The Lowdown on Low-Resistivity Pay

Austin Boyd Evaluating low-resistivity pay requires interpreters to discard the notion Harold Darling that water saturations above 50% are not economic. Various tools and Jacques Tabanou Sugar Land, Texas, USA techniques have been developed to assess these frequently bypassed zones, but there are no shortcuts to arriving at the correct petrophysi- Bob Davis Bruce Lyon cal answer. New Orleans, Louisiana, USA Lamination of beds Shale clasts Clay-lined burrows ■ Charles Flaum Clays are the primary cause of low- Ridgefield, Connecticut, USA resistivity pay and can form during James Klein and after deposi- ARCO Exploration and tion. They are distributed in the for- Production Technology 0.5 in mation as laminar Plano, Texas shales, dispersed clays and struc- Pore fillings Pore linings Clay grains Robert M. Sneider tural clays. Other Robert M. Sneider Exploration, Inc. causes of low-resistivity pay include Houston, Texas small grain size, as in intervals of Alan Sibbit igneous and meta- Kuala Lumpur, Malaysia morphic rock fragments, and conductive minerals Julian Singer Burrowed sand Ash shards Conductive pyrite like pyrite. New Delhi, India

For help in preparation of this article, thanks to Jay Tittman, consultant, Danbury, Connecticut, USA; Bar- bara Anderson, Ian Bryant, Darwin Ellis, Mike Herron, Bob Kleinberg, Raghu Ramamoorthy, Pabitra Sen, 0.25 mm Chris Straley, Schlumberger-Doll Research, Ridgefield, Connecticut; David Allen, Kees Castelijns, Andrew Kirk- wood and Andre Orban, Schlumberger Wireline & Test- When Conrad and Marcel Schlumberger low oil prices driving the reexploration of ing, Sugar Land, Texas, USA; Steve Bonner and Trevor Burgess, Anadrill, Sugar Land, Texas; Dale Logan, invented the technique of well logging, low- mature fields, methods of interpreting low- Schlumberger Wireline & Testing, Roswell, New Mexico, resistivity pay was, practically speaking, a resistivity pay have proliferated. USA; and Pierre Berger, GeoQuest, Bangkok, Thailand. contradiction in terms. Their pioneering This article examines the causes of low- research hinged on the principle that gas- or resistivity pay in sands, then explores the oil-filled rocks have a higher resistivity than tools and techniques that have been devel- water-filled rocks. Through the years, how- oped to evaluate such zones. A case study ever, low-resistivity pay has become recog- shows how log/core integration helps pin- nized as a worldwide phenomenon, occur- point the causes of low-resistivity pay in the ring in basins from the North Sea and Gandhar field in India. Indonesia to West Africa and Alaska. With Generally, deep-resistivity logs in low- resistivity pay read 0.5 to 5 ohm-m. “Low

4 Oilfield Review Lowstand basin floor fan complex

Leveed channel Overbank complex deposits Overbank ■The most com- The cation exchange capacity, or CEC, mon depositional expressed in units of milliequivalent3 per environments for low-resistivity pay: 100 grams of dry clay, measures the ability A) Lowstand basin of a clay to release cations. Clays with a floor fan complexes high CEC will have a greater impact on low- B) Deep water ering resistivity than those with a low CEC. levee-channel For example, montmorillonite, also known complexes and B overbank deposits as smectite, has a CEC of 80 to 150 Transgressive C) Transgressive- meq/100 g whereas the CEC of kaolinite is marine sands marine sands only 3 to 15 meq/100 g. D) Lower parts Clays are distributed in the formation (toes) of delta front deposits and lami- three ways: nated silt-shale- • laminar shales—shale layers between sand intervals in sand layers the upper parts of • dispersed clays—clays throughout the alluvial and dis- sand, coating the sand grains or filling the tributary channels. (Adapted from Dar- pore space between sand grains ling HL and Sneider • structural clays—clay grains or nodules in RM: “Productive Low the formation matrix. C Resistivity Well Logs Laminar shales form during deposition, Alluvial Distributary Delta front “toes” and of the Offshore Gulf interspersed in otherwise clean sands (left). channel channel shingled turbidites of Mexico: Causes and Analysis,” in In the Gulf Coast, USA, finely layered sand- reference 1.) stone-shale intervals, or thin beds, make up

In this article, AIT (Array Induction Imager Tool), ARC5 (Array Resistivity Compensated), CBT (Cement Bond Tool), CDR (Compensated Dual Resistivity tool), CMR (Combinable Magnetic Resonance tool), CNL (Compen- sated Neutron Log), DLL (Dual Laterolog Resistivity), ELAN (Elemental Log Analysis), EPT (Electromagnetic Propagation Tool), FMI (Fullbore Formation MicroIm- ager), Formation MicroScanner, GeoFrame, GLT (Geo- chemical Logging Tool), Litho-Density, IPL (Integrated Porosity Lithology), MicroSFL, NGS (Natural Gamma Ray D Spectrometry tool), Phasor, RAB (Resistivity-at-the-Bit tool), SFL (Spherically Focused Resistivity), SHARP (Synergetic High-Resolution Analysis and Reconstruction contrast” is often used in conjunction with clay contributes to low-resistivity readings for Petrophysical Parameters) and TDT (Thermal Decay low resistivity, indicating a lack of resistivity depends on the type, volume and distribu- Time) are marks of Schlumberger. Sun is a mark of Sun Microsystems, Inc. contrast between sands and adjacent shales. tion of clay in the formation. 1. Moore D (ed): Productive Low Resistivity Well Logs Although not the focus of this article, low- Clay minerals have a substantial negative of the Offshore Gulf of Mexico. New Orleans, contrast pay occurs mainly when formation surface charge that causes log resistivity val- Louisiana, USA: Houston and New Orleans Geologi- waters are fresh or of low salinity. As a ues to plummet.2 This negative surface cal Societies, 1993. 2. Scala C: “Archie III: Electrical Conduction in Shaly result, resistivity values are not necessarily charge—the result of substitution in the clay Sands,” Oilfield Review 1, no. 3 (October 1989): low, but there is little resistivity contrast lattice of atoms with lower positive valence 43-53. between oil and water zones. —attracts cations such as Na+ and K+ when 3. One milliequivalent equals 6 x 1020 atoms. Because of its inherent conductivity, clay, the clay is dry. When the clay is immersed and hence shale, is the primary cause of in water, cations are released, increasing the low-resistivity pay (previous page).1 How water conductivity.

