SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 A NEW CAPTURE AND INELASTIC SPECTROSCOPY TOOL TAKES GEOCHEMICAL LOGGING TO THE NEXT LEVEL R. J. Radtke, Maria Lorente, Bob Adolph, Markus Berheide, Scott Fricke, Jim Grau, Susan Herron, Jack Horkowitz, Bruno Jorion, David Madio, Dale May, Jeffrey Miles, Luke Perkins, Olivier Philip, Brad Roscoe, David Rose, and Chris Stoller, Schlumberger (Odom et al., 2008), and innovating on a traditional design (Galford et al., 2009). Copyright 2012, held jointly by the Society of Petrophysicists and Well Log Analysts (SPWLA) and the submitting authors. This paper was prepared for presentation at the SPWLA 53rd Annual Logging Symposium held in Cartagena, Colombia, June 16-20, 2012 Existing spectroscopy tools suffer from several technical shortcomings. The first of these is measurement precision, which is inversely proportional to the square root of logging speed. The precision of measured elements is a function of count rate, and currently the maximum count rate is limited by the available scintillator, photomultiplier tube (PMT), and electronics technologies. Precision and accuracy of elemental concentrations are also compromised at borehole temperatures because, for the scintillation detectors currently in use, the spectral resolution and performance degrade rapidly with increases in temperature. Tools using a BGO detector, for example, require some form of downhole temperature control and have a limitation of time at temperature. Lastly, systems using PNGs must cleanly separate the capture and inelastic spectra, which can be very challenging. ABSTRACT The increasing complexity of today’s reservoirs demands an accurate understanding of formation composition and mineralogy. This is particularly true for unconventional reservoirs, in which quantification of both mineralogy and organic carbon is critical for resource evaluation. The new geochemical spectroscopy tool described here combines the advantages of inelastic and capture gamma ray spectroscopy, opening new avenues for detailed description of complex reservoirs. Capitalizing on advances in technology, the new service provides higher precision and improved accuracy for the analysis of key elements in rock formations and simultaneously offers a standalone quantitative determination of total organic carbon (TOC). The measurements are offered at faster logging speeds. Eliminating the americiumberyllium (241AmBe) radioisotopic source makes combination with traditional measurements a much more attractive and viable logging option for both conventional and unconventional markets. These shortcomings are largely overcome by the tool described in this paper. The new tool improves the precision and accuracy of all elements traditionally measured by capture spectroscopy. In addition, it successfully integrates capture and inelastic gamma ray spectroscopy, making it possible to measure carbon accurately, account for inorganic carbon from carbonate minerals, and determine total organic carbon (TOC), which is essential for the evaluation of many unconventional plays such as shale gas and shale oil. Merging capture and inelastic data also significantly improves precision, accuracy, and interpretation consistency. This is particularly true for magnesium, a key element for differentiating calcite from dolomite. The tool’s measurements, including TOC, do not require a calibration to core or the use of complex, local, empirical, single- or multiple-tool interpretation models. INTRODUCTION Geochemical logging was introduced over 30 years ago, starting with a wireline tool based on a pulsed-neutron generator (PNG) and a thallium-doped sodium iodide (NaI(Tl)) scintillation detector (Hertzog, 1980). The emergence of new scintillation detectors led to tools based on gadolinium oxyorthosilicate (GSO) (Scott et al., 1991) and bismuth germanate (BGO) (Herron and Herron, 1996a), both of which are still in active use. Ruggedization of PNG technology enabled the introduction of nuclear spectroscopy to logging-whiledrilling (LWD) environments (Weller et al., 2005). Recent developments include combining capture and inelastic spectroscopy (Pemper et al., 2006; Herron et al., 2011), introducing new scintillation detectors This paper provides details on the technology that makes these measurement breakthroughs possible and shows several field examples highlighting the performance. 1 SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 TOOL DESCRIPTION The scintillator also produces much more light per incident photon than other materials. The light yield is approximately 50% larger than NaI(Tl), which has historically been considered the benchmark for high light output in a scintillator. High light output translates directly into improved spectral resolution, because the counting statistics associated with the light collected from the scintillator is one of the main factors that determine resolution. The basic layout of the tool is shown in Figure 1. The outer diameter is 114 mm (4.5 in.), which is smaller than that of previous generation openhole spectroscopy tools. It is rated to 138 MPa (20,000 psi) and 175°C. The measurement section contains a deuterium-tritium (d-T) PNG, which emits 14-MeV neutrons. The neutrons interact with the formation to produce gamma rays that are detected by a scintillation detector. Tungsten shielding reduces the direct passage of neutrons and gamma rays from the generator to the detector. The housing near the detector is surrounded with a thermal neutron shield containing boron to reduce the number of capture gamma rays produced from the tool. Table 1 Summary of some essential room temperature performance parameters of scintillators used for downhole spectroscopy applications The combination of a state-of-the-art scintillation detector, a high-output pulsed neutron generator, and a very fast pulse processing system opens up new possibilities in downhole spectroscopy logging. The following sections provide details on the technological advances that underlie the unprecedented tool performance. Property NaI(Tl) BGO Density (g/cm3) 3.67 7.13 5.29 Effective Atomic Number 50.8 75.2 46.9 Primary Decay Time (ns) 230 300 25 8.2 61 Light Yield 43 (photons/keV) Source: Lecoq et al., 2006. 