9 Doveton, J. H.

, 2002, The geological application of wireline logs:

A keynote perspective, in M. Lovell and N. Parkinson, eds.,
Geological applications of well logs: AAPG Methods in
Exploration No. 13, p. 115–122.

The Geological Application of Wireline

Logs: A Keynote Perspective
John H. Doveton
Kansas Geological Survey, University of Kansas, Lawrence, Kansas

ABSTRACT

W
ireline logs have served generations of geologists as the principal medium for lithostratigraphic
correlation in the subsurface. However, the expanding range of logging measurements, novel
techniques of analysis and display, and integrated geological studies have dramatically broad-
ened the scope of geological applications of wireline logs. Subsurface investigations can gain major new
insights into geology from these measurements that supplement traditional geological methods. Modern
astronomy has benefited enormously from the expansion of observation beyond the narrow bounds of the
visible spectrum to the range from gamma-ray to radio wavelengths. To some extent, our studies of geol-
ogy in the subsurface may progress in an analogous manner, but they will require innovative methods of
database creation and data display that map raw digits into imagery at all scales, with immediate geologic
impact.

INTRODUCTION

The phrase the geological application of wireline log is something of a tautology; what other applications are
there? Because wireline tools make measurements of properties of rocks in the subsurface, their logging
records are intrinsically geological. To be sure, a semantic shift could be attempted to confine wireline logs to
geophysics. However, even this proves increasingly difficult because, for example, the properties recorded by
some nuclear tools measure elemental concentrations by techniques similar to those used in geochemical lab-
oratories. As another example, the resolution of borehole electrical imagery has long passed the point at
which small-scale sedimentary features and trace fossils are observed and studied routinely. Fundamentally, a
host of rock properties is measured by logging tools, and these are part of the larger enterprise called geology.
But can logging data and their analysis by petrophysical methods be considered an integral part of geol-
ogy? Petrophysical logs are a vital source of geologic information from the subsurface (where the vast bulk of
rocks resides), which otherwise would be a geophysical world pierced by occasional geological glimpses pro-
vided by core and fragmentary drill cuttings. However, if assessments of subsurface geology from wireline
logs are included within the fold of geology, it would probably be useful to make some distinction from the
classical geology of outcrop studies.

115
116 Doveton

THE ROLE OF LOGS IN that are invisible to conventional astronomy in a way analo-
SUBSURFACE GEOLOGY gous to that of the gamma-ray spectrum, which reveals pat-
terns of striking geological significance in visually monoto-
Science, by its very nature, is an agent for change. New nous shales.
models or ways of looking at the world can gain ground only Interestingly enough, there appears to be no widely ac-
through rethinking in an expanded synthesis. As in geology, cepted collective name that distinguishes these newer astro-
developments in astronomy have changed both our under- nomical methods from classical telescopy, other than the
standing of the universe and the methods used to study it. But “new astronomy” (Henbest and Marten, 1996). Clearly, this
is an astronomer still identified by his use of an optical tele- term will seem increasingly absurd as “new” ages to “old.”
scope? How does a specialist in stellar X rays relate to core However, there are already signs that what is now novel and
astronomy? What is the appropriate name for astronomical strange will be absorbed as an organic part of mainstream
studies outside the visible spectrum? In truth, there has been astronomy. Any distinction of a “new astronomy” will simply
a radical change in astronomy in this century. Ground-based fade away as its usefulness comes to an end. Perhaps some
telescopes have been limited to a rather “weathered” view of term such as “virtual geology” might serve a similar role, as a
the universe because of atmospheric fluctuations and light term that promotes the use of logs in subsurface geology
pollution from natural and man-made sources that perturb beyond providing formation tops for a lithostratigraphic cor-
the night sky. The operation of the orbital Hubble telescope relation framework. Ultimately, methods that would now be
has moved optical observation above the disruptive atmos- termed petrophysical should become such an integral part of
phere, for fresh pictures that are rich in new detail. Perhaps subsurface geological study that it would be hard to imagine
the excitement felt by an astronomer looking at Hubble the practice of subsurface geology without them. At that
images is akin to that of the field geologist who works for the point, the geological application of wireline logs would be a
first time with pristine core from a formation in the subsur- core discipline of subsurface geology and would automati-
face that was seen previously only in its weathered outcrop cally be associated with that name. Until that time, “virtual
expression. geology” or other unsatisfactory terms probably will be used
More important, in this analogy with wireline-logging to describe newer methods that are receiving limited treat-
tool sensors, the range of astronomical observation has ment in subsurface geology manuals and textbooks.
expanded dramatically from the narrow keyhole of the visible Finally, how should the spread of petrophysical methods
spectrum (Figure 1). Studies in the broader spectrum can of subsurface geological investigation be incorporated within
help astronomy only in its fundamental purpose of a descrip- traditional geology? In considering an answer, it is important
tive mapping of the universe. So, for example, the orbiting to consider the evolution of geological investigations in this
Compton Gamma Ray Observatory has done much to revo- century. Until about forty years ago, comparatively little was
lutionize our understanding of the universe from its observa- known of the geology in the subsurface of southern England,
tions across an extraordinary six decades of the electromag- although an impressive amount had been inferred from clas-
netic spectrum. Gamma-ray images show celestial features sical field studies. On the face of it, this is a remarkable situa-

