High-Definition Spectroscopy—Determining Mineralogic Complexity Neutron-induced capture spectroscopy tools measure the concentrations of specific elements in downhole rocks. From these data, petrophysicists can derive mineralogic, lithologic and matrix properties. Early spectroscopy tools lacked the spectral sensitivity to derive total organic carbon—an important measurement for understanding unconventional resource plays. A new tool delivers total carbon, from which organic carbon concentrations can be determined. This tool also has the ability to resolve complex lithology with a degree of accuracy never before possible. Manuel Aboud Rob Badry Calgary, Alberta, Canada Jim Grau Susan Herron Cambridge, Massachusetts, USA Farid Hamichi Jack Horkowitz Sugar Land, Texas, USA James Hemingway Houston, Texas Robin MacDonald Saudi Aramco Al-Khobar, Saudi Arabia Pablo Saldungaray Al-Khobar, Saudi Arabia Don Stachiw Northern Cross (Yukon) Ltd. Calgary, Alberta Christian Stoller Princeton Junction, New Jersey, USA Richard E. Williams BHP Billiton Houston, Texas Oilfield Review Spring 2014: 26, no. 1. Copyright © 2014 Schlumberger. CMR-Plus, ECS, ELANPlus, GST, Litho-Density, Litho Scanner, Minitron, Platform Express, RST, SpectroLith and TerraTek HRA are marks of Schlumberger. LECO is a mark of the LECO Corporation.
34
Rocks comprise an assortment of minerals and fluids. Many processes combine to form the complex mixtures found in the subsurface, including the transport mechanisms that delivered sediments and rock fragments to their current resting place, heat and pressure applied during burial and subsequent lithification and a myriad of internal and external forces acting on the rocks. Using downhole spectroscopy tools, also referred to as geochemical tools, geologists can unravel the composition of sedimentary, metamorphic and igneous formations and better understand their stratigraphy, mineralogy, diagenesis and hydrocarbon potential.
In the early days of well logging, geologists and petrophysicists developed models to help identify the presence of hydrocarbons, estimate their quantities and determine production potential. Saturation models such as those described in equations proposed by Gus Archie, later modified to account for the influence of shale, usually assume homogeneous, isotropic formations.1 These methods provide reasonable results when computing hydrocarbon saturations in conventional reservoirs; however, to determine the oil and gas potential in complex reservoirs and unconventional resource plays, petrophysicists have replaced simple models with techniques
Oilfield Review
Pulsed Neutron Generator
Highvoltage supply
n Controls
Ion source
Americium-Beryllium Source Reaction
γ (60 keV)
On-Off switch Main power
Target 241
Am
237
237
Np*
Np
γ (4.4 MeV)
α (5.5 MeV) n p+
+
Deuterium 2 H
n n p+
n n p+ p+
Tritium 3 H
Helium 4 He
+
n Neutron n
+
Kinetic energy E (17.6 MeV)
9
Be
13
C*
12
C*
n (4 MeV, average) 12 C
> Neutrons from a pulsed neutron generator (PNG) and an americium beryllium [AmBe] radioisotopic source. PNGs (top left) are self-contained particle accelerators that produce neutrons from a fusion reaction (bottom left). The neutrons leave with high kinetic energy of approximately 14 MeV of the total 17.6 MeV released. Typical PNG output is 3 × 108 neutron/s. AmBe sources, on the other hand, generate neutrons as by-products of nuclear reactions (right). The AmBe source contains a mixture of americium [ 241Am] and beryllium [ 9Be]. When 241Am decays to an excited state of neptunium [237Np*]—the * denotes an excited state—it emits 5.5-MeV alpha particles (α). To reach its final ground state, the excited 237Np* emits a 60-keV gamma ray (γ). A small fraction of the alpha particles from the 241Am react with the 9Be, resulting in the formation of an excited state of carbon [13C*], which emits 4-MeV neutrons (n) to reach an excited state of 12C*. The 12C* reaches its stable state through the emission of a high-energy gamma ray (approximately 4.4 MeV). A typical AmBe source generates 4 × 107 neutron/s.
that require greater understanding of the composition and mineralogy of the rocks. In the laboratory, scientists have an array of instruments at their disposal to peer into the rock structure. Using these tools, they can determine the chemical and mineral composition of the rocks, hypothesize about their origins and diagenesis and establish empirical relationships of rock properties that affect generation, accumulation and production of hydrocarbons. In the downhole environment and in words familiar to most petrophysicists, “All interpretations are opinions based on inferences from electrical or other measurements.” 2 However, as technologies and techniques advance, service companies are providing a number of laboratory-grade measurements from tools at the end of a wireline cable or attached to drillpipe. Spectroscopy measurements, which are crucial to understanding complex reservoir rocks and unconventional resource plays, have been used by scientists in the laboratory for several decades. Downhole spectroscopy tools have been available since their introduction in the 1980s, but the recently introduced Litho Scanner highdefinition spectroscopy service delivers geochemical data at a level of precision and accuracy that has never before been available downhole. The tool acquires measurements of a greater number of elements than were available from earlier tools and includes an accurate measurement of carbon, from which total organic carbon (TOC) can be derived. For understanding unconventional resources such as oil- and gas-bearing shales, TOC is crucial.
Spring 2014
This article reviews the basic theory of spectroscopy measurements and the development of neutron-induced capture spectroscopy tools, including advances in spectroscopy measurements introduced by the Litho Scanner tool. Case studies from an Arctic wildcat well, an oil-bearing resource play in the US and an unconventional resource play with complex lithology in the Middle East demonstrate various applications of spectroscopy data. Spectroscopy—Capturing Complexity Two families of downhole spectroscopy tools are used in the oil and gas industry: spectral natural gamma ray tools and neutron-induced gamma ray spectroscopy services. Geoscientists primarily use spectral gamma ray tools to quantify the concentrations of naturally occurring thorium, potassium and uranium in rocks by measuring the energy level of gamma rays emitted as these radioactive elements decay. From these data, log analysts can estimate clay type, quantify radioactive mineral effects on natural gamma ray measurements and identify radioactive deposits. Neutron-induced gamma ray spectroscopy, which is a more comprehensive measurement technique than that of spectral gamma ray tools, delivers concentrations of the most common elements found in the minerals and fluids of reservoir and source rocks. A neutron-induced spectroscopy tool records transitory effects—lasting a few microseconds to a few milliseconds—from formations bombarded with neutrons from a source: either an electronic pulsed neutron generator
(PNG) or an americium [241Am] and beryllium [9Be] radioisotopic source [AmBe] (above).3 The AmBe chemical sources used for downhole logging output a relatively stable number of neutrons with a predictable energy level. Compared with AmBe sources, PNGs generate many more neutrons and at much higher energy levels, but their output can vary with temperature, tool power and PNG age. Unlike AmBe sources that are always generating neutrons, when electrical power is removed from PNGs, neutron generation ceases. Laboratory spectroscopy tools such as X-ray diffraction (XRD) and X-ray fluorescence (XRF) spectrometers bombard rock samples with X-rays or gamma rays and measure the resulting emissions. To determine mineralogy, technicians use XRD devices; to perform elemental analysis, they use XRF equipment. The XRF equipment in the laboratory can measure more elements than can downhole tools. However, the subset of elements measured downhole includes the common mineral-forming elements, which are sufficient for geologists to determine the mineralogic composition of most reservoir and source rocks. 1. For more on the Archie water saturation equation: Archie GE: “The Electrical Resistivity Log as an Aid in Determining Some Reservoir Characteristics,” Petroleum Transactions of AIME 146 (1942): 54–62. 2. For many years, these words, or a similar statement, appeared on the printed logs provided by most service companies. 3. For more on PNGs used as neutron sources: Allioli F, Cretoiu V, Mauborgne M-L, Evans M, Griffiths R, Haranger F, Stoller C, Murray D and Reichel N: “Formation Density from a Cloud, While Drilling,” Oilfield Review 25, no. 2 (Summer 2013): 4–15.
35
Inelastic Neutron Scattering
Electronic source High energy
Excited nucleus
Traditional source 10 6
Neutron energy leaving source
n
Inelastic region
Deexcited nucleus
n
Neutron energy, eV
Intermediate energy 10 4
Inelastic gamma rays 10 2
Epithermal energy
Neutron Capture Capture gamma ray emitted
10 0
Excited nucleus
Deexcited nucleus
n Average thermal energy 0.025 eV
10 –2
Thermal neutron
Neutrons with thermal energy Capture gamma ray 200
400
Time, μs
> Life of a neutron and neutron scattering. Both electronic (PNG) and traditional (radioisotopic) sources emit high-energy neutrons. Neutrons from the PNG used in the Litho Scanner tool have approximately 14 MeV initial kinetic energy, whereas AmBe sources emit neutrons with around 4.4 MeV (left). These fast neutrons rapidly reach thermal energy level (approximately 0.025 eV). During those first few microseconds, before their energy falls below about 1 MeV, the neutrons experience inelastic interactions (top right). Inelastic neutron scattering occurs when high-energy, fast neutrons collide with, pass closely by or are absorbed by atomic nuclei. The now excited nucleus emits inelastic gamma rays to return to a deexcited state. Neutron capture (bottom right) occurs when thermal neutrons are absorbed by atomic nuclei. The capturing atom generates gamma rays to return to a deexcited state.