1.8 LaBr3:Ce Relative light yield 1.6 Fig. 1 Measurement section of the spectroscopy tool. 1.4 1.2 NaI(Tl) 1 0.8 0.6 0.4 BGO 0.2 New Spectroscopy Detector - The tool uses a large cerium-doped lanthanum bromide (LaBr3:Ce) gamma ray detector (Van Loef et al., 2001; Saint Gobain Crystals, 2006; Stoller et al., 2011), which is coupled to a state-of-the-art high-temperature photomultiplier. LaBr3:Ce is a very fast scintillator with high light output and excellent spectral resolution. In addition, LaBr3:Ce has outstanding high-temperature performance with only a minimal loss in light output and resolution at temperatures up to 200°C. LaBr3:Ce 0 0 50 100 150 200 Temperature (°C) Fig. 2 Temperature dependence of scintillator light yield for LaBr3:Ce, NaI(Tl), and BGO. LaBr3:Ce has higher light output over the entire range of logging temperatures, unlike NaI(Tl) and BGO. Of great importance in a logging tool, the performance of LaBr3:Ce is maintained at high temperatures. This can be seen by comparing the light output of LaBr 3:Ce, NaI(Tl), and BGO as a function of temperature (Figure 2). While the light output of BGO is already low at 25°C, it drops significantly with small temperature increases. Above 60°C, the light output is too low to make acceptable logging measurements. This makes it necessary to use thermal protection in BGO-based A comparison of the properties of LaBr 3:Ce with other scintillators commonly used in downhole spectroscopy applications is shown in Table 1. The decay time of the light from the scintillator is an order of magnitude faster than NaI(Tl) and BGO. The fast decay makes very high counting rates possible and consequently improves measurement precision. 2 SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 spectroscopy tools and limits the available time in most boreholes for useful measurements. This limitation excludes long-duration operations such as drillpipe conveyance or tractoring. In contrast, LaBr3:Ce maintains its high light output over the entire temperature range. clearly sharper and better defined. This allows any measured spectrum to be more easily decomposed into its component parts, resulting in improved precision of spectroscopic answers. As described previously, a decrease in light output translates into undesirable spectral degradation. The elemental standard spectra for BGO (Figure 3, orange curves) show significant broadening and loss of definition as temperature increases from room temperature to the BGO maximum operating temperature of 60°C. On the other hand, the standard spectra from LaBr3:Ce (Figure 3, blue curves) are almost identical between room temperature and the much higher temperature of 150°C. New-Generation Pulsed Neutron Generator - Neutrons are produced by a new, high-performance d-T PNG developed to meet the needs of this advanced spectroscopy tool. The PNG eliminates the need for a radioisotopic source such as 241AmBe, thus reducing operational, transportation, and safety risks. The high neutron output of the PNG (capable of 3 × 108 neutrons/s nominal and higher) enables the tool to make the best use of the fast LaBr3:Ce scintillator by providing a very high counting rate and consequently better measurement precision and faster logging speeds. This output is about a factor of 8 higher than that of present radioisotope sources. The PNG is designed to allow clean separation of inelastic and capture gamma rays, improving the quality of both measurements. A necessary condition for this separation is a well-defined, repeatable neutron burst shape, which enables optimum timing of the inelastic and capture measurements and a capture measurement uncontaminated by inelastic gamma rays. To ensure that this condition is met, the PNG uses a hot cathodebased Minitron* neutron tube. The hot cathode technology produces a crisp 8.0-µs burst with rise and fall times faster than 400 ns (Figure 4). The timing of the burst is stable and predictable, making it possible to begin clean capture spectra acquisition very close to the burst for maximum count rates. Fig. 3 Elemental standard spectra measured with LaBr3:Ce and BGO detectors at room temperature (upper panel) and at higher temperatures (lower panel). The PNG has been designed for reliable operation over extended periods at temperatures up to 175°C. The technology builds on what was developed for the successful LWD PNG (Weller et al., 2005). Extensive test facilities were constructed and used to evaluate multiple PNG design generations. To date, over 3,600 LaBr3:Ce has spectral resolution unmatched in a downhole tool. An example is shown in the top panel of Figure 3, where elemental standard spectra for LaBr3:Ce are compared with those of BGO at room temperature. The features in the LaBr3:Ce standards are * Mark of Schlumberger. 3 SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 job-equivalent runs at maximum environmental conditions have been completed with several dozen PNG units. The extensive test results, accumulated over several years of development, were used for continuous design improvements to provide high reliability and performance at tool introduction. decomposition are called yields. The elemental sensitivities relate the yields to the elemental weight fractions which are of interest; these sensitivities are largely a function of neutron capture cross section. Fig. 4 Neutron burst shape used for spectroscopy. Electronics - The combination of the fast LaBr3:Ce scintillator and the high-output PNG produces a very high counting rate, often in excess of 2,500,000 counts/s. Processing such fast signals requires the use of specialized electronics. The electronics must provide excellent spectroscopy performance and outstanding pile-up rejection to limit spectral distortion due to nearly coincident gamma ray arrivals. Fig. 5 Schematic of the spectroscopy