Autumn 1995 5 Evaluated Gas Pay Potential Gas Pay Spherically Focused Density Porosity Short Normal Resistivity MDEN=2.68 Resistivity Spontaneous Potential 60 0 0.2 ohm-m 20 -160 40 0.2 ohm-m 20 Compensated Depth, m Spontaneous Potential 6FF40 Induction Total Gamma Ray Deep Induction Neutron Porosity -160 40 0.2 20 0 GAPI 150 0.2 ohm-m 20 60 p.u. 0

X100

X200

■Left: Induction Electrical Survey logs run in 1960 in a thinly bedded, gas-bearing section of the Vicksburg formation in south Texas, USA. Net pay is 7 ft. Right: Conventional triple combo—neutron, density and gamma ray tools—run in 1993 in a well offset 100 ft from the original 1960 well. Net pay is 14 ft.

about half the low-resistivity zones.4 Many can be produced (see “Low-Resistivity Pay fragments—all fine grained— mimic the log logging tools lack the vertical resolution to in the Gandhar Field,” page 8). signature of clays, featuring high gamma ray, resolve resistivity values for individual thin Structural clays occur when framework low resistivity and little or no spontaneous beds of sand and shale. Instead, the tools grains and fragments of shale or claystone, potential (SP). Unlike thin beds, this type of give an average resistivity measurement over with a grain size equal to or larger than the low-resistivity pay can vary in thickness the bedded sequence, lower in some zones, framework grains, are deposited simultane- from millimeters to hundreds of meters. higher in others. ously. Alternatively, in the case of selective Finally, sands with more than 7% by vol- Intervals with dispersed clays are formed replacement, diagenesis can transform ume of pyrite, which has a conductivity during the deposition of individual clay par- framework grains, like feldspar, into clay. greater than or equal to that of formation ticles or masses of clay. Dispersed clays can Unlike dispersed clays, structural clays act water, also produce low-resistivity readings.5 result from postdepositional processes, such as framework grains without altering reser- This type of low-resistivity pay is considered as burrowing and diagenesis. The size differ- voir properties. None of the pore space is rare. ence between dispersed clay grains and occupied by clay. The challenge for interpreting low-resistiv- framework grains allows the dispersed clay Other causes of low-resistivity pay include ity sands hinges on extracting the correct grains to line or fill the pore throats between small grain size and conductive minerals measurement of formation resistivity, esti- framework grains. When clay coats the sand like pyrite. Small grain size can result in low mating shaliness and then accurately deriv- grains, the irreducible water saturation of resistivity values over an interval, despite ing water saturation, typically obtained from the formation increases, dramatically lower- uniform mineralogy and clay content. The some modification of Archie’s law.6 ing resistivity values. If such zones are com- increased surface area associated with finer Improved vertical resolution of logging pleted, however, water-free hydrocarbons grains holds more irreducible water, and, as tools and data processing techniques are with clay-coated grains, the increasing helping to tackle thin beds. Nuclear mag- water saturation reduces resistivity readings. 6 Intervals of igneous and metamorphic rock Oilfield Review Potential Gas Pay 2 ft [0.6 m] and 4 ft [1.2 m]. The FMI tool images the borehole with an array of 192 Spontaneous Potential button sensors mounted on four pads and four flaps.8 It has a vertical resolution of -160 MV 40 0.2 in. [5 mm]. Successive improvements in resolving thin APS Capture Density Porosity Cross-Section (SIGF) MDEN=2.68 (DPO) beds are strikingly visible in a series of logs 10 c.u. 40 60 p.u. 0 made 33 years apart in adjacent wells in the south Texas Vicksburg formation (previous Depth, m HNGS Potassium Content AIT Resistivities Neutron Porosity page and left).9 In 1960, induction/ short (HFK) 10-90 in. Sandstone (APSC) normal logs indicated 7 ft of net gas pay and 0 p.u. 5 0.2 ohm-m 20 60 p.u. 0 only two beds with resistivity greater than 2 HNGS Thorium Content Differential Caliper ohm-m. In 1993, a new well was drilled (HTHO) within 100 ft [30 m] of the original well and ppm 0 45 -20 20 logged with conventional wireline tools. The induction/SFL Spherically Focused X100 Resistivity logs doubled the estimated pay to 14 ft [4.3 m], with seven beds above 2 (continued on page 11)

