Allen, L.S., Tittle, C.W., Mills, W.R., and Caldwell, R.L., 1967, Dual-spaced neutron logging for porosity, Geophysics, 22(1), 60–68.
The primary contribution of this paper was to show that a two-detector neutron logging tool could be developed that was sensitive to formation porosity and less sensitive to borehole and salinity effects than previous neutron logging tools. The authors found that placing the neutron detectors at sufficiently long spacing, using high-intensity neutron sources, and using the ratio of the counting rates obtained at the two detector locations would allow the logging instrument to have good precision for formation porosity measurement, and minimum effects from other, adverse influences, such as borehole salinity, mudcake, washouts, etc. This led to the widespread application of two-detector neutron logging tools for measuring formation porosity.
The authors used one- and two-group neutron diffusion theory to calculate the response of a two-detector neutron tool. This is also a significant contribution as diffusion theory has been used for many years to calculate compensated neutron log responses in the borehole/formation environment.
This work is considered to be a seminal work for dual-detector neutron logging. Today, all logging companies provide two-detector neutron logging services (commonly called compensated neutron logs) and the tool is one of the most important tools for the measurement of formation porosity. There are literally thousands of compensated neutron instruments in the industry and most are designed as described in this paper.
This paper has been cited many times through the years, and books and training courses commonly include a review of the concepts and illustrations presented in the original paper.
This paper showed that a dual-detector neutron logging tool could be used for a quantitative measurement of formation porosity. Combined with other logging tools, such as formation density, the tool could be used to calculate formation lithology and identify gas bearing zones.
These tools are used in most logging provinces in the world. They are almost always included in any logging run. The tools are not only used in wireline logging but tools with very similar configurations are also used in logging-while-drilling.
Prior to the development of two-detector neutron logging tools the neutron log was considered to be principally qualitative. With the development and understanding of two-detector tools provided by this paper the compensated neutron tool became an important instrument for measuring formation porosity.
This paper provides a clear and concise explanation of the use of diffusion theory to understand and characterize the response of two detector neutron tools, and clearly describes the conclusions reached from the work performed by the authors. 10/10.
Anderson, B., 1986, The analysis of some unsolved induction interpretation problems using computer modeling, The Log Analyst, 27(5), 60–73.
David Kennedy/Teruhiko Hagiwara
Early "electrical survey" logs exemplified by short normal, long normal, and long lateral arrays were plagued by various eccentricities in their responses, including borehole and thin-bed effects. The apparent-resistivity response of the deep-sensing long lateral array is asymmetric and continually varying across an ideal thick bed of constant resistivity. Thus, the wiggles visible on the recorded responses could be more reflective of the instrument impulse responses than of the resistivity variation in the formation. Corrections to the logs recorded in the field were effected by laborious application of "departure curves", but the departure curves were designed by the analytic solution of boundary value problems in geometries of unrealistically high symmetry; other departure curves were created using analog computer resistor networks. There was no practical method for correcting conventional electrical survey logs for complicated formation resistivity distributions. Early in the 1950s H.G. Doll introduced two types of "focused" resistivity logs: the laterolog and the induction log. The apparent-resistivity responses of these instruments were designed to be symmetric in symmetric beds, with the apparent resistivity being approximately equal to the formation resistivity in thick-enough formations. In the evolution of the induction log, a 1957 publication proclaimed "The induction log accordingly gives the value of Rt without correction" (Dumanoir et al., 1957, Transactions AIME). Although this statement was properly qualified in the paper, it entered the logging lexicon as "for induction logs, Ra = Rt." Of course, this was never true, but for thick-enough beds in conductive-enough rocks and in vertical boreholes, it was approximately true, and came to be universally believed. However, in cases where resistivities are high, where beds are thin and laminated with conductive shales, and where boreholes are deviated, apparent resistivities depart significantly from formation resistivities. In the decades following the 1950s, electronic computers were becoming increasingly capable, common, and affordable to large companies. In the late 1970s mathematical analysis of resistivity instrument responses began to appear in the technical literature, and with increasing frequency into the 1980s.