The first geochemical logs were created by combining measurements from several existing tools. In the late 1980s, scientists at the Schlumberger Houston Product Center, supported by researchers at the Schlumberger-Doll Research Center in Ridgefield, Connecticut, USA, combined the data from an NGT natural gamma ray tool, a GST gamma ray spectrometer tool and an AACT aluminum activation clay tool.4 From these data, they computed simple elemental concentrations for the following: aluminum [Al], calcium [Ca], iron [Fe], gadolinium [Gd], potassium [K], sulfur [S], silicon [Si], thorium [Th], titanium [Ti] and uranium [U]. These elemental concentrations provided information about mineralogy and rock composition. Although early geochemical tools provided geologists with information about the geochemistry of the rocks, first-generation tools suffered from some inherent limitations. These limitations include slow logging speeds, lack of combinability with other logging tools, degradation of both quality and resolution of the measurements in downhole environments, an inability to differentiate organic carbon from inorganic carbon and a lack of sensitivity for some elements that are essential for understanding complex lithology. For instance,
36
geologists use magnesium [Mg] to differentiate dolomite from calcite, and an accurate Mg measurement was difficult to obtain with older generation tools. Many geologists and petrophysicists consider geochemical logging data crucial for accurately characterizing reservoir rocks, but the tools were not universally included in traditional evaluation suites for many reasons. Among these were the facts that the tools were long, could not be combined with other services and had to be run slowly; also, the information could be obtained from core data. The application of the ECS elemental capture spectroscopy tool for shale gas exploration revolutionized the service.5 Because of its ability to provide the mineralogic composition of the rocks, a geochemical tool was frequently included in logging programs for unconventional reservoir evaluation and completion design. Elements of Neutron Capture Spectroscopy Of the many types of nuclear radiation, two are of particular interest for spectroscopy measurements—gamma rays and neutrons. Gamma rays are similar to X-rays and visible light and are the highest energy form of electromagnetic radiation. Visible light has a wavelength range of about
400 to 700 nm; gamma rays, with wavelengths much smaller than 1 nm, have a range of frequencies. Wavelengths typically encountered in downhole measurements are roughly 0.001 nm. However, gamma rays are not usually described by their wavelength, but by their energy level, expressed in electron volts (eV) or the larger units of keV (thousand eV) and MeV (million eV). Induced neutron spectroscopy tools count gamma rays over a range of discrete energy bins—the gamma ray spectrum. In essence, they measure the energies of artificially induced gamma rays emitted by elements in the formation that have been bombarded by high-energy fast neutrons supplied by the tool. These neutrons collide with other particles and rapidly lose energy until they eventually reach thermal energy level of about 0.025 eV. Because neutrons are similar in mass to hydrogen’s single proton, the maximum energy transfer and the neutron’s most rapid slowing occur from collisions between neutrons and hydrogen atoms (above).6 Thermal neutrons are eventually absorbed— captured—by the atomic nuclei of various elements found in the formation, the borehole and the tool. These now excited nuclei emit gamma
Oilfield Review
Spring 2014
Si Inelastic
Probability
Si Capture
0
2
4
6
8
10
0
2
4
Gamma ray energy, MeV Capture Gamma Ray Spectrum H
Gd
Cl
6
8
10
Gamma ray energy, MeV Inelastic Gamma Ray Spectrum
K
Si
Fe
CTB
Al
Ca
Fe
Ca
S
ITB
O
Counts
rays—referred to as capture gamma rays because they are a product of neutron capture— to return to their lowest stable energy state. Capture gamma rays have energy levels that are characteristic of the element from which they are emitted. Elastic scattering and eventual capture can take place over a few tens to hundreds of microseconds. Most downhole neutron capture spectroscopy tools rely on neutron-induced capture gamma rays for their measurements. Prior to reaching thermal energy level, highenergy neutrons that have not yet been significantly slowed may cause inelastic reactions. Inelastic reactions differ from elastic scattering and occur in about a microsecond after neutron bombardment. These interactions are characterized by atomic nuclei that become excited by encounters with neutrons with energy levels above 1 MeV. During inelastic interactions, neutrons may collide with an atomic nucleus, transfer energy to that nucleus and then emerge with reduced energy, or the fast neutron may be absorbed after first knocking a subatomic particle from the nucleus. As in neutron capture, nuclei become excited by these encounters and emit one or more gamma rays to return to a deexcited state; however, the gamma rays resulting from inelastic reactions have specific energy levels that differ from those of neutron-induced capture gamma rays for the same element (right). Only PNG-based tools can accurately distinguish between the effects of capture and inelastic neutron interactions, but not all PNG-based tools can make this measurement. To measure inelastic interactions, the neutron generator must be turned on and off rapidly, emitting pulses of high-energy neutrons. Furthermore, for accurate measurements, the pulse must have a welldefined, repeatable burst shape, meaning the neutron emissions have a constant and identical output for each pulse of neutrons. Most spectroscopy tools, including the ECS tool, detect gamma rays from inelastic reactions but cannot accurately determine elemental yields from these measurements. Some downhole logging tools may offer qualitative inelastic scatter data, but without hardware and measurement techniques to take advantage of the inelastic interactions, quantitative measurements are not possible. Measurements from inelastic interactions are less sensitive to environmental effects than those from capture interactions. For instance, chlorine [Cl] has a high thermal neutron capture cross section and can significantly reduce the number of thermal neutrons available for capture by other elements.7 Reducing the pool of available
C Mg
S
Ti
Al
Gamma ray energy
Si
Mg
Gamma ray energy
> Gamma ray spectra. Most neutron capture gamma ray spectroscopy logging tools rely on capture gamma rays to determine elemental yields. After absorbing thermal neutrons, atomic nuclei emit capture gamma rays with characteristic energies. For example, silicon [Si] (top left ) emits gamma rays with several emission energies, but the highest probabilities are approximately 3.5 and 4.8 MeV. The full capture gamma ray spectrum (bottom left ) is the combination of contributions from all the elements generally found downhole. Inelastic gamma rays are generated when fast neutrons—those with energies above 1 MeV—interact with nuclei in the formation, mud and the tool and result in the emission of gamma rays. These inelastic gamma rays have an energy spectrum (bottom right ) that looks similar to the capture gamma ray spectrum, but the characteristic energies differ. The Si inelastic gamma ray energy (top right ) is about 1.8 MeV. The Litho Scanner tool takes advantage of both spectra, which gives enhanced resolution to some elements, such as Mg and Fe, and adds others such as C, which is not available from the capture spectrum. Capture tool background (CTB, bottom left ) and inelastic tool background (ITB, bottom right ) are contributions to the measurement from the tool and the borehole environment detected during spectral acquisition.
thermal neutrons for capture increases the statistical variability of the measurement. Because the inelastic measurements are not affected by neutron absorbers, they can serve to enhance the resolution or precision of some capture data in the presence of high Cl levels.
The Litho Scanner tool utilizes capture gamma rays to determine concentrations of Al, Ca, Fe, Gd, K, S, Si and Ti, as other tools have done, but also quantifies concentrations of barium [Ba], Cl, hydrogen [H], Mg, manganese [Mn], sodium [Na] and metals such as copper [Cu]
4. Hertzog R, Colson L, Seeman B, O’Brien M, Scott H, McKeon D, Wraight P, Grau J, Ellis D, Schweitzer J and Herron M: “Geochemical Logging with Spectrometry Tools,” SPE Formation Evaluation 4, no. 2 (June 1989): 153–162. 5. For more on the ECS tool: Barson D, Christensen R, Decoster E, Grau J, Herron M, Herron S, Guru UK, Jordán M, Maher TM, Rylander E and White J: “Spectroscopy: The Key to Rapid, Reliable Petrophysical Answers,” Oilfield Review 17, no. 2 (Summer 2005): 14–33. 6. Radioisotopic neutron sources emit neutrons with energy levels on the order of 4 million eV and typically output 4 × 107 neutron/s. PNGs emit neutrons with energies
around 14 million eV and typically output 30 × 107 neutron/s and higher. Thermal neutrons are defined as those with energy of 0.025 eV. 7. Neutron capture cross section is a relative measurement of the probability of a nucleus capturing a neutron and has the unit of barns (1 barn = 10–24 cm2). Of the elements commonly encountered downhole, Cl is one of the most receptive for absorbing thermal neutrons, thus has a high capture cross section of 35 barns. Thermal neutron capture cross section is low for most other common downhole elements such as O (0.00019 barns), C (0.0035 barns), Si (0.17 barns) and Ca (0.43 barns). Its low capture cross section is one reason C concentrations are determined using inelastic interactions.