acquisition system. Signals from the detector pass through an integrator and analog-to-digital converter (ADC) and are then processed to form the pulse-height spectrum. The pulse height analyzer system that accomplishes these tasks is shown schematically in Figure 5. Gamma rays are detected in the large LaBr3:Ce scintillator described previously, which is coupled to a hightemperature spectroscopy photomultiplier with an integrated high-voltage supply and preamplifier. The signals from the preamplifier enter an integrator. The integrated signal is digitized by an analog-to-digital converter (ADC) and processed to obtain the pulse height, which is accumulated in a pulse height histogram that provides a spectrum of counts versus pulse height. Characterization for this tool relies primarily on extensive measurements in real and laboratory formations. Many of the formations used were designed specifically for geochemical applications. In addition to the elements conventionally measured by capture (H, Na, Cl, K, Ti, Cr, Ni, Ba, and Gd), inelastic (C, O), or both types of spectroscopy (Mg, Al, Si, S, Ca, and Fe), the characterization includes other elements, such as Mn and several other metals (see Table 2). Figure 6 shows some of the formations used to measure elemental standards and sensitivities. Mathematical modeling has been employed to optimize the design of the tool for both capture and inelastic spectroscopy and to construct representative formations for characterization. The modeling is based on the industry-standard Monte Carlo N-Particle (MCNP) transport code (Pelowitz, 2008). The MCNP code was adapted to simulate spectroscopy measurements by adding the capability to record detected gamma rays by the isotope and nuclear reaction that created them and Tool Characterization - To translate the spectra acquired by the tool into petrophysically relevant elemental concentrations, standards and sensitivities are needed. Elemental standards represent the distinct spectral signature of each element detected with the tool. As described in the next section, measured spectra can then be represented by a linear combination of the standards after correction for environmental and electronic factors. The coefficients from the 4 SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 by the region where they were produced. Converting the gamma ray response to a pulse height spectrum is accomplished through the GAMRES code (Evans, 1981), which was tailored with the measured nonlinear light output of the LaBr3:Ce scintillator. The modeling results are in very good agreement with measured spectra. However, any simulation of this kind will suffer from inaccuracies in even the most up-to-date neutron-gamma cross sections and in realistically simulating light collection in the scintillator, PMT response, and pulse processing. Experimental measurements for derivation of elemental standards and sensitivities thus remain the procedure of choice for obtaining unbiased spectroscopy answers. Fig. 6 Some of the geochemical formations used in the tool characterization. Table 2 List of some elements that can be determined through capture or inelastic gamma ray spectroscopy. Element Description Capture Inelastic Al Ba C Ca Cl Cu Fe Gd H K Mg Mn Na Ni O S Si Ti Aluminum Barium Carbon Calcium Chlorine Copper Iron Gadolinium Hydrogen Potassium Magnesium Manganese Sodium Nickel Oxygen Sulfur Silicon Titanium √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ Nuclear Interactions, Gamma Ray Measurements, and Data Acquisition - The well-defined burst of the PNG creates a population of high-energy (14-MeV) neutrons around the tool which interact and moderate to thermal energies on the scale of microseconds. These neutrons induce the emission of gamma rays from nuclei in the formation, borehole, and tool via two primary interactions: inelastic scattering and thermal neutron capture. In the case of inelastic scattering, which can occur only above specific neutron energy thresholds, a portion of the incident neutron energy is transferred to the target nucleus, which de-excites by emitting gamma rays at one or more characteristic energies. In thermal neutron capture, neutrons near equilibrium temperature are absorbed by a target nucleus, which de-excites by emitting a different set of characteristic gamma ray energies. These “fast” and “slow” interactions are illustrated in Figure 7. The gamma rays themselves undergo scattering as they travel toward the detector, where they deposit some or all of their remaining energy. The principle of the geochemical logging measurement is that the total detected gamma ray spectrum can be deconstructed based on each element’s characteristic standard spectrum, as discussed below. √ √ √ √ √ PHYSICS AND INTERPRETATION Transforming the spectra acquired by the tool into petrophysical quantities is an involved process. In brief, neutrons are emitted from the PNG and produce gamma rays. The pulse height spectra of the detected gamma rays are recorded as a function of time relative to the pulsing sequence of the PNG. Each spectrum is decomposed into a linear combination of standard spectra from individual elements. The coefficients of the linear combination are converted to elemental weight fractions, which may be further analyzed to produce mineralogy. These steps are described in more detail in the following sections. Elements that are typically found in the downhole environment exhibit a wide range of cross sections for thermal neutron capture. One advantage of inelastic interactions is that they are less affected by environmental factors including the presence of elements such as chlorine, which have high thermalcapture cross sections and can significantly reduce the population of thermal neutrons available for capture by elements of interest in the formation. Table 2 presents a list of some of the elements that can be measured through capture or inelastic gamma ray spectroscopy. For some elements, capture and inelastic gamma rays 5 SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 