In the early 1980s, some oil companies and service companies started modeling resistivity log responses using computer models to characterize logging tool responses as well as to help log interpretation by analyzing the modeled tool responses. Barbara Anderson is one of the early investigators of resistivity log modeling. In this paper, she discussed induction log response to thinly laminated sand/shale formation, the shoulder-bed effect in dipping formations, and a case study of anomalous induction log response that may be attributed to a nearby conductive anomaly. The paper illustrates how mathematical modeling of tool responses could be useful to understand subsurface formation and improve formation evaluation, particularly in explaining anomalous responses that could not be explained by the standard models
This landmark paper was quickly followed by many others, e.g., "Strange Induction Logs" (Anderson and Barber, 1988), that led to the understanding of induction log responses that had been misunderstood or unexplained for the previous three decades. Moreover, while the 6FF40 array remained the industry standard, commercially available induction log modeling programs became de rigueur. Although the introduction of proprietary array induction tools put an end to routine resistivity modeling for formation evaluation, such modeling is still available through university consortiums, and is still used to better understand log responses of the array induction (and other) logs.
Prior to the introduction of the numerical modeling of log responses in general, and resistivity log responses in particular, logs displayed many features that went unexplained. Paradoxically, other features of logs that at the time were considered to be well understood turned out to be due to different causes than had been believed. The modern response to unexplained log responses is "let us numerically model this strange response to see if we can understand the cause."
Resistivity log modeling has been used to understand anomalous (or "strange") resistivity logs of all vintages, recorded on all continents beneath all oceans. Resistivity modeling is "global" not only in the geographic sense of the word, but also in the sense of its being extended to the modeling of all logging instrument responses.
The paper was a harbinger of modern interpretation methods that rely heavily on computers for viewing field logs, computing derived logs, and tool response modeling as an adjunct to standard interpretation methods. The paper was not the cause of the shift in paradigm, but rather heralded that a shift was about to begin.
The use of computer modeling to analyze logging tool responses was the paradigm change already underway during early 1980s. Barbara Anderson was one of leading figures in that effort. This paper documented some results of this paradigm change. The "parallel conductance model" of thinly bedded formation that was the result of modeling induction log response in thinly bedded formations discussed in the first part of this paper, had been already in practice in Shell Oil Company and US patent 4,739,255 had been filed earlier. The "macroscopic anisotropy model" had been also developed to analyze dipping thinly bedded formations in the mid-1980s using computer modeling of tool response in deviated boreholes. The more prominent dip effect had been identified for higher frequency induction-type (propagation) LWD resistivity tools as the polarization-horn effect at bed boundaries and was explained by computer modeling, though this paper did not address in the second section. However, this paper provided a good example of using computer modeling to understand some anomalous log data in the third section. This paper showed that computer modeling was not only useful for tool developers in an oilfield service company to characterize tool responses but also for petrophysicists and log analysts to use to understand log data and to help improve formation evaluation.
The paper, like all of Barbara Anderson's publications, is written in tutorial style, and is quite accessible to those already familiar with the jargon of resistivity logging.
Doll, H.G., 1949, Introduction to induction logging and application to logging of wells drilled with oil base mud, Paper SPE-949148-G, Journal of Petroleum Technology, 1(6), 148–162.
Resistivity logging was invented in 1927. The first instruments were passive galvanic arrays that measured varying potential drops in the borehole due to various geometrical arrangements of current electrodes injecting and returning current as the electrode array was pulled up the wellbore. The resulting log responses were asymmetric even in symmetric formations, and in randomly varying formations could be interpreted only qualitatively. In 1941, G.E. Archie's invention of a method of quantitatively interpreting resistivity logs provided an impetus to produce resistivity instruments capable of producing more interpretable responses. However, the development of new logging technology remained in limbo for five years while World War II was fought and won, but within the four years following victory in 1945, H.G. Doll invented the two principal resistivity logging technologies that remain the cornerstone of resistivity logging to this day. The first introduced was the induction logging instrument. The response of this instrument was symmetric in symmetric formations, did not depend upon conductive mud in the borehole to couple its transmitter signals to the conductive formation, and seemed to promise a means to improve both vertical and radial resolution of the instrument response by judicious design of its antenna array. This inaugural model for the induction instrument established that the response can be modeled as the convolution of the formation conductivity with a geometrical factor, a model that remains in use, with refinement, to this day. From the late 1950s, induction logging, continually—if episodically—refined, became the principal technology used for resistivity logging.
