5 - Spe 153004 MS PDF

SPE 153004
Unconventional Reservoirs: Basic Petrophysical Concepts for Shale Gas

Juan C. Glorioso, Aquiles Rattia, Repsol
Copyright 2012, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE/EAGE European Unconventional Resources Conference and Exhibition held in Vienna, Austria, 20-22 March 2012.
This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been reviewed
by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or
members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is
restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.
Abstract
Unconventional reservoirs have burst with considerable force in oil and gas production worldwide. Shale Gas is one of
them, with intense activity taking place in regions like North America. To achieve commercial production, these reservoirs should
be stimulated through massive hydraulic fracturing and, frequently, through horizontal wells as a mean to enhance productivity.
In sedimentary terms, shales are fine-grained clastics rocks formed by consolidation of silts and clays. In log
interpretation of conventional reservoirs, it is very common to observe that the clay parameters used to correct porosity and
resistivity logs for clay effects are in fact read in shaly intervals rather than in pure clay. Although no considerable deviation have
been observed in shaly sandstones, anyway these concepts and procedures must be reviewed to run log analysis in shale gas.
Organic matter deposited with shales containing kerogen that matured as a result of overburden pressure and temperature,
giving rise to source rocks that have yielded and expulsed hydrocarbons. Shale gas reservoir type is a source rock that has retained
a portion of the hydrocarbon yielded during its geological history so that to evaluate the current hydrocarbon storage and
production potential it is necessary to know the kerogen type and the level of TOC - total organic carbon - in the rock. Produced
gas comes from both adsorbed gas in the organic matter and "free" gas trapped in the pores of the organic matter and in the
inorganic portions of the matrix, i.e. quartz, calcite, dolomite.
In these unconventional reservoirs, gas volumes are estimated through a combination of geochemical analysis and log
interpretation techniques. TOC, desorbed total gas content, adsorption isotherms, and kerogen maturity among other things can be
measured in cores, sidewall samples and cuttings, in the laboratory. These data are used to estimate total desorbed gas content and
adsorbed gas content which is part of the total gas. Also in laboratory, porosity, grain density, water saturation, permeability,
mineral composition and elastic modules of the rock are measured. Laboratory measurement uncertainty is high and consistency
between different providers appears to be low, with serious suspicions that procedures followed by different laboratories are the
source of such differences. The permeability is one of the most important parameters, but at the same time, one of the most
difficult to measure reliably in a shale gas. Core calibrated porosity, mineral composition, water saturation and elastic modules can
be obtained through electric and radioactive logs. All these information is used to estimate log derived total gas volume which
results are also subject to a high degree of uncertainty that must be overcome.
Once this key information is obtained, it is possible to estimate different gas in-situ volumes. Indeed, an estimate of
porosity-resistivity based total gas in-situ and, on the other hand, geochemical based adsorbed gas in-situ can be performed. Log
total gas in-situ can be, and it is advisable to do, compared with adsorbed gas estimations and also with another gas measurement
called direct method - total gas desorption performed on formation samples. The difference between log total gas in-situ and
adsorbed gas in situ should be the "free" gas in situ. Free gas occupies the pores of kerogen and matrix; also it can be stored in
open natural fractures if such fractures are present.
The main objective of this paper is to discuss the state-of-the-art in petrophysical evaluation of shale gas reservoirs, to
summarize the experiences of operators and researchers, and to bring some views on the criteria and techniques for the evaluation
of cores and logs. An inventory of laboratory tests and results, log responses in the presence of kerogen, log interpretation
techniques and estimation methods for different volumes of gas in-situ, together with important aspects of the use of analogy in
SPE 153004
shale gas reservoirs has been done. At the end, a basic petrophysical workflow is outlined for the volumetric determination of gas
in situ.
Shale
The term shale suggests a laminar and fissile structure present in certain rocks (Fig.1 and 2). On the other hand, it is
also used to refer to fine-grained, detrital rocks, composed of silts and clays. Mudrocks are also considered shale and, hereafter,
will indiscriminately be referred to as silty argillites or shale. They are fine-grained rocks, although they can also be mud masses or
mudstones, which do not have any fissile properties.
In petrophysical terminology, shale has been generically used to distinguish medium to coarse clastic components like
sand and conglomerates from finer elements like silts and clays that often make up the matrix of rocks, independent of its texture;
mudstones that form thin sheets and interstratified with sandstones are also considered shale, as are those that are present in the
nodules that form part of the rock matrix and shale within interstitial pore space. The shale volume (%) estimated with logs has,
and is currently, often used to calculate shale corrected log porosity, pretending that it represents the porosity free of clays effects,
paradoxically, without using the volume (%) of pure clay in the calculation. The main reason for this is the insufficient application
of techniques that use Spontaneous Potential, Gamma Ray and Density-Neutron logs amongst others that has resulted in obtaining
the volume (%) of clay + silt (shale) prior to that of pure clay. In brief, the main cause of this deficiency is the choice of apparent
pure clay values in formation intervals that are, in fact, shale; as a result, those values are not representative of pure clay. Thus, it is
common to observe that the volume deducted from total log porosity is that of the log porosity of the shale components. This
approximation has been useful when evaluating conventional reservoirs, and it has not created a level of considerable uncertainty.
However, it is obvious that for unconventional reservoirs, like Shale Gas, this simplification must be revised.
Figure1: Evidence of fissility in shale outcrops (Questerre Energy 2009 & Repsol YPF Vaca Muerta Fm).
Figure 2: Core in laminated shale (left:

Passey, SPWLA 2010, right: Repsol
YPF shale core). Massive mudstones
rich in organic matter can also contain
and produce gas.
Another aspect that must be revised, and will be discussed in this paper, is that rock composition in shale gas reservoirs is
highly variable, containing clay and carbonate or quartzitic silts, in which often the % of clay is very low.
Geological Environment
Sediments that give rise to unconventional shale gas reservoirs generally have either marine or lacustrine origins. The
analyses performed in this paper are based on South America and USA cases and they involve moderate-to-deep marine sediments,
fine grained detritus that have been mixed with organic matter and concurrently deposited in anaerobic environments. However,
organic matter can be found in both distal and proximal sediments and in different sedimentary systems. Currently, it is possible to
find regions and vast lands with thick silt-rich clay formations of different hydrocarbon generation potential. These have
historically been known as "source rocks", as they have been responsible for generation and expulsion of hydrocarbons towards
more porous and permeable strata by different courses and mechanisms. Recently, however, it has been showed that they can
produce "in-situ" hydrocarbons commercially; or, that they can simultaneously serve as source, reservoir, and seal rocks. They are
illite-rich, and often contain chlorite. Mica is often present in these formations. Non-clay-rich fine detrital fractions can be, in some
cases, formed by predominantly quartz-rich silts containing other components such as feldspars, plagioclases and diagenetic
products like silica, added to dense minerals like pyrite and siderite. In other instances, in the non-clay-rich fractions, fine-grained
carbonates, calcites and dolomites are predominant. Establishing the predominant component is essential in designing well
stimulation. In the Neuqun Basin, Argentina, thick clay-rich bituminous formations are called "margas bituminosas" (bituminous
SPE 153004
marls) and react to chlorhydric acid given their high lime content. Other formations that generate hydrocarbons due to their organic
matter content, like carbonates, will not be discussed in this work.
After overburden maturity takes place controlled by pressure and heat to which the organic matter rich-strata are
subjected. It is during this stage that kerogen and bitumen are produced and hydrocarbons are generated. The origins of organic
matter mark the different types of kerogen found in the rock, as will be discussed later in the report. Later, geologic history has
provided the current form to reservoirs that, by definition, are not affected by hydrodynamic forces and which reservoir limits are
often difficult to establish prior to developing the field. The most favourable and productive areas are difficult to detect before
drilling and, once the process has begun, their production potential is established mainly through long-term production or well
testing.
Petrophysical Model
Below, a petrophysical model (Fig.3) that would serve as a basis for resolving the total gas content calculation, either by
core analysis measurements or, more indirectly, by interpreting electric and radioactive logs is presented.
Conceptually, rock's matrix is comprised by inorganic detritus and kerogen. These formations, that are similar in
appearance, tend to be unique and lithologically complex, containing clay, quartz, carbonate, and feldspar minerals, in addition to
other dense minerals. Pyrite, when present, can affect electric and radioactive logs. Furthermore, in reduced environments the
presence of uranium raise Gamma Ray (GR) log readings above normal levels, in addition to restricting the ability to estimate clay
volume with a total GR curve. The matrix may be clay-rich or contain less than 50% clay with quartz-rich silts in some cases, or
carbonate minerals in others. When their carbonate and quartz content is high, rocks are often more brittle; it has been reported
(Rickman et al. 2008; Buller 2010) that this can positively impact the outcome of hydraulic fractures.
Matrix
KEROGEN
INORGANIC MATRIX
(Dry Clay + Nonclay
Minerals)
Fluid
Free (!kFF)
Adsorbed (!kad)
Gas
Free (!FF)
Gas
Pc bound (!PcBW)
Water
Clay bound (!CLBW) Water
Figure 3: Schematic Petrophysical model of the rock without natural fractures. The term
"inorganic matrix" is used to differentiate kerogen from other non-organic components of the
matrix.
Figure 4: SEM showing pores in organic matter.

This can be considered secondary porosity as it
results after kerogen maturity and the consequent
expulsion of hydrocarbons (Reed, R, BEG
2008).
Kerogen forms parts of the matrix (Fig.4); the main features for consideration in petrophysical evaluations is its low
density, close to that of water, long travel-time, high neutron porosity and high resistivity; for these reasons it is essential to
establish the kerogen content in weight and volume and its properties, although it is somewhat difficult to do so. Porosity can be
found in both inorganic fraction and kerogen and pore sizes tend to be small; either micropores or nanopores can be found in these
kinds of rocks. Historically, in conventional reservoirs it was considered that such pore sizes contained only immobile fluids,
namely water; commonly, water adsorbed by clay fractions has been called clay bound water (CLBW) and different abbreviations
have been used to refer to the irreducible water in the inorganic matrix. PcBW will be used here for the capillary effects in the
inorganic matrix. This can change in instances where the rock is of mixed wettability or oil wet. Initially, the absence of natural
fractures, a water wet inorganic component and kerogen mature and porous will be assumed for this analysis. Kerogen is
considered hydrophobic, however, there is scarce information regarding the effects that this may have on fluid distribution. Rocks
containing natural fractures will be discussed later, although the effect of natural open fractures on storage capacity is considered
relatively low, in contrast to its significant influence on the evaluation of system permeability or the orientation of horizontal wells
and hydraulic fractures design. Matrix permeability is so low that is very difficult to accurately measure it in formation core
samples.
SPE 153004
Geochemical Measurements
Kerogen
Rocks present in shale gas reservoirs can generate and store hydrocarbons as they contain kerogen. Kerogen is a mixture
of organic compounds that make up the organic matter of some sedimentary rocks and is insoluble in normal organic solvents
given the massive molecular weight of its complex compounds. Soluble elements are known as "bitumen". When subjected to
intense heat, some types of kerogen release crude oil or natural gas, together known as fossil fuels. High concentrations of
sufficiently mature kerogen may turn the rocks into source rocks. Kerogen rich shales with insufficient grade of thermal maturity
to release hydrocarbons can give rise to superficial deposits called Oil Shales. To extract oil from oil shales, the rock must be
subjected to high temperatures and special processes. To establish the type of kerogen present in the rock and its hydrocarbon
generation potential, laboratory analysis of present formation samples is required; this process focuses on methods of combustion,
pyrolysis, chemical degradation, spectral fluorescence or petrography. Details regarding these laboratory techniques can be found
in many geochemistry bibliographies. Based on this sort of analyses, the type of kerogen, its thermal maturity and the TOC (Total
Organic Carbon) content of a rock can be determined. These are all necessary elements in assessing unconventional reservoirs,
such as shale gas reservoirs.
In geochemistry, kerogen is divided into four main types:
o
o
o
o
Type I: rich in algal and amorphous organic matter

