GeoNeurale Newsletter


                                                                  July-August 2008


1.  GeoNeurale Courses Program  2008                                

2.  Two for One    (by Gene Ballay)

3.  Bibliograpy

4.  The Compact Theory

5.  Determinismus and / or Artificial Intelligence for Pattern Recognition?

     Artificial Intelligence Methods for Pattern Completion and Interpretation,

     AI Methods for Applications in the field of Geosciences 

     (by Hansruedi Frueh)

6.  Carbonate Petrophysics Consulting

7.  References and Links


1-5 September  2008        Applied Carbonate Stratigraphy                                                                                   PROGRAM

                                          Instructor:  Franz Meyer

9-11 September  2008     The Logic of Neural Networks for the Petrophysical, Seismic and Facies Estimation     PROGRAM

                                         Instructors:  Hansruedi Frueh  & Tino Perucchi                   


6-10 October 2008            Geostatistical Applications and Petrophysical Analysis                                                 PROGRAM

                                          Instructors:  Gene Ballay  & Jacques Deraisme

17-21 November 2008       Geostatistical Modeling for Petroleum Reservoir Characterization                               PROGRAM

                                          Instructors:  Hans Wackernagel &   Olivier Jaquet

24-26 November  2008       Advanced Carbonate Petrophysics                                                                              PROGRAM

                                            Instructor:  Gene Ballay 

8-12 December 2008          The Fundamentals of Upstream Petroleum Economics and Risk Analysis                   PROGRAM

                                           Instructor:  Tim James

October - November  2008   3D Seismic Attributes for Prospect Identification

                                             Kurt Marfurt  ( SEG )

October - November  2008   Seismic Fluid Detection, Reservoir Delineation and Recovery Monitoring:
                                            The Rock Physics Basis

                                             Gary Mavko  ( SEG )

Actual Courses

GeoNeurale Courses Schedule


2 for 1 or 1 + 1 = 3

by  R.E. (Gene)  Ballay  PhD


A routine suite of open hole logs, that includes both porosity and resistivity, has the potential to provide
not one, but two, independent evaluations of the formation.

When a full suite of OH logs are available, one typically estimates Sw according to

                                    Sw n =  a R w / ( Phi m R t)

In a newly drilled area  “ a , Rw and  m "  may not be well known.
And if pore geometry is possibly changing (such as in a carbonate, vuggy  <-> intergranular / intercrystalline),
the associated variable “m" presents an additional challenge in even well-drilled fields.

If a water leg is present, the so-called resistivity ratio can yield (first) an estimate of Rw @ FT, followed
by (second) Sw, higher in the column, both independent of a porosity log and / or Archie’s cementation

                                      Rw = Rmf * ( Rdeep / Rxo )

                              Sw8/5 = ( Rw / Rmf ) * ( Rxo / Rdeep )

The formulation (the 8/5 ratio) can be calibrated to locally specific behavior, yielding a yet better estimate.

Quick Look interpretations, in general, serve a number of useful purposes:

QL methods nicely illustrate fundamental petrophysical relations

- Executing QL analyses will reinforce the basics

QL results provide integrated QC on newly acquired data

- Do the composite results ‘make sense ‘

QL provides benchmark against more sophisticated results

- Is computer model properly specified ?

- Particularly important with probabilistic models

The resistivity ratio estimate of Sw is easily coded into any s/w package, and is a routine part of my
evaluations whenever an Rxo tool is available.

                                   The Resistivity Ratio

Archie’s equation applies in both the invaded and non-invaded, zones.

                           Swn = Rw / (Phi m * Rdeep)

                          Sxon = Rmf / (Phi m * Rxo)

If these two equations are ratio'ed, one has

                Swn / Sxon = ( Rw / Rmf )* ( Rxo / Rdeep )

n = 2  and   
Sxo = Sw1/5    then yields

      Sw2 / Sxo2 = [ Sw / Sw1/5 ] 2 = [ Sw4/5 ] 2 = Sw8/5 = ( Rw / Rmf ) * ( Rxo / Rdeep )

While the “n" ~ 2.0 is obvious, the
Sxo = Sw1/5   assumption may raise an eyebrow, but is in fact a surprisingly
robust assumption.