37
Element
Element Name
Al
Aluminum
Ba
Barium
C
Carbon
Ca
Calcium
Cl
Chlorine
Cu
Copper
Fe
Iron
Gd
Gadolinium
H
Hydrogen
K
Potassium
Mg
Magnesium
Mn
Manganese
Na
Sodium
Ni
Nickel
O
Oxygen
S
Sulfur
Si
Silicon
Ti
Titanium
Capture
Inelastic
> Elements determined through capture and inelastic gamma ray spectroscopy. (Adapted from Radtke et al, reference 9.)
and nickel [Ni]. The tool uses inelastic data primarily to quantify carbon [C] and Mg (above). With an accurate Mg measurement, petrophysicists can differentiate calcite [CaCO3] from dolomite [CaMg(CO3)2]. The accurate C measurement is crucial for determining TOC levels. Hidden in the Spectra Most downhole gamma ray logging tools use scintillation crystal detectors. When a gamma ray encounters the detector crystal, the energy of that
Gamma rays
Scintillation crystal
Light output
gamma ray is converted into a flash of light— hence the name scintillation—and the magnitude of the light pulse is proportional to the energy transferred to the crystal by the incident gamma ray. A photomultiplier tube converts the flash of light to a current, which it amplifies many times before sending it to additional electronics, where the analog signal is further amplified and converted to a digital value. The amplitude of the signal is determined by a pulse height analyzer, and these data are combined with all the other pulses Photomultiplier tube (PMT)
Amplification, pulse shaping and pulse height analyzer
Dynodes Gamma ray counts
Photo cathode
Anode
100 10 1 0.1 0.01 0
2
4
6
8
Gamma ray energy, MeV
> Scintillation detector. Gamma rays enter the scintillation crystal (top left ) causing a flash of light. The intensity of the flash is directly related to the energy transferred to the crystal by the incident gamma ray. The photomultiplier tube receives the light, converts it to a current, amplifies the current through a series of dynodes and passes the signal along for additional amplification, shaping and pulse height analysis (top right ). The information from all the gamma rays is combined, and counts are plotted versus discrete energy levels (bottom right ).
38
that arrive at the detector to produce a gamma ray spectrum (below left). Sodium iodide [NaI] crystals doped with thallium [Tl] are used as detectors in most conventional gamma ray logging tools, including some neutroninduced spectroscopy tools. Although the NaI crystal is robust, it is not efficient enough, nor is its resolution great enough to separate the spectra of all the desired elements. The ECS tool uses a bismuth germanate [Bi4Ge3O12], or BGO, crystal, which, because of its high density and atomic number, produces a unique gamma ray spectrum. However, the BGO scintillator is temperature sensitive; its spectral response broadens and loses definition or resolution, at elevated temperatures. The Litho Scanner tool uses a cerium-doped lanthanum bromide [LaBr3:Ce] crystal, which has a fast decay time that permits high counting rates as well as stable yields up to 200°C [400°F]. The light output of the crystal is 50% brighter than that of NaI crystals, the benchmark for scintillation crystals; at room temperature, its brightness is an order of magnitude higher than that of BGO crystals. The use of the LaBr3:Ce scintillator marks a significant increase in the ability to detect and count gamma rays, and thus when combined with the high neutron output of a PNG, constitutes a major advance in spectroscopy logging. To be useful to petrophysicists, the gamma ray spectrum measured by spectroscopy tools must be translated into relevant mineralogy, a multistep process. The first step is the acquisition of the gamma ray spectrum, which is a measure of gamma ray counts versus energy bins as determined by the scintillation detector. After the spectral response has been recorded, the spectrum must be converted to elemental yields. Each element detected by the tool has a unique signature, or elemental standard (next page, top). 8. Sedimentary minerals contain single or multiple oxides. Examples are quartz [SiO2], calcite [CaCO3] and dolomite [CaMg(CO3)2]. Clay minerals can also be treated as complex mixtures of oxides. Examples are illite {(K,H3O) (Al,Mg,Fe)2(Si,Al)4O10[(OH)2,(H2O)]} and montmorillonite [(Na,Ca)0.33(Al,Mg)2(Si4O10)(OH)2·nH2O]. Concentrations are expressed in weight %; the mass and not the volume of any given element contributes to the spectrum. For more information about the oxide closure method: Grau JA, Schweitzer JS, Ellis DV and Hertzog RC: “A Geological Model for Gamma-Ray Spectroscopy Logging Measurements,” Nuclear Geophysics 3, no. 4 (1989): 351–359. 9. Radtke RJ, Lorente M, Adolph B, Berheide M, Fricke S, Grau J, Herron S, Horkowitz J, Jorion B, Madio D, May D, Miles J, Perkins L, Philip O, Roscoe B, Rose D and Stoller C: “A New Capture and Inelastic Spectroscopy Tool Takes Geochemical Logging to the Next Level,” Transactions of the SPWLA 53rd Annual Logging Symposium, Cartagena, Colombia, June 16–20, 2012, paper AAA. 10. Radtke et al, reference 9.
Oilfield Review
Capture Standards
Capture Standards
Counts, arbitrary log scale
Fe Ca
Inelastic Standards
Cl
O
Na
Mg
K
Al
S Ti
Si
Si
Mn
S
Al
Ba
Ca
Mg
Gd
Fe C
H Gamma ray energy, MeV
Gamma ray energy, MeV
Gamma ray energy, MeV
> Elemental standards and tool calibration. Engineers characterized the Litho Scanner tool at the Schlumberger Environmental Effects Calibration Facility in Houston. The tool was placed in slabs of formation rocks (left ) and laboratory-prepared, simulated formations (right ) with known geochemical and lithologic composition. Standards were derived for 18 elements using capture spectroscopy and 13 elements using inelastic spectroscopy (center, not all shown). These standards are the basis for computing elemental yields.
tists apply the oxide closure model to the dataset.8 The closure model assumes that dry rock consists of a set of oxides or compounds, and the sum of the proportions of all the oxides must equal 100%, or unity. This closure requirement produces a unique normalization factor at each depth level, which in turn is applied to the relative spectral yields to produce the dry weight concentrations of specific elements.9 Dry weight elemental yields are then converted to mineralogy and lithology using software modeling programs. SpectroLith lithology pro-
These elemental signatures can be used to decompose the total measured spectra—which are first corrected for environmental and electronic factors that distort them—into the contributions from the elemental standards. In the case of the Litho Scanner tool, these standards were established in test formations at the Schlumberger Environmental Effects Calibration Facility in Houston. To obtain elemental weight fractions and produce realistic mineralogic models of the formation, scienSpectral Acquisition • Inelastic • Capture
Spectral Stripping • Elemental yields
Oxide Closure • Elemental weight fractions
Interpretation • Minerals • Total organic carbon (TOC) • Matrix properties C t d Corrected Density
Inelastic Normalized counts
Normalized counts
Al
Ca
Fe
S
ITB
Si
Ca
Fe Mg
S
Al
K
Na Mn
Ti
Gd
C
Illite Quartz K-feldspar Na-feldspar Calcite Dolomite Anhydrite Pyrite Kerogen
Inelastic
cessing for spectroscopy tools is one example. It is an empirical model developed from hundreds of laboratory measurements in known rock types.10 ELANPlus advanced multimineral log analysis is another technique. This analysis program computes the most probable formation mineralogy and pore volume based on inputs from several tools, including Litho Scanner yields (below). Geologists may use knowledge of expected rock types to guide the modeling software toward the correct mineralogic solution.
2 g/cm3 3
Matrix Density TOC
2 g/cm3 3
O C Si
Mg
Energy channel
Energy channel
Capture
Capture
Normalized counts
Normalized counts
H
Energy channel
Gd
Cl
K
Si
Fe
CTB
Ca
Mg
S
Ti
Al
Energy channel
> From acquisition to interpretation. Capture gamma ray and inelastic data (left ) are acquired with the Litho Scanner tool. Using elemental standards established for the tool, spectral stripping converts data to elemental yields (center left ). Software computes elemental weight fractions from these elemental yields based on the oxide closure model (center right ). Elemental analysis programs convert the yields or weight fractions to mineralogy (right, Track 1). The Litho Scanner tool also directly measures carbon, from which TOC is computed (Track 2). Petrophysicists can use matrix density computed from the elemental weight fractions and corrected for TOC (Track 3) to improve computed properties such as density porosity.