are available, but both are not necessarily used or useful for the determination of elemental concentrations. Spectra-to-Yields Processing - The purpose of spectrato-yields processing is to obtain an accurate suite of relative elemental yields based on decomposition of the recorded neutron-induced gamma ray spectra. An inherent assumption is that the measured spectra can be correctly represented by a linear combination of known elemental standard spectra. Given this assumption, the desired yields can be produced by a linear weighted least-squares (WLS) fit of the measured spectrum with a set of standard spectra within a specified energy range. However, temperature variations and electronics effects cause spectral distortions, which must be compensated for by transforming the measured spectrum’s energy calibration to match that of the standards. The energy resolution of the standards also must be matched to that of the measured spectrum. This methodology is formulated as a nonlinear least-squares problem and solved utilizing the Levenberg-Marquardt method (Gill et al., 1981; Grau and Schweitzer, 1987, 1989). With the measured spectrum at the proper energy calibration and the standards adjusted to match the resolution of the measured spectrum, the elemental relative yields can be derived using a linear WLS method. Fig. 7 Conceptual sketches of the two primary neutrongamma interactions in geochemical logging, showing a fast neutron undergoing inelastic scattering (top) and thermal neutron capture (bottom) The timing sequence of the PNG and its connection with the spectral data is shown in Figure 8. The inelastic spectrum is acquired during an 8.0-s-long neutron burst in a time gate denoted SBUR. The capture spectra come from early (SEAR) and late (SLAT) capture gates immediately following the burst plus a capture gate (STAU) that follows a series of 50 bursts. The STAU gate is also used for a formation neutron capture cross-section measurement (). This sequence is repeated 62 times, after which a quiescent period of approximately 8 ms is used to acquire a background spectrum (SBKG) to measure radiation from the tool, formation activation, and background radiation. The roles of these gates in the processing are described in the next section. The cycle takes 125 ms to complete and is repeated 3 times to form each spectroscopy data frame. The spectral data channels are recorded and used by the acquisition system for gain regulation of the PMT, then depth-gated and written out for further processing. Major benefits of the well-defined neutron burst are that there is no contamination by inelastic reactions during the “capture” gates and that capture data can be collected very soon after the end of the burst. Fig. 8 Timing sequence used for data acquisition. 6 SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 The spectra-to-yields processing is performed independently on appropriate time gate spectra to produce capture and inelastic relative yields. For capture processing, spectra from the three capture gates (SEAR, SLAT, and STAU) are summed to create a total capture pulse height spectrum. The net capture spectrum to be decomposed is obtained by subtracting the background spectrum (SBKG) while properly accounting for duty factors and counting losses. For inelastic processing, the contributions from capture must first be removed. This is accomplished by subtracting the early capture gate (SEAR) from the burst gate (SBUR) using a subtraction factor determined from neutron-capture physics, duty factors, and counting losses. The result is a net inelastic spectrum and associated variance (Figure 8). C and Mg are of particular value. An important limitation of the inelastic physics is that its set of yields is not complete enough to form a closure model. Instead, the normalization factor that converts these inelastic yields into elemental concentrations is derived by requiring that the derived dry weights of some of the elements, which occur in both inelastic and capture measurements, are consistent. The method ensures that all the resulting dry weights are mutually consistent for elements with both capture and inelastic yields. As a consequence the analysis gains the enhanced precision of the inelastic Mg measurement and achieves better accuracy overall. A second benefit of the inelastic analysis is the determination of a total carbon concentration. By using common association factors for carbonate minerals, the amount of inorganic carbon present can be quantified and subtracted from the total inelastic carbon to compute TOC (Herron and LeTendre, 1990). The combined analysis of inelastic and capture measurements is part of the generational breakthrough represented by this tool. By definition, the elemental yields are “relative” in that the sum of either the capture or the inelastic yields for each spectrum is separately equal to unity. Elemental relative yields are a function of the volumetric proportion of an element in the measurement region, as well as the sensitivity of the tool to each element. These yields are the starting point for determining quantitative elemental concentrations and mineralogical volumes. Dry-Weight Elements to Minerals - Mineralogy and/or lithologic fractions can be derived from elemental concentration logs through either a sequential processing method such as SpectroLith* (Herron and Herron, 1996a, 1996b) or through an inversion approach (Mayer and Sibbit, 1980; Quirein et al., 1986; Cannon and Coates, 1990; Peeters and Visser, 1991). Yields to Dry-Weight Elements - The conversion of relative spectral yields from neutron capture into absolute elemental concentrations is accomplished via a modified geochemical oxides closure model (Grau and Schweitzer, 1989; Grau et al., 1989) or through an iterative inversion technique such as ELAN* elemental log analysis (Quirein et al., 1986). Oxides closure models take advantage of common mineralogical associations to relate the concentration of unmeasured elements to the concentration of