4. Thin beds have a thickness of 5 to 60 cm [2 in. to 2 ft] and laminae are less than 1-cm [0.4-in.] thick, commonly 0.05 to 1 mm [0.002 to 0.004 in.]. Bates RL and Jackson JA (eds): Glossary of Geology. Falls Church, Virginia, USA: American Geological Institute, 1987. Dictionary of Geological Terms. New York, New York, USA: Doubleday & Co., 1984. 5. Clavier C, Heim A and Scala C: “Effect of Pyrite on Resistivity and Other Logging Measurements,” Transactions of the SPWLA 17th Annual Logging Symposium, Denver, Colorado, USA, June 9-12, 1976, paper HH. X200 6. In 1942, Gus Archie proposed an empirical relationship linking a rock’s resistivity, R , with its porosity, φ t , and water saturation Sw :

R w R t = m n . φ S w Other terms in the equation are the formation water resistivity Rw , and the cementation and saturation exponents, m and n. For further reading: “Archie’s Law: Electrical Conduction in Clean, Water-Bearing Rock,” The Technical Review 36, no. 3 (July 1988): 4-13. “Archie II: Electrical Conduction in Hydrocarbon- Bearing Rock,” The Technical Review 36, no. 4 (October 1988): 12-21. For a discussion on the numerous versions of Archie’s law that have been developed to handle a variety of shaly sand environments: Worthington PF: “The Evolution of Shaly-Sand Con- cepts in Reservoir Evaluation,” The Log Analyst 26 (January-February 1985): 23-40. ■AIT Array Induction Imager Tool and IPL Integrated Porosity Lithology logs run in the same well as conventional triple combo on previous page. The improved 7. Barber TD and Rosthal RA: “Using a Multiarray Induction Tool to Achieve High-Resolution Logs with vertical resolution of AIT logs and the enhanced sensitivity of the IPL-derived Minimum Environmental Effects,” paper SPE 22725, neutron porosity have increased net pay to 63 ft. presented at the 66th SPE Annual Technical Confer- ence and Exhibition, Dallas, Texas, USA, October 6-9, 1991. netic resonance (NMR) logging shows One obvious method for resolving the resis- Hunka JF, Barber TD, Rosthal RA, Minerbo GN, promise for assessing irreducible water satu- tivity of thin beds is to develop logging tools Head EA, Howard AQ Jr and Hazen GA: “A New ration associated with clays and reduced with higher vertical resolution, deeper depth Resistivity Measurement System for Deep Formation Imaging and High-Resolution Formation Evaluation,” grain size (see “Nuclear Magnetic Reso- of investigation, or both. Two logging paper SPE 20559, presented at the 65th SPE Annual nance Imaging—Technology for the 21st devices that have proved especially helpful Technical Conference and Exhibition, New Orleans, Century,” page 19). And because the most in evaluating thin beds are the AIT Array Louisiana, USA, September 23-26, 1990. opportune time to measure resistivity occurs Induction Imager Tool and the FMI Fullbore 8. FMI* Fullbore Formation MicroImager. Houston, Texas, USA: Schlumberger Educational Services, during drilling, when invasion effects are Formation MicroImager tool. The AIT tool 1992. minimal, resistivity measurements at the drill uses eight induction-coil arrays operating at 9. Olesen J-R, Flaum C and Jacobsen S: “Wellsite bit also play an important role in diagnosing multiple frequencies to generate a family of Detection of Gas Reservoirs with Advanced Wire- line Logging Technology,” Transactions of the 7 low-resistivity pay. five resistivity logs. The logs have median SPWLA 35th Annual Logging Symposium, Tulsa, Thin Beds depths of investigation of 10, 20, 30, 60 and Oklahoma, USA, June 19-22, 1994, paper Y. 90 in. and vertical resolutions of 1 ft [0.3 m], Autumn 1995 7 Low-Resistivity Pay in the Gandhar Field

The Gandhar field, on the western coast of India, is the largest on-land field in the country (left).

er Most hydrocarbon production comes from deltaic Khambhat Riv aga Mahis sands of the Hazad member, three of which con- tain low-resistivity pay. Dabka One of these sands, called GS-11, has resistiv-

r ity values of 2 to 6 ohm-m, but contains wells ive r R that produce clean oil on the order of 50 m3/d ha ad Dh [315 B/D] (next page). A detailed study of GS-11, integrating core and log data, allowed inter- Gandhar preters to unravel the low-resistivity phenomenon and formulate a reliable mineralogical model and water saturation estimates. er Riv ada Narm G U L F O F Core Studies C A M B A Y Sixty core samples from three GS-11 wells pro- vided thin sections for study of texture and miner-

0 miles 15.5 alogy. Polished sections helped reveal the pres- 0 km 25 ence of metallic minerals. Scanning electron microscope (SEM) and X-ray diffraction (XRD) Delhi studies of cores identified clay minerals. In addi- ■ Gandhar field on the western tion, laser and sieving methods were used to I N D I A coast of India. analyze grain size. The core investigations showed several mech- anisms contributing to high conductivity. Medium- to fine-grained sands ranged from gray to green-gray, with green indicating chloritic

Clay Coating Quartz Overgrowth

200 µm 20 µm ■SEM photographs showing coated grains and clay matrix (left) and quartz overgrowth with chlorite coating on quartz grains (right).

8 Oilfield Review clays. Bioturbation created thin, fine clay lamina- Gamma Ray tions over clean sands. Quartz was the most 0 GAPI 150 prominent mineral, with minute opaque SP Deep Resistivity Density 3 minerals—pyrite or magnetite—occurring in -25 MV 125 1.95 g/cm 2.95

bioturbated sections. Pyrite, which increases Depth, m Caliper Shallow Resistivity Neutron Porosity the formation conductivity, was limited to the 6 in. 16 0.2 ohm-m 2000 45 p.u. -15 clayey part of the matrix and constituted less than 5% by volume. Clay, primarily chlorite, coating the grain sur- faces was indicated by SEM pictures and XRD studies (previous page, bottom). Smaller grains were coated more than larger grains. Laser analysis of samples shows the GS-11 sand to XX80 be in the silt range, with grain sizes averaging 22 to 32 microns.