Spring 2014
39
Litho Scanner tool standard ECS tool standard
Room Temperature Fe
Nal (Tl)
BGO
LaBr3:Ce
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
Property
43.0
Light yield, photon/keV
Gamma ray counts
Si
Ca
S
H
LaBr3:Ce at 150°C, BGO at 60°C 1.8
LaBr3:Ce
1.6
Fe
1.2
Si
1.0
Nal (Tl)
0.8
Gamma ray counts
Relative light yield
1.4
0.6 0.4
BGO
0.2 0
0
50
100
150
Ca
S
200
Temperature, °C H
0
2
4
6
8
10
Gamma ray energy, MeV
> Crystal scintillator comparisons. Several types of scintillation crystals are used in gamma ray logging tools; the NaI crystal is the most common because of its ruggedness and low cost. A BGO scintillator is used in the ECS tool. For the Litho Scanner tool, engineers chose the LaBr3:Ce scintillator because of its superior qualities compared with those of other scintillators. The quick response time of the LaBr3:Ce scintillator—based on primary decay time—compared with that of other detectors (top left ) translates into greater efficiency and much higher counting capability. The relative light yield is stable from 0°C to 175°C [32°F to 350°F] (bottom left ), a clear improvement over the BGO scintillator, which can operate up to only about 60°C [140°F] before the output drops below a usable level. The light yield of the LaBr3:Ce detector is higher than that of either the NaI or BGO crystals. The LaBr3:Ce crystal detector is also more immune to thermal degradation than other detectors (right ). The clearly defined peaks for elemental standards at room temperature (top right, green) are similar to those at 150°C (bottom right ). The elemental standards response for the BGO crystal used in the ECS tool (red) broadens and loses definition at 60°C.
Litho Scanner Development Neutron-induced capture spectroscopy data have proved their value in characterization of complex lithologies in both conventional reservoirs and unconventional resource plays. However, petrophysicists who use these data have recognized some of the limitations of early spectroscopy tools. Engineers and scientists at Schlumberger worked for many years to develop a spectroscopy tool to address these concerns and correct issues that affect data accuracy and precision.
40
Since the raw spectra measured by the tool are the foundation upon which all other information is based, engineers searched for an alternative to BGO detectors used in the ECS tool, the gadolinium orthosilicate (GSO) [Gd2SiO5] detector used in the RST reservoir saturation tool and NaI detectors used in many other tools. One of the major operational reasons for replacing BGO detectors is their temperature sensitivity. BGO crystals are sealed in a Dewar flask and cooled with carbon dioxide [CO2] to keep the
tool internal temperature below 60°C [140°F] for the entire logging operation. The BGO crystal output drops dramatically with temperature—light output when the crystal temperature is greater than 60°C is too low to make acceptable logging measurements. This severely limits the use of the ECS tool for long duration logging such as drillpipe-conveyed or tractoring operations. Schlumberger design engineers chose a largediameter LaBr3:Ce gamma ray detector for use in the Litho Scanner tool. Compared with NaI and
Oilfield Review
Neutron output, relative counts
Litho Scanner Tool, Zero-Porosity Limestone Example
-2
0
2
4
6
8
10
Time, μs
> Stable and rapid neutron output. The hot cathode method used by the Minitron PNG delivers a rapid response when current is applied to the PNG and an even faster decay when power is switched off. This repeatable, controlled output allowed design engineers to develop the inelastic measurement that complements traditional neutron capture gamma ray spectroscopy.
Gamma ray count, arbitrary scale
High count rate Low count rate Pileup corrected
0
1
2
3
4
5
6
7
8
9
10
Gamma ray energy, MeV
BGO crystals, this scintillator has an order of magnitude faster response time. The faster response enables high counting rates, improving the tool’s precision compared with that of other devices. The brighter output compared with that of NaI and BGO scintillators translates to improved spectral resolution. The LaBr3:Ce scintillator has a stable response from 0°C to 150°C [32°F to 300°F], and even above 150°C, the light yield is not significantly reduced (previous page). The engineers also focused on the neutron source when developing the Litho Scanner tool. The PNG in the Litho Scanner tool includes a Minitron PNG tube that uses a proprietary hot cathode technology to produce crisp 8-μs bursts with 400-ns rise and fall times (above). The rapid response of this neutron generator allows precise separation of inelastic and capture interactions. Rated to 175°C [350°F], the PNG is capable of producing 3 × 108 neutron/s; this high output makes full use of the LaBr3:Ce scintillator’s fast counting capabilities, as the count rate can exceed 2.5 million count/s. Engineers designed a new state-of-the-art photomultiplier tube that is able to handle the output from the high count rates now possible from combining the LaBr3:Ce scintillator and the new PNG. The Litho Scanner tool also incorporates specialized electronics to process the high rate signals to avoid pileup, a condition in which more counts arrive than can be separated by the detector or the electronics (above right).11 Using fast signal processors to handle the load avoids spectral distortion caused by nearly coincident
Spring 2014
> Pileup distortion. When more gamma rays arrive at the detector than can be counted, pileup occurs, and the result is spectral distortion. The problem is more evident during high count rates (red) than low count rates (blue). Because the Litho Scanner tool utilizes a high neutron output PNG and an efficient LaBr3:Ce detector, pileup is most pronounced during inelastic spectrum measurements. Algorithms have been developed to remove the pileup degradation from the field spectrum based on the count rate (purple).
gamma ray arrivals. Unprecedented spectral resolution and precision are attained with the coupling of the scintillator, PNG, downhole electronics and signal processing. The combination of these enhancements results in the Litho Scanner tool as a high-definition, third-generation neutron capture gamma ray spectroscopy service. Spectroscopy, Rocks and TOC Because of the increased development of unconventional resources, the ability to quantify TOC in organic-rich rocks may be one of the most important features of the new tool. TOC is the weight % of organic carbon that resides within the pore space of rocks. TOC includes carbon in kerogen, bitumen and other solid, volatile and liquid hydrocarbons trapped within the pore space. Kerogen is the insoluble organic matter from which hydrocarbons are generated. Kerogen density is slightly higher than that of fluids that fill the pore space; using only bulk density measurements, petrophysicists have difficulty differentiating between liquid-filled pore volume and the presence of immovable bitumen in pores or kerogen in the rock framework. Computing the true porosity of organic-rich
shales requires the removal of solid hydrocarbon from the porosity measurement, which can be accomplished with accurate TOC data combined with other measurements such as those from magnetic resonance tools. For organic-rich shale exploration, geologists and petrophysicists target formations that have TOC values between 1.5 and 10 weight %. Rocks with more than 10 weight % TOC from kerogen only are usually considered too immature for development.12 TOC values are typically derived from core samples using a combustion technique in which inorganic carbon is removed using phosphoric acid. The remaining sample material is combusted in an oxygen-rich environment, and the resulting CO2 is measured in an infrared detection cell such as the LECO carbon analyzer. A limitation of determining TOC from cores is that 11. Pileup occurs when more gamma rays arrive at the detector than can be resolved by the system. Because of the high output of the PNG used in the Litho Scanner tool, pileup can be problematic during inelastic processing. If the response of the system to pileup can be characterized, the condition may be correctable. 12. Alexander T, Baihly J, Boyer C, Clark B, Waters G, Jochen V, Le Calvez J, Lewis R, Miller CK, Thaeler J and Toelle BE: “Shale Gas Revolution,” Oilfield Review 23, no. 3 (Autumn 2011): 40–55.
41
Litho Scanner TOC 0
%
20 0
%
Core TOC Depth, 0 m
%
Schmoker TOC
ΔLogR TOC
Litho Scanner TOC 20 0
Core TOC 20 0
%
Litho Scanner TOC
%
0 20
Core TOC 20 0
%
ΔLogR TOC
%
Schmoker TOC 20
Schmoker TOC 0
%
20
ΔLogR TOC
20 0
%
20
Calculated TOC, weight %
Schmoker TOC
20 15 10 5 0 -5
0
10 Core TOC, weight %
20
Measured TOC, weight %
Litho Scanner TOC XX,000
20 15 10 5 0 -5
XX,025
0
10 Core TOC, weight %
20
Calculated TOC, weight %
ΔLogR TOC
XX,050
20 15 10 5 0 -5
0
10 Core TOC, weight %
20
> Comparison of methods to determine TOC. Several techniques have been developed to quantify organic carbon indirectly from well logs. The Schmoker method utilizes density logs, and ΔlogR is based on sonic and resistivity data. The logs (left ) compare continuous outputs for Schmoker TOC (Track 1, blue), Litho Scanner TOC (Track 2, purple) and ΔlogR TOC (Track 3, tan) with core-derived TOC values (red dots). The three methods are shown together for direct comparison (Track 4). The crossplots (right ) compare calculated TOC weight % with core-derived TOC weight % values. The TOC data from the Litho Scanner tool (center right ) compared most favorably with core-derived TOC values, especially in rocks with high TOC weight %.