Type II: rich in algae and herbaceous matter (a variant containing levels of sulphur also exists)
Type III: rich in wood and humic matter
Type IV: made up of decomposed organic matter
It has been established that during thermogenic maturity of type I kerogen, liquid hydrocarbons tend to be generated
(Appendix I), type II generates gas and oil, type III generates gas, coal (often CBM, or coal bed methane) and, in extreme
conditions, oil. It is generally considered that type IV kerogen is not capable of generating hydrocarbons. Shale gas is generated by
type III and type II kerogen in post maturity stages. These are the general characteristics of kerogen that give rise to thermogenic
shale gas.
Figure 5: Carbon, hydrogen

and oxygen ratios obtained
during a geochemical
analysis; this assists in
establishing kerogen types
(modified Van Krevelen).
There are also gas accumulations of biogenic origin1 in which organic matter has not been subjected to geothermal
gradients strong enough to generate hydrocarbons; however, there has been sufficient bacterial action to generate biogenic gas that
has been adsorbed by organic matter. TOC levels are high in biogenic shale and relatively low in thermogenic shale (>2 w %) in
which the majority of kerogen has been converted to hydrocarbons.
The adapted van Krevelen diagram (Fig.5) demonstrates the relationship between the hydrogen index (HI) and the oxygen
index (OI); this highlights the different types of kerogen and their respective maturity products and windows. The general trend in
the thermal transformation of kerogen to hydrocarbon is generally characterised by generation of non-hydrocarbon gases, and the
SPE 153004
progresses to oil, wet gas and dry gas. During this progression, the kerogen loses oxygen primarily as it gives off CO2 y H2O;
later, it begins to lose more hydrogen as it yields hydrocarbons2.
Kerogen/TOC
In order to establish the source potential of rocks, it is essential to know their original kerogen (Kero) content; on the
other hand, in petrophysical assessments of shale gas establishing the present volume of kerogen (Kerpd) is necessary in order to
estimate adsorbed gas content and porosity with reasonable accuracy. The volume percentage of kerogen is fundamental in log
interpretation as far as they are performed based on volume. However, establishing the present volume of kerogen by log analysis
is not a simple task, and it currently requires the use of both conventional logging and geochemical laboratory techniques, the latter
used to calibrate log derived kerogen content as will be discussed later. From kerogen volume, it is possible to derive the TOC.
TOC, or total organic carbon, is the amount of carbon bound in organic compounds of the rock; it is the remnant of
organic life preserved in sedimentary rocks subjected to chemical and bacterial degradation that have later been modified by heat
and pressure over time. This final stage, maturity, is a result of long periods of burial and proximity to heat sources. Chemical
reactions resulting from maturity of organic matter are responsible for producing gas, oil, bitumen, pyrobitumen and coal that
together contribute to total carbon content. Frequently the terms TOC and kerogen are used interchangeably, and it can be difficult
to differentiate through laboratory analyses or log interpretation kerogen from ordinary and soluble hydrocarbons such as oil,
bitumen or pyrobitumen that, as discussed above, also form part of TOC. In other words, if kerogen matures and produces gas and
oil, the amount of kerogen will decrease, whereas the TOC in the rock remains constant until, for example, the hydrocarbons are
expelled to other reservoirs.
TOC has primarily been used to evaluate the quality of source rock, although it is now also used in the evaluation of
unconventional reservoirs. Generally, TOC and kerogen are closely associated with shales and silt-rich claystone; however, they
can also be present in many carbonates.
TOC is fundamental in estimating adsorbed gas and is a property of the rock. Fig.6 shows average values reported in
some areas of active shale gas reservoirs. Certain elements present in kerogen, like hydrogen, oxygen, nitrogen and sulphur are not
included in the TOC so that conversion factor must exist between the two (kerogen and TOC) that accounts for both the type of
kerogen and its maturity stage. Some authors suggest that a value of 1.2 is sufficient when expressing TOC as a percentage of
weight (w %)3.
Marcellus
Barnett
Haynesville
Fayetteville
Vaca Muerta
Argillaceous
mudstone
Siliceous
mudstone
Argillaceous
calcareous
mudstone
Siliceous
mudstone
TOC avg: 6%
TOC avg: 5%
TOC avg: 4%
TOC avg: 4%
TOC avg:
2.5 - 3.5%
!avg: 6%
!avg: 5%
!avg: 10%
!avg: 6.5%
!avg: 8%
Argillaceous
siliceous &
calcareous
mudstone
Figure 6: Weight content (w %) of TOC of currently active shale gas reservoirs (composite from Lyle 2009 & Repsol YPF proprietary
data).
That is to say, if the kerogen volume can be estimated, the TOC content can be established by using a formula like the one
presented below. For that reason, it is also necessary to ascertain kerogen density and rock density, the former being difficult to
establish.
TOC =
Vol ker ! " ker

"b ! k
Where:
TOC
Volker
!ker
= total organic carbon (lbf/lbf)

= kerogen volume (vol/vol)
=kerogen density (g/cc)
(1)
!b
k
SPE 153004
=formation density (g/cc)

=conversion factor (~1.2)
Some authors also estimate that, inversely, the direct relationship between kerogen and TOC may be between 2 and 2.54.
The two following approaches can be used to estimate kerogen (vol%):
Ker (vol %) = TOC ( wt %) ! f
"
f # 2 - 2.5
(2)
' (1 ) ( )! * ma $
Ker(vol %) = %
" ! TOC ( wt %)
& (* ker ! Cker ) #
(3)
Where:
$
!ma
!ker
Cker
= porosity
= matrix density
= kerogen density
= w% of carbon in kerogen
It is often assumed that the product !ker * Cker=1

To evaluate the current adsorbed gas content of the rock, it is essential to first establish the current TOC content (TOCpd).
This can be obtained from the % of kerogen, as outlined above, or can be measured in laboratory by employing the same
measuring techniques employed in geochemistry (Fig.7). Given that samples available for analysis are limited, many authors have
studied the correlation between TOC measured in laboratory and other parameters also measured in laboratory, such as, the current
rock density (!b). Then, using the !b of density logs, they have obtained a regression that assists in establishing log TOC (Fig.7
bis).
Once this calibration has been completed, adjustments must be performed to compensate that TOC and !b are measured in
laboratory using a dry and fluid-free sample, whereas log !b includes gas and water adsorbed in clay. Therefore, some techniques
use the !ma (matrix density) correlation as a basis. The advantage of achieving such a correlation stems from the fact that logging is
performed in the majority of wells, whereas core samples are obtained with less frequency.
2,8
X800
X900
X000
X100
Density (g/cc)
2,7
y= -0.0302x + 2.5887
R2= 0.7156
2,6
2,5
2,4
2,3
2,2
2,1
2
0
X200
Figure 7: Plot that shows increase in

measured TOC's content with depth. (source:
Repsol YPF)
10
15
20
Total Organic carbon (TOC w%)
Figure 7 bis: Kerogen reduces rock density so that a connection between TOC content
(w %) and !b can be established. To use the information obtained by density logging,
the correlation must be adjusted if the apparent laboratory density has been measured in
a dry sample (adapted from Jarvie, D. 2011)
SPE 153004
TOC can also be derived from correlations performed with other kinds of logs, such as GR, U (uranium) from NGT and
sonic travel time (%t) logging, to name just a few examples. Strong correlation between %t and TOC have been registered in the
Neuqun basin; later, a particular technique, %LogR (Passey, Ref. 4) will be discussed - the technique was first utilised by the
industry some years ago and it also utilises travel time, in this case combined with resistivity.
Thermal Maturity of Kerogen
Thermal maturity of organic matter represents a critical element of thermogenic shale gas. The maturity process gradually
lessens kerogen content, but only produces changes in TOC when expulsion occurs. The pictures below5 demonstrate changes in
TOC (gas/oil, kerogen, inert carbon) as a result of the maturity and expulsion process (Fig.8).
Low maturity
Mature
Total Organic Carbon (TOC)

Kerogen
Non-source Kerogen
HC
HC Expelled
HC
Figure 8: Maturity and expulsion

produce a reduction in kerogen
and TOC given the amount of
carbon expelled as oil/gas; inert
carbon has increased (adapted
from Jarvie, D. 1999).
Kerogen
Non-source Kerogen
Total Organic Carbon (TOC)
Vitrinite reflectance is key for estimating the level of thermal maturity of organic matter and can be measured in
laboratory on formation samples. This method must be applied with the type of kerogen analysed in mind. Sampling, analysis and
data interpretation are all carried out by specialists. This is an industry standard, although there are also other methods of
establishing kerogen maturity. Below are some examples of the results obtained through these measurements (Figs.9 and 10).
Wet
Gas
Shale Oil Play

Shale Gas/Oil Play
Reflectance (Ro)
12
11
0.7
LOM
Inmature
Oil
Early
Moderate-Early
Moderate
Moderate-Peak
Peak
Oil
Dry Gas
Level of Maturity (LOM) - Vitrinte
Biogenic Shale Gas
10
9
8
1.0
1.4
Gas
Shale Gas Play
Figure 9: Vitrinite reflectance correlation with depth in shale (Repsol

YPF data interpretation).
0,25
0,50
0,75
1,00
1,25
1,50
Ro
Figure 10: Relationship of Vitrinite reflectance and the level of

organic metamorphism (LOM) (adapted from Passey, 2011)
Calibration of LOM parameters can be completed based on different sample analyses, such as Vitrinite reflectance itself,
thermal alteration index (TAI) and Tmax. By applying known laboratory techniques the following can be obtained; hydrogen
index (HI), oxygen index (OI), free hydrocarbons (S1), hydrocarbons generated by thermal cracking or residual hydrocarbon
potential (S2), CO2 released by kerogen (S3), temperatures at which maximum release of hydrocarbons is registered (S2 peak)
representing thermal maturity index (Tmax), and production index (PI) which is indicative of kerogen-hydrocarbon conversion
(ref. 5). Hydrogen and oxygen indexes facilitate the classification of different types of kerogen; thermal maturity is an indicator of
potential hydrocarbon generation.
As will be seen later, in some cases, to estimate log TOC it is necessary to establish the kerogen maturity index through
vitrinite reflectance (Ro) or the level of organic metamorphism (LOM). The tables below (Figs.11 and 12) represent Ro ranges
within maturity windows and quality of TOC content depending on the source potential of the rock. The interpretation of maturity
stages must be performed for each type of kerogen.
Thermal maturation
Inmature
Oil window
Wet gas window
Dry gas window
Gas destruction
Ro
<0.6%
0.6-1.1%
1.1-1.4%
1.4-~3.2%
>~3.2%
Figure 11: Interpreted maturity stages; this depends on the

type of kerogen.
Quality
Poor
Fair
Good
Very good
Excellent
SPE 153004
TOC (w%)
<0.5
0.5 to 1
1 to 2
2 to 4
>4
Figure 12: TOC content quality.
Based on the information above, it is fair to state that the mere presence of kerogen in a rock does not necessarily equate
to the generation of hydrocarbons. Thermal maturity, as outlined above, or biogenic maturity are essential conditions for the
generation and adsorption of gas in shale gas, oil in shale oil, on each type of kerogen and level of maturity.
Gas Content
The best method for establishing gas content in these reservoirs stems from decline curve analysis in productive wells.
However, the development of this resource is relatively new in many areas and there is not sufficient information to establish a
reliable performance trend. Volumetric analysis is one alternative method and is better when it is supplemented by the best extent
of analogy established with the behaviour of more mature wells from nearby geographic areas. Within volumetric calculations,
those that, in part, directly or indirectly measure the gas content of rocks and minerals (coals) are increasingly accepted by the
industry and the have been adapted from the coal industry as a suitable methodology for shale gas. Direct and indirect methods of
measuring gas content in formation samples are described below.
Adsorbed Gas - indirect method
Part of the gas stored in these reservoirs consists of gas adsorbed by organic matter and, according to many authors (Ref.
18), absorbed in clay; the other part consists of free gas that inhabits the pore space of the rock. The industry uses the Langmuir
isotherm to measure adsorbed gas content in the form of methane gas adsorbed by the surface of kerogen. In 1916 Irwin Langmuir,
Nobel Chemistry Prize in 1932, published an equation to measure adsorbed gases in solids. Adsorption is a phenomenon that
occurs when gas accumulates in the walls of a solid, resulting in the creation of a molecular or atomic film. This must not be
confused with absorption, when a substance spreads in a liquid, resulting in the creation of a solution, or it is trapped within a
solid. The term desorption is used in both cases to refer to the inverse process, or the expulsion of gas.
Lewis, 2004 (Ref.3) has presented the general expression of the Langmuir isotherm used in petrophysics:
gc =
Vl ! p
( p + Pl )
(4)
Where:
gc
P
Vl
Pl
= adsorbed gas content (scf/ton)

= reservoir pressure (psia)
= Langmuir volume (scf/ton)
= Langmuir pressure (psia)
The shape of the curve for a given temperature or TOC level depends on Langmuir volume (Vl; volume of gas adsorbed to
infinite pressure) and Langmuir pressure (Pl; pressure where one-half of the gas at infinite pressure has been desorbed). The curve
(Fig.13) defines the equilibrium between adsorbed and free gas as a function of reservoir pressure at the isotherm temperature. It
must be stressed that this ratio has been developed for methane gas, and the presence of other gases may impact on the method's
reliability; to this end, measuring techniques to compensate for the effect of different gases have been developed (Hartman et. al.,
2011).
SPE 153004
Figure 13: Langmuir isotherm; it is widely

used to estimate adsorbed gas content.
Langmuir volume and pressure for methane
gas must be adjusted in the presence of other
gases. Performing tests on different samples
and within a relatively small geographical
area is recommended (adapted from OFR,
autumn 2006).
The isotherm is measured at a set temperature and TOC level; therefore, in order to apply this equation to logging, it is
necessary to introduce corrections to account for temperature variations and different TOCs found in the reservoir.
Temperature correction:
Vlt = 10 (" c 3!(T +c 4 ))

Plt = 10 (" c 7!( P+c8 ))
c4 = logVl + (c3 ! Ti )
c8 = log Pl + (c7 ! Ti )
Where:
Vlt
Plt
C3
C7
T
Ti
= Langmuir volume at reservoir temperature (scf/ton)

= Langmuir pressure at reservoir temperature (psia)
= 0.0027
= 0.005
= reservoir temperature (C)
= Langmuir isotherm temperature (C)
TOC correction:
As with temperature, the isotherm is measured at a set TOC level; however, the formation interval of interest may contain
variable TOC levels:
Vlc = Vlt !
TOClog
TOCiso
(5)
Where:
= Langmuir volume at TOC corrected reservoir temperature (scf/ton)
Vlc
TOClog = TOC level as established by logging (w%)
TOCiso = TOC level used to measure Langmuir isotherm (w%)
10
SPE 153004
Therefore, the final equation used to establish adsorbed gas content is as follows:
gc =
Vlc ! p
( p + Plt )
(6)
These calculations are complex and for them to be used reliably there must be necessary to obtain sufficient experience
and amount of data.
The isotherm may be obtained from a sample as received (AR), or from a sample that has been washed and dried. The former is
preferable when analysing gas initially in-place (GIIP) and to simulate reservoir conditions; however, the latter can be very useful
for comparing adsorption capacity of different shales, although it must be stressed that it will be very optimistic if employed in
GIIP estimation. The use of various Langmuir calculations focused on as restricted an area as possible is recommended.
Estimating Adsorbed Gas Initially In-Place (GIIPad)
Accurately measuring the above mentioned parameters, that is to say the % volume or weight of kerogen, vitrinite
reflectance, LOM, TOC and adsorbed gas content, provides the means of estimating the volume of adsorbed gas initially in-place
as follows (ref. 3), generally named in this paper as adsorbed gas estimated with Langmuir isotherms:
GIIPad = gc ! Den ! Area ! h ! C
(7)
Where:
GIIPad
gc
Den
Area
h
C
= adsorbed gas initially in-place (Bcf)