Formation Evaluation Logs were analyzed to identify clay types and

heavy minerals. Thorium-potassium crossplots of XX90 the NGS Natural Gamma Ray Spectrometry logs identified predominant clays as chlorite in the sands and kaolinite/chlorite in the shales. The density-neutron crossplot showed a trend toward

high density (low porosity) with little increase in GS-11 sand the neutron. The particles associated with this X100 behavior, which included fine-grained quartz and heavy minerals such as siderite, pyrite and ilmenite, were collectively called silt. From core- and log-derived information, a mineralogical model of kaolinite, chlorite, quartz and silt was chosen for the GS-11 sands. Validation for the model came from geochemical analysis of X110 21 core samples from different wells. A few samples were analyzed to determine the weight per- ■Log response from Well Z shows an average resistivity reading of 3 to 4 ohm-m over the GS-11 sand, which produced clean oil during conventional testing. cent of oxides, such as silicon dioxide [SiO2], using X-ray fluorescence (XRF) and the results were interpolated between samples. The percentages were then converted into weight percent of elements using standard tables and processed

Autumn 1995 9 with a mineralogical model to give weight per- Weight % of Minerals from Volume % of Minerals Log Analysis XRF Analysis of Oxides from XRF Weight % cent of minerals. The model based on geochemi- from Cores and Log Porosity cal analysis was constrained to include only Free Water Free Water quartz, kaolinite, chlorite and ilmenite. This

Quartz Quartz Quartz constraint allowed the weight percent of minerals to be converted to volume percent using the Silt Ilmenite Ilmenite total porosity from log interpretation and the Bound Water Bound Water Bound Water mineral densities.

Chlorite Chlorite Chlorite Comparison of the log and XRF mineral analy- Depth, m Core Description ses shows agreement between the total clay per- Kaolinite Kaolinite Kaolinite 1:100 m centage and the relative volume of kaolinite and chlorite (left). The silt and ilmenite percentages XX54 do not agree, as might be expected since the silt was defined to include finer grained quartz.

XX56 Conclusions The composite results from the extensive log- core analysis show agreement between core- and log-derived parameters (next page). Water satu- XX58 ration values computed from the Waxman-Smits equation compare well with those derived from capillary pressure measurements.1 Because little XX60 water had been produced from existing GS-11 wells, the log-derived water saturation values were considered to represent irreducible water

XX62 saturation values. The core studies showed that the low-resistiv- Sandstone Coarse Bioturbated Silty carbon- Laminated Shale sandstone sandstone aceous shale silty shale ity measurements in the GS-11 sand have two sources. First, individual sand grains are coated ■Comparison between log and XRF mineral analyses of Well Y. A mineralogical model of kaolinite, chlorite, with clay. Second, the silt-sized formation grains quartz and silt was chosen. lead to higher irreducible water saturations in the formation.

1. Waxman and Smits modified Archie’s law to account for the increased conductivity of shale by introducing a shaliness parameter based on cation exchange capacity (CEC). See: Waxman MH and Smits LJM: “Electrical Conductivi- ties in Oil-Bearing Shaly Sands,” Society of Petroleum Engineers Journal 8, no. 2 (1968): 107-122.

10 Oilfield Review Quartz ohm-m. Later the same year, the second well was logged with a combination of AIT Silt m from Q from Logs Sw from Moved and IPL Integrated Porosity Lithology tools.10 EPT/MicroSFL v Waxman-Smits Hydrocarbon Bound Water The high resolution of the AIT tool—1 ft ver- 1.0 3.0 0 2.0 100 0 Chlorite sus 2 ft for the induction—and the Moved Water m= enhanced sensitivity of the IPL-derived neu- Qv from Sw from f(Q from Logs) Kaolinite Bad v Co vs Cw Archie Water tron porosity increased net pay to 63 ft [19.2 Hole 1.0 3.0 0 2.0 100 0 Combined Model m] and showed 13 beds with resistivity Flag Oil p.u. 0 100 greater than 2 ohm-m. m from Q from S from Fluid Analysis 5.0 0.0 v wirr C vs C Wet Chemistry Cap Studies φ o w p.u. from Core Resistivity Measurements at the Bit 1:200 m 1.0 3.0 0 2.0 100 p.u. 0 50 100 100 0 Improvements in measurements-while- drilling (MWD) technology have not only boosted the efficiency of directional drilling, x780 but also enhanced thin-bed evaluation.11 Two tools‚ the RAB Resistivity-At-the-Bit tool and the ARC5 Array Resistivity Compen- GS-11 sated tool—are especially useful in thin-bed environments by providing resistivity data before invasion has altered the formation. The RAB tool provides five different resis- x790 tivity readings plus gamma ray, shock and tool inclination measurements. Configured as a stabilizer or a slick collar, the RAB tool is run behind the bit in a rotary drilling assembly and above the motor in a steerable drilling assembly. One resistivity measurement, called “bit resistivity,” uses the drill bit as part of the x800 transmitting electrode. With the RAB tool attached to the bit, alternating current is cir- culated through the collar, bit and formation before returning to the drillpipe and drill collars above the transmitter. In the case of oil-base mud, which is an insulator, the current loop is complete only when the collars and stabilizers touch the borehole wall. The x810 vertical resolution of the RAB bit resistivity is GS-10 only 2 ft and it gives the earliest possible warning of changes in formation resistivity. Four additional resistivity measurements, with 1-in. vertical resolution for thin-bed applications, are made with three button ■Composite log-core analysis of Well X. Core results are shown for the cementation exponent m; the CEC electrodes and a ring electrode. The shallow φ normalized for pore volume, Q v, irreducible water saturation Swirr and porosity . Q v, the CEC per volume of depths of investigation—3, 6 and 9 in. for pore fluid, was calculated from cores, by measuring resistivity at different water salinities, and from logs.