This saturation ratio assumes (for example) that if Sw ~ 20 %, then Sxo ~ 70 %, and while local rock may vary
from this somewhat, it is often at least a “get started" estimate : Figure 1



As experience in a particular area is gained, a spreadsheet may be used to “adjust" the exponent to match local rock behavior : Figure 2


 Key assumptions in the use of the resistivity ratio, in addition to the exponent ratio itself,
are as follows.

.      Formation generally clay-free

o      Archie’s equation is only applicable to ‘clean’ formations

o      As formation brine becomes saltier, the clay conductivity issue diminishes in importance,
       and in the salty waters of the Middle East for example, Sw(Ratio) ~ Sw(Archie) in even
       relatively shaly formation.

·        Significant, but not excessive, invasion has occurred

o      Rxo &  Rdeep must represent the invaded & non-invaded, rock; ie correspond to Rmf and Rw, respectively.

o      This constraint may in practice, be more or less satisfied, up and down the formation.


                                                             Application Examples


The first application example is drawn from Schlumberger’s 1975 Arabian Well Evaluation Conference : Figure 3



In this particular well, there are both water and hydrocarbon intervals. Looking to the water leg,
one deduces the Rmf / Rw  ratio from the measured  Rxo / Rdeep responses :
Rmf / Rw  ~ Rxo / Rdeep  ~ (1.0 / 0.23 ) ~ 4.35

Once the Rmf / Rw  ratio is established, one moves up the formation, estimating Sw from the measured
Rxo / Rdeep ratio.

                Sw8/5 = ( Rw / Rmf ) * ( Rxo / Rdeep ) = (1/4.35)*(1.2 /1.5 ) = 0.184

                Sw = 0.35

The second application example was found via Google, posted to the www but without Author / Publication
details, and so we are unable to fully credit this nice article.

                                    Kansas City – Lansing, Anadarko Basin

                      Take “n" = 2, and define the Moveable Hydrocarbon Index as :

                            Sw / Sxo = [( Rw / Rmf ) * ( Rxo / Rdeep  )] ^ (1/2)

Schlumberger (1972) guidelines are that if the ratio of Sw / Sxo > 1.0, then no hydrocarbons
were moved during invasion. This is true regardless of whether the zone contains hydrocarbons.

Whenever Sw / Sxo < 0.7 for sandstones or Sw / Sxo < 0.6 for limestones, then moveable hydrocarbons
are indicated.

If a carbonate reservoir has a Moveable Hydrocarbon Index < 0.6, you can conclude that
hydrocarbons are present (although not necessarily in commercial quantities), and that the
reservoir has enough permeability so that hydrocarbons have been moved during the invasion
process by mud filtrate.

The ratio water saturation is calculated with the assumption : 
Sxo = Sw1/5

       Sw2 / Sxo2 = [ Sw / Sw1/5 ] 2 = [ Sw4/5 ] 2 = Sw8/5 = ( Rw / Rmf )* ( Rxo / Rdeep )

                              Sw (Ratio) = [(Rw/Rmf ) * (Rxo/Rdeep )] 5/8 

Sw(Archie) is calculated independently per the routine Archie equation

If Sw(Archie) ~ Sw (Ratio) the assumption of a step-contact invasion profile is correct and
all calculated values:

Sw, Rt, Rxo, di

are correct (reasonable).

If Sw(Archie) .NE. Sw (Ratio), additional considerations are necessary.

Consider the interval displayed in Figure 4. This zone is likely wet, even though Sw(Archie, m=2)
is not excessively high.





