‘THE GEOLOGICAL INTERPRETATION OF WELL LOGS MH Rider Petroleum Exploration Consultant Halsted Press, a Division of John Wiley and Sons ‘New YorkPreface This book describes, discusses and illustrates with examples the interpretation of geological data from logs obtained in wells drilled for hydrocarbon explo- ration. Itis written for those who use logs or require an understanding of the basics of logs. Traditionally, the analysis of well logs has been left to the petrophysicist, whose brief is to quantify re- servoir parameters and, in particular, the amount of indicated hydrocarbons. The petrophysicist may state that a reservoir bas 20% porosity and a water satu- ration of 35% However, the logs can also tell us that the sandstone is in a fining upwards sequence and was probably deposited in a finvial channel associated with a cycle of deltaic progradation. It is the latter aspects, rather than quantitative petrophysics, that are con. sidered in this book. ‘The subject is dealt with logically and concisely. The carly chapters consider each of the principal open hole logging tools. Measurements made by each tool are discussed in terms of simple theory and acquisition of data, The geological interpretation is then «scribed, using many field examples, logs being accoripss ved by a relevant interpretation wherever possible. The ater chapters bring together the log/data from earlier pi of the book to form single data sets. This complete interpretation is shown to provide geological infor- mation which is as detailed as that at outcro} I wish to acknowledge the stimulation provided by my discussions with colleagues of the Compagnie Francaise des Pétroles over the years, Early inspiration came from the work of Elf-Aquitaine (SNEAP), and especially Oberto Serra, However, I take full res- ponsibility for what iswritten. Credit must be given to the reviewers of the MS, who made valuable sugges- tions which Jed to notable improvements. MHRAcknowledgements Permission to cepraduce the following material i gratefully acknowledged, Figures 22nd 5.12: modied and redramn from Slain. Ptson, Handbook of Weil Log “Analysis: For Oil and Gas Formation Evaluation, © 1963, pp.39, 72, reprinted by permission of rent Hall, Englewood Cie, Ni fits 26 modified nd terawa om 17, Dewan, srt of lrn Open Hole Log Interpretation, 1983, p11, by pes OT sees rk gir 12 8:modifed nd ralawa om FP ol Sand and Sandstone, 1972, Figue 5.1, by permission of Springer Verly Meidelbere. He Se by pemiion Springer Was rs 89, 820: modifed and redrawn after RE Shei Slamic Swatgraphy, 1950, Figure 6.1, . 118, and Figure 612, p-132, bY permission of IHRDC Press, Boston, Mass. Figure 125: modified and redrawn from TJ. Burnie al, in Petroleum Geology of the Continental Shelf of NW Europe, 1381, Figure, p.252, by peruission of The insite of Petroleum, London. Figures 44, 7.13 and 9.4: modified and redrawn from Dresser Alles Wireline Services Cotclogue, 26, 2d Dresee Alls, ell Logging and Interpretation Techriques, 1982, Figures 8 and 10.1, p.6 and 89, by permission of Dresser UK Lid, London. pbs UAL: reproduced fom Drees Ais Well Logging and Interpretation Technigues 1982, Figure 281, p. 235 by permission of Interpretation Techrqes 198, igure 28,0, 25,0) Ps Figure 42: redrawn after C. Comnford, in Introduction 19 the Petroleum Geology of the North Sea (ed. K.W. Gleanieh, 1984, Figure 99, p.179,by permission of Blackwell Scientific Publications Lid, Oxford Figures 48, 622, 627,628,633, 635, 814,816, 9.12.919,920, 925,105, 126.136, 149 and 144: modified and redrava alter Landes (i967, Keith and’ Pitman (983), MeCeessan’ (960, Schade (1973) Meyer and Nederlot (1984 Sarg and Skjold (1982) Burst (1969), Magara (1968), Gilat (1966), Schmoker and Hester (1982, Asche (1950, Seley (1978), Rawson and Riley (1982), Van Sis 0950 and Cabesin 84). n Ball Ase, Pa eo. P-839; 676) 9.1398; 4544) pp.d53, 464 512) pp.373-374% 80) 108 AAPG Meno Sh pos Bul mAs Per Geol ‘SX 9.82; 5212) pp.2474, 2483: AAPG Memoir Bp. 37; Bull Are ‘Assoc. Pir. Geo. 612) p.2170, 343) p.959, 602) p.186, 612) 1.2637; 4248) p. 1899, 688) p. 1092, by permision of the American ‘Asoriation of Petroleum Geologists, alsa, Oklahoma, Fpure 722: reproduced from Scherer (1980), AAPG Memoir 30, p46, by permission of the American Assocation of Petroleum Geologists, Tulsa, Oklaboma. Figure G7: mediied-and redrawn afer M.RJ. Wyle, The Pindamentals of Wel Log tnterpretrion, 38 edn, 1963, Figure 4 p17, by permission of Academie Pres Ine. Figures 510, 67 and. 725 modifed and redrawa after M Gonduinet al, Trans AIME 210, 1957, 8-72, 1957 SPE-AIME, Figuce 3, p.60, C- Cavier et al, in 1977 Amul Techical Conference and Exhibuion, Deaver, Oct. 9-12, 1971, paper SPE. 6259, © 1977 SPE-AIME, Figure p.7.JA, Quirein eal, in 1982 Annual Tecbricel Cosferesce and Exhibition, New Orleans, Sep 36281982, paper SPE 1143, (0 1982SPE-AIME, Figute 2.0.2; by permision ofthe Society of Petroleum Engineers, Dallas, Teas. Figure 1120 reproduced from A. Poupon etal, J. Per. Tech, ‘ug: 1971 995-1005, paper SPE 2025, 61971 SPE-ATME, Figure, iby permission ofthe Society of Peroizum Engineers, Dalia, Texas Figures 14,814, 615,617 1010nd 1026 modied and redrawn from Log Interpretation, Vol. 11-Aplicatins, The Essentials of Log Interpretation Practice and Log Iterpretation Chars. © 19,1972, 1979, Schlumberger Inland Series Ine Figures 1.1, 819 (om Allaud, Land Martin. M. Schanberge Iistoire dun tecnigue, 1976, pp. 12,315, by permistion of Berger: Levratlt SA, ars Figures 443,922,924, 1029, 11.10. (rom Evans (1977, Exmyre (1981, Laver ant Seis (1977), Ransom (1977) and Sanyal eal. {0980 in Log Analyst XVII (2) pp. 7.1; XI (6) p. & XVI (2) PLIOXVITUC) p51, XXI UD) pe Figures 35, 45,611, 74,88 and Ho.1,36,1034 1023 11.11, 112115, 11.16and 1.19, 11.26 123: from Cox (1983, Benoit ¢¢ el (1980, ‘Thveadgold (197), Hassan tf a, (1979) MeFadzea (1973), Sherman and Locke (1975), Heslop (0979, Nelson and Glenn (1975), Kowalebuk etal (974, Walters (1968), Burk ea. (196), Hestop (1974, Wolf tal (1982) and Serra nd Supic (1975, SPWLA Ana. Sp. Trans 24, pape, p18; 2I,paper AA. p40 12, paper W, p12, (7 paper Hyp. 10, paper Y, 15:16, paper Q, pp. 6 7 16 paper Fy. 8516, paper Mp. 15, Paper Hp. 168, paper Fp i, 10.3% 15, paper Me pp. 8.3; 23, paper FF, p. 2, 16, paper W. p. 8 Figures 725, 825-from Fert 0379) and Thomas 197) in SPHVLA Eur Symp. Trans. 6 paper Q. £25 5: paper 1 p-9, by pemisson of the Society of Profesional Well Log Analy Houston, Texas oarsContents 1 Introduction LI Well logs~a definition 1.2 Well logs—the necessity 13 Well loge —the making 14 Log runs 15 Log presentations 16 The logging companies 17 Welltog interpretation and uses U8 This book—aims and content 2 The logging enviroument 24 roduetion 22 The pressure environments of borehole logging and invasion 23 Temperature environment of borehole logging 2% Logging toot capabilites 255 Conelusion | 3 Caliper logs ~ 31 Mecnanical caliper—the tool 32 Log presentations 33 Interpretation and uses 34 Three-dimensional calipers 4 Temperature logging 41 Grotemperatures 42 Borehole temperature measurement 43 True formation temperatures 44 Significance of gecterperatures: interpretation 5 Sclf-potential or SP logs 51 Generates 52 Principles of measurement 53 Log characteristics 54 Quantitative uses 55 Qualitative uses 6 Reseviey and conductivity age | 61 Generates 62 Theoretical considerations 63 Zones of invasion and resistivicy 64 Resistivity tools 65 Induction tools 666 Log characteristics 67 Quantitative uses ofthe ress 68 