10. “Neutron Porosity Logging Revisited, ”Oilfield Review 6, no. 4 (October 1994): 4-8. 11. Bonner S, Burgess T, Clark B, Decker D, Lüling M, Orban J, Prevedel B and White J: “Measurements at the Bit: A New Generation of MWD Tools,” Oil- field Review 5, no. 2/3 (April/July 1993): 44-54. Allen D, Bagersh A, Bonner S, Clark B, Dajee G, Dennison M, Hall JS, Jundt J, Lovell J and Rosthal R: “A New Generation of Electrode Resistivity Mea- surements for Formation Evaluation While Drilling,” Transactions of the SPWLA 35th Annual Logging Symposium, Tulsa, Oklahoma, USA, June 19-22, 1994, paper OO.

Autumn 1995 11 100 ■ Evaluating inva- the buttons and 12 in. for the ring electrode sion with the RAB Laminated wet sands tool. In laminated —allow interpreters to characterize early- wet sands, the RAB time invasion (left). logs made after The recently-introduced ARC5 tool pro- drilling and while vides five phase and attenuation resistivity drilling anticorre- MicroSFL late, showing pref- measurements, like the AIT tool, with a verti- erential invasion. cal resolution of 2 ft. With a 4 3/4-in. diame- ter, it is especially useful for formation evalu- AIT ation in slim holes typical of deviated drilling (next page, top). The measurements and spacings of the RAB ring ARC5 and AIT tools are comparable, after drilling although not identical, making petrophysi- Resistivity, ohm-m Resistivity, cal evaluation with either tool in the same well or between wells seamless (below left). The multiple measurements of the ARC5 RAB ring tool also allow interpreters to radially map while drilling out the invasion process. The additional phase and attenuation measurements provide a better characterization of electrical 0.2 anisotropy than existing MWD tools. 570 580 590 600 Distance, ft Improving Thin-Bed Evaluation Through Data Processing Despite the emphasis on developing high- ARC5 Phase Shift Resistivities at CAT Well resolution resistivity logging tools, many 34 in. 102 openhole tools still have a vertical resolution 28 in. of 2 to 8 ft [0.6 to 2.4 m]. Several data pro- 22 in. 16 in. cessing techniques have been developed to 10 in. enhance the vertical resolution of these tradi- tional tools (next page, bottom). All methods use at least one high-resolution measurement to sharpen a low-resolution one and 101 require a strong correlation between the two. Resistivity, ohm-m Resistivity, An existing technique helpful in interpreting low-resistivity pay is Laminated Sand Analy- sis (LSA), a computer program for evaluating the shaliness, porosity and water saturation in beds as thin as 2 in. [4 cm].12 A newer approach for identifying and AIT Resistivities at CAT Well 90 in. evaluating thin beds is the SHARP Syner- 102 60 in. getic High-Resolution Analysis and Recon- 30 in. struction for Petrophysical Parameters soft- 20 in. ware. SHARP processing improves the 10 in. resolution of log inputs to the ELAN Elemen- tal Log Analysis module, thereby improving saturation and reserve estimates. Currently, 101 SHARP software exists as an interactive, Resistivity, ohm-m Resistivity, stand-alone prototype application for Sun workstations but a second generation ver- sion will be incorporated into the GeoFrame reservoir characterization system by the end 400 450 500 550 of 1995. Depth, ft

■ Comparison of ARC5 log with the AIT log at 2-ft vertical resolution. The logs were run 12. Allen DF: “Laminated Sand Analysis,” Transactions in the Customer Acceptance Test (CAT) Well in Houston, Texas, USA. of the SPWLA 25th Annual Logging Symposium, New Orleans, Louisiana, USA, June 10-13, 1984, paper XX.

12 Oilfield Review ARC5 Phase Shift Resistivity CCL Borehole Corrected -19 1.0 Inelastic Count Rate Far Detector 34 in. (P34H) Borehole Sigma Far Detector Count Rate c.u. cps cps 28 in. (P28H) 100 0 1500 0 1200 0 Near Detector 22 in. (P22H) Far Detector Background TDT Porosity Count Rate cps ROP5 cps 0.6 0 3000 0 16 in. (P16H) 500 ft/hr 0 Gamma Ray Gamma Ray ( ) Formation Sigma Depth, ft 10 in. P10H 0 GAPI 150 0.2 ohm-m 200 0 GAPI 100 60 c.u. 0

X800

X900

■ARC5 log run in wash down mode in front of thin gas stringers. Rough hole conditions precluded running wireline logs in the well except for a CBT Cement Bond Tool log and a TDT Thermal Decay Time log. The TDT log confirmed the presence of gas indicated by the ARC5 log.