 Porosity from the density-neutron is 25 pu. With an Rt of 3.5 ohm-m, the Archie estimate of Sw is 38 %
 (Figure 5),  lower than the Sw (ResRatio) of 53 %




The Moveable Hydrocarbon Index is slightly greater than the carbonate 0.60 limit and the Bulk Volume Water is considerably larger than common carbonate values ( Figure 6 ) for either vuggy or intergranular / intercrystalline pore systems. Note that were BVW to be calculated with Sw(RR), it would be yet larger.


As data and experience build, in a particular province, the boundary values may be adjusted, but
lacking that detail, one may consult the Kansas Geological Survey site for this information in a
variety of formats (by reservoir, etc).

Petrography rounds out the evaluation, revealing the pore system is oomoldic : “vuggy" systems
require “m" larger than two for proper evaluation.

In essence, at a specific porosity, and for a fixed resistivity, the vuggy system will require more
(conductive phase) water present because the pore system is more tortuous. The resistivity ratio
estimate, which is porosity and “m" independent, correctly identifies the high Sw.

The author of this article (Combining Water Saturation, Moveable Hydrocarbon ….) points out that a
similar evaluation approach is described in the literature for the oomoldic Smackover reservoirs,
with the following comments.

These reservoirs are similar to the oomoldic - 4,810 feet - zone in the Lansing-Kansas City, and 
Mitchell-Tapping (1983) conclude that the Smackover cannot be properly evaluated by the standard
Archie technique. When Moveable Hydrocarbon Index, water saturation Ratio Method, and bulk volume
water are used, then correct judgements about the productive potential can be made.

The Smackover can, in fact, present a different signature, as related in GCAGS Transactions
Volume 29 (1979).

Predicting the production from wells drilled in certain Smackover reservoirs is often difficult.
Production history from the field, core analysis and log data have not always proved to be helpful.

The Smackover reservoirs for which log interpretation is difficult fall into two categories.

The first is oolitic limestone characterized by low resistivity, moderate porosity (10-20%) and
permeability (100-500 md). High water saturation calculated from logs does not necessarily
preclude hydrocarbon production.

The second is oomoldic limestone typically having high resistivity, high porosity (20-35%) and
low permeability (< 100 md). Although log interpretation indicates low water saturation, no
hydrocarbons are produced.

An example of these ideas, with locally specific boundary values, is found in: G.M. Hamada.
Hydrocarbon Moveability Factor: New Approach to Identify Hydrocarbon Moveability and Type from
Resistivity Logs. Emirates Journal for Engineering Research, 9 (1), 1-7 (2004)

Always remember the underlying assumptions, and importance of locally specific guidelines

                                                   1 + 1 = 3

A complete suite of OH logs, to include both resistivity and porosity, has the potential to offer
two independent evaluations, 1 + 1 = 2, each with strengths and weaknesses.

When  faced with a variable pore system, the resistivity ratio can possibly identify the transition
from intergranular / intercrystalline to vuggy porosity, and thereby flag when an alternative “m" ,
in Archie’s equation, is appropriate.

Each time I use this approach, fond memories come to mind of my friend Lou McPherson, who ran Shell’s
Bellaire Petrophysics Training in the 1970s and gave me this idea.

                             1 [Sw(Archie)] + 1 [Sw(RR)] + 1 [Memories] = 3


Combining Water Saturation by Ratio Method, Moveable Hydrocarbon Index, Bulk Volume Water and Archie Water Saturation. Found with Google. Author, date and publication details n/a.

Hamada, G. M. Hydrocarbon Moveability Factor: New Approach to Identify Hydrocarbon Moveability and
                       Type from Resistivity Logs. Emirates Journal for Engineering Research, 9 (1), 1-7 (2004)

Gulf Coast Association of Geological Societies Transactions Volume 29 (1979)
Available from

Focke, J. W. and D. Mun. Cementation Exponents in ME Carbonate Reservoirs. 
                                          SPE Formation Evaluation,   June 1987

Wyllie, M. R. & A R Gregory: Formation Factors of Unconsolidated Porous Media: Influence of Particle
                                           Shape and Effect of Cementation, Petroleum Transactions of the AIME 198
                                           (1953):  103-110.