measured elements. For capture elements, the weight fraction (Wi) of element i at a given depth is a function of a closure normalization factor (F) determined at that depth, the measured relative yield (Yi ) of element i at that depth, and the tool’s sensitivity (Si) to element i for capture reactions: The SpectroLith technique is based on the derivation of empirical relationships between accurate elemental concentrations and mineral concentrations. A highquality core database of sedimentary rock samples characterized by dual-range Fourier transform infrared (DRFT-IR) mineralogy (Herron et al., 1997) and chemical analyses has been used to develop and evolve the processing algorithms. Chemical analysis measurement techniques include X-ray fluorescence, sulfur by combustion infrared detection technique (LECO), TOC by elemental analyzer, and trace elements primarily by induction-coupled plasma mass spectrometry. In combination, these techniques measure virtually all of the elements that influence nuclear well logs or are of interest for formation evaluation. The DRFT-IR library includes more than 50 mineral standards representing 28 minerals. Wi = F(Yi /Si). ............................................... (1) The normalization factor F is determined at each depth by solving the simple closure relation: F (AiYi/Si) = 1, ........................................... (2) where Ai is a factor accounting for all the unmeasured elements that are associated with element i. The quantitative elemental concentrations and processed lithology can be used directly with other log data in petrophysical interpretation programs to provide more accurate petrophysical answers (Mayer and The inelastic measurement offers a complementary, independent set of yields from which to extract elemental concentrations. The inelastic sensitivities to 7 SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 Sibbit, 1980; Quirein et al., 1986; Cannon and Coates, 1990; Peeters and Visser, 1991). fractions must therefore be validated to provide confidence in any mineralogy or other formation properties derived from them. TOOL APPLICATIONS Figures 10-13 provide four examples that compare the elemental weight fractions measured by the tool with those analyzed from core. Each figure displays a mineralogy column and elemental concentrations of silicon, calcium, iron, magnesium, sulfur, potassium, aluminum, titanium, manganese, and total organic carbon. The log data are displayed as solid black lines with shaded areas representing the uncertainties; core data are displayed as red dots; all values are presented in weight fraction. The core data are used for validation only; they are not used for calibration of elemental concentration logs in any example. The tool performance described in this paper has been validated by extensive field testing performed in several locations, mostly in North America, and covering the main unconventional plays in the United States and Canada. Over 50 successful jobs were performed accumulating more than 500 hours of tool operation with no PNG or other hardware failures. Repeatability and Precision – An example of tool repeatability is shown in Figure 9. The left eight tracks show two logging passes made at 900 ft/h. The curves overlay very well in all cases, even for the historically challenging measurement of Mg and the new quantification of C. The right eight tracks compare two passes acquired at 3,600 ft/h. Here, the repeatability is slightly worse, as expected, but the major rock-forming elements Si, Ca, Fe, and even Mg would still produce a good quantitative lithological description of the formation at much faster logging speeds than existing tools. Each of the examples comes from a different basin in North America. The agreement between core and log elemental concentrations is very good across a wide variety of lithologies. Of particular interest is the new log of TOC, which closely matches the organic carbon weight fraction measured on core samples. This information can be extremely valuable when evaluating organic-rich shale formations as well as conventional reservoirs. Examples of the accuracy of the TOC measurement are given in Figure 14, which presents logs from four wells in unconventional resource plays from North America. TOC ranges from 0 to 12 weight percent (displayed as weight fraction in the figures). Log and core data agree well over the entire dynamic range in both water-based and oil-based mud systems. In oil-based muds, TOC can be biased unless the contribution to the C signal from the oil in the borehole is properly removed. The logs were acquired at speeds between 600 and 1,200 ft/h, considerably faster than traditional carbon logging. Measurement precision is a function of the tool design, logging speed, depth averaging, and logging environment. Borehole diameter and salinity, formation porosity and salinity, mud and formation composition, and neutron output are all contributing factors. A tool planner is available to estimate the precision of elemental concentrations and interpreted properties such as matrix density or lithology for a given environment. The algorithms used in the tool planner are based on MCNP modeling that was benchmarked to experimental data. The recommended logging speed will depend on the required precision for a given measurement of interest. As suggested by Figure 9, if only basic mineralogy is needed and wellsite efficiency is critical, logging speeds of up to 3,600 ft/h may be achieved. A slower logging speed may be required for more demanding applications such as carbon quantification, precise dolomite determination, or logging in high-salinity borehole fluids. DISCUSSION Enabled by recent technological advancements, the measurement capabilities of the new elemental spectroscopy tool represent an important step beyond the previous generation of tools. Early field results highlight the considerable potential of this new service that enables a range of answer products and applications not possible with previous tools or tool combinations. Accuracy – The elemental weight fractions measured by the tool can be used to evaluate the mineral composition of a formation. Interpretation of log data to determine porosity, permeability, saturation and geomechanical properties generally depends on mineralogy. The accuracy of the elemental weight For the first time, the precision and accuracy of measured Mg and S enable a solution for mineralogy in carbonates at standard wireline logging speeds. 