= average formation density in h; !b average (g/cc)
= area (acres)
= average usable depth (ft)
= units conversion factor; 1.3597*10^-6
These equations have been established by the CBM industry. For CBM, average formation density varies depending on
the condition of the sample analysed to obtain gas content. This is, analyses carried out on an "as received" or on a dry ash-free
sample will bring two different density data according to its carbon purity (ref. COGEH, Vol.3, section 4.4.1 b.ii y iv). Necessary
adjustments should be made so that logging density is equivalent to laboratory density. The same applies to shale gas; if the test is
carried out on a dry sample, log density must be adjusted accordingly.
Later, it will be discussed that both the inorganic matrix (clay, silt, carbonates etc.) and the organic matter can be porous in these
types of rocks which in turn means that in the organic matter adsorbed and free gas can coexist.
Total Gas Content and Estimating Total Gas Initially In-Place (GIIPtot) direct method
The results of applying the Langmuir isotherm may be complemented by measuring desorbed gas in formation samples
(cores, sidewall cores). This type of analysis was designed primarily to measure the gas content in coal samples 6 and can be
referred to as either the direct method, or canister desorption analysis.
Full diameter samples are placed in a sealed recipient, canister in (Fig.14) and are sent from the well to the laboratory,
where the timeline of gas desorption is measured as the sample is heated to the reservoir temperature. The experiment continues by
also measuring the volume of gas desorbed in crushed samples. Afterwards, a correction is made to estimate lost gas, gas lost
between the time 0 (t0) prior to isolating the sample and sealing it in the canister. t0 is the moment at which the sample starts to
desorb gas due to imbalances with hydrostatic pressure during its tripping out of the hole and up to the surface after it was cut and
removed from bottomhole. It is difficult to correct lost gas precisely, although this factor is less important when dealing with shale
as opposed to coal (Decker, 1993); proportionately, however, it can represent an important percentage of the gas content estimated
SPE 153004
11
by this method. Lost gas can be significant if the rock has natural open fractures that were gas saturated, or if gas is widely spread
in the matrix (moderate permeability); furthermore, certain measures must be applied to avoid excessive dead space in the canister,
between the upper surface of the sample and the lid of the recipient, as this can cause gas expansion and shrinkage issues produced
by heat and pressure together with the risk of oxidation of part of the organic matter. The test provides data on total gas basis and it
cannot account for the adsorbed gas percentage on the total gas content. Some authors have correlated total gas content and TOC,
thereby establishing a direct correlation (Fig.15).
Therefore, the formula representing total gas content is:
Gc = 32.0368 !
(8)
= total gas content (scf/ton)

3
= volume of lost gas, (cm )
3
= volume of measured gas, (cm
3 )
= volume of crushed gas, (cm )
= air-dry gas volume (g)
Total Gas Content, scf/ton
Where:
Gc
Vl
Vm
Vc
Mad
Vl + Vm + Vc
mad
100
80
60
y=3,684x + 26,93
R2=0,841
40
20
0
0
10
12
14
Total Organic Carbon, w%
Figure 14: Canister apparatus used to

hold core sample (Weatherford
laboratories).
Figure 15: Correlation between total gas content and TOC

(Luffel, D., et al.).
To demonstrate the magnitude of the results coming from these calculations, gas contents of about 15 to 350 scf/ton have
been recorded in established shale gas fields active in the US.
Using the same equation as (7), but replacing the term gc with Gc, we can estimate total gas initially in-place GIIPtot.
These tests also analyse gas composition, which can be important in calibrating the parameters of adsorption tests.
Diffusivity data is obtained that consists of relating the slope of gas data measured as a function of the square root of the elapsed
time from extraction of the sample, with the total content of gas obtained. Gas content of crushed sidewall cores has also been
measured, confirming the viability of said technique; given that sidewall cores are often easier to obtain, they represent a more
economical means of testing than extracting core samples.
It is therefore reasonable to state that if there are sufficient formation samples (from cores and sidewall cores) and
comprehensive laboratories tests as described above, it is possible to complete various estimations on gas initially in-place.
However, these tests are not exempt from uncertainty (Ref. 21) and, furthermore, as normally there are insufficient cores and
laboratory tests it is frequent to alternate with electric and radioactive log interpretation techniques for total gas in-place
estimations. Log information is usually abundant and serves to extend the estimations to other intervals of the reservoir. Supported
by sate-of-the-art laboratory analysis of porosity, grain density, water saturation, mineralogy and elastic modulus of core and
sidewall cores, log interpretation techniques are fundamental tools in assessing unconventional reservoirs and are in constant
evolution.
12
SPE 153004
Petrophysical measurements for Shale Gas

Laboratory Measurements
Commercial laboratories have strived to find methods for accurately and specifically measure porosity, grain density,
permeability, mineralogy and elastic modulus of these types of rock. Currently, to analyse porosity, grain density, water saturation
and permeability, the Gas Research Institute (GRI) method, reported by Luffel and Guidry in 1992, is commonly used7. GRI
primarily consists of a conventional analysis of total porosity, grain density and Dean Stark water saturation in standard conditions
on crushed samples. Additionally, it presents a new method for estimating permeability. The GRI method establishes that by using
crushed samples, better helium expansion is obtained in the sample during the test period. If there is insufficient helium expansion
during the test, lower levels of porosity and lower grain density are obtained; GRI researchers affirm that this factor often occurs in
tests performed on whole cores in which during the first 30 minutes of the experiment there has been insufficient gas penetration of
all pore spaces. The Gas Research Institute has widely published information of its methodology.
Porosity
Porosity of these systems, natural fractures aside, is composed of: a) porosity of the non-clay matrix, b) clay porosity and
c) kerogen porosity
To adequately discuss about different measurements, defining porosity nomenclature (phi, !) that will be used in this
report is fundamental (Fig.16):
In laboratory measurements, pore space is as follows:
Total !
Effective !
As received (AR) !
Gas Filled ! (AR)
Connected and unconnected of a crushed sample

Connected of a whole core or plug sample under humidity control*
Measured by helium expansion in a preserved sample
Measured in two separate pieces of the sample; the gas is measured in an unwashed and un-dried sample by
injecting mercury, this is known as "gas-filled phi"; liquids are measured in the other piece by distillation, also
known as "summation of fluids"; "gas-filled porosity" is also found as gas saturation (Sg), derived from helium
expansion porosity and Dean Stark extraction in an AR sample. In shale gas, it does not include adsorbed gas.
*If samples are adequately dried under controlled humidity, water adsorbed in the clay and clay hydration water will be preserved. On the other
hand, if the drying process is not controlled and is completed at very high temperature, the measurement of porosity will include said pore space.
In electric and radioactive logging, pore space is as follows:

Total phi
Effective phi
Shale phi
Clay bound water phi
Capillary bound water phi
Free fluid phi
Connected and unconnected matrix porosity, added to the shale porosity

Total phi minus shale porosity
Estimated on shale volume basis
Occupied by water adsorbed in clay and clay hydration water8
Occupied by capillary trapped water (water-wet rocks) in the non-clay matrix
Occupied by mobile fluids (phi FF = phi total - clay bound water phi - capillary bound water phi); free gas would
occupy this pore space when So~0%.
SPE 153004
13
Figure 16: Types of

porosity measures in
laboratory and with
logging; in a shale
gas-type reservoir,
clay bound water is
proportional to the
type of clay and its
volume percentage.
Non-scaled volumes
in this picture
(adapted from API
RP-40 Recommended
Practices for Core
Analysis).
Currently, the Gas Research Method (DOE/GRI) of measuring porosity, grain density and water saturation in laboratory is
calling the attention to the petrophysicist community; however, alternative measurements are being tested by the industry (Ref.
21). They all need to overcome similar issues as they are affected by factors such as:
o
o
o
o
o
Removal of water and hydrocarbons from the pore system

Problems for full access of gases (helium, nitrogen, methane) and liquids (mercury, water) to the porous system due to the low
permeability of the system.
Adsorption effects
Sample size, crushing of the sample, and weight of the crushed sample
Presence of fractures and micro fractures (natural or induced)
The GRI method measures porosity through helium expansion in crushed samples, rather than full-diameter cores. It uses
around 100 grams of crushed samples, with 15-30 grams of the sample sieved (20/35 mesh) put aside to measure porosity. Luffel,
et al., have reported a 0.5% level of uncertainty regarding porosity in duplication experiments they have performed.
As mentioned above, the sample drying process, amongst other factors, can affect the final results; the temperature used in the
process and control of humidity determine whether the results represent effective porosity or total porosity. Some authors claim to
have noted significant differences in experiments from different laboratories using the original GRI method, although these issues
may stem from the use of "twin" samples which are not entirely identical, or have been processed in different ways. The GRI
thinks that his method measures total connected porosity because the sample is completely dried, removing clay bound water. With
the caveat that temperatures used in the drying process have been suitable ones, porosity therefore represents the total porosity of
kerogen, the inorganic matrix and clay together. It must be stressed that current kerogen is the organic matter remnant of the rock's
maturity process; it is not altered by the drying process provided that adequate temperatures and conditions are used.
The images below clearly show kerogen porosity, in addition to its coexistence with pores of the inorganic matrix, Fig.17. The
perfection of imaging techniques is undergoing constant evolution, and there are currently laboratories performing experiments to
obtain porosity and permeability by using sophisticated DRP imaging tools (Digital Rock Physics, Ingrain Inc.), Fig.17 bis.
14
SPE 153004
Figure 17bis: 3D pores, organic matter and minerals

obtained using focused ion beam scanning electron
microscopy (FIB SEM, Ingrain 2011).
Figure 17: Kerogen porosity, left image (SEM BEG 2008) and the inorganic matrix, right image.
(Quirein et al, SPWLA Shale Gas workshop, Colorado Springs, 2011).
Other porosity measuring techniques for these reservoirs include Mercury Injection Capillary Pressure (MICP) and
Nuclear Magnetic Resonance (NMR). To measure porosity, MICP9 measures total rock volume by immersion methods, and total
porous volume in dried samples by injecting mercury at a pressure of 60.000 psia. Olsen and Grigg (2008) have made positive
remarks regarding the correlation between MICP and AR (as received) porosity.
NMR magnetic resonance techniques are useful for measuring porosity in water-saturated rocks independently from the
matrix. When paramagnetic minerals are present in the matrix, or when the hydrocarbon phase is present, interpreting relaxation
time becomes more complex. Diffusion effects that facilitate identification of fluid quality when combined with T1 (longitudinal
relaxation) or T2 (transverse relaxation) do not seem to work in shale gas. In samples of low porosity, as is the case in this report,
the weakest signal (low signal-to-noise ratio) and the hydrogen-defect of the gas are not considered ideal conditions for accurately
measuring NMR porosity. Even so, the author believes that it would be beneficial to continue experimenting to better understand
the effects of mud filtrate and, furthermore, observe whether gas trapped in free pores, mainly in the hydrophophic kerogen,
remains without being displaced in the area depth of investigation of the tool, as hydrogen deficiency and high diffusivity may
irreversibly alter the result of these measurements. A highly valuable aspect of this technique is that it can be transported from the
laboratory to the well via NMR logging tools and facilitates to obtain porosity independent from the composition of the matrix.
This is especially important given that estimating porosity with conventional logging is difficult due to the high levels of
uncertainty in establishing lithological components and kerogen volume of the rock. The main limitations and characteristics of the
technique for a shale gas are outlined below:
Hydrogen-defect of the gas in regards to water (resulting in inaccurately low porosity levels)
Low signal-to-noise ratio given the low level of porosity
Stationary measurements or very low logging speed
Mixed wettability
Potentially low diffusion effects and surface relaxation control; difficult to separate gas-water signals
5,0
NMR T2 Distribution
Saturated Incremental Porosity

0,45
15,0
Incremental
0,45
Saturated Cumulative Porosity
0,40
0,50
4,5
Desaturated Incremental Porosity
4,0
3,5
0,30
3,0
0,25
2,5
0,20
2,0
0,15
1,5
0,10
1,0
0,05
Incremental Porosity (%)
0,35
Cumulative Porosity, %
Desaturated Cumulative Porosity
13,5
Cumulative
0,40
12,0
0,35
10,5
0,30
9,0
0,25
7,5
0,20
6,0
0,15
4,5
0,10
3,0
0,05
1,5
Cumulative Porosity (%)
0,50
Incremental Porosity, %
o
o
o
o
o
0,5
0,00
0,00
0,1
1,0
10,0
100,0
1000,0
0,0
10000,0
Figure 8: T2 transverse relaxation time distribution in Devonian shale with <