Since the early 1960s, induction logging instrument technology been the principal means of estimating formation resistivity in formations with resistivities less than 100 to 500 Ω∙m, comprising (as a guess) at least two thirds of all resistivity logs that have ever been, and are being, run.
The introduction of this technology was the first step on the journey that has led inexorably, if slowly, to the state-of-the-art that is enjoyed today. Although induction logging antennas are mere coils of wire, the elaboration of the technology had to wait upon the invention of transistors and integrated circuits, computer hardware and software, telemetry and materials science, and accurate modeling of formation resistivity as a biaxial, second order tensor. In the 1950s it was widely believed in the logging community that all formations of interest were isotropic, if not necessarily homogenous and infinite; the modern view, engendered by having to explain induction log responses, is that most formations are anisotropic and heterogeneous. This change in thinking was thrust upon us by induction log responses.
It is safe to say that on any given day, and possibly at any given hour, an induction log is being run somewhere on planet earth. They have sampled formation resistivity on every continent, and beneath every ocean, for 65 years.
The induction instrument technology introduced in this paper is the poster child for paradigm shifting in technology. In 1949, when the paper appeared, the only logging technologies available were the so-called "conventional electrical survey" comprising electrode arrays called "normals" and "laterals"; by the 1970s it was hard to find any conventional electrical survey hardware left in field locations. In 20 years the preferred technology changed entirely.
Probably few have thought, and written, so clearly on induction logging principles and technology as H.G. Doll in this 1949 paper. It is true that the theory presented in the paper is a low-frequency, low-conductivity approximation to the actual physics involved in the response of the tool to the formation, but the mathematics, which is kept to a minimum, needed only to be corrected by expanding certain terms in the Doll theory to include the effects of frequency and formation conductivity. If I cannot say that it is a genuine pleasure to read, at least I can say it is much more pleasurable than most of the papers on induction logging theory that have followed.
Doll, H.G., 1951, The Laterolog: A New Resistivity Logging method with electrodes using an automatic focusing system, Petroleum Transactions, AIME, 192, 305–316.
The induction log, which was introduced earlier by Doll, is a conductivity log that measures formation conductivity and works well for highly resistive, i.e., oil-based mud systems and low formation resistivity (high conductivity). However, the induction log is not suited for conductive mud and high-resistivity formations. Galvanic resistivity tools, e.g., normal and lateral logs, had been used for such formations, but mudcake and mud-filtrate invasion caused significant problems for reading formation resistivity. Because of their asymmetric electrode configuration, the interpretation of log response was very difficult. Doll introduced the laterolog galvanic resistivity tool that works better in high-resistivity formations with conductive mud. The tool achieves its deep depth of investigation by focusing currents into highly resistive formation using autocontrolled guard electrodes.
The laterolog is the tool of choice in most carbonate reservoirs with high formation resistivity and water-based mud. The dual laterolog is commonly used to better profile mud-filtrate invasion into the formation. Recently, array laterologs have been developed for improved invasion profiling.
The laterolog is the tool of choice in most carbonate reservoirs with high formation resistivity and water-based mud systems.
The laterolog made resistivity well logging possible in highly resistive formation and very conductive mud without being impeded by mudcake and near-borehole effects.
Duesterhoeft, William C., 1961, Propagation effects in induction logging, Geophysics, 26(2), 192–204.