the core sample may not be representative of the rest of the reservoir; TOC can vary considerably across a reservoir section, which can be tens or even hundreds of meters thick. The Litho Scanner tool offers a continuous carbon measurement from which TOC data can be computed. A continuous dataset of TOC is a more cost-effective and statistically accurate option than measuring TOC on hundreds of core
samples. Many log-derived techniques—such as the Schmoker and ΔlogR methods—have been used to estimate TOC.13 The uncertainty can be high for the various indirect measurement techniques and most require calibration to core data (above). Log analysts use the carbon component from the Litho Scanner inelastic spectral measurements to quantify TOC. The carbon measurement from the formation includes both inorganic (car-
bon in minerals) and organic carbon. The inorganic carbon can be quantified by assigning it to the calcium and magnesium measurements, which are associated with calcite and dolomite; the amount of carbon bound up in these rocks can be computed by first quantifying these elemental weight fractions. Some calcium and magnesium may be associated with minerals other than carbonates. To address this situation, an
13. Gonzalez J, Lewis R, Hemingway J, Grau J, Rylander E and Schmitt R: “Determination of Formation Organic Carbon Content Using a New Neutron-Induced Gamma Ray Spectroscopy Service That Directly Measures Carbon,” Transactions of the SPWLA 54th Annual Logging Symposium, New Orleans, June 22–26, 2013, paper GG. For more on the Schmoker technique: Schmoker JW: “Determination of Organic-Matter Content of Appalachian Devonian Shales from Gamma-Ray Logs,” AAPG Bulletin 65, no. 7 (July 1981): 1285–1298. For more on the ΔlogR method: Passey QR, Bohacs KM, Esch WL, Klimentidis R and Sinha S: “From Oil-Prone Source Rocks to Gas-Producing Shale Reservoir—
Geologic and Petrophysical Characterization of Unconventional Shale-Gas Reservoirs,” paper SPE 131350, presented at the CPS/SPE International Oil and Gas Conference and Exhibition in China, Beijing, June 8–10, 2010. 14. Gonzalez et al, reference 13. 15. Several varieties of calipers are used to measure borehole diameter. An X-Y caliper measures the borehole diameter with two sets of arms positioned 90° apart and can more accurately describe the borehole geometry than can single-axis calipers. 16. “Canada’s Arctic,” Alberta Online Encyclopedia, Canada’s Petroleum Heritage, http://www.albertasource. ca/petroleum/industry/historic_dev_canada_arctic.html (accessed March 24, 2014).
17. Some historians consider Norman Wells—discovered around 1910—at 65° 16’ 52” N latitude in Northwest Territory, the first arctic oil field in Canada; however, it is located just south of the Arctic Circle, which is the defining line of the Canadian Arctic at 66° 33’ 44” N latitude. For reference, the Eagle Plain basin, located in Yukon, Canada, straddles the Arctic Circle. 18. For more on Arctic exploration: Bishop A, Bremner C, Laake A, Strobbia C, Parno P and Utskot G: “Petroleum Potential of the Arctic: Challenges and Solutions,” Oilfield Review 22, no. 4 (Winter 2010/2011): 36–49.
42
Oilfield Review
extensive set of Litho Scanner rock matrix measurements has been developed. Other less common minerals with inorganic carbon that might be encountered in oil and gas exploration include siderite [FeCO3], rhodochrosite [MnCO3] and ankerite [Ca(Fe, Mg, Mn)(CO3)]. Litho Scanner tools measure the elemental concentrations necessary to correct for the presence of these carbonbearing minerals.14 Remaining carbon can then be considered organic in nature and is equivalent to TOC. The organic carbon determined using this technique includes carbon in the kerogen, bitumen and any hydrocarbons—solid, oil and natural gas—in the pore volume. Correcting for Borehole Fluids Borehole fluids are another potential contributor of carbon to computed TOC. Determining TOC in wells drilled with water-base mud (WBM) systems is fairly straightforward. In the absence of organic-based additives, the organic carbon computed from tool measurements can be associated with solid, liquid or gaseous hydrocarbons. Additives in a WBM system may contribute to the total carbon measurement, and a constant correction is often applied to compensate for it. Oilbase mud (OBM) systems present a different challenge, and applying a constant offset may not always account for the borehole contribution, which is sensitive to borehole size and shape and to environmental effects. Scientists at Schlumberger-Doll Research, working in collaboration with engineers in the field for a solution to the OBM contribution to TOC, discovered that the correlation between the borehole carbon contribution and TOC is not a simple linear relationship. Because the composition of mud in the borehole can vary considerably from TD to surface, application of a simple offset correction may not be valid. The researchers were, however, able to develop a correction algorithm that has been successfully tested in both OBM and WBM systems. This new method computes an empirical carbon offset from the Litho Scanner carbon measurement as a function of borehole geometry determined from caliper data. Software then determines a correction factor to normalize results for the specific mud system. For the purpose of computing this final correction, an X-Y caliper is preferred, especially in hole sections prone to ovality or enlargement.15 The correction is applied at each depth frame (right). This technique recently proved its value in an Arctic exploration well in Yukon, Canada.
Spring 2014
Arctic Exploration Indigenous peoples in the Canadian Arctic were aware of oil seeps in that region for centuries and used pitch from these seeps to waterproof their fishing boats.16 But it wasn’t until 1974 that the
first Canadian Arctic oil field was discovered.17 In the recent past, oil, rather than natural gas, has often been the target of exploration in the Arctic because of oil’s portability; today, however, both natural gas and oil are viewed as targets.18
Litho Scanner TOC (Borehole Offset) 0
%
20
Litho Scanner TOC (Constant Offset) Litho Scanner TOC (Constant Offset) –2.5
Depth, m
%
Effective Hole Size 200
mm
0
20 0 325
%
20
Core TOC %
20
Correction Difference
X,X00
X,X50
X,Y00
> TOC correction for borehole contribution. Early methods of compensating for borehole fluid TOC contributions applied a constant offset to the TOC output; however, this method is sensitive to changes in borehole geometry. For example, the TOC computed with a constant offset (Track 1, black) generally follows the effective borehole size (magenta) when the hole is washed out. Because borehole integrity is often difficult to maintain while drilling in shales, data quality problems may be encountered. Recognizing this limitation, Schlumberger scientists developed a more effective method to compensate for TOC contributions from the mud system. This method computes the TOC contribution in an in-gauge hole section, uses X-Y axis calipers to model enlarged boreholes more accurately and applies a realistic depth-by-depth offset. The TOC computed with the new method (Track 2, blue) no longer reflects borehole geometry. The yellow shading indicates the difference between the constant offset correction (gray curve) and the borehole offset correction (blue curve) methods.
43
Ar
ct
ic
Cir
cle
Eagle Plain
Y u k o n
Arctic Circle
C
A
N
A
D
A
> Exploration in Arctic regions. Northern Cross (Yukon) Ltd. is exploring an area near the Arctic Circle in Yukon, Canada. Only 34 wells had been drilled in the company’s 5,000-km2 lease in the Eagle Plain basin prior to the operator’s recent activity. Harsh conditions in and around the Arctic Circle limit the drilling season and can potentially increase exploration and development costs. (Photograph courtesy of Don Stachiw.)