Figure 1. The electromagnetic spectrum.

 The Geological Application of Wireline Logs: A Keynote Perspective 117

tion for one of the most densely populated regions on the automated methods. Regardless of whether done visually or
planet, but it reflected the needs and concerns of society. Ear- by computer methods, the end result of correlation between
lier in the century, field geological methods were instrumental well logs is a set of formation tops that, when linked between
in the location and exploitation of coal resources and metallic borehole control, describe surfaces. The geometrical infor-
ores that fueled the Industrial Revolution. With the shift to mation in maps and cross sections is rich in its geological
liquid and gaseous fossil fuels, the higher demands on implications regarding not only tectonic history but also
groundwater, and the increasing concern about subsurface regional sedimentary architecture (Figure 2). Automated
contaminants, borehole penetration of British rocks both off- correlation methods have an additional advantage of being
shore and onshore has increased substantially. Clearly, the able to consider several wells simultaneously in a 3-D synthe-
additional information from the subsurface has advanced the sis that precludes geometrical impossibilities in their search
cause of traditional geology. However, at the same time, many for the most reasonable solution (Olea et al., 1997). A by-
critical rock properties of pore networks and fluid contents, product of this approach can be the generation of a giant
which are target parameters of subsurface investigation, are “dipmeter log” among boreholes (Figure 3); its patterns of
not accessible to the eye or hammer. Even if they were, their dip vectors show unconformities and records of compaction
systematic analysis would require numerical estimates rather and tectonic movements.
than visual descriptions. The older electrical logs responded primarily to pore vol-
Therefore, in satisfying the needs of contemporary soci- umes and fluids; this inhibited more imaginative geological
ety, tried and true field methods should be maintained vigor- applications, beyond the correlative tracing of unit bound-
ously, while new techniques and goals are incorporated. Of aries. However, in sedimentology, the study of vertical grain-
course, much of this is already happening, with field studies size differences as an indication of depositional facies encour-
of reservoir- and aquifer-analog exposures at a hierarchy of aged the recognition of signatures in log-shape profiles, and
scales that cannot be matched in subsurface studies. Even more detailed interpretation was made possible by electrical
the richest petrophysical data set is currently limited to de- borehole imagery. This depositional-environment analysis has
scription of a very narrow borehole, while the coarse scale of been termed the “second age of log analysis” by Selley (1992),
seismic resolution is still largely insensitive to important fea- who also considered a third age to be the use of logs in the
tures that can be seen on the outcrop. Reliance on core stud- analysis of diagenetic changes of leaching and cementation.
ies for subsurface geology has always been limited by the Everyone has a different reading of history, but it could
expense of recovery but, sad to say, the perceived expense of be argued that a major shift in geological content occurred
housing existing core is now leading to the loss of large col- with the introduction of nuclear logs. Up to that time, litho-
lections through cost-cutting measures, much precipitated logical typing tended to be restricted to a binary distinction
by mergers of oil companies. By contrast, the costs of digital between rocks with low resistivity (typically shales) and rocks
data storage continue to fall, so that the preservation of pet- with relatively high resistivity (“nonshales”). To be sure, the
rophysical measurements is not under immediate threat, introduction of the first nuclear log, the gamma-ray log,
other than from simple neglect—a failure of geologists to ap- maintained the focus on differentiation of shales from other
preciate the rich information content of logs concerning sub- rock types, but subsequent refinement of the gamma-ray log
surface geology. to spectral estimates of potassium, uranium, and thorium
sources has markedly extended its range of geological appli-
cation. In particular, the widespread proliferation of neutron-
SUBSURFACE VIRTUAL
density- and photoelectric-logging measurements has made
GEOLOGICAL MAPPING possible the indirect estimation of mineral composition in
lengthy sections and in multiple locations within many sedi-
Lang (1986) noted that the overwhelming use of geo- mentary basins. Computer processing of these compositional
physical logs was as an aid for lithostratigraphic correlation. profiles (e.g., Figure 4) is relatively straightforward, using
Almost certainly, this observation holds true today, but corre- inversion methods that follow the pioneering work of Savre
lation is done increasingly on geologic workstations using (1963). Although geochemical logs are much rarer in occur-
raster logs rather than by grappling with logs on paper copy rence, extensions of this inversion procedure applied to mea-
(Montgomery, 1997). Raster logs are merely electronic sured elemental concentrations led to mineralogical profiles
images of the paper copy, so correlation remains entirely a of igneous, metamorphic, and sedimentary sequences that
pattern-recognition task for the geologist. Vector logs are cur- can be considered to be normative rock compositions (Dove-
rently more expensive than raster logs but contain the digital ton, 1986).
information that can be processed for compositional analysis. Compositional mapping of formations across a region in
In addition, the digits of the vector log allow correlation by the subsurface poses special problems, the resolution of
 118
Doveton
Figure 2. Regional lithostratigraphic section from automated correlation of gamma-ray logs of the Muncie Creek (Pennsylvanian) genetic unit in Kansas. The section crosses the
carbonate-dominated Upper Pennsylvanian shelf. The shelf has a pronounced carbonate-bank margin that downlaps into shales in the northern end of a flysch foredeep in the
Ouachita Basin of Oklahoma. Adapted from Olea et al. (1998). Lithostratigraphy is indicated by gamma-ray log; graphic patterns represent percentage of shale content ranges,
with 0 to 20 shown as black, 20 to 60 as gray stipple, and greater than 60 percent as dark gray. White marks are intervals that are not correlated.
 The Geological Application of Wireline Logs: A Keynote Perspective 119