Prior to the Duesterhoeft paper the only treatment of induction log response characteristics had been H.G. Doll's inaugural introduction to the technology in 1949. The theory included in the Doll paper turned out to be valid only for very low frequencies or very low conductivities; consequently there was no recognition in the Doll theory of the existence of skin effect, borehole effects, or shoulder-bed effects. In the Doll approximation, every formation element contributed to the total response only according to its location with respect to the receiver coil and its conductivity. Duesterhoeft's paper was first to formulate the problem of induction logging in terms of Maxwell's equations for electromagnetic wave propagation in penetrable, conducting media. Duesterhoeft's solution is in terms of the Hertz potentials. He gives the solution to three important special cases; the induction instrument's response in (1) an infinite, homogeneous, conducting medium without borehole or layering; (2) in a horizontally layered one-dimensional formation without borehole; (3) in a radially layered medium without layering. Although this Duesterhoeft article was for two-coil sondes only, it was followed 1961 by a companion article by Duesterhoeft et al., "The effect of coil design on the performance of the induction log (SPE-1558-G)" which elucidated the application of the two-coil theory to multicoil, "focused field" induction logging instruments.
More than a year following the 1961 publication of the Duesterhoeft papers, Moran and Kunz of Schlumberger-Doll Research, published essentially identical results in Geophysics (Moran, J.H., and Kunz, K.S., 1962, Basic theory of induction logging and application to study of two-coil sondes, Geophysics, 27(6), 829–858.) The formulation of the electromagnetic boundary value problem in the Moran and Kunz paper is in terms of the electromagnetic vector potential (as opposed to the Hertz potential used by Duesterhoeft) but the resulting electromagnetic fields (derivatives of the potentials) are the same, confirming the Duesterhoeft analysis. The three boundary-value problems were used for three decades to compute (1) the apparent resistivity response and skin effect for the multicoil 6FF40 antenna array in infinite, homogeneous, isotropic media (this is the apparent resistivity presented on a log), (2) thin-bed corrections, and (3) tornado-chart invasion corrections.
The ability to use the Duesterhoeft formulation to model the apparent resistivity response of the 6FF40 antenna array (ubiquitous in every service company in the 1960s, 1970s, 1980s, and early 1990s) taught us that our understanding of the response of induction logs was seriously flawed. For example, the apparent resistivity response in the center of a 100-ft thick, infinitely resistive reservoir sandwiched between 1 Ω∙m shoulder beds is about 90 Ω∙m, with the entire signal coming from the shoulder beds 50 ft from the center of the antenna array. It also revealed many pitfalls in the interpretation of thin beds and low-resistivity pay, instrument responses in deviated wellbores, and the egregious effects of rugose boreholes combined with saline mud and high-resistivity reservoirs, to name just a few.
The three boundary-value problems that Duesterhoeft first presented analytical solutions for are approximations when applied to any real logging job. For example, every horizontally layered medium that is logged is penetrated by a borehole, which gives each layer its own radially layered conductivity distribution due to invasion. Every radially layered medium is bounded on the top and bottom by a layer of different conductivity. To achieve a thorough understanding of all of the effects when combined requires that all of the effects must be simultaneously included in the model of the electromagnetic field. The two radially layered boundary value problems are used to provide the basis for solutions of so-called hybrid methods, which use the analytic solutions for part of the solution, and numerical transforms for the remainder of the solution. It must be said that this is still an area of active research to this day.
The development of the modern generation of induction logging instruments featuring their multicoil asymmetric arrays and software focusing for different depths of investigation and vertical resolutions, and triaxial coils for interpretations of anisotropic media follow directly from the first analytic solution of the induction log boundary value problems. It was modeling based upon these solutions that brought the realization that no real understanding of the conductivity structure of the earth is possible without use of more elaborate antenna arrays that fully sample all possible components of the fields induced by sources capable of exciting all possible modes of response. All modern wireline and LWD induction and propagation instrumentation owe their designs and interpretations to the solutions of the boundary values problems first introduced by Duesterhoeft.