Northern Cross (Yukon) Ltd. has recently begun a campaign to actively explore the Eagle Plain region in northern Yukon, Canada, a basin covering more than 5,000 km2 [2,000 mi2] (above). Northern Cross speculates that Eagle Plain has the largest oil and gas potential of any onshore basin in the Yukon. The Arctic region is a harsh environment for drilling and exploration operations. Unlike locations in more temperate climates, vast areas in the Arctic have seen little to no drilling activity because of logistical difficulties. Across the expanse of the Eagle Plain basin, only 34 wells had been drilled prior to the Northern Cross exploration campaign, and these were drilled mostly in the 1960s and 1970s. The existing seismic data were 2D legacy surveys acquired before many of the recent advances in high-resolution 3D techniques. From previous drilling programs, engineers with Northern Cross knew that the basin was geologically complex, and drilling through some sections, including organic-rich shales, posed operational difficulties. Northern Cross targeted formations that include conventional reservoirs and unconventional resource plays. A strong potential for structural and stratigraphic traps in the basin exists,
44
and these traps may provide opportunities for conventional hydrocarbon production. For the initial exploration phase, the operator planned six wells, of which four had been drilled by the end of 2013. Because of their close proximity to the Dempster Highway, three wells are accessible year-round and have been drilled. Typical of many wells drilled in northern Canada, the other three locations are accessible only during winter months; one of these locations was drilled during the 2012–2013 drilling season. In addition to the logistical problems attributed to weather, operators exploring in the Arctic face other challenges. In developing petrophysical analysis programs, geologists must decide which tools and techniques should be utilized to best evaluate exploratory wells. These geologists face a daunting task, especially in complex reservoirs such as those of the Eagle Plain basin because there are few wells with legacy datasets for correlation and little state-of-the-art information about subsurface geology. Acquiring as much data as is economically feasible is the norm, which often includes taking conventional core.19 However, these are rank wildcat wells, and there are no offset wells to offer guidance in determining which intervals to core. To avoid the expense
of coring rock that has no production potential, engineers with Schlumberger suggested a traditional logging suite augmented with data from the Litho Scanner tool. These data could then be processed using the TerraTek HRA heterogeneous rock analysis service to determine optimal sidewall core points, which could be taken using a rotary coring tool.20 The output of the TerraTek HRA software is commonly used for determining geomechanical rock properties, but it also groups similar rock types.21 Engineers and geologists used the rocktyping feature to pick rotary core depths, thus ensuring desired rock types were represented in the sampling program while avoiding oversampling in rocks with similar properties. The geologists also used TOC data from the Litho Scanner tool to help further define core points. Because the wells were drilled with a water-base mud system, any conventional zones that displayed elevated TOC values should correspond to hydrocarbons in the pore space and be further evaluated. Because the processing was conducted in real time, the geologists were able to cross-reference rocks identified from Litho Scanner data as having high TOC content with superior reservoir
Oilfield Review
quality rock types identified from TerraTek HRA software (right). Rotary sidewall cores were taken, and recovery was considered excellent in both quality and quantity. Data from the high-graded rotary coring program helped confirm results from the Litho Scanner tool and provided lithology information similar in quality to that obtainable from whole core without the expense and operational inefficiencies associated with cutting conventional core. In addition, the operator avoided costly conventional coring over intervals of little interest. While processing the Litho Scanner data during the initial log evaluation, petrophysicists observed some puzzling results; a few intervals exhibited elevated TOC values where none were expected. These intervals generally corresponded to borehole washouts, pointing to the mud system as the source of the organic carbon. A review of the mud report revealed the culprit of the elevated TOCs. In some wells, the mud engineer occasionally used a lignite-based additive to improve drilling performance. Lignite, a low-rank coal, is a source of organic carbon, and its presence explained the elevated readings. The additive was not uniformly dispersed in the wells and was not present across all intervals. Schlumberger researchers had developed a borehole correction technique to account for organic carbon in oilbase mud systems. Engineers used the technique to correct for the presence of lignite, resolving the problem. In addition to the effects of mud additives encountered by log analysts evaluating these Arctic wells, operational issues related to drilling affected the logging programs. During the course of drilling two of the exploration wells, openhole logs were acquired prior to a planned casing point. Drilling deeper, the operator encountered difficulties in a shale section that necessitated drilling with a technique referred to as casing drilling, in which the drill bit and mud motor are attached to the casing. The interval is drilled, and rather than being pulled from the well when the rig reaches TD, the casing is cemented in place.22 Schlumberger and Northern Cross petrophysicists and geologists acquired Litho Scanner data in the cased section. Although spectroscopy data can be acquired in cased hole, the influence of the steel and cement behind the casing create data offsets that require corrections. No openhole logs over the casing-drilled section existed for comparison, but portions of the cased section overlapped some previously logged openhole intervals. By comparing openhole data to well logs obtained inside the casing, engineers were
Spring 2014
Litho Scanner Mineralogy Anhydrite Density Porosity
Siderite Core TOC
Pyrite Dolomite
0
Calcite Depth, m
30
Quartz+Feldspar+Mica 0 Clay
%
12 30
Litho Scanner TOC %
TOC
%
–10
Neutron Porosity %
–10
Corrected Porosity
12 30
%
–10
TerraTek HRA Rock Types
Rotary Core Depth
X,700
X,750
> The Litho Scanner tool as an alternative to conventional coring. Because of cost and drilling efficiency, conventional coring may not be an ideal choice for Arctic exploration wells; sparse offset well data may provide little guidance for determining coring intervals. Northern Cross geologists used the continuous mineralogy data from the Litho Scanner tool (Track 1) and TOC content computed from carbon data (Track 2, gray shading) to identify zones with hydrocarbon potential. They then applied TerraTek HRA software to identify similar rock types (Track 4) and determine the best depths for rotary sidewall coring (Track 5, black dots). TOC measurements from those cores (Track 2, red dots) compare favorably with Litho Scanner TOC measurements. The integration of these various data types resulted in sampling that provided representative cores without needless oversampling. Neutron porosity (Track 3, blue), density porosity (red) and Litho Scanner corrected porosity (black) computed using the true mineralogy are also presented; the lithology-corrected porosity demonstrates how Litho Scanner data can enhance petrophysical measurements.
19. For more on conventional coring: Andersen MA, Duncan B and McLin R: “Core Truth in Formation Evaluation,” Oilfield Review 25, no. 2 (Summer 2013): 16–25. 20. For more on rotary sidewall coring: Agarwal A, Laronga R and Walker L: “Rotary Sidewall Coring— Size Matters,” Oilfield Review 25, no. 4 (Winter 2013/2014): 30–39. 21. For more on the TerraTek HRA technique: SuarezRivera R, Deenadayalu C, Chertov M, Hartanto RN,
Gathogo P and Kunjir R: “Improving Horizontal Completions on Heterogeneous Tight Shales,” paper CSUG/SPE 146998, presented at the Canadian Unconventional Resources Conference, Calgary, November 15–17, 2011. 22. For more on the casing drilling technique: Fontenot KR, Lesso B, Strickler RD and Warren TM: “Using Casing to Drill Directional Wells,” Oilfield Review 17, no. 2 (Summer 2005): 44–61.
45
0
Openhole Mineralogy
Cased Hole Mineralogy
%
%
100 0
Anhydrite Cased Hole Gamma Ray 0
Depth, m
0
100
Pyrite
Pyrite
Dolomite
Dolomite
Calcite
Calcite
Openhole Gamma Ray
Quartz+Feldspar+Mica
Quartz+Feldspar+Mica
gAPI
Clay
Clay
gAPI
150
150
Cased Hole Litho Scanner TOC
Anhydrite –3
%
12
Openhole Litho Scanner TOC –3
%
12
TOC
X,600
X,650
> Spectroscopy data from inside casing. While drilling an exploration well in the Eagle Plain basin in Yukon, Canada, Northern Cross drilling engineers encountered hole problems that necessitated drilling with casing to reach TD. The cased interval included sections previously logged in open hole and sections not logged before setting casing. Geologists decided to acquire data inside casing with the Litho Scanner tool and compare it with the data from openhole runs. The gamma ray logs (Track 1) from the openhole (magenta) and cased hole (black) passes were corrected for casing and cement effects. Lithology and mineralogy data from the Litho Scanner tool run in open hole (Track 2) and cased hole (Track 3) have good agreement. The TOC data from openhole (Track 4, magenta) and cased hole (black curve, gray shading) measurements differ to some degree but are within the statistical limits of the measurement precision.
able to apply offsets and correct for the contributions from the steel and cement (above). Satisfied with the comparison of data from the previously logged openhole section and logs from the now cased section, Northern Cross had confidence that the data faithfully represented the lithology and TOC in the newly drilled portion. Northern Cross plans to continue its exploration program in the Yukon and is in the process of acquiring 3D seismic data across its lease position. Interpretation of log data indicates both oil and natural gas potential in the basin. What’s in a Name? When referring to resource plays, some industry professionals broadly apply the term shale to unconventional reservoirs. Although many unconventional reservoirs may not necessarily meet the
46
standard geologic definition of shale, the term is used to describe reservoir rocks that are often rich in clay and have very low permeability.23 The targets for exploration are generally referred to as organic-rich shales because they have relatively high volumes of kerogen, a source of hydrocarbons. To have the potential for hydrocarbon production, these rocks must possess the proper mineralogy, porosity, hydrocarbon saturation, organic content and thermal maturity.24 One other aspect of most successful plays is the presence of large volumetric quantities of nonclay components such as quartz, feldspar and carbonate. In contrast to clay, which tends to possess low strength and may be highly ductile, these nonclay minerals have high strength and contribute to a rock’s ease of fracture.