which is needed. Individual boreholes could be thought of as that is used in individual wells. In this procedure, inverse
“outcrops,” but wireline logs of them have very limited lateral matrix coefficients are applied to map grids of log properties
ranges. Logs used in compositional estimation must be cor- rather than to log responses of zones in depth. When applied
rected for environmental conditions and normalized for mea- to the compositional estimation of relatively thick formations,
surement error. Broad trends in simple (one-component) these interpretive maps (e.g., Figure 6) have been validated
compositional changes can be mapped in the three dimen- by “ground truth” provided by the more limited core control
sions of depth and geographic axes, by interpolating statisti- (Bornemann and Doveton, 1983 ).
cal moments of logs (Figure 5). Alternatively, kriged esti- Advances in technology are interesting in their own right,
mates can be mapped areally, based on semivariogram but in a cost-conscious age, practical applications and bene-
(autocorrelation) analysis of spatial variability of the logged fits are of primary concern. The Kansas Geological Survey
property, between wells. Mapping of multiple components continues to map surface geology, but it is asked more often
can be done by a 2-D extension of the inversion procedure to address problems of the subsurface, concerning fossil

Figure 3. Regional-dip plot from automated correlation of gamma-ray logs among three wells on the downthrown block of the Hum-
boldt fault, northern Kansas. The increased throw with depth shows the Humboldt fault to be a growth fault, active over long geo-
logic time; breaks in the dip pattern show an excellent match with known regional unconformities. Adapted from Olea et al (1997).
120 Doveton

