some independent source (like petrography or XRD) but the tool responses are not.

You will:

• Reload and set up the data.

• Model a "pseudo-mineral".
• Optimize poorly known responses using the NIMBLE module.

Exercise 1

Before you begin:

1. Using Connect, delete your existing HALFWAY_TEST well and then reload it.

2. Run the Precalc module.

3. Open the model editor and open the MM_HALFWAY_ALT model.

4. Save the model as HALFWAY_NIM.

5. Run the model HALFWAY_NIM over the HALFWAY_TEST interval.

Geolog 6.6.1 - Multimin Tutorial NIMBLE Example 73

 Exercise 2

Modelling a "Pseudo-Mineral"
Often there will be heavy, possibly radioactive accessory minerals associated with
clays (or non-clay minerals). In some wells, there might be sufficient logs and
auxiliary data to be able to model these secondary minerals. However, there might
be other wells in the same field where less data is available and the extra volumes
required for the secondary mineral(s) could not be accommodated. NIMBLE could
possibly be used to good effect to model a composite "pseudo-mineral" which
would have some combined response of the clay(s) (or non-clay minerals) and the
accessory mineral(s).

1. Open the multimin_recon layout and check the existing analysis.

2. For the HALFWAY_TEST well, make a pseudo-mineral that will be composed

of quartz and dolomite.

3. Run General > Evaluate to create a VOL_NEW log that is the sum of the
VOL_QUARTZ and VOL_DOLOM logs.

4. Delete the original VOL_DOLOM and VOL_QUARTZ logs.

5. Rename VOL_NEW to, for example, VOL_QUARTZ.

Since quartz is the main mineral volume in this analysis, we would expect that
the parameters of the composite mineral would be between those of quartz
and dolomite but closer to quartz.

6. Edit the model that was used to produce the current evaluation to remove
VOL_DOLOM. Leave VOL_QUARTZ in, as this now represents the
composite mineral whose response parameters we wish to optimize.

7. To determine the optimum parameters for the "pseudo-mineral", select

Petrophysics > Multimin > Inverse Analysis (NIMBLE).

8. You will need to nominate the model that was used to produce the current
evaluation for parameter "MODEL_INVERSE" (i.e., HALFWAY_NIM).

9. Save the optimized parameters back to your model.

Geolog 6.6.1 - Multimin Tutorial NIMBLE Example 74

 10. Rerun Multimin with the updated model and examine the results. How has the
"pseudo-mineral" (or composite mineral) performed?

Exercise 3

Volumes Well Known But Some Parameters Poorly Known

If volumes are well-known from sources other than the wireline logs (e.g., from
XRD/IR data) then poorly known responses can be optimized using NIMBLE.

1. Use the XRD (or other data) in a Multimin model with normal uncertainties set
on their response equations.

2. Wireline and other equations that are not to be subjected to parameter

calibration should be included with their normal uncertainties at this time.

3. Response equations that will be subjected to parameter calibration must

either be omitted or have their uncertainties increased by a tenfold factor.

4. Optimize the model with normal Multimin passes, producing mineral volumes
in the wireline set.

5. Edit the model to remove the XRD equations and to include all other
response equations with their normal uncertainties.

6. Calibrate for desired response parameters using NIMBLE.

At the time of writing of this course, a suitable example of this application was not
available.

Geolog 6.6.1 - Multimin Tutorial NIMBLE Example 75

 Carbonate Example

Procedure
This is an example that is designed to examine how to model the effects of shallow invasion
on tool response, and how to model secondary porosity.

Exercise 1

Review of Available Well Information

1. Open the SECPHI well.

This well is a Middle East carbonate example. The upper levels were
described as predominantly dolomite with minor amounts of calcite and clay.
Lower levels became primarily calcite.
The intervals you want to analyze are HYD and WATER in the TOPS.TOPS
set.
The mineral model should include calcite, dolomite, illite and kaolinite. This is
a gas well.
The following parameter values were taken from the log header:

Values Values Values

Name
(Imperial) (Metric) (Mixed)
SHT 80 (degF) 26.7 (degC) 80 (degF)
BHT 167 (degF) 75 (degC) 167 (degF)
TD 5185 (feet) 1580.4 (m) 1580.4 (metres)
DFD 11.16 (lb/g) 1370 (K/m3) 11.16 (lb/g)
KCl 0 %
RMS 0.3 (ohmm) 0.3 (ohmm) 0.3 (ohmm)
MST 98 (degF) 36.7 (degC) 98 (degF)
RMFS 0.28 (ohmm) 0.28 (ohmm) 0.28 (ohmm)
MFST 84 (degF) 28.9 (degC) 84 (degF)
RMCS 0.45 (ohmm) 0.45 (ohmm) 0.45 (ohmm)
MCST 85 (degF) 29.4 (degC) 85 (degF)
BS 8.5 (ins) 215.9 (mm) 8.5 (ins)

Fluid properties: Rw 0.044 ohmm @ 75 degF (23.89 degC)

Geolog 6.6.1 - Multimin Tutorial Carbonate Example 76

 2. Run a Multimin analysis using the above simple model.
How do the results look?

What about lithology and porosity consistency?