Most shale developments, such as the Barnett Shale, the Marcellus Shale and the Haynesville Shale, focus on rocks with a large proportion of quartz, feldspar and mica (QFM)—an assemblage of silicate minerals common in sedimentary rocks. An abundance of these minerals within the shale matrix may translate into successful unconventional wells. An exception to the QFM-rich reservoir model is the Eagle Ford Formation—or Eagle Ford shale—in south Texas, USA. This formation, which is the source rock for the prolific Austin Chalk formation, has produced both liquids and gas in relatively large volumes. The Eagle Ford Formation differs from many other shale plays because of its high carbonate content. As a result, the formation is amenable to hydraulic fracture stimulation.25 The Eagle Ford Formation extends from south Texas into northeast Mexico and is roughly 80 km [50 mi] wide and 644 km [400 mi] long (next page, top left). The average thickness is 76 m [250 ft] at reservoir depth, which is approximately 1,220 to 3,660 m [4,000 to 12,000 ft] deep. The Eagle Ford is sandwiched geologically between the Austin Chalk and the Buda Limestone formations. In some areas, the Maness Shale may lie between the Eagle Ford and the Buda Limestone. Results from a well recently drilled by BHP Billiton demonstrate the value of spectroscopy data for evaluating the complex mineralogy of the Eagle Ford Formation, especially when these data are combined with information from the CMR-Plus combinable magnetic resonance tool. The CMR-Plus tool was operated in a newly developed 50-burst enhanced precision mode that resolves small pores typically found in unconventional reservoir rocks.26 The TOC computed from 23. Shales are fine-grained rocks that form from the compaction of silt and clay-sized particles. Because they are formed from mud, they are also referred to as mudstones. Shales are differentiated from other claystones and mudstones in that they are laminated— finely layered—and fissile, which means they can be broken or split into sheets along their laminations. For more on shales and shale exploration: Alexander et al, reference 12. 24. For more on characteristics for targeting organic shale: Glaser KS, Miller CK, Johnson GM, Toelle B, Kleinberg RL, Miller P and Pennington WD: “Seeking the Sweet Spot: Reservoir and Completion Quality in Organic Shales,” Oilfield Review 25, no. 4 (Winter 2013/2014): 16–29. 25. For more on oil-prone source rocks and their evaluation: Passey et al, reference 13. 26. For more on the 50-burst enhanced precision mode: Hook P, Fairhurst D, Rylander E, Badry R, Bachman N, Crary S, Chatawanich K and Taylor T: “Improved Precision Magnetic Resonance Acquisition: Application to Shale Evaluation,” paper SPE 146883, presented at the SPE Annual Technical Conference and Exhibition, Denver, October 30−November 2, 2011.
Oilfield Review
Water Oil TOC Pyrite Dolomite
Eagle Ford Formation
UNITED STATES
0 0
km
300
miles
300
Depth, ft
Formation Name
T e x a s
Calcite Pyrite Quartz+Feldspar+Mica 0 Dolomite Bound Water Calcite Montmorillonite Kaolinite Quartz+Feldspar+Mica Clay Illite
Total CMR-Plus Porosity %
25
Bound Water Free Water Oil Kerogen
E
X
X,450
I
C
O
> The Eagle Ford Formation. The Eagle Ford Formation is the oil and gas source rock for the prolific Austin Chalk formation. In Mexico, it lies along the Mexican border with the US (red) and then extends north through central South Texas (green). Several E&P companies are evaluating the Eagle Ford for both oil and gas production.
Austin Chalk
M
X,475
the Litho Scanner tool’s carbon measurements consists of all forms of organic carbon, including kerogen, bitumen, coal and oil. Magnetic resonance measurements are sensitive only to liquids. The integration of fluid property measurements from the CMR-Plus tool with TOC data from the Litho Scanner tool allows geologists to distinguish between solid and liquid hydrocarbons and quantify oil potential for the reservoir.
Spring 2014
X,525
X,550
X,575
X,600
Maness Shale
. Optimizing liquids production from the Eagle Ford Formation. Operators developing the Eagle Ford (blue shaded interval) have found that oil can be produced economically. Based on the SpectroLith mineralogy data (Track 1), the Eagle Ford Formation is rich in calcite (light blue), unlike the clay-rich (tan shading) Maness Shale that lies below it; the Austin Chalk that lies above it is almost pure calcite. The calcite in the Eagle Ford facilitates hydraulic stimulation treatments. As seen in the ELANPlus mineralogy data (Track 2), the Eagle Ford has significant TOC content (Track 2, maroon shading)—the source of its oil; the TOC is composed of both oil and kerogen—the nonproductive solid hydrocarbon portion. Petrophysicists used the results from a combination of tools to determine the optimal interval for landing the lateral, locating the well in the better rock type for stimulation while also taking advantage of the liquids-rich section. For example, the clay component of the Eagle Ford is composed of varying amounts of montmorillonite, kaolinite and illite (Track 2); illite may be less ductile than other clay types and thus a target for fracture stimulation. Engineers also landed the lateral in the intervals with stiffer rocks such as the calcite-rich sections. To determine oil-bearing intervals, the density porosity was first corrected for matrix density from Litho Scanner mineralogy. This porosity (Track 3) is the sum of the volumes of all liquids and the solid hydrocarbons (kerogen). The CMR-Plus total porosity (Track 3, thick black curve) is the sum of all liquid volumes—clay-bound water (light blue), free water (blue) and oil (green). The difference between the CMR-Plus total porosity and the mineralogy-corrected density porosity is the unproductive kerogen portion of the TOC volume (Track 3, maroon shading). The remaining TOC volume, not associated with kerogen, must be the liquid oil volume.
Eagle Ford
X,500
X,625
Operators can use this information to plan the placement of lateral sections and make completion decisions. For the formation evaluation program, BHP conventionally cored the Eagle Ford section; sample plugs were taken at 1- to 5-ft [0.3- to 1.5-m] intervals and analyzed for TOC weight % using a LECO carbon analyzer. The wireline logging program included a traditional Platform Express
triple combo suite along with the CMR-Plus and Litho Scanner tools. Litho Scanner mineralogy data clearly differentiate the compositions of the Maness Shale and the Eagle Ford Formation (above). Compared with that of the Eagle Ford, the Maness section has a large volume of illite and smectite, which are ductile clays not suitable for hydraulic fracture stimulation. However, the most telling difference
47
between the two formations is the high TOC volume in the Eagle Ford, which is absent in the Maness. High organic carbon volume in the Eagle Ford Formation makes it a target for exploration. In the Eagle Ford Formation, TOC weight % from core analysis and from the Litho Scanner processed data ranges from 2 to 7 weight %. Organic carbon may be associated with both kerogen and oil in the formation; therefore, without more information about the composition of the TOC, fully evaluating the resource potential of this reservoir would be difficult. Nuclear magnetic resonance (NMR) data from the CMR-Plus tool helped resolve this uncertainty. NMR tools respond to liquids in formation rocks. If the pore space is filled with oil or water, the NMR porosity should replicate the porosity determined from the Litho-Density tool. Because gas has a low density and kerogen is a solid, the NMR porosity measured in rocks containing these substances will be lower than that computed from density tools.
Pores in unconventional reservoirs such as the Eagle Ford Formation are small, and therefore most NMR tools are incapable of properly measuring the total liquid volume. The CMR-Plus tool has the shortest echo spacing in the industry, which translates into the ability to resolve small pores and compute a more accurate liquid volume than other tools can in similar environments, especially when the tool is operating in the enhanced precision mode. The NMR porosity measurement includes water—both free and bound—and oil. In clay-rich rocks, most of the water measured by the CMR-Plus tool is bound water associated with the clays. For liquids-rich unconventional resource plays such as the Eagle Ford, petrophysicists can compare fluid volumes computed from CMR-Plus data with the Litho Scanner TOC volume and derive a volumetric oil component. Reservoir engineers can then use this information to determine the
Nafud Basin Stratigraphic Column Chronostratigraphy Lithostratigraphy Epoch Wenlockian
Stage Homerian
Silurian
Sheinwoodian Telychian Llandoverian
Aeronian
Qalibah Formation
Period
Rhuddanian
Zarqa facies
Sandbian
Llandeilian Darriwilian Llanvirnian
Quwarah member Qasim Formation
Ordovician
Katian
Ra’an member Kahfah member Hanadir member
Dapingian Floian
Tremadocian
Hot shale
Sarah Formation
Himantian
Arenigian
Qusaiba member Hot shale
Ashgillian
Caradocian
Sharawra member
Saq Formation
Tremadocian
> Nafud basin stratigraphic column. Geologists consider the organicrich Qusaiba Shale member of the Silurian-period Qalibah Formation the hydrocarbon source rock for many Middle East oil and gas fields. Because the gamma ray logs from the Qusaiba Shale have very high counts, the shale is considered a “hot shale.” High gamma ray counts indicate organic-rich shales, and geologists target these formations for exploration. (Adapted from Al-Salim et al, reference 27.)