Data Processing Methods for Enhancing Vertical Resolution

Technique Measurements Method Improvement in Resolution

Enhanced Phasor Phasor Induction log Medium-induction From 7 ft to about 3 ft Processing (1988) measurement used to enhance [2 to 1 m] deep induction measurement

Enhanced Resolution Litho-Density log Near-detector measurement From 18 in. to 4 in. Processing (1986) used to compensate for far [45 cm to 10 cm] detector measurement

CNL Compensated Near-detector measurement From 24 in. to 12 in. Neutron Log used to compensate for far [61 cm to 30 cm] detector measurement

Laminated Sand Triple combo Computes bound water saturation Down to 2 in. [5 cm] Analysis (1984) (gamma ray, neutron (shaliness) from EPT tool, used and density), EPT with dual-water model to Electromagnetic redistribute the measured induction Propagation Tool resistivity, yielding estimates of the resistivity of thin beds. Effective porosity, water saturation and permeability are computed.

Autumn 1995 13 ■Establishing bed SHARP analysis relies on high-resolution Average Resistivity boundaries and inputs, such as Formation MicroScanner, Formation MicroScanner Images Square Average Resistivity modes (right) with FMI or EPT Electromagnetic Propagation 1.0 ohm-m 100 the SHARP program Tool logs to define a layered model of the and a Formation formation (right). The program looks at the MicroScanner log (left). SHARP analy- zero crossings on the second derivative of sis determines bed the high-resolution log, where the slope boundaries from changes sign, to indicate bed boundaries. In inflection points the case of a Formation MicroScanner or XX20 on the second FMI log, the SHARP program examines the derivative of an average Formation second derivative of an average resistivity MicroScanner reading from all button sensors. resistivity reading. With bed boundaries established, SHARP Modes are estab- analysis plots a histogram of the frequency lished by grouping resistivity measure- of a particular resistivity value within the ments on a his- logged interval of interest. By studying how togram (not shown). resistivity values cluster, an interpreter can XX40 A square represen- group the values into different populations, tation of the aver- or modes. SHARP analysis assumes that all age Formation MicroScanner resis- resistivity data in a particular mode come tivity curve shows from the same kind of formation, and further the bed boundaries that the resistivity value in a particular mode and modes. is constant. In addition, SHARP evaluation assumes that petrophysical parameters such XX60 13. Ramamoorthy R, Flaum C and Coll C: “Geologically 1 Consistent Resolution Enhancement of Standard Petrophysical Analysis Using Image Log Data, 2 paper SPE 30607, to be presented at the 70th SPE Annual Technical Conference and Exhibition, 4 6 Dallas, Texas, USA, October 22-25, 1995. 14. Chapman S, Colson JL, Everett B, Flaum C, Herron 3 5 M, Hertzog RC, La Vigne J, Pirie G, Quirein J, Schweitzer JS, Scott H and Wendlandt R: “The Emergence of Geochemical Well Logging,” The XX80 Technical Review 35, no. 2 (April 1987): 27-35. Mode number

■Enhancing the reso- Formation lution of deep lat- MicroScanner Refined erolog measurements Average Resistivity Model Reconstructed LLD Resistivity Model Reconstructed LLD (LLD) of the DLL Dual Resistivity Laterolog Resistivity tool. The Formation Depth, ft Square MicroScanner aver- Average Original LLD Original LLD Original LLD Original LLD Resistivity age resistivity curve ohm-m ohm-m ohm-m ohm-m ohm-m is shown with its 1.0 100 1.0 100 1.0 100 1.0 100 1.0 100 XX40 SHARP-generated average resistivity square log (far left). The bed boundaries and number of modes established by SHARP is used to generate an XX60 enhanced-resolution LLD curve. The modeled LLD measurement of the DLL tool is refined by compar- ing the original DLL log with the recon- XX80 structed LLD log (middle) and interactively adjusting the bed boundaries and mode values to achieve a better match (right).

X100

14 Oilfield Review as density, neutron porosity and sonic veloc- Bed Boundaries Enhanced Density ity are also constant in a given mode. After establishing the number of beds and Original Density Original AIT 10-in. 1.65 g/cm3 2.65

modes in the logged interval—the “square Depth, ft Enhanced log”—the SHARP program calculates a set Gamma Ray Enhanced Neutron Enhanced AIT 60-in. Porosity of mode values that minimizes the differ- DCAL Formation in. Original -1 4 Gamma Ray MicroScanner Images Original Neutron ence between the original and square logs. Original AIT 60-in. Porosity GAPI This model, a squared resistivity log of bed 1:120 ft 0 150 Pad 1 Pad 2 Pad 3 Pad 4 0.2 ohm-m 200 60 p.u. 0 boundaries and mode values, is filtered with XX900 the response function of a logging tool to produce a synthetic, or so-called reconstructed, log (previous page, bottom). The model is refined by minimizing the difference between the measured log and the reconstructed log. At a workstation screen, the log interpreter can interactively adjust the boundaries and bed values of the modes to achieve a match. When the synthetic and measured logs match, the model can be used as a high-resolution input into the ELAN interpretation. To sharpen the resolution of other logs, such as the gamma ray, the model of bed boundaries determined previously is utilized to reconstruct other squared, enhanced logs for high-resolution formation evaluation. A low- resistivity example from the Gulf of Mexico shows how SHARP analysis improved reserve estimates by 28%, even when applied to AIT measurements and a high-resolution triple combo of density, neutron and gamma ray logs (right and next page). Rather than reconstruct logs using SHARP analysis, Raghu Ramamoorthy and Charles Flaum, of Schlumberger-Doll Research, Ridgefield, Connecticut, USA have developed a simpler technique to enhance pro- ducibility and hydrocarbon content estimates made with conventional petrophysical analyses in thin beds.13 Working with logs from the GLT Geochemical Logging Tool, they picked a high-resolution clay indicator, either the FMI or EPT log, and calibrated it to the clay volume derived from the GLT measurement. In addition to clay volume, the GLT tool combines nuclear spectrometry logging measurements to determine mineral X1000 concentrations and cation exchange capacity of the formation.14