Borai, A. M. A New Correlation for the Cementation Factor in Low-Porosity Carbonates, SPE Formation
                    Evaluation 2 (1987): 495-499

Brie, A. & D.L. Johnson & R.D. Nurmi. Effect of Spherical Pores on Sonic and Resistivity Measurements.
                                                              SPWLA 26th Annual Logging Symposium. June 17-20, 1985

Watfa, M. Seeking the Saturation Solution. Middle East Well Evaluation Review No. 3 (1987)

Schlumberger Technical Review, Volume 36 Number 3

Lucia, F. Jerry. Carbonate Reservoir Characterization. Published by Springer, 1999

Mazzullo, S. J. Overview of Porosity Evolution in Carbonate Reservoirs

Bureau of Economic Geology, University of Texas

Ross Crain's On-line Tutorial

Kansas Geological Survey (John Doveton) Tutorial

United States Geological Survey Rock Catalogue

Kansas Geological Survey Gemini Rock Catalogue

Kansas Geological Survey Abyss Rock Catalogue




Determinismus and/or Artificial Intelligence for Pattern Recognition?

Artificial Intelligence Methods for Pattern Completion and Interpretation,

AI Methods for Applications in the field of Geosciences  ( Hansruedi Frueh )



This is an interview with  Dr. Hansruedi Frueh, the founder of the company

Neuronics AG ( in Zurich and  instructor for the course:

“ The Logic of Neural Networks for the Petrophysical, Seismic and Facies Estimation “ 

at GeoNeurale in Munich.


1)    GeoNeurale –

           What is the definition of "Artificial Intelligence" and which
           are its most direct technological applications ?

Hansruedi Frueh - 

Artificial Intelligence (AI) has many definitions, depending on the
 viewpoint. There is a consense, however, that solutions which include
learning, reasoning based on optimisation algorithms, data mining,
language understanding and deduction under uncertainty are located in
the domain of AI. Soft computing technologies, such as neural networks,
evolutionary algorithms, and fuzzy logic belong to it as well as
Bayesian networks (based on probability models) and hard computing
learing principles (e.g. reinforcement learning, support vector
machines, hidden Markov models etc.).
Technical applications which include AI are in use everyday everywhere.
Whether it is the coffee machine, the car, or the search engine in the
internet. And of course in robotics, a field which I work on since more
than a decade.
In the deep subsurface exploration, it also will cover an interesting potential
for the interpretation and spatial distribution analysis of petrophysical and
seismic attributes.

2)   GeoNeurale –

     To what extent do your robots take advantage of Neural Networks ?

Hansruedi Frueh - 

We are using neural networks preferrably when sensor values from
different modalities have to be integrated and related to a desired
behavior. Such a "learning machine" which performs sensor data fusion
and adaptive categorization of the input is exactly what is needed for
many applications, especially in mobile robotics. But we also use
Kohonen networks (a non supervised type of neural networks) to improve
the precision of the robot arm when it has to pick up objects from a
grid. In that case, not every point of the grid is teached in
separately, but only the corners. After an approximative calculation of
all points based on direct and inversed kinematics, there remain small
errors which have different causes (gravity, mechanical deviations,
unperfect models etc.). The integrated Kohonen network allows to learn
how to correct all remaining points from a few examples.

3)  GeoNeurale –

You are a pioneer in the field of applied Artificial Intelligence in Switzerland,
for applications in Robotics and Mechatronics.
The robot measures features of the real world  through its sensors. And performs
a topological representation based on the measurements.
In this worklflow of processing and interpretation, Neural Networks
play a very important role.