48
volume of oil in place, estimate the oil production potential and make better informed decisions on where to land the lateral. Saudi Arabia Unconventional Reservoir Saudi Aramco utilized the Litho Scanner tool to evaluate formations in the Nafud basin and determine their potential as unconventional resource plays.27 The basin is characterized by a thick sequence of Paleozoic rocks from the Cambrian through Devonian periods. The Silurian-aged Qusaiba Shale—the target for these wells—is a member of the Qalibah Formation (below left). The organic-rich Qusaiba Shale is a prolific source of hydrocarbons, generating an estimated 90% of the Paleozoic light oil and gas found in the Middle East, and it is the source rock for many major oil and gas fields. The Qusaiba Shale is characterized by high gamma ray readings, which result from precipitated uranium in the reducing environment where the shale was deposited. The deepest shale intervals are Rhuddanian stage and typically have 8 to 9 weight % TOC on average. Younger Aeronian- and Telychian-aged intervals have lower TOC values. To evaluate the Litho Scanner tool’s ability to characterize the mineralogy of the formation and quantify TOC, Saudi Aramco ran the tool in two wells, one drilled with 10-lbm/galUS [1,200-kg/m3] OBM and another with 9.2-lbm/galUS [1,100-kg/m3] WBM. Saudi Aramco did not cut a conventional core in the first well drilled with OBM because LECO TOC core data were available from a well located about a mile away. These data compared favorably with TOC from the Litho Scanner tool run in the new OBM well. For a more direct comparison between log data and core measurements, the operator ran the Litho Scanner tool in a second well and cut core over the kerogen-rich zone of interest. The target formation in this case was the Rhuddanian hot shale. The operator conducted a special study on core samples. To minimize the effects of rock heterogeneity on core measurements and obtain measurements more representative of the volume probed by the spectroscopy tool, technicians took 1 ft [0.3 m] long trim core slabs. These samples were then crushed into a homogenized powder for analysis. Technicians used XRF to analyze a portion of the powder for elemental concentrations and a LECO total carbon analyzer to determine TOC.28 Schmoker TOC was computed from the formation
Oilfield Review
TOC from Core Mineralogy
0
Pyrite Dolomite
Depth, ft
Calcite Quartz+Feldspar+Mica Clay
0
Caliper 6
in. 16 0
Al from Core
Si from Core
Fe from Core
S from Core
%
%
%
%
20 0
50 0
20 0
10 0
Ca from Core
Mg from Core
Na from Core
%
%
%
20 0
10 0
Litho Scanner Dry Weight Al
Litho Scanner Dry Weight Si
Litho Scanner Dry Weight Fe
Litho Scanner Dry Weight S
Litho Scanner Dry Weight Ca
Litho Scanner Dry Weight Mg
Litho Scanner Dry Weight Na
%
%
%
%
%
%
%
20 0
50 0
20 0
10 0
20 0
10 0
K from Core 5 0
%
Litho Scanner Dry Weight K 5 0
%
%
20
Litho Scanner TOC
5 0
%
20
Schmoker TOC 5 0
%
20
X,600
X,650
X,700
X,750
> Dry weight yields and TOC from a Middle East well. To confirm the quality of downhole spectroscopy data, Saudi Aramco petrophysicists compared core-derived elemental yields from XRF measurements (Tracks 2 to 9, black dots) with Litho Scanner dry weight yields (red curves). The elemental concentrations show good agreement, except around X,600 ft, where there are high concentrations of pyrite (Track 1, orange) and TOC (Track 10). The recovered core from that zone was fractured and fragmented, possibly causing some depth mismatch when the core was analyzed. TOC computed from the Litho Scanner data (Track 10, red) was compared with core TOC (black dots) and TOC computed from the Schmoker technique (blue); Litho Scanner TOC matched core results better than did the Schmoker technique. (Adapted from Al-Salim et al, reference 27.)
density log as a third source for comparison.29 The results of the core laboratory measurements compared favorably with the Litho Scanner elemental dry weights and TOC data (above). Petrophysicists combined the Litho Scanner dry weight data with other logging data and computed the reservoir mineralogy; these data were then compared with dual range Fourier transform infrared (FTIR) spectroscopy measurements from the core (next page). The mineralogy analysis is a model-dependent computation, and application of the appropriate model is crucial for correct results. Engineers with Saudi Aramco and Schlumberger made several findings from the analysis of data from the Litho Scanner tool. The Litho Scanner TOC data closely matched core TOC data without empirical calibration. Using core-derived TOC values as the baseline for comparison, they determined that the Schmoker
Spring 2014
technique TOC was not as accurate as the TOC computed from the Litho Scanner carbon output. Because the Schmoker technique was developed specifically for Appalachian Devonian shales and the Bakken Formation, whose characterizations differ from those of the Nafud basin, the results are not surprising. Further refinement or calibration is necessary to apply this technique in formations other than the ones for which it was developed. The Litho Scanner tool provides reliable information for developing or refining petrophysical models in formations with complex lithologies. The improved accuracy in measuring certain elements allows petrophysicists to include more minerals in formation evaluation models to describe the reservoir rocks and better understand depositional environments. Correctly characterized mineralogy translates into more accurate matrix properties and, consequently,
more accurate porosity and water saturation computations. These benefits may be achieved in a fraction of the time and at a fraction of the cost of cutting and analyzing whole core. This information is especially important during exploration and early development stages when core data may be scarce or cover a limited area of a new prospect. 27. Al-Salim A, Meridji Y, Musharfi N, Al-Waheed H, Saldungaray P, Herron S and Polyakov M: “Using a New Spectroscopy Tool to Quantify Elemental Concentrations and TOC in an Unconventional Shale Gas Reservoir: Case Studies from Saudi Arabia,” paper SPE-SAS-312, presented at the SPE Annual Technical Symposium and Exhibition, Al-Khobar, Saudi Arabia, April 21–24, 2014. 28. X-ray fluorescence is a measurement technique that bombards materials with X-rays to ionize the atoms. Ionization results in the emission of characteristic fluorescent radiation in a manner similar to the element-specific gamma ray emissions from neutron capture. Individual elements in complex mixtures can be accurately measured in the laboratory using this technique. 29. Schmoker, reference 13.
49
Depth, ft
XRD-Derived Mineralogy
Litho Scanner Mineralogy
Biotite Ca-Feldspar Orthoclase Siderite Pyrite Muscovite Dolomite Calcite Ankerite Na-Feldspar Quartz Smectite Kaolinite Illite Chlorite
Siderite Pyrite Muscovite Dolomite Calcite Ankerite Na-Feldspar Quartz Smectite Kaolinite Illite Chlorite
Illite from Core 0
%
Kaolinite from Core 100 0
Litho Scanner Dry Weight Illite 0
%
100 0
%
Na-Feldspar from Core 100 0
%
Muscovite from Core 50 0
%
Siderite from Core 50 0
%
Pyrite from Core 20 0
%
Dolomite from Core 25 0
%
50
Litho Scanner Litho Scanner Litho Scanner Litho Scanner Litho Scanner Litho Scanner Litho Scanner Dry Weight Kaolinite Dry Weight Quartz Dry Weight Feldspar Dry Weight Muscovite Dry Weight Siderite Dry Weight Pyrite Dry Weight Dolomite
100 0
Illite
%
Quartz from Core
%
Kaolinite
100 0
%
Quartz
100 0
%
Na-Feldspar
50 0
%
50 0
Muscovite
%
Siderite
20 0
%
Pyrite
25 0
%
50
Dolomite
X,000
X,100
X,200
> Mineralogy comparison. Scientists at the Schlumberger-Doll Research Center performed FTIR spectroscopy analysis on cores from a well drilled with WBM and compared the XRD-derived core mineralogy (Track 1) with the mineralogy computed from Litho Scanner data and other log inputs (Track 2). Accurate mineralogy data are crucial for computing many petrophysical properties such as porosity and fluid saturations. In this well, mineralogy data helped petrophysicists make proper analyses; for instance, high K levels in sands can be attributed to orthoclase (K-feldspar) or muscovite (K-mica) (Track 7). The matrix density values of these minerals are 2.57 g/cm3 and 2.76 g/cm3, respectively. In this case, geologists have local knowledge of the rock types, and all the K was attributed to muscovite. The correct mineralogy results in a more accurate matrix density and, consequently, more accurate density porosity and water saturation computations. In addition, a better quality Na measurement from the Litho Scanner tool can be used to quantify concentrations of Na-bearing minerals such as albite—Na-plagioclase feldspar (Track 6)—with less uncertainty. (Adapted from Al-Salim et al, reference 27.)
Ultimate Answers Downhole spectroscopy is just one method petrophysicists use to determine the complex nature of reservoir rocks. Spectroscopy tools provide bulk measurements but are not able to determine rock fabric. For instance, the Litho Scanner tool can identify zones with pyrite but cannot determine how the mineral is dispersed. Similarly, the percentage of clay in one zone may be identical to that in another, but the tool cannot determine spatial distribution of the clay
50
particles, specifically whether they are structural, laminated or pore filling. Certain questions about mineral composition can be answered only from core analysis. Many mineralogic and lithologic conditions affect log responses, especially those of resistivity and nuclear tools. In this age of unconventional reservoir development, petrophysicists must rely on the integration of multiple data sources to understand the rock composition and fabric.
In earlier times, simple models sufficed to identify hydrocarbon productive zones and quantify production potential. Wells in which simple conditions prevail are becoming more uncommon. To characterize hydrocarbons in complex rocks and reservoirs, petrophysicists now have more and better tools and techniques at their disposal. Geologists and petrophysicists are using these new tools and techniques to help operators find and produce more oil and gas from increasingly complex plays. —TS
Oilfield Review