3. Allow the core porosity to be presented. What has happened?

Independent Analysis

4. Review the section of the Geolog online help which covers modelling of vugs
and fractures, and how to model shallow invasion.
What is the most likely situation here?

Build models for each case and check the results against core again.

Geolog 6.6.1 - Multimin Tutorial Carbonate Example 77

 Appendix A - Multimin Aliases

Overview
This appendix is intended to review the philosophy and application of aliases within Multimin.
The concept of aliases is discussed, and how aliases are structured and used within a Multimin
framework.

Alias Concepts
Aliases allow the mapping of one or more alternative log mnemonics to a desired log
mnemonic. It is also possible to alias alternative constant names to a desired constant name.

Multimin and other Geolog programs use as inputs particular log or constant names that are
expected to be present in the database. If, on processing, these are absent from the database,
the program will use an alias list to find the next most desirable curve or constant. An example
of the practical use of aliases would be the choice that must be made in using a gamma ray
log. The usual gamma ray curve is GR (a simple total gamma ray log), but it would be
appropriate to use SGR (the spectral total gamma ray log), or CGR (the spectral computed
gamma ray log). To accomplish this, it would be necessary to alias first SGR, then CGR to the
desired GR.

The Geolog6 project specs directory contains the Paradigm supplied alias.alias file.
This file can be viewed at the Unix level or by using an NT text file editor. The user can make
changes or additions to this file but these changes should be stored in the specs directory
at site or project level (otherwise, new releases of Geolog will overwrite your changes). The
alias.alias files are cumulative—the file at the project level is read first, then the site,
then the Geolog6 project level. Only previously undefined aliases are added at each
consecutive level. It is, therefore, only necessary to include changes in your project or site
alias.alias files.

Multimin Alias Conventions

The conventions that are applied in using aliases in Multimin are generally self-explanatory
and best understood by studying the alias.alias file. A few points of interest:

• The alias list for GR allows you to name gamma ray curves with names that
reflect the tool that they were measured with or from different service
companies. Later aliases pick up curves measured with a spectral gamma ray
sonde.

Geolog 6.6.1 - Multimin Tutorial Appendix A - Multimin Aliases 78

 • NPHI has two aliases. The TNPH measurement reflects the advances in
processing neutron measurements made by Schlumberger. NPOR represents a
neutron curve of enhanced vertical resolution.

• RHO has the alias NRHO representing a bulk density curve of enhanced vertical
resolution.

• RT has aliases that represent the various resistivity and induction tools, and
enhanced resolution processing.

The alias.alias file shows an interpretation of the plethora of naming schemes available
for the calipers measured with various sondes. For practical purposes, Geolog only uses one
caliper for environmentally correcting porosity curves, and one for correcting resistivity curves.

It is assumed that depth-shifted/edited curves will have the higher version number, and these
will be chosen by Geolog unless an alternative choice is forced by the user.

Most layouts, etc. call for environmentally corrected curves. If the curve is missing, the
priorities for aliases are:

1. An uncorrected curve of the same name.

2. Corrected, then raw curves of a similar type.

The same priorities are assigned to predicted curves with the only difference being the
addition of the predicted suffix _PRED. Note the following exception to the general table:

• U_PRED does not include an environmental correction designation because it is

based in the curve U, calculated when running the Precalc or density
environmental correction modules.

Geolog 6.6.1 - Multimin Tutorial Appendix A - Multimin Aliases 79

 Appendix B - THOR Minerals from
Rock Element Catalogue

Geolog 6.6.1 - Multimin Tutorial Appendix B - THOR Minerals from Rock Element Catalogue 80
Geolog 6.6.1 - Multimin Tutorial Appendix B - THOR Minerals from Rock Element Catalogue 81
Geolog 6.6.1 - Multimin Tutorial Appendix B - THOR Minerals from Rock Element Catalogue 82
Geolog 6.6.1 - Multimin Tutorial Appendix B - THOR Minerals from Rock Element Catalogue 83
Geolog 6.6.1 - Multimin Tutorial Appendix B - THOR Minerals from Rock Element Catalogue 84
Geolog 6.6.1 - Multimin Tutorial Appendix B - THOR Minerals from Rock Element Catalogue 85