40% clay; very compact (Repsol YPF).
0,1
10
100
1000
0,0
10000
T2 Relaxation Time (ms)
Figure 19: T2 transverse relaxation time distribution

in a Carbonniferous sandstone with high levels of
Swirr (Repsol YPF).
SPE 153004
15
Above, a practical NMR example shows laboratory results for two different types of rock. The graph in Fig.18 shows
laboratory measurements for shale with clay content < 40%, very compact whilst the graph in Fig.19 shows laboratory
measurements in low porosity sandstone with high levels of total irreducible water (capillary and clay bound fluids) that retain
medium sized pores saturated in free fluid. 3.4% helium and 3.9% NMR have been measured in the shale, whereas 5.5% helium
and 6% NMR have been measured in the sandstone, both of which are considered acceptable. As can be seen, the signal in shale
rock is present and offers porosity data which is compatible with other techniques under the same test conditions. Widespread use
will determine whether or not these porosity measurements are useful for shale gas.
Grain density
This is a very important parameter in the petrophysical assessment of such reservoirs, since kerogen has a strong
influence on grain density (GD or !ma). Kerogen decreases grain density in shale when compared with other similar, kerogen-free
shale. This parameter is useful to calibrate log interpretation results; however, as will be discussed below, this is complicated given
the difficulties experienced when estimating accurate kerogen volume and density in formations by using logs. Some authors have
reported differences in comparative measurements from different laboratories (Sondergel, et al., 2010; Passey, et al., 2010); in
large part they have been attributed to the procedures followed for sampling and sample preparation (crushing, sieving, washing,
and drying). Grain density can also be estimated by correctly interpreting mineralogical analysis, such as XRD, XRF or FT-IR.
Some of these analyses, like XRD, are unable to detect kerogen; however, a separate TOC w% measurement can be integrated to
the XRD analysis to estimate GD. If sample frequency is high, curve continuity obtained in laboratory will also be high and this
will facilitate core- to-log correlations.
Water saturation
The technique used by GRI to measure water saturation (Sw) is Dean Stark on crushed samples. Sw can also be
established by the retort method in laboratory. As is the case in conventional reservoirs, water salinity in the rock has a significant
impact on these analyses and can often result in high levels of inaccuracy. Luffel, et al., have found saturation differences of up to
3.5% in duplicate experiments they have performed. Volumes of water, and oil extracted with toluene are measured to obtain
saturations. Gas saturation is then estimated by difference. There are no great variants in regards to standard procedure, except the
use of a crushed sample and the unconventional characteristic of the rock. Differences between the results of different laboratories
have been reported. Some authors suggest that the procedures followed by each laboratory may differ, which would explain the
cause of these differences.
Other parameters
Electric properties
There are few references to electrical parameters measurements in samples of this type of rocks. The cementation factor
(m), saturation exponent (n) and the coefficient (a) normally employed in resistivity models are the parameters in question.
Patchett10 (1975) studied shale conductivity, reaching the conclusion that clay conductivity is highly dependent on formation
temperature and the cation exchange capacity (CEC) of the clay mineral; he also asserted that water salinity in clay does not
significantly affect conductivity. In the same study, he suggests that 2 is the most appropriate value to express m. Rather than
investigating the estimation of water saturation, Patchett's work focussed on detecting the type of clay present through resistivity
logging and the correlation of resistivity and other parameters with CEC. Experts today may disagree with Patchett's conclusions,
especially in regards to the value of 2 being the most appropriate to express m. The lack of modern investigation into laboratory
measurements of electric parameters of shale gas means that possible disagreements on the m value may remain unresolved.
Furthermore, the same lack of resolution apply to the values that saturation exponent (n) must take in case of more than one fluid
phase is present.
Quirein (2010) reported an empirical m value of 1.6, calibrated to correlate with $*Sg measured in cores, without
specifying the resistivity-water saturation model used to establish this.
Permeability
Matrix permeability determines the feasibility of maintaining long-term production in these reservoirs; measuring
permeability is therefore of great interest. The limitations of traditional methods, many of which are only able to measure levels up
16
SPE 153004
to .01 mD under acceptable accuracy, in addition to the undesirable effects in many cores samples, such as natural or induced
micro-fissures, has oriented to the search for innovative alternatives. The challenge is to reliably measure in the microDarcy range,
or even less. GRI applies a method that takes advantages of the helium expansion cycle in crushed samples to interpret the
pressure-time curve and analytically derive a result for the permeability of the matrix. Given that the experiment is completed on
an as received sample, irreducible fluids will be present in the rock; therefore this permeability may represent the effective
permeability of gas @ Swirr and gas filled porosity can be derived. Likewise, experiment conditions are @ ambient conditions and
likely the chips are free of microfracture effects. There has been no Kinkenberg correction and net overburden stress is considered
negligible11. The use of this information must be limited to the analysis of matrix properties in a microscopic space/volume in
relation to the wide areal and vertical span of such sediments. Only a dense statistical sampling can be considered representative of
a higher volume of rock, however, such data density is often not available.
Other findings have been reported for crushed samples, plugs, and whole cores; however, they have not been sufficient to
conclusively recommend one technique or another. Disparity is commonplace in comparative results with greater interaction
required between operators and laboratories to create standard and universally accepted procedures (Sondergeld, 2010).
Mineralogy
To choose a minerals model, it is first necessary to obtain core laboratory measurements to ground-truth data.
Identifying the mineral composition in a shale gas is essential in order to complete a quantitative evaluation of the reservoir's
properties and gas content through core analysis and electric and radioactive logging. Various authors have reported that the
mineral composition of a shale gas can vary from one metre to the next, or even more in some instances. Establishing the
composition of the rock facilitates, for example, the estimation of matrix density, a fundamental parameter in calculating logging
porosity. Laboratory analysis is essential to calibrate geochemical logging, the success of which in predicting rock composition
depends on the model chosen.
X-ray diffraction is one of the most commonly used and accepted methods, although it is only considered semiquantitative as it does not detect amorphous elements, such as kerogen. This type of analysis can be complemented by other
techniques and the use of at least two data sources is usually recommended. Some of the available methods are listed below:
o
o
o
o
o
o
SEM-EDS (scanning electron microscope energy dispersive spectrometry)

XRD (X-ray diffraction)
XRF (X-ray Fluorescence)
FT-IR - DRFT-IR (Fourier transform infra-red spectroscopy diffuse reflectance infra-red Fourier transform)
LIBS (Laser Ionization Breakdown Spectroscopy)
EA (elemental analyser)
To the date, each of these technologies has been spread well in the literature by numerous specialists / suppliers and its
detailed description is not inside the scope of this report.
The key in shale gas is to identify quartz, calcite, dolomite, types of clay, heavy minerals like pyrites, and kerogen
amongst others. The proportions of minerals are normally expressed in weight (%w); when they are also expressed in volume
(%vol), it is essential to establish the density of the mineral used in laboratory to correlate with log interpretation volumes.
Operators have reported that hydraulic fractures are more effective in shale that has lower clay content and a high brittleness index
(BI, see Appendix III). BI can also be estimated through mineralogical analysis and/or with the elastic modulus of Young and
Poisson. Establishing clay type is important, as they have different levels of elasticity that affect formation stimulation.
Elastic modulus
Clays are elastically anisotropic; therefore, in order to characterise them various elastic constants must be measured.
Elastic constants are any physically measurable features that characterise the elastic behaviour of an elastically deformable solid.
Although there are many elastic moduli, the most frequently used for shale gas are Young's modulus and Poisson's ratio. The
elastic modulus, or modulus of elasticity, is the mathematical description of an object's or substance's tendency to be deformed
elastically, non-permanently, when a force is applied to it. Young's modulus is a measure that characterises the behaviour of an
elastic material according on the direction in which a force is applied. Poisson's ratio is an elastic modulus that measures the crosssectional stretching of an isotropic or linear elastic material when it stretches lengthwise and contracts perpendicularly to the
stretching.
SPE 153004
17
Some authors state that in shale, vertical anisotropy is greater than horizontal anisotropy, depending on the level of
porosity and compaction of the shale itself. The moduli measurements can be performed either statically, or dynamically. Static
measurements are performed on cores whereas dynamic measurements can be performed either on the core, or using data obtained
from well logging. Methods for measuring cores must be well designed to replicate subsoil conditions and even that, still they may
not be able to successfully capture the entire complexity of the system due to limitations of the methods, or due to the limited
sampled volume not capturing all formation features within the extent of hydraulic fractures and their respective drainage areas.
Natural fractures may be out of the cored interval and, furthermore, kerogen content and the contact characteristics of shale sheets
can widely vary in just a few metres from the sample point. There are therefore two key points that must be considered when
designing stimulations and planning horizontal or deviated drilling. 1) Core measurements may not be representative of the entire
system and 2) efforts may need to focus on calibrating elastic modulus derived from logging with well stability and productivity
data.
Well log information that may assist in obtaining geomechanical data is available. This includes dipole sonic logging,
density logging, lithological logging and well image analysis. Pressure measurements in clayey formations using wireline tools are
often unsuccessful and imprecise as it is very expensive to achieve pressure stabilisation within a reasonable period of time. This
data should ideally be obtained through well tests.
Geomechanical data can be important in estimating propagation and closure of hydraulic fractures and the geometry of
the drainage area caused by such induced fractures. Non-clay-rich lithological components establish the brittleness index (BI) of
the rock, which in turn establishes the most favourable conditions for hydraulic fracturing. The BI can be derived from the clay/silt
ratio (quartzitic a/o carbonate silt), or by applying Young's modulus and Poisson's ratio. This is another highly complex specialty
for field studies, which is not the subject of this report.
Net Thickness
It is believed that the current % of mature kerogen, the current TOC level, hydrocarbon saturation, permeability, porosity
(i.e., matrix and natural fractures) and, in some cases, lithologies define the net thickness of these reservoirs.
o
o
o
o
o
o
Kerogen/TOC content
Gas saturation
Permeability
Porosity and fracture porosity
Rock density (! )
bulk
Lithology - Brittleness Index (BI)
Rock density and lithology aside, many of these factors are difficult to determine precisely in shale gas, even when using
the best core sampling and analysis, and electric logging techniques. Luffel et al. (1992) suggest that combining core porosity and
hydrocarbon volume it is possible to obtain a porosity cut-off (Fig.20). If this works in the reservoir of interest, the analyst can
apply such limits to obtain net thickness.
Figure 20: Laboratory experiment data suggests
that when porosity is less than 2.5%,
hydrocarbons have not been generated. If it can
be demonstrated that these experiments are
reliable, this level of porosity could be used as a
cut-off to establish net thickness (adapted from
Luffel et al., 1992).
14
BVO + BVG (%)
12
10
8
6
4
2
0
0
Porosity (%)
10
12
14
18
SPE 153004
With the expressed hitherto, combining the comments on geochemical and petrophysical laboratory measurements it can
be inferred that unconventional reservoirs require special analytic attention. In addition, they demand the permanent adjustment
and refinement of the technologies and the constant evolution of the investigation, to obtain with major precision volumetric key
parameters of the reservoir. The table below (Fig.21) shows the minimal values required to reach commercial viability in these
reservoirs, as reported in different technical and business papers. These minimum values must be used as general references, and
must be verified on a case-by-case basis.
Parameter
TOC
Ro
Porosity
Water saturation
Oil saturation
Permeability
Minimum value
>2%
>1%
>4%
<45%
<5%
>100ndarcies
Figure 21: Values establishing a reservoir as commercially viable; this

is a general reference tool, amongst others, in ref. 2.
Oil saturation must be low as to not interfere with the relative mobility of gas; if oil saturation is high or very high, it is
probably that a shale oil is present (Appendix I). Certain characteristics of shale oil differ from those of shale gas. For example, the
type of kerogen and/or the generation window may be different; the oil of interest is oil that has not migrated, and for which
information can be obtained through geochemical analysis of kerogen given that adsorption is not a feature of oil; there may be
two phases (not including water) so that GOR and relative permeability, which is difficult to measure, must be considered in
estimations; PVT analysis is also highly important, as is the oil in natural fractures that may be more mobile.
Electric and radioactive log responses for shale gas