■Comparison of original and SHARP- enhanced logs for a low-resistivity pay example from the Gulf of Mexico. The interval was logged with a high-resolution triple combo. The original 10-in. and 60-in. depth-of-investigation curves from the AIT log are shown with the enhanced AIT 60-in. log. Only the 60-in. curve was enhanced because the SHARP prototype software does not yet have the modeled tool response for other AIT measurements. An ELAN interpretation (next page) using 10.00 1.66 0.20 the enhanced logs shows a 28% increase Resistivity, ohm-m in estimated reserves.

Autumn 1995 15 ELAN with High-Res Inputs ELAN with Enhanced Inputs respond to clays, the GLT-derived clay volume—a low-resolution measurement—is Irreducible Water Irreducible Water enhanced by looking at local variations of Moved Water Moved Water the high-resolution FMI measurement. The Water Water low-resolution GLT clay volume is adjusted by the difference between the FMI-derived Hydrocarbons Hydrocarbons clay volume and its value averaged over the Bound Water Bound Water resolution of the GLT tool, which is 2 ft: Depth, ft Quartz Quartz V clay, high-res= V clay, low-resGLT+ Montmorillonite Montmorillonite

Gamma Kaolinite Kaolinite [V clay, high-res FMI- < V clay, high-res FMI>]. Ray Illite Illite GAPI Sw With well data, an empirical relationship 0 150 Volume Scale p.u. Volume Scale p.u. is established between clay volume and 1:120 ft 0 p.u. 100 Image 100 0 0 100 porosity. This relationship is applied to the XX900 enhanced GLT clay volumes to derive high- resolution porosity values. Enhanced GLT clay volumes and porosity values are then processed with calibrated FMI resistivity values to boost the resolution of hydrocarbon saturation estimates. Applying this technique to GLT and FMI logs from a well in Lake Maracaibo, Venezuela reveals overlooked reserves. The FMI image shows the highly laminated nature of the formation, with beds on the order of 1 ft. A comparison of standard-resolution and high-resolution ELAN interpreta- tions shows that potential pay zones have been completely masked in the conventional processing (next page, top).

Using Electrical Anisotropy to Find Thin-Bed Pay James Klein and Paul Martin of ARCO Exploration and Production Technology in Plano, Texas, and David Allen of Schlum- berger Wireline & Testing in Sugar Land, Texas are modeling electrical anisotropy to detect low-resistivity, low-contrast pay such as thin beds.15 The researchers found that a water-wet formation with large variability in grain size is highly anisotropic in the oil leg and isotropic in the water leg. They attribute the resistivity anisotropy to grain-size variations, which affect irreducible water saturation, between the laminations. They tested their theory by modeling the thin, interbedded sandstones, siltstones and mudstones of the Kuparuk River formation A-sands of Alaska’s North Slope, located 10 miles [16 km] west of Prudhoe Bay. The model, based on a Formation MicroScanner X1000 interpretation, contains layers of low-permeability mudstone and layers of permeable sandstone with variable clay content. The simulated resistivity data are described as either perpendicular—measured with current flowing perpendicular to the bedding—or parallel—measured with current flowing parallel to the bedding.

16 Oilfield Review Water Plotting perpendicular versus parallel resis- Bound Water Hydrocarbon tivity for a given interval shows how hydrocarbon saturation influences electric Formation MicroScanner Siderite Calcite Images anisotropy (below left). Simulated resistivity Orthoclase Quartz data in the oil column curve to the right, 0 180 360 but simulated resistivity data in the water Pyrite Rutile leg are nearly linear. The position of data

Depth, ft Montmorillonite Muscovite along the oil column arc indicates the lithology of the formation. Illite Kaolinite Today, this technique works only with

p.u. p.u. 2-MHz MWD tools such as the CDR Com- 0 100 0 100 pensated Dual Resistivity tool. The CDR phase and attenuation measurements provide a unique response to anisotropy that allows the perpendicular and parallel resis- X80 tivities to be determined. The technique requires that the logging tool be parallel to the beds so that differences in the phase and attenuation of resistivity measurements can be used to establish anisotropy. Although the technique cannot yet be applied at other angles, its originators believe some opera- tors will value it enough to tailor the devia- tion of their wells so that logging tools can run parallel to beds of interest.

Nuclear Magnetic Resonance Logging X85 Although thin-bed evaluation is challenging, the tools and techniques described so far provide answers in most cases. More trou- blesome to interpreters than thin beds is another prominent cause of low-resistivity pay, reduced grain size, which contributes to high irreducible water saturations. The CMR Combinable Magnetic Resonance tool shows potential for measuring irreducible water saturation and pore size.16 ■Using a high-resolution measurement to enhance a low-resolution one. The clay volumes derived from a GLT log of well in Lake Maracaibo, Venezuela were enhanced The CMR tool looks at the behavior of with an FMI log from the same well. The enhanced ELAN interpretation (track 3) fea- hydrogen nuclei—protons—in the presence tures several pay zones that were missed on the standard ELAN interpretation (track 2). of a static magnetic field and a pulsed radio