For students of geophysical wave propagation in general, and electromagnetic methods and wave propagation in particular, Duesterhoeft's formulation of the horizontally layered medium in terms of Hertz potentials follows simply from the reduction of the Maxwell vector fields to derivatives of potentials. The formulation is formally very similar to the propagation of acoustic waves in layered media, and as such, gives geophysics students a touchstone for understanding the method and interpreting its results. Although originally developed with vertical wells and isotropic homogeneous layered beds in mind, the Duesterhoeft formulation generalizes very easily to the case of deviated boreholes penetrating anisotropic homogeneous layered beds.
Ekstrom, M.P., Dahan, C.A., Chen, M.Y., Lloyd, P.M., and Rossi, D.J., 1987, Formation imaging with microelectrical scanning arrays, The Log Analyst, 28(3), 294–306.
This paper introduced borehole electrical imaging with a clear explanation of the technology and a variety of examples of immediate geological interest. The authors pointed out that the dramatic resolution of the images at a scale of millimeters made them comparable with core photographs in the evaluation of both clastics and carbonates. The graphic output of image logs encouraged geologists to look at other logs as sources of lithological information rather than simply as curves for correlation.
Electrical imaging is routinely run in many wells around the world for a variety of purposes. As an extension of dipmeter technology, it provides the orientation of sedimentological, stratigraphic, and fracture systems. Continuous advances in technology improve image resolutions at a much finer scale than conventional logs, and so have been a major factor in thin-bed sand analysis of turbidite plays.
Although the physics of conductivity imaging differs from visual core observation, the similarity between image logs and cores in clastic successions meant that no special training was needed for experienced geologists to interpret them. In contrast, image logs of carbonates require a different perspective because of their sensitivity to pore volume and size rather than matrix textural features used in core description. Consequently, developments in carbonate image interpretation promote new ideas in interpretation and analysis that supplement conventional core description.
Image logging is now available from all major service companies and finds application in both conventional and non-conventional reservoirs all over the world.
The technology was introduced at a time of significant industry downturn, so that its relative expense discouraged initial commercial application. However, its widespread use in the Ocean Drilling Program opened the eyes of a generation of academic geologists to the capability of image logs, and by extension, to other logs as tools for geology rather than simply as correlation frameworks.
This paper is clearly written and provides a balanced overview of tool technology together with data acquisition and display, illustrated by useful geological examples. 10/10.
Miller, M.N., Paltiel, Z., Gillen, M.E., Granot, J., and Bouton, J.C., 1990, Spin echo magnetic resonance logging: porosity and free fluid index determination, Paper SPE-20561, presented at the SPE Annual Technical Conference and Exhibition, 23–26 September, New Orleans, Louisiana.
This paper reported on the introduction of a working nuclear magnetic resonance (NMR) logging tool (given the trade name MRIL) that did not require any special borehole treatment, provided a continuous record of lithology-independent porosity and irreducible water saturation. The description of the methodology was sufficiently detailed to explain why the tool was blind to any contribution for the borehole fluid itself, a grave failing of prior attempts to create a working tool.
The paper provided well-documented and favorable comparisons between the MRIL log derived porosity and both core and conventional neutron-density porosity from a Conoco test well in Oklahoma. Further discussion in the paper extended the interpretation of the basic T1, T2 and diffusion measurement to the estimation of permeability, fluid viscosity and formation resistivity factor.
Prior to publication of this paper the use of the NMR technique had been limited to benchtop laboratory equipment or a seldom-used logging tool that only made stationary measurements and required doping of the borehole mud system.
The lasting impact of these disclosures has been the adoption of the MRIL technology by two of the three major wireline service companies to provide a frequently used service filling an important technological niche. Subsequent to the publication this paper, the technology has been refined to provide wellsite porosity, permeability, pore-size distribution, pore-fluid type, viscosity and moveability.
Tittman, J., and Wahl, J.S., 1965, The physical foundations of formation density logging (gamma-gamma), Geophysics, 30(2), 284–294.
There are several significant contributions brought by this paper.