Electric and radioactive logs responses are particular in shale gas intervals (Fig.22). It is worth noting that this is also the
case when logging shale oil. The presence of kerogen and hydrocarbons can result in alterations in the responses of resistivity,
density, sonic, neutron, spectral and natural gamma ray logging and microresistive imaging in comparison to responses from
intervals in which there is no significant kerogen presence. The experience with magnetic resonance logging is immature at this
stage.
Resistivity: for the clay type and volume %, the presence of kerogen and hydrocarbons increases the resistivity of the rock. These
rocks may contain a variable % of clay, between 30-70% according to studies carried out to date; the greater the % of clay, the
lower the resistivity of the rock. However, the property of clay that most affects resistivity of the rock is the cation exchange
capacity (CEC) that varies depending on the type of clay and mixed layers; the higher the CEC, the lower the resistivity of the
rock. Shale gas intervals are more resistive than adjacent clay-rich -strata that report very low levels of kerogen, or are kerogen
free. Within a single shale gas interval, the presence of kerogen can be gradational resulting in bell-shape and sequential electric
log responses as reported by Passey (Ref. 4, 2010). In instances of over-matured kerogen and when graphite has been produced,
the effect consists in a drastic reduction in resistivity as graphite is conductive.
Natural and Spectral Gamma Rays: the presence of shale increases the level of natural radioactivity of the formation. It is
normally easy to distinguish between clay-rich-strata and sandy or chalky layers. Abnormally high natural radioactivity in shale
gas is mainly caused by the presence of concentrated uranium, or uranium-ions, e.g. as a result of reductive conditions in the
marine environment. Therefore, it is highly recommendable to use a spectral gamma ray tool (NGT) to evaluate such formations.
The ability to separate levels of thorium (TH), potassium (K) and uranium (U) radiation is highly advantageous in regards to total
GR. It is widely believed that illite, mica and feldspar are the main contributors of potassium; detrital (adsorbed) clay minerals and
micas are the main contributors of thorium12, and uranyl ions are the main contributors of uranium. The effect of montmorillonite,
chlorite and kaolinite on radiation levels is less significant. The relationships between TH-K-Pef, adjusted by laboratory analysis,
are some of the options used to distinguish clay-rich minerals.
SPE 153004
19
Density and the Photoelectric Factor: kerogen and gas are low density (!b) and low photoelectric factor (Pef); as a result, high
levels of the aforementioned components reduce the density and photoelectric factor of the rock. Also, kerogen matrix density
(!ma) is very low, similar to the density of water; it is for those reasons that if the kerogen volume is not estimated accurately, the
subsequent calculation of porosity will be incorrect, providing values higher than the actual ones. On the other hand, some authors
have made correlations between !b and TOC by exploiting this characteristic, reporting satisfactory results in deriving TOC from
rock density as measured in laboratory, the application of which could be extended to log interpretation. The borehole diameter
and the good shape of the wellbore walls are critical for this type of logs; the mud filtrate invasion depth and the type of invading
fluid must be known in order to use this technique, although in this type of rock, changes in formation near to the borehole walls
may be due to other factors rather than mud filtrate invasion.
Sonic: kerogen and gas have high acoustic transit-times (%t); a high volume of those components therefore increases the
compressional transit-time, %tc. As a result, sonic logging yields exaggerated porosity estimations when gas and kerogen are
present when it is not possible to counteract the effects of those components. Some publications also report on a likely impact on
the shear wave response (Passey, 2010), and about the correlation that exists between %tc and TOC which appears reliable in many
cases. The %logR method as presented by Passey will be discussed later; this method is used to detect potentially productive areas
and to estimate TOC.
Figure 22: Log responses of some regions, in shale gas reservoirs with
high level of mature kerogen.
Compensated Neutron: this is one of the conventional logs less used in detecting and evaluating the organic content and
productive potential of shale gas. Neutron logs are affected by the hydrogen in organic matter, the hydrogen in clays OH+, and
furthermore by the hydrogen in water and hydrocarbons present that also affect neutron porosity measurement. It is believed that
the neutron porosity of kerogen is high, but its use in detecting and estimating TOC has been limited. It is also believed that the
hydrogen index of kerogen is lower than that of water13; therefore, neutron porosity tends to reduce according to the amount of
kerogen present. The gas in-place tends to reduce neutron porosity as it has a lower hydrogen index than water, although
quantifying the effects of reduced porosity due to lack of hydrogen in gas and kerogen is quite complex in this rock type. In effect,
neutron-density combined logs are of limited use in locating gas and kerogen due to the strong impact of clays that increase
neutron porosity or due to the presence of calcite and dolomite that result in complex and contrasting neutron log responses.
Specifically, calcite tends to reduce the neutron porosity of a shale gas, and this can be confused with gas effects. This type of log
is less affected by the condition of the wellbore walls, although any other factor that may alter the formation in areas close to the
borehole walls can impact on the quality of this measurement.
Magnetic Resonance (NMR): In these reservoirs, calculating porosity with conventional logs is elusive given, amongst other
factors, the difficulty in accurately determining the matrix parameters such as the mineral composition and kerogen content, in
addition to uncertainty regarding the properties of clay minerals and kerogen (density, travel-time, etc.). NMR techniques look
highly useful for estimating porosity in unconventional reservoirs given that they do not depend on matrix characteristics in the
absence of paramagnetic minerals. However gas content, even in cases of low porosity, may adversely affect the accuracy of NMR
porosity. There is still insufficient evidence regarding these effects and their impact on the final NMR porosity figure; therefore, in
using this technique, it is important to bear in mind that it is a process continually in evolution. The work of Sigal (2011) is of
20
SPE 153004
particular interest in this respect. The borehole diameter and the good shape of the wellbore walls are critical for this type of logs;
the mud filtrate invasion depth and the type of invading fluid must be known in order to use this technique, although in this type of
rock, changes in formation near to the borehole walls may be due to other factors rather than mud filtrate invasion.
Microresistive Imaging: In water-base mud, the colours in a microresistive image of an area with high kerogen content tend to be
clear, whereas in adjacent shaly areas of low or nil kerogen content the colours are darker. Personal communications have
highlighted the use of microresistive imaging to corroborate the detection and quantification of kerogen with Rt (high) and P ef
(low). The responses of kerogen must not be confused with those of tight calcareous or siliciclastic layers, coal beds, or those of
highly resistant mineral concretions. It has been reported that in oil-based mud, this imaging technique is still reliable, provided
that the mud does not disturb or deeply invade the formation, and there are no natural open fractures that could make the
interpretation process more difficult. The borehole diameter and the good shape of the wellbore walls are critical for this type of
logs; the mud filtrate invasion depth and the type of invading fluid must be known in order to use this technique, although in this
type of rock, changes in formation near to the borehole walls may be due to other factors rather than mud filtrate invasion.
Log Interpretation
As of here, various comments have been made in regards to certain petrophysical characteristics of these reservoirs. Use
of core, logging and special testing information to study reservoir properties and take operational and management decisions
regarding the resources are another aspect that want to be considered. Geological and petrophysical interpretation of logs is a key
piece in this operational framework. Hereon, we will discuss petrophysical interpretation of logs, the objectives of which are to
provide useful data to:
o
o
o
o
Detection of potentially productive areas

Estimation of gas content
Establish net thickness and rock properties
GIIP, TRV and reserve volumetric estimations
Detection of Potentially Productive Areas

Potentially productive intervals can be detected by log responses, mud logs, cuttings, core samples and sidewall cores.
Following the response pattern usually print by kerogen in logs (Fig.22), as has been mentioned above, it is feasible to locate the
top and base of the intervals of interest. In doing so, it is also worth to be supported by mud log gas detection and the description
and visual inspection of cuttings. The extraction of cores and sidewall cores can provide valuable information, on their bituminous
features for example, even before laboratory analysis.
Estimation of gas content with logs
Adsorbed Gas
Estimation of TOC with logs
As described above, quantifying the amount of adsorbed gas without needing to know porosity and water saturation of the
rock is a possibility. However, establishing adsorbed gas is not sufficient to estimate total gas content unless the former is the only
gas present in the reservoir, given that total gas is the sum adsorbed gas and free gas filling the pores. As described before, in order
to estimate adsorbed gas content it is first necessary to establish TOC % present in the rock; therefore, it is necessary to analyse the
various forms of determining TOC with logs to calculate adsorbed gas content through the Langmuir isotherm.
To achieve this estimate, it is first necessary to perform certain studies on formation samples of the interval of interest, such as:
o
o
o
o
Maturity Index of organic matter (Ro, LOM)

Total Organic Carbon Content (TOC) and Hydrogen Index (HI)
Correlation between TOC and other petrophysical properties
Langmuir's isotherm
SPE 153004
21
At least two ways of determining in-place TOC figures through logging have been reported: for the purposes of this study,
one will be referred to as Density-TOC, and the other as the Passey method, or %logR-TOC.
Density vs TOC
This method consists of correlating rock density (!b) and TOC content, both measured in formation samples (see fig.7 bis
above). If a strong correlation can be established through laboratory analysis, density logging can therefore be used to establish
TOC. Similarly, TOC can be established by correlations obtained using laboratory grain density (GD), then using the matrix
density value (!ma) obtained in complex log interpretations. The use of gamma ray, uranium, sonic and neutron logs has also been
reported in establishing similar correlations.
Once reliable TOC data has been obtained, adsorbed gas content can be estimated by using one or more appropriate
Langmuir isotherms.
!logR vs TOC
Passey et al., (ref. 14) discuss this method in great detail, and it consists on:
1. Calibrate the graphical difference of a porosity log curve (linear scale) with respect to the resistivity log curve (logarithmic
scale) assigning a value to this crossover, commonly called %logR,
2. Calibrate %logR against TOC according to the LOM (maturity index); TOC and LOM should be obtained from formation
samples retrieved from the widest range of locations and intervals of interest as possible,
3. Establishing a correlation between LOM and depth or temperature in the zone of interest,
4. Using these correlations to derive TOC from %logR obtained in wells where cores or cuttings are not available.
Figure 23: Resistivity

and sonic curve
fitting to estimate
%LogR (similar to
the described by
Passey, 1990).
Sonic logging has been extensively used together with resistivity logging (Fig.23) as it is widely available in most wells,
especially older ones.
22
SPE 153004
In order to derive TOC from logs, Passey14 has established correlations between TOC, LOM and %logR (Figs.24 and 25).
These correlations must be revised to ensure that they also serve in other fields of study areas, or new correlations must be found
that serve each individual case. This author has established a correlation between LOM and depth, in a specific area; this suggests
that similar features could be established for other fields of studies. It must be noted that there are also correlations between Ro
(vitrinite reflectance index) and LOM, as has been discussed earlier in this report.
) R &
$$ + (P # (!t " !tbaseline ))
!LogR = Log10 ''
( Rbaseline %
(9)
P= Resistivity logarithmic cycle over transit time cycle. Normally equal to .02
.02 =
1
50 ms / ft
TOC - LogR - LOM - Ro
13
12
11
10
9
8
7
6
5
4
3
2
1
0
Calibration Limit
LOM 5
20
for LOM < 7
Calibration Limit
for LOM > 10,5
LOM 8
18
LOM 10
16
14
TOC
LOM
Idealized LOM/Depth Plot
Ro0,42
12
LOM 11
Ro>0,9
10
8
6
LOM 12
4
2
500
1000
1500
2000
2500
3000
Depth
Figure 24: Correlation between LOM and depth for a
productive shale gas area (adapted from Passey, 2010).
TOC (w%) = #LogR !10 $ (^2.297 " .1688 ! LOM )
0
0
0,5
Area of original calibration
1,5
2
DeltaLogR
2,5
3,5
Modified from Passey et al. 1990
Figure 25: Relationship of Ro/LOM with %LogR used to derive

TOC content (adapted from Passey, SPWLA 2011).
(10)
Passey asserts that the method works well for LOM between 6 and 12 and in recent presentations the upper limit has been
reduced to 10.5. Also, he warns out that it is important to consider the relationship between TOC and S2 (type of kerogen) of the
geochemical analysis. For further information we recommend to read Passey of ref. 14 and other works completed by him and
others co-authors15. Other authors have commented that it has sometimes been necessary to adjust Passeys equation (10) with a C
factor, the value of which depends on the calibration performed in the area under study. One reason behind the occasionally failure
of the method has been the high percentage of carbonate hindering the presence of clay in formations studied. Sondergeld16
establishes 4 as a reliable C value.
Once reliable TOC data has been obtained, adsorbed gas content can be estimated by using one or more appropriate
Langmuir isotherms (Eqs. 4 to 7).
Total Gas
Estimating Total Gas with logs and resistivity models
It is essential to discuss the difficulties in estimating total gas with logs and resistivity models, the results of which also
contribute to the determination of free gas by analysing the difference between total gas estimated with these models and adsorbed
SPE 153004
23
gas content obtained by the above mentioned procedures. For practical purposes, it will be assumed that the quality of logs used is
good, without borehole effects or other undesirable alterations, aspects that should be detected prior to using the information.
As can be seen in the comments made above, the main limitations are related to problems in resolving the lithological
composition of the rock, including the % of kerogen and its properties, low porosity of the rock and high uncertainty regarding the
determination of the electric parameters a, m and n to estimate water saturation.
The goal here is to resolve the estimation of porosity and water saturation using logs, with the special characteristic of
evaluating such reservoirs properties without seeking to neutralise neither the porosity of clay in the rock, nor clay bound water
saturation and its effect on resistivity; this is not the case in conventional reservoirs. Therefore, if it is possible to establish an
adequate and empirical use of these models, total porosity and total water saturation can be derived.
Total log porosity
In these systems, very small pores, micropores and nanopores are predominant; therefore, capillary effects were
considered so important that they were capable of inhibiting commercial production flows. Irruption of commercially productive
shale gas reservoirs has resulted in the revision of these concepts. Quantities of gas and oil have often been reported when drilling
through these levels; however, previously they had not been studied with a view to commercial production. The way of functioning
of these reservoirs has been re-analysed regarding storage capacity and, most importantly, transport and fluid production.
Historically, it has been recognised that these source rocks have expelled hydrocarbons which have been trapped in high quality
reservoir rocks by means of seal rocks. Today, the remnants of hydrocarbon that has not been expelled, trapped gas and liquid and
adsorbed gas in the rock are extracted in some areas by using well designs and stimulation techniques that in many cases increase
the production capacity of these reservoirs.
Total porosity of these systems (Fig.26), without taking natural fractures into account, is primarily composed of:
o
o
o
Non-clay matrix porosity