Oil Column Water Leg 100 50% sand: Increasing 15. Anisotropy is the variation of a property with 50% shale oil saturation direction. In this case, it is variation of resistivity in the vertical (perpendicular) versus horizontal (parallel) planes. For a review of electrical anisotropy: 10 Anderson B, Bryant I, Helbig K, Lüling M and Spies B: “Oilfield Anisotropy: Its Origins and Electri- cal Characteristics,” Oilfield Review 6, no. 4 100% sand (January 1995): 48-56. Allen DF, Klein JD and Martin PR: “The Petrophysics of Electrically Anisotropic Reservoirs,” Transactions of 1 the SPWLA 36th Annual Logging Symposium, Paris, Perpendicular resistivity France, June 26-29, 1995. 16. Morriss CE, MacInnis J, Freedman R, Smaardyk J, 100% shale Straley C, Kenyon WE, Vinegar HJ and Tutunjian PN: “Field Test of an Experimental Pulsed Nuclear Mag- 100% netism Tool,” Transactions of the SPWLA 34th Annual shale 100% sand Logging Symposium, Calgary, Alberta, Canada, June 0.1 13-16, 1993, paper GGG. 0.1 1 10 100 0.1 1 10 100 Chang D, Vinegar H, Morriss C and Straley C: Parallel resistivity, ohm-m Parallel resistivity, ohm-m “Effective Porosity, Producible Fluid and Permeability in Carbonates from NMR Logging,” Transactions of ■Effect of saturation on electrical anisotropy in the Kuparuk River formation, Alaska, the SPWLA 35th Annual Logging Symposium, Tulsa, USA. Resistivity data taken in the oil column curve to the right, but resistivity data Oklahoma, USA, June 19-22, 1994, paper A. taken in the water leg are nearly linear. The position of data along the oil column arc indicates the lithology of the formation. Strong anisotropy may be present in the oil column, depending on the saturation in the more resistive component. In the water leg, the same formation might display little or no anisotropy. 17 frequency (RF) signal (right). A proton’s size distributions. The area under a spec- magnetic moment tends to align with the trum of T2 times is called CMR porosity. static field. Over time, the magnetic field Unlike previous NMR tools, the CMR gives rise to a net magnetization—more pro- tool is a pad-mounted device. Permanent Static tons aligned in the direction of the applied magnets in the tool provide a static mag- B magnetic field than in any other direction. netic field focused into the formation. The field Applying an RF pulse of the right fre- CMR tool’s depth of investigation, about 1 in. quency, amplitude and duration can rotate [2.5 cm], avoids most effects from mudcake Net magnetization the net magnetization 90° from the static field or rugosity. Its vertical resolution of 6 in. direction. When the RF pulse is removed, the [15 cm] allows for comparison with high- protons precess in the static magnetic field, resolution logs. emitting a radio signal until they return to A low-resistivity example from the their original state. Because the signal Delaware formation in West Texas shows strength increases with the number of mobile how the NMR response allows log inter- protons, which increases with fluid content, preters to measure residual oil saturation B the signal strength is proportional to the fluid directly from the CMR log (below). NMR Net content of the rock. How quickly the signal measurements on core samples from the Radio frequency decays—the relaxation time—gives informa- Delaware formation show that the NMR pulse magnetization tion about pore sizes and, to some extent, response will decay within the first 200 mil- the amount and type of oil. liseconds (msec) if the pores are filled with A CMR log displays distributions of relax- water. If the pores are filled with oil, how-

ation, or T2 times, which correspond to pore ever, the signal decays after about 400 msec.

210-msec CMR Oil Show oil/water line T2 Log Laterolog Deep 0.08 p.u. -0.02 B .003 sec 1 CMR Free Fluid Bound Fluid Volume Radio Caliper from 30 p.u. -10 33-msec line signal Litho-Density tool Laterolog Shallow sec CMR Porosity 6 in. 16 30 p.u. -10 Gamma Ray MicroSFL Log T2 Distribution Depth, ft Mud Log Show 0 GAPI 200 0.1 ohm-m 1000 0 gas units 1000 0.001 1.50 ■Principle behind the CMR tool. Permanent magnets in the CMR tool create a static X250 magnetic field B that gives rise to a net magnetization among hydrogen nuclei (top). A pulsed radio frequency signal rotates the net magnetization 90° away from the static magnetic field (middle). After the RF pulse is removed, the protons precess back to their original state, emitting a radio signal whose strength is proportional to the fluid content of the rock (bottom).

The T2 distributions in track 4 have been divided into three parts. The area under the X300 T2 curve to the left of the first cutoff, shown as a blue line at 33 msec, represents irreducible water saturation. The area under the curve from 33 msec to 210 msec (red line) represents producible fluid. Above 210 msec, the area under the curve represents oil, presented as a CMR oil show in track 3. This measurement of oil actually refers to residual oil saturation since the CMR tool

X350 looks only at the flushed zone. With the introduction of the CMR tool, ■Early field test of the CMR tool in the Delaware formation, West Texas, USA. Based on log interpreters are gaining the upper hand NMR measurements of core samples from the Delaware formation, the T2 distributions in the struggle to assess low-resistivity pay. in track 4 have been divided into three parts. The area under the T2 curves to the left of Although there are no easy answers when the 33-msec cutoff is irreducible water saturation. From 33 msec to 210 msec, the area evaluating low-resistivity pay, the tools and under the curve represents producible fluid. Above 210 msec, the area under the curve represents oil. In track 3, the mud log show curve was derived from the total gas mea- interpretation techniques are in place to sured on the mud log. It indicates that there is an oil-water contact halfway through more efficiently find these frequently the interval, at about X320 ft. bypassed zones. —TAL

18 Oilfield Review