8 SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 Logging for TOC becomes a reality for unconventional resource plays such as shale gas, shale oil, and in situ thermal processing of oil shales. Improvements in the precision and accuracy of other measured elements such as Na and K and quantification of new elements such as Mn permit a more thorough and accurate mineralogy interpretation. Transactions of the 31st Annual SPWLA Logging Symposium, 2427 June 1990, Paper V. Evans, M. L., 1981, A computer model for calculating gamma-ray pulse-height spectra for logging applications, LA-UR-81-400, Los Alamos National Laboratory, Los Alamos, New Mexico. Galford, J., Truax, J., Hrametz, A., and Haramboure, C., 2009, A new neutron-induced gamma ray spectroscopy tool for geochemical logging, Transactions of the SPWLA 50th Annual Logging Symposium, 21–24 June 2009, The Woodlands, Texas, USA, Paper X. Operational aspects also distinguish the new tool. The improved precision of the measurement translates directly into faster logging. The electronic source means that no additional radioisotope source or logging runs are required to obtain geochemical information. The higher temperature rating of the detector allows very long jobs at temperature as with drillpipe or tractor conveyance. The reduced tool diameter enables logging in slimmer boreholes. The tool is also combinable with most wireline openhole services, including triple combo, nuclear magnetic resonance (NMR), dielectric, sonic, and imaging tools. Gill, P. E., Murray, W., and Wright, M. H., 1981, Practical Optimization, Academic Press. Grau, J., and Schweitzer, J. S., 1987, Prompt γ-ray spectral analysis of well data obtained with NaI(T1) and 14 MeV neutrons, Nuclear Geophysics, Vol. 1, No. 2, 157–165. The combination of all these features allows the simultaneous acquisition of neutron, density, resistivity, NMR, and capture and inelastic spectroscopy in demanding environments with a single toolstring. The availability of real-time analysis of the measurements supports enhanced real-time log quality control and intime decision-making. Grau, J., and Schweitzer, J. S., 1989, Elemental concentrations from thermal neutron capture gammaray spectra in geological formations, Nuclear Geophysics, Vol. 3, No. 1, 1–9. Grau, J. A., Schweitzer, J. S., Ellis, D. V., and Hertzog, R. C., 1989, A geological model for gamma-ray spectroscopy logging measurements; Nuclear Geophysics, Vol. 3, No. 4, 351359. Last but not least, the tool is a key step in eliminating radioisotope sources. The benefits in connection with health, safety, and the environment and with the costs of transportation, security, and liability are clear. Herron, M. M., Matteson, A., and Gustavson, G., 1997, Dual-range FIR mineralogy and the analysis of sedimentary formations, Proceedings of the 1997 Society of Core Analysts, 7–10 September, Calgary, Paper SCA-9729. ACKNOWLEDGMENTS The authors thank Energen Resources Corporation and Whiting Oil and Gas Corporation for supporting field testing of this tool in their wells and for allowing the use of their well log data. The authors also gratefully acknowledge Saint-Gobain Crystals, particularly Peter Menge and the engineering team, for their contributions in developing the detector technology described in this paper. Lastly, we thank the many people on the Schlumberger development team in Sugar Land, Princeton, and Cambridge for bringing this tool from concept to reality. None of the results reported here would have been possible without their expertise, dedication, and hard work. Herron, M. M., Grau, J. A., Herron, S. L., Kleinberg, R. L., Machlus, M., Reeder, S. L., Vissapragada, B., Burnham, A. K., Day, R. L., and Allix, P., 2011, Total organic carbon and formation evaluation with wireline logs in the Green River oil shale, SPE 147184, SPE Annual Technical Conference and Exhibition, 30 October–2 November 2011, Denver, Colorado, USA. Herron, S. L., and Herron, M. M., 1996a, Quantitative lithology: An application for open and cased hole spectroscopy, Transactions of the SPWLA 37th Annual Logging Symposium, 16–19 June, New Orleans, Louisiana, USA, Paper E. Herron, S. L., and Herron, M. M., 1996b, Quantitative lithology: Open- and cased hole applications derived from integrated core chemistry and mineralogy database, Geological Society Special Volume on Core- REFERENCES Cannon, D. E. and Coates, G. R., 1990, Applying mineral knowledge to standard log interpretation, 9 SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 Log Integration, Geological Society of London Meeting on Core-Log Integration, 1718 September, Burlington House, Piccadilly, London. models, Transactions of the 27th Annual SPWLA Logging Symposium, 913 June, Paper DD. Saint Gobain Crystals, 2006, BrilLianCe scintillators performance summary. Herron, S. L., and Herron, M. M., 2000, Application of nuclear spectroscopy logs to the derivation of formation matrix density, Transactions of the SPWLA 41st Annual Logging Symposium, 4–7 June, Dallas, Texas, USA., Paper JJ. Scott, H. D., Stoller, C., Roscoe, B. A., Plasek, R. E., and Adolph, R. A., 1991, A new compensated throughtubing carbon/oxygen tool for use in flowing wells, Transactions of 32nd Annual SPWLA Logging Symposium, 16–19 June, Midland, Texas, USA, Paper MM. Herron, S. L., and Le Tendre, L., 1990, Wireline source rock evaluation in the Paris Basin, AAPG Studies in Geology #30, Deposition of Organic Facies (AAPG, Tulsa), p 57-71 