Clay porosity
Kerogen porosity
If natural open, or partially open, fractures are present the fracture void space is measured by logs as porosity and is
included in total log porosity. Natural fractures can be detected in cores and well image logs; natural open fractures can also be
inferred from well test interpretations. Full diameter laboratory porosity analysis may be macro fracture sensitive; however, it must
be noted that the GRI (Gas Research Institute) does not consider this an appropriate measure for shale gas, given the difficulty in
spreading helium in the sample can result in underestimations regarding matrix porosity.
Matrix
KEROGEN
INORGANIC MATRIX
(Dry Clay + Nonclay
Minerals)
Fluid
Free (!kFF)
Adsorbed (!kad)
Gas
Free (!FF)
Gas
Pc bound (!PcBW)
Water
Clay bound (!CLBW) Water
Figure 26: Petrophysical model without natural fractures. If natural open or partially
open fractures were present, they would likely contain free fluids. This pore space is
measured by logging as log porosity, and will be included in the total log porosity.
Some authors recommend evaluating gas content in terms of bulk volume gas (BVG). To do so, they suggest estimating
total porosity logs and multiplying this value by total gas saturation (1-SwT) to reach the volume % of gas in the rock. If there is no
oil present, this gas volume should be representative of total gas volume. Thus, free gas volume is obtained as the difference
between BVG and the adsorbed gas volume estimated by using Langmuir isotherms.
We do not recommend estimating shale rock properties on an effective porosity basis. If someone claims to estimate
free gas with log effective porosity and log effective water saturation, they must account for a gas free clay volume, and a
resistivity of clay minerals free of gas effects. Log effective porosity has historically been understood as total porosity minus shale
24
SPE 153004
porosity, as explained at the start of this report. However, in shale reservoirs it would be deemed precarious to use effective
porosity in volumetric estimations. On the other hand, if using effective porosity, deducting only the clay porosity component
instead of shale component, it must remain clear that the model thereby assumes that only irreducible fluids are present in the clay,
and that free, absorbed, and adsorbed gas can only be stored by the non-clay matrix porosity and kerogen respectively. Even so,
this fails to clearly define the treatment of absorbed gas in clays, if any, that affects porosity and resistivity logs. It is therefore
recommendable to estimate gas volume using total porosity as opposed to effective porosity, because uncertainty can be lower.
The evolution of Magnetic Resonance techniques in these reservoirs is highly anticipated. Certain groups17 are
investigating the possibility of using said techniques to reliably establish total porosity in addition to identifying Free Gas filled
NMR porosity, which would in turn facilitate the direct formulation of BVG independent from parameters such as matrix density
or water salinity of the formation. These works have also created doubt regarding the contribution of adsorbed gas to NMR signals
and that diffusion effects can be neglected; the last being contrary to the desired objective of separating gas and liquid signals.
These experiments are receiving the attention of many experts and log analysts18,19.
Anyway, volumetric estimation of total porosity can be resolved with conventional logs if the following can be
determined accurately:
1. The volume % of lithological components
2. The matrix values of lithological components (including kerogen)
3. Fluid parameter values
1. At least the principle lithological components, those that most affect the results, must be identified and quantified. Evaluating
the % of kerogen, isolated carbonate components (calcite, dolomite), dense minerals and clay minerals represents the biggest
challenge as they are difficult to identify. Pulsed neutron techniques represent an advance in improving estimations of calcite,
dolomite, clay-rich components, dense minerals and quartz-feldspar-mica. Establishing the volume percentage of special minerals
with these techniques still involves high levels of uncertainty, according to communications held with some users of these services.
The percentages of special dense minerals tends to be very low and difficult to quantify, as is distinguishing calcite from dolomite,
factors that introduce high levels of uncertainty. Kerogen volume cannot be obtained from geochemical logs; therefore, the
interpretation must be combined with the kerogen value obtained by the above mentioned methods. The adoption of a mineral
model based on laboratory analysis and the calibration of log results are very important and necessary to apply these techniques.
The incorporation of detailed description of cuttings, cores and sidewall cores to the analysis can provide excellent added value
because they are a direct source of information.
2. The great difference between matrix density in calcite and dolomite (2.71 against 2.85g/cc) results in considerable errors if the
percentages of each mineral are not accurately estimated. Nevertheless, as an advantage factor, the matrix values included above in
parenthesis can be reliably used as indicators of extremes for each mineral. On the contrary, matrix properties of pure and dry clay
minerals are generally unknown, and they may be highly variable at reservoir conditions. The matrix properties of kerogen are not
well known and, normally, it is assumed that density is around 1 1.5 g/cc; information on matrix travel-time, %tma, is scarce as is
matrix neuron porosity, $Nma, in regards to both kerogen and clay minerals.
3. In regards to the properties of hydrocarbons, it is believed that in these reservoirs analysts face the same uncertainties as in
conventional reservoirs. However, the relationship between porosity and fluid mobility appears to be highly affected by viscosity,
capillarity and relative permeability; little is known about this and low prediction levels results in higher levels of uncertainty. The
extreme low capacity of shales to produce water makes formation water salinity difficult to measure due to a lack of representative
production samples. If considerable amounts of water were produced (by natural fractures, for example) it would be likely free
water. Moreover, free water, and water trapped by capillarity in the non-clay matrix, may differ in terms of electrical conductivity
from water adsorbed in clay. Electrical conductivity of clay adsorbed water is also a function of clay CEC and not only of the
salinity of clay adsorbed water.
Finally, although it has not been possible to verify that all laboratory methods are sufficiently accurate to use them as
conclusive reference, controlling the results with laboratory porosity analysis will always prove useful. Furthermore, it is important
to remember that laboratory analysis on core plugs do not measure porosity of natural (meso-macro) fractures, whilst logs estimate
total system porosity.
Given the complexity described above, analysts have exerted considerable efforts in using advanced tools and techniques
to interpret porosity with logs. The most commonly used applications are those based on probabilistic models using special
(geochemical) logs and conventional logs mentioned before in point 1. If the results achieved are coherent and the result of the
SPE 153004
25
operations demonstrates that the parameters established by using these techniques can be considered reasonably reliable, it may be
recommendable to apply them in field work, despite their complexity and high costs.
However, on the other hand, there will always be the possibility of using a simpler model to resolve the reservoir porosity.
These models are more important in instances of limited log availability. Below, one of the most used variants, like the calculation
of porosity from density logs, will be discussed.
The most simple and common form for this tool (eq. 11) produces total porosity values that are too high and unrealistic
when using a matrix density value that does not consider the effect of kerogen, as kerogen density is very low and will affect
formation density values (!b) of this tool.
"T =
(!m # !b )
(!
"T =
(11)
# !f )
(!ma # !b )
(!
ma
# !f )
(11bis)
Where:
$T
!m
!ma
!b
!f
= total porosity
= matrix density without TOC
= matrix density with TOC
= formation density as established through logging
= fluid density
In order to use this model, it is first necessary to obtain information from additional logs that facilitate the estimation of
matrix density that include the kerogen effect (!ma, eq. 11bis). NGT potassium and thorium curves, added to other data such as the
volumetric photoelectric factor U (!b*Pef) and, in general, the entire measured or computed lithological curve that is not porosity
dependant, will facilitate the compilation of multi-mineral modules to estimate matrix density including kerogen. With these
models, and based on density logs, the following is established:
!ma= (vol Min1*!min1) + (vol Min2*!min2) + (vol MinN*!minN) + (vol Ker*!ker)
Kerogen volume must be estimated separately, with TOC weight % data (w%TOC) and the models presented before in
eq. (1), (2), and (3) or with the following equations (ref.16):
Ker (vol %) =
wTOC
!TOC
(!
# "T ! f )
(12)
Where:
Ker = kerogen volume (%)
= TOC matrix density, theoretic values should be between 1 and 1.5 (g/cc)
!TOC
WTOC =% TOC weight
Equation 12 proposes a way to correct the density log value given that w%TOC is measured in dry conditions, whereas !b
includes formation fluids. If imbalances are observed when using w%TOC estimated with logs, the result of the equation can be
adjusted by a k coefficient; in the interests of simplicity, here it is assumed that this value is equal to 1. A fixed fluid density figure
can be used, or it can be estimated if Sw is known, by using the following equation:
! f = ! g (1 " SwT ) + ! w SwT

Where:
SwT
!g
!w
= total water saturation (%)

= gas density (g/cc)
=formation water density (g/cc)
(13)
26
SPE 153004
It would be recommended to calibrate this estimation with grain density measured in laboratory and with matrix density
achieved through XRF, XRF, SEM analysis, including TOC.
If there is insufficient log information to estimate mineral percentages, equation 14 may be used as it corrects the kerogen
effect on density measured by logs (!b):
"T =
wTOC
!TOC
(! m # ! f )
! m # ! b '' ! m
&
# wTOC + 1$$
%
(14)
Where:
=matrix density without TOC
!m
For this purpose, a log interval that is almost unaffected by kerogen is selected (fig.27) to estimate matrix density taking
into account the effect of shale porosity; as a result, the matrix density must include that of dry clay. If porosity is higher than zero,
wet shale log density will be lower than dry shale.
If w%TOC information and core grain density is available, matrix density without TOC can be estimated to later be
applied to the entire logged interval of interest, as follows:
(m =
( ma * (1 ' wTOC )
(15)
#
&w
1 ' $$ TOC * ( ma !!
"
% (TOC
Where:
=matrix density with TOC
!ma
!ma < !m " w%TOC>0
Figure 27: Shale gas logs; upper intervals with low presence of kerogen.
SPE 153004
27
Total log water saturation

Prior to estimating water saturation with porosity and resistivity logs, certain key features of the rock must first be
established, in addition to identifying the appropriate model in order to do so. The estimation of water saturation with porosity and
resistivity logs, demands for the knowledge of certain key parameters of the rock, beside to find the appropriate model. It has been
demonstrated that in shale reservoirs, gas and oil can be produced; therefore, a hypothesis must be established regarding porous
system distribution and the preferential place where the pore saturating fluids are retained. In a hypothetical shale gas scenario,
different fluids are distributed between kerogen pores and the inorganic matrix as shown in Fig. 28.
Figure 28: Fluid distribution in porous shale; includes natural

fractures.
Fig. 29 shows the image of a core with natural fractures in which mineralisation over the open fractures walls
demonstrates that these are not induced fractures. In Fig. 30, the interpretation of Passey (2010) regarding fluid distribution uses a
composite figure of colour fluid patterns overlying a microscopic image.
Figure 29: Naturally fractured shale

core with mineralisation on the
fracture walls (Gale et al., aapg
bulletin n4, 2007)
Figure 30: Fluid distribution in the porous system according to

Passey; free gas (dark red) in kerogen pores, adsorbed gas in the
kerogen pore wall (light red) and water (light blue) in the inorganic
matrix that looks microfractured (Passey, 2010).
As stated before, if it is possible to estimate total water saturation, then an estimate of free gas content can be achieved by
deducting the adsorbed gas content (as calculated by using TOC, rock density and the Langmuir isotherm) from total gas (1-SwT).
To achieve this, application of the widely known Archie model for the determination of total water saturation of rocks, based on
porosity and resistivity logs, will be analyzed as follows:
28
SwT = n
a ! Rw
"T m ! Rt
SPE 153004
(16)
To apply Archie, in addition to total porosity, the following information is necessary:

1.
2.
3.
Resistivity of the rock (Rt)

Rock electric parameters (a,m,n)
Salinity of the formation water (Rw)
1. Resistivity of the formation is measured with well known industry logs. Either Induction or Laterolog well logging tools can be
used, depending on the drilling fluid type and level of rock resistivity. Vertical resolution of these logs is not ideal; however, the
industry has had to overcome this limitation, which can affect correct correlation with porosity logs or gamma ray logs, just to
name some cases.
2. It is not well known the existence of large databases of laboratory measurements of electrical parameters a, m and n of shale gas
reservoirs. Cementation exponent (m), saturation exponent (n) and coefficient (a) have been discussed at depth, but rarely in
regards to unconventional reservoirs. It is therefore understandable that a trial and error process has been used to obtain this
information to date.
3. It is likely that the predominant water volumes stems from water adsorbed in clays and irreducible water or capillary water in
the non-clay-rich matrix. The Archie model does not differentiate the electrical contribution of water saturating pores of the rock,
and uses a sole value for Rw, resistivity of formation water. Evidently, this simplicacin can turns out to be erroneous when
different electrical contribution exists from adsorbed water and free water. Rw value can be obtained by an Archie derived method
that can be applied in porous and permeable reservoirs which have a bottom water leg. Apparent water resistivity, Rwa, can
therefore be calculated for that water bearing section; then, an average, or in other cases minimum, Rwa value can be used as Rw
in the hydrocarbon interval. For shale gas, a Rwa calculated over non-kerogen intervals (intervals with no kerogen content) should
be adopted.
Archie is a simple and easy method to use, however, it may be deemed too simple to be appropriately applied to the
complexity presented by unconventional reservoirs. On the other hand, it is advisable to evaluate or question whether the input
parameters of the extended solutions of the Archie method used in shaly formations, like Indonesia or Simandoux amongst others,
can be estimated with reasonable certainty, or whether its application on shale gas reservoirs make sense; that is to say, the
pretension to deduct the conductivity of clay minerals from which pure conductivity is unknown. For example, it would be
essential to reconsider the use of the term shale in the saturation equations; it must be replaced by clay and, concurrently, it would
be necessary to find a value for the little known clay resistivity instead of use shale resistivity selected from an adjacent shale, as is
common practice. Finally, after such an effort, it is likely that the electrical parameters a, m and n are going to be selected by trial
and errors, until reliable data bases or laboratory experiences exist to determine those parameters.
If log calculations and gc (gas content) with the Langmuir isotherm are completed accurately, they should demonstrate
the following:
GIIPTot (derived from SwT) & GIIPad (derived from Langmuir gc)
t
When using porosity and Sw parameters derived from logs, it is essential to remember that density of adsorbed gas may
be greater than that of free gas. This depends on the state of both at reservoir conditions as they can differ.
Determination of Net Thickness (h)
Once the potentially productive intervals have been identified, TOC, porosity and water saturation parameters have been
established based on log information, it is necessary to establish the net thickness of the reservoir to estimate in-place and
technically recoverable resources to progress with reserves calculations. Net thickness (h) is an extremely important parameter,
and together with the reservoir area, it facilitates the estimation of net rock volume. Specialists use the following to obtain h:
SPE 153004
o
o
o
o
o
o
29
Kerogen/TOC content
Hydrogen Index
Total gas saturation (1-SwT)
Porosity
Rock bulk density (!b)
Lithology - Brittleness Index (BI)
Many of these parameters are difficult to precisely determine in shale gas, even when using the best core sampling and
electric/radioactive logging techniques. Some operators do not use porosity and water saturation cut-off; in personal
communications with the author, some operators have expressed that they do not use any cut-offs whatsoever. Despite the
difficulties in estimating accurate net thicknesses, identification of top and base of the intervals of interest is often possible using
logs, as expressed above. Given that unconventional resources require artificial improvement of the reservoir quality for
commercial production, many authors maintain that it is necessary to study the mechanical properties of the rock as they can assert
an important influence on stimulation design and the resulting improvement in post-stimulation production.
The most productive and best quality areas are known as "sweet spots". These areas can be detected in wells by cuttings and
mudlogging and well tests and logs; however, it is later necessary to verify their areal or lateral continuity using seismic or
geological modelling tools, although this in turn gives rise to a number of technical issues that will not be discussed in this report.
However, the results of seismic experiments are highly anticipated as they can have an enormous impact in evaluating the areal
extension of net reservoir properties.
Having established the potentially productive intervals at the well level, it is necessary to gain information on the above
mentioned core, log and testing parameters to more accurately establish net pay through application of cut-offs, if considered
necessary or technically correct. Regarding well testing, it must be remembered that, normally, these reservoirs are commercially
productive post-stimulation; therefore, pre-stimulation information achieved via testing is important, but not entirely conclusive.
Remember this when analysing the information of old wells` tests. Given the current design of horizontal wells or highly deviated
wells to produce some of these reservoirs, establishing the quality of net thickness intervals to locate and drill the well in the
potentially most productive zone is a key factor. To this end, obtaining the best quality and quantity of raw and processed
information is essential to derive the parameters used in net thickness and quality determination; these should in turn be integrated
with seismic and geostatistical studies.
GIIP, TRV and reserve volumetric estimations
In unconventional reservoirs it is recommendable to estimate recoverable volumes and reserves by interpreting reservoir
decline. However, this information is rarely available during the initial stages of field development. Therefore, volumetric methods
are used; furthermore, it is always recommendable to complement performance analysis with volumetric estimations. Volumetric
information is also an essential requirement of simulation studies in initialising the simulator. Total gas in-place based on log
interpretation and adsorbed gas in-place based on adsorbed gas content derived with Langmuir isotherms are estimated using the
following equations:
Total gas in-place, established with logs:
) 1 &
!9
GIIPTot = 43560 # A * h **T # (1 ! SwT )# ' $ #10 "
'B $
( g%
Where:
GIIPTot
SwT
$T
Bg
A
h
= total gas initially in-place (Bcf)