The Geoscientist measures the subsurface parameters through indirect and direct
methods, trying to represent the reality of an external world, the subsurface,
which can be measured only with spatial discontinuity and physical approximation.
3D Seismics and Petrophysics are still difficult to be correlated, petrophysical
properties are difficult to be realistically upscaled.
The Geoscientist is therefore continuously confronted with interpolation
issues at the search of the right analyitical methods for the topological laws
to be applied on the property space and on the geological / structural space.
Do you see a parallelism between the Robot´s Logic and the Geosciences
in their measurements and interpretation workflow ?

Hansruedi Frueh - 

I see the parallelism in the sensorics which is important for both
fields (although quite different sensors) as well as in the uncertainty
under which many decisions have to be made. Both need intelligent
algorithms in order to optimize behavior of the mobile robot or
learnings, respectively.

4) GeoNeurale –

To what extent do you think  Neural Networks applications can
help in the property estimation and classification in the Geosciences
Applications referred to the Subsurface Analysis ?
Could you give some examples ?

Hansruedi Frueh –

Neural networks can help to learn nonlinear data efficiently, if the
appropriate network type as well as a fine-tuned learning procedure is
applied. As mentioned above, neural networks are a good choice for tasks
which include sensor data fusion and categorization. Also, neural
networks can reduce the dimensions of data and provide models which can
be used, e.g., for pattern completion.

A typical example is parameter estimation for controllers or for errors
(as in the precision improvement example mentioned above).
Classifications are used in nearly all fields. Neural networks are often
combined with other methods, e.g. fuzzy logic or adaptive probability
(Bayesian) networks which can learn in terms of structures as well as by
modification of the events' probablities.

5) GeoNeurale –

We think now about specific petrophysical properties and log analyses.
A deterministic approach could  involve the application of the Archie´s
Equation or a modified version of it. If this doesn´t seem to be fully
applicable (as the measured logs show a deterministic incongruence but
at the same time we observe a statistical repeating pattern),
what are the main (AI) applications that we could use to perform such a
log analysis ?
Hansruedi Frueh –

In my view, questions as you put them are a matter of an
interdisciplinary approach. Neural networks can serve as "learning
machines" to relate input vectors to the most appropriate category which
may be a member of a model created by the network itself or by another
methodology. Alternatively, the correct approach may be a combination of
two different neural network types (e.g., Backpropagation and Kohonen).
It is for sure a very interesting topic, and the potential gain from a
solution based on Artificial Intelligence seems to be high.

                                  CARBONATE PETROPHYSICS CONSULTING


                                           Integrated Consulting Services

                                          Carbonates Petrophysical Analysis,
                                          Log Interpretation

                                          RESEARCH & OPERATIONS

                                          Light / Heavy Oil / Gas / Geothermal
                                          Single / Multi-Well
                                          Vertical and Horizontal Orientations

                                         SPECIAL APPLICATIONS

                                         - Carbonate Petrophysics

                                         - Petrophysics-Geostatistical Applications

                                         - Imaging interpretation techniques

                                         - Petrophysical Input to the Static Geological Modeling




Predicting Reservoir System Quality and Performance.
Dan J. Hartmann and Edward A. Beaumont

Public Domain Databases
National Energy Technology Laboratory Public Database

United States Geological Survey Rock Catalogue
Kansas Geological Survey Gemini Rock Catalogue

Kansas Geological Survey Abyss Rock Catalogue

Reference Material

Bureau of Economic Geology, University of Texas

Ross Crain's On-line Tutorial

Kansas Geological Survey (John Doveton) Tutorial

Schlumberger (OilField Review, ME Well Review, Oil Field Glossary)

Baker Hughes (InDepth Magazine, Oil Field Glossary)

Interactive Periodic Table


  GeoNeurale Contact


     Lichtenbergstrasse 8

    85748  Munich-Garching


    T +49 ( 0 ) 89 5484  0

    T +49 ( 0 ) 89 5484  2240

    T +49 ( 0 ) 89 89 69 1118

    F +49 ( 0 ) 89 89 69 1117