fuels, water, and plumes of pollutants. The principal source

of subsurface data is the survey’s collection of hundreds of
thousands of paper wireline logs. Increasingly, this passive
archive of analog traces is being converted to digital data that
are stored in a dynamic medium. The Digital Petroleum
Atlas and the Dakota Aquifer Program are two examples of
projects in which both logging data and petrophysical images
are immediately accessible on the World Wide Web
(www.kgs.ukans.edu). In a more ambitious initiative, the sur-
vey is compiling a statewide petrophysical database of mod-
ern digital logs as the Kansas Virtual Geology Program.
Although the core of the completed database will consist of
raw digits from individual wells, the data will be coupled with
software for analysis and display of regional geology. Kansas
Virtual Geology will be available on the Web and will serve as
a virtual observatory of Kansas subsurface geology.
Special problems—such as interpolation and normaliza-
tion—are to be surmounted, but Kansas Virtual Geology will
be used to reach beyond formation surface topology to map-
ping of a wide variety of petrophysical properties and their
transforms. Have these activities distracted the survey from
its historic mission of mapping and surveying the rocks of
Kansas? It hardly seems so, when its core library is expanding
to be one of the largest in the U.S. Midcontinent. With the
increasing availability of digital cameras, digital core images
and logging data probably will be integrated closely in a sub-
surface geological synthesis. Clearly, the road forward calls
for close integration of traditional and newer methods and
information in the mapping of subsurface geology. The expe-
rience of the Kansas Geological Survey is being replicated by
other geological agencies, with variations dictated by local
needs and availability of subsurface information, principally
wireline logs.

CONCLUSIONS
As demands of society increase for more sources of
water, more energy resources, environmental cleanup, and
controls of pollution sources, studies of subsurface geology
will increase and prosper. The geological application of wire-
line logs will be a key component of this work. Some geolo-
gists might regret this development as being somewhat alien
to the subject they learned as students. However, it is to be
Figure 4. Example of a compositional profile of a sequence of
hoped that informed reflection will show this to be an exciting
Middle Ordovician strata in northern Kansas, computed from
opportunity to learn new skills—skills that open facets of
inversion of gamma-ray, photoelectric-factor, neutron, and
density logs. geology that were unimaginable a generation ago.
The Geological Application of Wireline Logs: A Keynote Perspective 121

Figure 5. Quartic shale-

proportion trend profile
of the Simpson Group
(Middle Ordovician) in
southern Kansas. Con-
tours are spaced at decile
intervals. Shale propor-
tions greater than 0.3 are
shaded; proportions
greater than 0.7 are heav-
ily shaded. The onlap
pattern captures the
major Middle Ordovician
transgression over car-
bonate rocks of the
Arbuckle Group at the
Kansas-Oklahoma bor-
der, in its passage north-
ward across the North
American craton (Dove-
ton et al., 1984).

Figure 6. Simplified
petrophysical lithofacies
map of the Viola Lime-
stone in southern Kansas,
generated by inversion of
neutron, density, and
sonic logs and imple-
mented as grid-to-grid
operations in an auto-
mated-contouring soft-
ware package. Adapted
from Bornemann and
Doveton (1983).
122 Doveton

REFERENCES CITED Olea, R. A., J. C. Davis, and D. F. Merriam, 1997, Automatic regional wire-
line log lithostratigraphic correlation in Kansas using CORRELATOR,
Bornemann, E., and J. H. Doveton, 1983, Lithofacies mapping of Viola in V. P. Glahn, ed., Proceedings of International Association of Miner-
Limestone in south-central Kansas based on wireline logs: AAPG Bul- alogic Geologists, 1997, p. 341–348.
letin, v. 67, p. 609–623. Olea, R. A., W. L. Watney, and J. C. Davis, 1998, High-resolution wireline
Doveton, J. H., 1986, Log analysis of subsurface geology: New York, Wiley correlation for sequence stratigraphic analysis in southwest Kansas, in
Interscience, 273 p. A. Buccianti, G. Nardi, and R. Potenza, eds., Proceedings of Interna-
Doveton, J. H., K. A. Zhu, and J. C. Davis, 1984, Three-dimensional trend tional Association of Mineralogic Geologists, 1998, p. 639–644.
mapping using gamma-ray well logs, Simpson Group, south-central Savre, W. C., 1963, Determination of a more accurate porosity and mineral
Kansas: AAPG Bulletin, v. 69, no. 6, p. 690–703. composition in complex lithologies with the use of sonic, neutron, and
Henbest, N., and M. Marten, 1996, The new astronomy, 2nd. ed.: Cam- density surveys: Journal of Petroleum Technology, v. 15, p. 945–959.
bridge University Press, 240 p. Selley, R. C., 1992, The third age of wireline log analysis: Application to
Lang, W. H., 1986, Correlation with multiple logs: The Log Analyst, v. 27, reservoir diagenesis, in A. Hurst, C. M. Griffiths, and P. F. Worthing-
p. 43–52. ton, eds., Geological application of wireline logs II: Geological Society
Montgomery, S. I., 1997, Raster logs may be basis for a geologic worksta- of London, Special Publication 65, p. 377–387.
tion: Oil & Gas Journal, April 7, p. 84–88.