= total initial water saturation (%)
= total initial porosity
= initial formation volumetric factor
= area (acres)
= average net thickness (ft)
Adsorbed in-place gas, derived with Langmuir isotherm:
(17)
30
GIIPad = gc ! Den ! A ! h ! C
SPE 153004
(18)
Where:
GIIPad
gc
Den
C
= adsorbed gas initially in-place (Bcf)

= units conversion factor; 1.3597*10^-6
The volume of free gas in-place is obtained by deducting adsorbed gas from total gas.
GIIPf = GIIPTot ! GIIPad
(19)
Some authors prefer to work with effective porosity and effective water saturation obtained from logs, rather than total, to
estimate free gas. Then, based on core analysis various authors state that free gas is the difference between effective pore space
and the fraction of effective porosity occupied by adsorbed gas.
It is the opinion of this author that estimating effective porosity and effective water saturation in these reservoirs can
create greater levels of uncertainty than the estimation of total porosity and water saturation estimation, the reason for which it is
recommendable to use the total porosity and water saturation models. Logs read "all" the pore space of the rock occupied by any
type of fluid; therefore, log total porosity represents the total volume occupied by fluids, including those adsorbed. However, it is
worth noting that the space occupied by gas may not react to certain porosity tools, due to physical limitations of the measuring
process. This can result in deficient estimations, therefore, log analysts must double the efforts to neutralise the effect of gas in
estimating porosity with logs. On the other hand, resistivity tools also read the whole resistivity of the rock and do not manage to
discriminate if the free gas produces different effect than the adsorbed gas in the conductivity of the rock. It is not possible to
differentiate between free gas and adsorbed gas by means of resistivity tools.
From GIIP established via volumetric methods, it is possible to establish the technically recoverable volume (TRV) by
applying a reliable recovery factor (RF) for the reservoir in question. This recovery factor may be established based on well or
field performance, and/or detailed simulation studies in the case of a mature field. However, in the early stages of production it is
difficult to obtain a reasonably certain recovery factor; therefore, analogies to nearby fields should be made. Even so, the great
heterogeneity in the productivity of these reservoirs can reduce the reliability of the analogy and the estimations may present
initially a high level of uncertainty. It is assumed that for high reservoir pressure, the volume of free gas is greater than the
adsorbed gas and, on the contrary, that the volume of adsorbed gas is more important at lower reservoir pressure. The initial
contribution of adsorbed gas to production is related to desorption pressure, with adsorbed recoverable volume also closely linked
to abandonment pressure. The recovery factor of free gas may differ from that of adsorbed gas. To investigate this in more detail, it
would be advisable to enter into a specific discussion on the matter with reservoir engineers.
It is essential to clarify that well or field reserves stem from an economic test performed over the technically recoverable
volume and considering the expiration date of the contract or concessions.
Analogy
During the evaluation phase, pilot plan or early stages of delineation and developing shale gas, it is highly likely that
reservoir parameters are selected from analogue reservoirs more mature and developed. However, in some areas, the experience is
scarce or absent given that exploitation is in the planning or limited pilot stage of previously unexploited territories.
SEC regulation number S-X 210.4-10 defines analogous reservoir in terms of reserve estimation. Analogous reservoirs,
as used in resources assessments, have similar rock and fluid properties, reservoir conditions (depth, temperature, and pressure)
and drive mechanisms, but are typically at a more advanced stage of development than the reservoir of interest and thus may
provide concepts to assist in the interpretation of more limited data and estimation of recovery. When used to support proved
reserves, an analogous reservoir refers to a reservoir that shares the following characteristics with the reservoir of interest: (i)
Same geological formation (but not necessarily in pressure communication with the reservoir of interest); (ii) Same environment
of deposition;
SPE 153004
31
(iii) Similar geological structure; and (iv) Same drive mechanism. Reservoir properties must, in the aggregate, be no more
favourable in the analog than in the reservoir of interest.
The industry is currently debating the application of distance criteria between the reservoir in question and the analogous
reservoir. In effect, this is an aspect that the SEC and the SPE have left open to some degree of interpretation, as their respective
regulations and guidelines fail to mention this topic. Specialists may adopt concepts from more mature reservoirs to estimate some
parameters of the reservoir of interest, giving priority to certain geological and petrophysical characteristics without prioritising the
distance between them. This process has led to compare reservoirs located in different and distant basins, in some cases, located in
different continents. To compare, for example, productivity, performance or definitive volumetric parameters, this is not
acceptable when the aim is to book a company's proved reserves, considering the SEC requirement that for proved reserves, the
analogous reservoir must share the same geological formation with the reservoir of interest. Therefore, it is essential that for
proved reserves the analogue must be located close to the reservoir of interest, or in some cases, that they are nearby reservoirs. It
is also widely believed that for unproved reserves (probable and possible reserves) it is convenient that the analogous reservoir be
in the same area or basin. The terms "nearby" and area" try to reflect a gradation of the distance with regard to the immediate,
contiguous or "neighbouring" nearness. To apply different criteria, the registering entity must demonstrate "reasonable certainty"
in regards to the estimations.
As is understandable, establishing a good analogy is not a simple task for shale gas. The areal extension of the formation
is often verifiable, whereas reservoir properties and well productivity have shown to be greatly variable in a relatively short
distance. These reservoirs have had to be stimulated, and the response to the stimulations has varied a great deal. To establish an
analogy and ensure that the post-stimulation results could be uniform, the stimulation method applied must be similar to the
stimulation method used in the reservoir of interest, and it must be demonstrated that its design has been tested, consistently and
repeatedly, in the analogous reservoir and provided reasonably reliable results. Comparing similar extraction methods is
paramount; the productivity of a horizontal well is rarely similar to that of a vertical well, the reason for which, to apply the
relationship horizontal / vertical productivity of a more developed area in new areas, there will be necessary to rely on repetitive
and reliable information that make it compliant with current regulations and industry usage.. The analogy has to be, in addition,
nearby and must assure the criterion of reasonable certainty in the continuity of production. Increasing consensus exists to consider
that the analog is the reservoir in itself, e.g., pilot projects. Certain data, such as the measurement of TOC content, Hydrogen
index, kerogen maturity, adsorbed gas content (Langmuir), total gas content (canister desorption) are important properties of these
reservoirs; therefore, they must be considered when establishing an analogy. Recently, SPEE has issued Monograph 3 where
analysts can find interesting concepts on statistics and a proposal of practical applications in undeveloped reserves determination
now under discussion in the industry20.
Conclusions
This paper has discussed the state-of-the-art of shale gas petrophysical evaluations; in addition, experiences of third
parties, operators and researchers, have been transmitted together with some own opinions about the criteria and the technologies
for the evaluation of cores and logs.
Petrophysical evaluation of a shale gas is often a difficult task that requires the use of special data and techniques to derive
reservoir parameters. These reservoir rocks must be carefully conceptualised in order to avoid miss calculation methodology, or to
avoid applying imprecise and less representative options. Some important conclusions that can be derived are:
o
o
o
o
o
o
o
Volumetric can be imprecise when estimating in-place shale gas, but it is one of the few techniques which is capable of
completing this task in the early evaluation and development stages.
Sample density must be high and appropriate to reduce the uncertainty inherent to petrophysical interpretations.
Laboratory techniques must evolve towards standards that may give greater confidence in petrophysical analysis results21.
In the absence of such standards, it is advisable to request the services based on protocols defined by operator and in
consensus with service providers.
If historic laboratory data from different providers is used, it is important to understand the laboratory protocols followed
by each one to ensure that the analysis is sufficiently comparative.
The more complex the interpretation, the higher the uncertainty in regards to the results achieved; therefore, it is
recommendable to adopt relatively simple and representative methodologies for the reservoir.
At least the principal lithological components of the rock must be determined, as this can considerably impact stimulation
design.
Locating the interval of interest can be relatively simple, but estimating net pay is more complex as, in addition to
32
o
o
SPE 153004
porosity or water saturation, it requires the use of unconventional parameters such as TOC or Ro. However, some
operators do not use porosity cut-off and/or Sw and others do not employ any cut-offs whatsoever.
In-place gas estimated by logs must be complemented by in-place gas established by TOC and the Langmuir isotherm;
both must be compared to the direct method of gas desorption, or, canister gas. Cross-check the estimation results.
To use analogies, certain basic rules must be followed in their creation. To that end, definitions provided by professional
entities and market regulators can be very useful.
Acknowledgements
We wish to thank Repsol YPF for proving permission to present this work. We also appreciate the opinions offered by
Tomas Zapata, Omar Aguirre and Aldo Montagna in regards to technical support of this work. We would like to thank Roberto
Redondo for his efforts in controlling the coherent structure of the document and assisting with certain technical definitions.
Glossary
COGEH
GIIP
GIIPad
GIIPtot
t
GIIPTot
t
Inorganic Matrix
K
Ker
Matrix
Matriz
Mbo
Logs
Reserves
SPEE
STOIIP
TOC
TRV
w%TOC
= Canadian Oil and Gas Evaluation Handbook

= gas initially in-place
= adsorbed gas initially in-place based on TOC and the Langmuir isotherm
= total gas initially in-place based on the direct method, or, canister gas
= total gas initially in-place based on log interpretation
= mineral part of the matrix composed of quartz, carbonates, clays and other minerals excluding kerogen
= permeability
= Kerogen
= solid part of the rock composed of various minerals, excluding the fluid contained in the pores (as commonly used in
petrophysical terminology)
= fine grained material located at the interstitial space of clastic rocks.
=one thousand barrels of oil
= electric and radioactive logging (e-logs)
= volume of technically and economically recoverable hydrocarbons (including cuts due to expiring contract/concession)
= Society of Petroleum Evaluation Engineers
=stock tank oil initially in-place
=Total Organic Carbon
= volume of technically recoverable hydrocarbon
= % TOC weight
About the Authors

Juan Carlos Glorioso currently consults the REPSOL YPF Reserve Control Management in Madrid, Spain. He is a graduate in
Chemistry at the Escuela Industrial Superior, Universidad Nacional del Litoral, Santa Fe, Argentina. He has 35 years
experience in different positions as log analyst and petrophysicist at YPF, Maxus and Repsol YPF.
Aquiles Rattia is currently the Corporate Reserves Director at Repsol YPF in Madrid Spain. He holds a BS degree in Petroleum
Engineering from Universidad Central de Venezuela, and an MSc and Ph.D from Pennsylvania State University in the same
speciality. He has more than 35 years of worldwide experience in different positions as Reservoir Engineer, Asset Manager,
R&D and Project Development.
SPE 153004
33
APPENDIX I - Shale Oil briefing