Stoller, C., Adolph, B., Berheide, M., Brill, T., Clevinger, P., Crary, S., Crowder, B., Fricke, S., Grau, J., Hackbart, M., Herron, S., Jorion, B., Lorente, M., Madio, D., Miles, J., Philip, O., Radtke, R. J., Roscoe, B., Shestakova, I., Ziegler, W., and Menge, P. R., 2011, Use of LaBr3 for downhole spectroscopic applications, IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC), 23–29 October, Valencia, Spain, 191–195. Hertzog, R. C., 1980, Laboratory and field evaluation of an inelastic-neutron-scattering and capture gamma ray spectrometry tool, SPE Journal, Vol. 20, No. 5, October, 327–340. Lecoq, P., Annenkov, A., Gektin, A., Korzhik, M., Pedrini, C., 2006, Inorganic Scintillators for Detector Systems: Physical Principles and Crystal Engineering (Springer-Verlag, Berlin), 23–26. Van Loef, E. V. D., Dorenbos, P., van Eijk, C. W. E., Kraemer, K., and Guedel, H. U., 2001, High-energyresolution scintillator: Ce3+ activated LaBr3, Applied Physics Letters, Vol. 79, 1573–1475. Mayer, C., and Sibbit, A., 1980, GLOBAL, a new approach to computer-processed log interpretation, SPE 9341, SPE Annual Technical Conference and Exhibition. Weller, G., Griffiths, R., Stoller, C., Allioli, F., Berheide, M., Evans, M., Labous, L., Dion, D., and Perciot, P., 2005, A new integrated LWD platform brings next-generation formation evaluation services, Transactions of the SPWLA 46th Annual Logging Symposium, 26–29 June, New Orleans, Louisiana, USA., Paper JJ. Odom, R., Dahlin, T., Dolliver, D., Elizondo, E., Seeds, E., Starnes, J., and Wilson, R., 2008, Design and initial field-test results of a new pulsed-neutron logging system for cased reservoir characterization, Transactions of the SPWLA 49th Annual Logging Symposium, 25–28 May 2008, Edinburgh, Scotland, UK, Paper O. ABOUT THE AUTHORS Peeters, M., and Visser, R., 1991, A comparison of petrophysical evaluation packages: LOGIC, FLAME, ELAN, OPTIMA, and ULTRA, Log Analyst, JulyAugust 1991, 350357. R. J. Radtke is a Principal Tool Physicist at Schlumberger’s Houston Formation Evaluation Integration Center (HFE) in Sugar Land, Texas. He has been involved in design, characterization, and algorithm development for nuclear LWD and wireline tools since joining Schlumberger in 1999. He graduated from The University of Chicago in 1994 with a PhD in physics. Pelowitz, D. B., Editor, 2008, MCNPX User’s Manual, Version 2.6.0, LA-CP-07-1473, Los Alamos National Laboratory, Los Alamos, New Mexico. Pemper, R., Sommer, A., Guo, P., Jacobi, D., Longo, J., Bliven, S., Rodriguez, E., Mendez, F., and Han, X., 2006, A new pulsed neutron sonde for derivation of formation lithology and mineralogy, SPE 102770, SPE Annual Technical Converence and Exhibition, 24–27 September 2006, San Antonio, Texas, USA. Maria Lorente is Wireline Product Champion for newgeneration nuclear tools, based in Sugar Land, Texas. She started her career as part of the engineering team in France and then moved to wireline field operations working in various assignments in the Middle East and Latin America. She holds a BS degree in electrical engineering from Universitat Politècnica de València, Spain, and an MS degree in electrical engineering from Supélec, France. Quirein, J., Kimminau, S., LaVigne, J., Singer, J., and Wendel, F., 1986, A coherent framework for developing and applying multiple formation evaluation 10 SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 Bob Adolph is an Engineering Advisor at Schlumberger’s Princeton Technology Center (PTC) in Princeton, New Jersey. Interests include wireline and LWD nuclear measurements, especially gamma ray spectroscopy, pulsed neutron technology, electronics, and signal processing. He joined Schlumberger in 1979 and holds a BS in electrical engineering from Rice University. a PhD in physics from the University of Pittsburgh in 1996. Dale May graduated in 1979 with a BS in physics from Texas Tech University, Lubbock, Texas. He joined Schlumberger in Midland, Texas, in 1980, as a field engineer. He has held positions as a field engineer, specialist engineer, sales engineer, log analyst, and center manager and is currently an Advisor Petrophysicist for Schlumberger Oilfield Services. Markus Berheide is a Principal Research Scientist in the Sensor Physics department at Schlumberger-Doll Research (SDR) in Cambridge, Massachusetts. His background is in nuclear measurements and enabling technologies. He holds a Doctorate in physics from Ruhr-Universität Bochum (Germany). Jeffrey Miles is a Senior Research Scientist at SDR. He received a PhD in physics from the Massachusetts Institute of Technology in 2007. His interests include the modeling of all aspects of nuclear physics in the oilfield, with emphasis on neutron-gamma spectroscopy and algorithms for fast modeling and inversion. Scott Fricke is a Principal Tool Physicist at HFE. He has been involved in software development, interpretation, and design of nuclear logging tools. He received a PhD in theoretical nuclear physics from the University of Minnesota in 1985. Luke Perkins is a Principal Engineer in the generators development group at PTC. He received his PhD in nuclear engineering from U.C. Berkeley and has worked since 1997 in Manufacturing, Sustaining and Research and Engineering positions within the oilfield radiation generators product line, including leading the development and commercialization of the EcoScope* pulsed neutron generator. Jim Grau is a Scientific Advisor at SDR. He has been involved for over 30 years in all aspects of borehole elemental analysis using nuclear spectroscopy techniques, including tool design, data acquisition software, and spectral analysis techniques. Jim received a PhD in experimental nuclear physics from Purdue University in West Lafayette, Indiana. Olivier Philip is a Principal Engineer in the detector development group at PTC. He received a PhD in nuclear engineering from Texas A&M University. He worked on nuclear tool design in the Schlumberger Sugar Land Product Center from 1996 to 2000. Since 2000, he has worked on photomultiplier and detector design and development for several downhole applications. Susan Herron is a Scientific Advisor in the Sensor Physics