Electrical measurements were introduced as early as 1927, and formation gamma ray in the 1940s. But, what the industry also needed was a quantitative evaluation of formation porosity. Many "porosity" measurements have been developed (acoustic, neutron, and even magnetic resonance) but as early as the 1960s, the density measurement brought a quantitative assessment of density from which, and through a simple, nonempirical equation, formation porosity could be derived.
Tittman and Wahl describe how a robust density measurement can be performed using Compton electron interactions. From this electronic density, a formation density that is very accurate in the common sedimentary rocks could be obtained. The apparatus can handle the photoelectric parasitic effect and also the impact of borehole mudcake.
The paper also clearly leads the way to further developments that involve spectral measurements and the use of multiple detectors. It has the merit to confirm sound physics with a good amount of experimental data. From this paper emerges the "spine-and-rib" approach that is still being much used by a number of logging companies.
These critical innovations have warranted an incomparable success. Many petrophysicists rely only on this density measurement for porosity derivation as other measurements have weak measurement-to-porosity transforms (e.g., acoustic or "sonic") or doubtful and poorly documented environmental corrections (e.g., neutron porosity). All logging companies have copied the design described in this paper.
Before the density measurement, porosity was qualitatively estimated from the spontaneous potential curve. In addition, the uncertainties of the density measurement are the best understood and the best managed.
Second only to the resistivity log, the density log is run in almost every openhole section. Logging companies have collected considerable revenues from running this measurement.
The proposed measurement enables the derivation of a quantitative porosity.
The paper is well written and easily read.
Zemanek, J., Caldwell, R.L., Glenn, Jr., E.E., Holcomb, S.V., Norton, L.J., and Strauss, A.J.D., 1969, The Borehole Televiewer—A new logging concept for fracture location and other types of borehole inspection, Journal of Petroleum Technology, 21(6), 762–774.
This paper was the first detailed description of the instrumentation, application, and results received from a new downhole logging system that provided acoustic images of the borehole environment. Called the Borehole Televiewer, the new system presented a "picture" of the borehole wall and gave a highly definitive view of the borehole environment. This paper provided a description of the logging tool and presented a large number of acoustic images that clearly showed fractures in open holes. In cased holes, the paper showed that the tools could be used to find perforations and to inspect and measure possible damage to the casing.
The technology introduced in this paper was primarily developed by research workers of what at the time was Mobil Research and Development Corporation. A fair measure of lasting impact of this technology is that all major logging service companies subsequently developed similar instrumentation for commercial applications. In most cases, these companies have now developed 2nd and 3rd generation borehole acoustic imaging tools.
This paper has been cited many times since it was published, and books and training courses commonly include a review of the concepts and illustrations presented in the original paper. The Mobil workers who developed this technology were familiar to and admired by most research workers in the logging industry whether or not they have read the original paper.
This paper showed that a pictorial description of the borehole environment could offer a clear method for locating and evaluating fractures in the formation. Prior to the development of the televiewer, other logs, such as resistivity, acoustic, or nuclear logs, were used to try to identify fractures but were difficult to interpret and seldom definitive. Other techniques, such as core analysis, impression packers, and downhole cameras were also used to search for fractures but were often unsuccessful. With this new technology, fractures could be clearly seen and their orientation quantified.
This technology was particularly successful in medium- to low-porosity carbonates due to the strong reflection of the acoustic signals in these environments. For higher porosity formations, the technology was less successful, but the tool had shown industry analysts the value of a pictorial representation of the borehole wall. This provided an impetus to look for technologies that could be used in other downhole environments and led to the development of electromagnetic imaging logging tools.
At first this technology was most commonly used in the low porosity, highly fractured limestones of West Texas, but subsequently has been used for applications all over the world. As mentioned previously, all the major logging service companies have developed acoustic imaging tools and these tools are used in many reservoirs worldwide.
The change this paper made was to show geologists and other industry analysts that natural and/or induced fractures could be clearly identified in situ and their orientations calculated. Applications of this technology in cased holes provided a direct image of the location of perforations and casing damage.
This paper is very well written and is easy to follow in both the technical description of the instrumentation and in the discussion of the application of the pictorial results shown by the log. 10/10.