Definitions
Many concepts discussed regarding shale gas can also be applied to shale oil. However, there are some key differences.
When discussing special characteristics of a shale oil in regards to a shale gas, it is necessary to use terminology that distinguishes
different shale oils according to the extraction method employed. Therefore, it is important to note that:
Oil shales: are shales containing a significant amount of immature organic matter (kerogen) from which significant amounts of oil
and combustible gas can be obtained by destructive distillation of rock extracted by surface mining means. When these rocks are
directly used as fuel, potential oil volumes that they may contain normally are not recognised as reserves, pursuant to international
accounting regulations that define gas and oil industries.
Shale oil: is in-situ produced oil from shale rocks rich in mature organic matter that simultaneously act as source, trap and
reservoir rocks for such oil. Production comes from wells drilled to different depths and, as matrix permeability of these rocks
normally measures only a few hundred nanodarcies, to produce economic and commercial rates, natural fractures and/or formation
stimulation by means of hydraulic fracturing is a prerequisite.
The comments below refer to shale oil and, in the interest of simplicity, they assume that the rocks analysed are water and
oil saturated.
Kerogen and TOC
As mentioned above, kerogen types I and II are those that tend to produce liquid hydrocarbons. Under extreme conditions,
this is also possible with kerogen type III. In order to detect the type of kerogen present, a modified Van Krevelen diagram,
presented in fig. 5 of the main report may be used. Geochemical analysis providing data to estimate thermal maturity, TOC and oil
content of the sample is used to estimate oil in-place in a shale oil. This geochemical estimation is contrasted with the estimation
of oil in-place based on logging data. To this point, at first glance, the process is conceptually similar to the process described in
the main report.
However, one of the differences is that with shale oil magnetic resonance (NMR) logging may be used to derive log TOC
in addition to other techniques already discussed. Effectively, this technique is based on the fact that NMR porosity treats kerogen
as it would be any other solid in the rock matrix, whilst density logs see it as any other fluid, and as part of the porosity of the
rock. At the same time, neither NMR nor density is affected by free gas which is absent. As a result of this, if an appropriate
matrix density is used, total NMR porosity will be lower than density porosity (fig.31) due to the presence of kerogen rather than
to the effect of gas.
Figure 31: Differences in

porosities derived from both
NMR and density logging,
caused by high kerogen
content in a shale oil interval
(Oil Field Review, Volume
22, n 4).
Kerogen volume obtained as a result of comparing NMR and density is converted into a weight % of TOC by means of
the equations presented in (20), calibrating the results with TOC data obtained in laboratory.
34
&) #
TOCw% = (*D ( *NMR )' $$ ker !!
% )b "
SPE 153004
(20)
Where:
TOC
$D
$NMR
!ker
!b
= total organic carbon (w%)

= density porosity (dec)
= magnetic resonance porosity (dec)
= kerogen density (g/cc)
=formation density (g/cc)
This parameter is then used to obtain oil content information, as described below.
Oil content
Fischer Method
One of the methods used to measure oil yields in organic matter-rich shale is the modified Fischer method (fig.32).
Figure 32: Kerogen volume

derived from logging
correlates with oil content
obtained from the Fischer
assay (Oil Field Review,
volume 22, n 4).
It is a retorting and distillation method suitable for oil shale as it brings information to evaluate the production potential of
immature kerogen. Oil content figures are often high as they represent both free oil and potential oil. For that reason, for shale oil
the following method is particularly useful.
Rock-Eval and Source Rock Analysis (SRA) methods
These methods, in practice by several laboratories and providers (IFP, Weatherford) seem to be more suitable for shale
oil. They are used in order to investigate the hydrocarbons source capacity of rocks, the type of kerogen, its level of maturity, and
TOC content among others; one of the providers has provided the following outline: The method was designed to help identify
and characterise source and reservoir rock by heating geological samples (e.g. outcrops, cuttings, conventional cores and
sidewall cores) to a programmed temperature in an inert atmosphere. This analyser quantitatively establishes the amount of free
hydrocarbons (S1) and the amount of hydrocarbons yielded by thermal cracking of non-volatile organic matter (S2) with a flame
ionization detector (FID). The amount of CO2 produced during kerogen pyrolysis (S3) is determined using an infra-red (IR)
detector (Fig. 3). The device also establishes the temperature at which maximum release of hydrocarbons from the cracked
kerogen during pyrolysis occurs (Tmax) and the Total Organic Carbon (TOC). The device can be used in the field or in laboratory
to evaluate organic properties, type of kerogen, thermal maturity, and source potential of an area or to identify potentially
productive gas and oil areas. (Weatherford).
SPE 153004
35
Figure 33: Source rock, maturity and

kerogen type analysis test cycle.
Peak S1 that represents the amount
of free hydrocarbons occurs at
350C; it can be altered by drilling
mud contamination (Pennsylvania
Geological Survey).
The amount of free hydrocarbon (S1) is used to estimate the oil content of the rock in shale oil evaluation. The data is in
mg of oil per gram of rock (mg/g) and here it will be identified with the following abbreviation: S1y (S1 yield). According to
publications, the S1y value tends to vary between 0.5 and 6 mg/g. It is important to consider and document TOC content and the
Ro maturity index of the sample.
S1y oil content and its corresponding TOC level, measured in laboratory, provide a correlation that can then be used to obtain log
derived S1 yields (S1yl), using log derived TOC (fig.34).
18
Figure 34: Correlation of TOC

content (w%) and S1 free oil content
of Rock-Eval or SRA laboratory
tests.
16
TOC (w%)
14
12
10
8
6
4
2
0
0
S1 mg/g
Estimation of oil in-place with oil content (STOIIPS1yl)

In-place oil derived from information obtained via geochemical testing can be estimated by employing a similar equation
to the one presented above for gas, adapting the constants and corresponding fluid content that will now represent the oil yield of
the sample.
S1
'
STOIIPS1 yl = %% .001* yl
(o
&
$
"" ! Den ! Area ! h ! C
#
Where:
STOIIPS1y
= oil initially in-place (Mbo)
(21)
36
S1yl
!o
Den
A
h
C
SPE 153004
= oil content (mg/g)

= oil density (g/cc)
= area (acres)
= average net thickness (ft)
= units conversion factor; 7.758
Estimation of oil in-place with logs (STOIIPTot)

As discussed in the main report, estimating recoverable volumes and reserves by interpreting reservoir decline is highly
advisable, although complementing the performance analysis with volumetric estimations is also recommended throughout the
development of the reservoir. Volumetric is employed when starting up, or even before, production. Total oil in-place can be
estimated by volumetric using the following equation:
& 1 #
STOIIPTot = 7.758 ' A * h *)T ' (1 ( SwT )' $$ !!
% Bo "
(22)
Where:
STOIIPTot
SwT
$T
Bo
A
h
=total oil initially in-place (Mbo)

=total initial water saturation (%)
=total initial porosity
= initial formation volumetric factor
=area (acres)
=average net thickness (ft)
In-place oil volume based on log interpretation (eq. 22) must be compared to the estimation made using oil content from
geochemical analysis (eq. 21). Any resultant differences must be investigated, and for each field analysts must establish the level
of tolerance that can be applied to the differential of both values for them to be considered coherent and reliable. This will likely
take time, but the results must be coherent in order for the method to be considered reliable. As production advances and decline
rates can be defined with reasonable certainty, reserves estimated by performance will provide an excellent control framework to
refine volumetric. This method will be less important in long-term developing and producing reservoirs, however, it will remain
important for immature reservoirs, for which volumetric of on-production mature reservoirs can serve as excellent analog for the
estimation of hydrocarbons in-place, recovery factors and reserves of an immature reservoir.
SPE 153004
APPENDIX II - Basic workflow
APPENDIX III - Brittleness index based on lithological components

BI= (Q+Carb) / (Q+Carb+Clay+TOC)
Where:
BI
Q
Carb
Clay
TOC
= brittleness index
= Quartz
= Carbonates
= Clay
= TotalOrganicCarbon
37
38
SPE 153004
SPE/AAPG/WPC/SPEE/SEG, Guidelines for Application of PRMS, November 2011.
Oil Field Review, Producing Gas from its Source, autumn 2006.
Oil Field Review, Producing Gas from its Source, autumn 2006.
3
Lewis, R., et al., New Evaluation Techniques for Shale Gas Reservoirs, Reservoir Symposium, 2004.
3
Lewis, R., et al., New Evaluation Techniques for Shale Gas Reservoirs, Reservoir Symposium, 2004.
2
Passey, Q. et al., From Oil-Prone Source Rock to Gas-Producing Shale Reservoir, SPE 131350, 2010.
Jarvie, D., Worldwide Geochemistry. LLC
Decker, A.D and Wicks, D.E., Log-Based Gas Content and Resource Estimates for the Antrim Shale, Michigan Basin, SPE 25910.
Luffel, D.L. and Guidry, F.K, New Core Analysis Methods for Measuring Reservoir Rock Properties of Devonian Shale, JPT, November
1992.
Asquith, G. B., Log Evaluation of Shaly Sandstone: A Practical Guide, AAPG.
Olson, R, et al., Mercury Injection Capillary Pressure (MICP) A Useful Tool for Improved Understanding of Porosity and Matrix
Permeability Distributions in Shale Reservoirs, AAPG 2008.
10
Patchet, J., An Investigation of Shale Conductivity, The Log Analyst, November, 1975.
11
11 Luffel, D.L., et al., Matrix Permeability Measurement of Gas Productive Shales, SPE 26633.
12
Serra, O., Fundamentals of Well Log Interpretation.
13
Mendelson, J.D., Petroleum Source Rock Logging, MIT, 1985.
14
Passey, Q., et. al., A Practical Model for Organic Richness from Porosity and Resistivity Logs, AAPG V.74, n12, December 1990.
15
Creney, S. and Passey, Q., Recurring Patterns of Total Organic Carbon and Source Rock Quality within a Sequence Stratigraphic
Framework, AAPG V77, n3, March 1993.
16
Sondergeld, C.H., et al., Petrophysical Considerations in Evaluating and Producing Shale Gas Resources, SPE 131768.
17
Sigal, R., et al., Laboratory NMR Measurements on Methane Saturated Barnett Shale Samples, Petrophysiscs, February, 2011.
18
Worthington, P., The Petrophysics of Problematic Reservoirs, JPT, December 2011.
19
Ramirez, T, Klein, J, Bonnie, R, and Howard, J, Comparative Study of Formation Evaluation Methods for Shale-Gas Reservoirs:
Application to the Haymesville Shale (Texas), SPE 144062.
20
21
SPEE Mongraph 3, Guidelines for the Practical Evaluation of Undeveloped Reserves in Resource Plays, December, 2010.
Handwerger, D:A:, Suarez-Rivera, R., Vaughn, K., Keller, J., Improved Petrophysical Core Measurements on Tight Shale Reservoirs
Using Retort and Crushed Samples, SPE 147456

5 - Spe 153004 MS PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

5 - Spe 153004 MS PDF

Uploaded by

Copyright:

Available Formats

SPE 153004

Unconventional Reservoirs: Basic Petrophysical Concepts for Shale Gas

Copyright 2012, Society of Petroleum Engineers

Unconventional Reservoirs: Basic Petrophysical Concepts for Shale Gas

Figure 2: Core in laminated shale (left:

Clay bound (!CLBW) Water

Figure 4: SEM showing pores in organic matter.

Unconventional Reservoirs: Basic Petrophysical Concepts for Shale Gas

Type I: rich in algal and amorphous organic matter

Figure 5: Carbon, hydrogen

Vol ker ! " ker

= total organic carbon (lbf/lbf)

Unconventional Reservoirs: Basic Petrophysical Concepts for Shale Gas

=formation density (g/cc)

Ker (vol %) = TOC ( wt %) ! f

It is often assumed that the product !ker * Cker=1

Figure 7: Plot that shows increase in

Total Organic carbon (TOC w%)

Total Organic Carbon (TOC)

Figure 8: Maturity and expulsion

Shale Oil Play

Level of Maturity (LOM) - Vitrinte

Biogenic Shale Gas

Shale Gas Play

Figure 9: Vitrinite reflectance correlation with depth in shale (Repsol

Figure 10: Relationship of Vitrinite reflectance and the level of

Unconventional Reservoirs: Basic Petrophysical Concepts for Shale Gas

Figure 11: Interpreted maturity stages; this depends on the

Figure 12: TOC content quality.

= adsorbed gas content (scf/ton)

Figure 13: Langmuir isotherm; it is widely

Vlt = 10 (" c 3!(T +c 4 ))

= Langmuir volume at reservoir temperature (scf/ton)

Unconventional Reservoirs: Basic Petrophysical Concepts for Shale Gas

GIIPad = gc ! Den ! Area ! h ! C

= adsorbed gas initially in-place (Bcf)

= total gas content (scf/ton)

Total Gas Content, scf/ton

Total Organic Carbon, w%

Figure 14: Canister apparatus used to

Figure 15: Correlation between total gas content and TOC

Unconventional Reservoirs: Basic Petrophysical Concepts for Shale Gas

Petrophysical measurements for Shale Gas

Connected and unconnected of a crushed sample

In electric and radioactive logging, pore space is as follows:

Connected and unconnected matrix porosity, added to the shale porosity

Figure 16: Types of

Removal of water and hydrocarbons from the pore system

Unconventional Reservoirs: Basic Petrophysical Concepts for Shale Gas

Figure 17bis: 3D pores, organic matter and minerals

Saturated Incremental Porosity

Saturated Cumulative Porosity

Desaturated Incremental Porosity

Incremental Porosity (%)

Desaturated Cumulative Porosity

Cumulative Porosity (%)

Figure 8: T2 transverse relaxation time distribution in Devonian shale with <

T2 Relaxation Time (ms)

Figure 19: T2 transverse relaxation time distribution

Unconventional Reservoirs: Basic Petrophysical Concepts for Shale Gas

SEM-EDS (scanning electron microscope energy dispersive spectrometry)

BVO + BVG (%)

Unconventional Reservoirs: Basic Petrophysical Concepts for Shale Gas