department at SDR. Her research interests include nuclear spectroscopy applications, mineralogy, and petrophysics. She holds a PhD in Geological Sciences from State University of New York at Buffalo. Jack Horkowitz is a Petrophysics Advisor with Schlumberger working at HFE on a number of projects including spectroscopy processing, integrated interpretation products and LWD calipers. Jack joined Schlumberger in 1995 and holds a PhD in geology from the University of South Carolina. He is a past President of SPWLA (2006–07). Brad Roscoe is a Scientific Advisor and the Nuclear Program Manager at SDR. His areas of interest include nuclear detectors, nuclear sources, gamma-ray spectroscopy, and measurement integration. He earned a PhD in nuclear engineering in 1981 from the University of Illinois at Urbana-Champaign. David Rose is Principal Petrophysicist and Manager of Interpretation Engineering for Nuclear Answer Products at HFE. He holds a Bachelors in geophysics from the Colorado School of Mines. Bruno Jorion is a Principal Engineer (Electrical) in the Engineering department HFE. He received a Diplôme d'Ingénieur from l'Institut Supérieur d'Electronique du Nord (ISEN, France) and has been working in several nuclear and NMR tool designs. Chris Stoller is a Scientific Advisor at PTC. He joined Schlumberger in 1986 and has worked on a variety of nuclear tool projects in wireline and LWD. He received a PhD in physics from the Swiss Federal Institute of Technology in Zurich in 1976. David Madio is a Senior Petrophysicist for Schlumberger at HFE. His interests include mineralogy, neutron porosity, and magnetic resonance. He received 11 SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 ADDITIONAL FIGURES 900 ft/h 3,600 ft/h Fig. 9 Example of elemental weight fractions repeatability at two different logging speeds, 900 ft/h in the left eight tracks and 3,600 ft/h in the right eight tracks. The well is located in Central Texas and was drilled with an 8-in. bit size and filled with fresh water. This interval consists of alternating zones of very pure limestone where Ca approaches 0.4, the value of calcite; dolomite where Mg approaches 0.13, the value in dolomite; and a small amount of sandstone (Si = 0.47 in quartz and about 0.3 in feldspars). Limestone and dolomite are easily distinguishable from the Ca and Mg concentrations even at 3,600 ft/h. The very low concentrations of Fe, S, K, and Al are also properly reproduced in the log. 12 Zone A SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 Zone B x100 Zone C x125 Zone D x150 x175 Fig. 10 Field example from a well in North Dakota. Zones A, B and C are part of the Bakken formation, and zone D corresponds to the Three Forks formation. The figure compares elemental weight fractions measured by the tool (black lines) to those derived from core analysis (red points). The yellow shaded areas along the black lines represent the uncertainties (which are very small in this example). The track to the left of the depth track shows the mineralogy, and the track to the right of the depth track shows an enhanced core photograph. Overall there is a very good match between the measurements and the core, especially the total organic carbon (TOC) measurement (far right track) ranging from nearly zero (zones B and D) to more than 12 weight percent (zones A and C). The core photograph shows the highly laminated nature of this formation, which can cause occasional scatter when comparing log data to standard core plug data. The logging speed was 600 ft/h. 13 SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 x100 x150 x800 x850 x900 Fig. 11 Field example from Western Canada showing two intervals from the same well. The figure compares elemental weight fractions measured by the tool (black lines) to those derived from core analysis (red points). Due to the low salinity mud system, the bit size, and a logging speed of 1,080 ft/h, the measurement statistical uncertainties are less than the thickness of the lines. The track to the left of the depth track shows the mineralogy. There is very good agreement between core and log for all elemental concentrations and total organic carbon (TOC). 14 SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 x100 x300 x500 Fig. 12 Field example from West Texas showing three intervals from the same well. The figure compares elemental weight fractions measured by the tool (black lines) to those derived from core analysis (red points). The yellow shaded areas along the black lines represent the uncertainties. The track to the left of the depth track shows the mineralogy. Overall there is a very good match between the measurements and the core. The logging speed was 900 ft/h. x050 x450 Fig. 13 Field example from the Southeastern United States showing two intervals from the same well. The figure compares elemental weight fractions measured by the tool (black lines) to those derived from core analysis (red points). The yellow shaded areas along the black lines represent the uncertainties. The track to the left of the depth track shows the mineralogy. Note the change in elemental concentrations and the good agreement between log and core data for all elements in both the clay-bearing and quartz-rich formations. The logging speed was 1,800 ft/h. 15 SPWLA 53rd Annual Logging Symposium, June 16–20, 2012 x075 x000 x000 x000 x100 x100 x050 x125 x100 x100 x200 x150 x150 x200 x300 x175 x200 x300 x400 x200 x250 x400 x500 x225 x300 x500 x250 x600 x350 Water-Based Mud Oil-Based Mud Fig. 14 Comparison of total organic carbon (TOC) measured by the new tool and derived from core from four wells in unconventional formations throughout North America. For each well, the first track shows the total carbon (TC, in red) derived from the inelastic measurement and the total inorganic carbon (TIC, in black) computed from carbonate minerals using capture spectroscopy elements. The separation between the two curves is TOC, which is displayed in the second track with core data (red dots) for validation. 16
© Copyright 2026 Paperzz