Qualitative uses ty logs 7 The gamma ray and spectral gamma ray logs 74 Generates 72 Natural gamena radiation 23 Tools 74 Log characteristics 25 Geochemical behaviour of potassium, thor and uranium and natural 16 Radioactivity of shales and clays 79 Quantitative use of the siaple gamma ray log 78 Quaniiative se ofthe spectral gamma ray log 789 Qualitative we ofthe simple gamma ray log, 7.10 Qualitative use of the special gamma ray or 8 Sonic or acoustic logs” 1 Generalities 82 Principles of measurement 83 Tools 4 Log characteristics 85 Quantitative uses u n 4 16 16 7 2 2 23 26 a » 2 3s 2 a 37 38 39 6 65 3 B 0 18 2vil CONTENTS 86 Qualitative uses £87 Scismic applications of the sonic log 9 The density log / 2. Generates 92 Princiles of measurement 93 Took 94 Log characteristics 95 Quantitative aes 9 Quaitate usce 10 The neutron log 101 Generalities 102 Principles of measurement 103 Tools 10d Log characteristics 10.5 Quantitative uses 105 Qualitative uses 10.7 Neutcon-density combination: lithology identification IT Lithology reconstruction from logs ALE Introduction 112 Lithology from drill data—the mud log 113 Lithology from cores —direet physical sampling 114 Lithology interpretation from wireline logs manual method ALS Computer aids to lithology interpretation 116 Mult-iog quantification of lithology 12 Facies and depositional environments from logs ~ 121 Iatrodvetion 122 Facies 123 The sequential analysis of logs 13 Stratigraphy and logs * 131 Introduction 132 Lithostratigraphy 13 Some aspects of correlation 134 Stratigraphie breaks and stratigraphic sequesees 135 Conclusions 14 Concluding remarks 141 The geologist’ problem 142 The lithology problem 143 The dipmeter problem 1444 Not a conclusion—an approach 1001 for sedimentological interpretation References Index 8E 3 96 7 8 165 16s 165 165 163 mL L [ Key LITHOLOGICAL SYMBOLS i Shale (alt.) Organic shale silt med. | Sandstone, fo SEDIMENTOLOGICAL FORMAT cal ese tr Limestone Dolomite feeE] sare crater Anhydrite RRRARAY Gypsum Volcanic Carbonate Sandstones silt Shale Qrganic shale CoalPlate Wa) tes 1 (a) & (b). Displays from the Volumetric Sexn Tool (a) Polar display horizontal tice) fan open hole section. Arrow indieates horehe! Image rotation. This example shows casing damage sicwed from the four compass points, Simlar results may be obtained in epen hole, ciopraphs by kind permission of LA. Broding1010 180° 90° 998 1010" 270° Plate 10)1 Introduction 1.1 Well logs—a definition The continuous recording of a geophysical parameter along a borehole produces a geophysical well log. The value of the measurement is plotted continuously against depth in the well (Figure 1.1). For example, the value os 1015 wire 130 pe ohm mV 140 150 160 AL we 160 depth in well m 190 200 210 220 pA Ja 220 240 v ‘Figure 1-1 A well log: Representation ofthe rst ‘log’ made at Peche!- bronn, Alsice, France, in 1921 byH. Doll (From Allaud and Martin, 1976 resistivity log is a continuous plot of a formation’s resistivity {rom the bottom of the well to the top and may represent over 4 kilometres (2 miles) of readings. The most appropriate name for this continuous depth-related record is a wireline geophysical well log, conveniently shortened to well log or log. It has often been called an ‘electrical log’ because historically the first logs were electrical measurements of electrical properties. However, the measurements are no longer simply electrical, and modern methods of data trans- mission do not necessarily need a wire line so the name above is recommended. This book therefore concerns wireline geophysical well logs. Tn France, where well logging was first invented by Schlumberger and Doll, the: original name was “Carottage Flectrique’ (electrical coring) as opposed to mechanical coring. Today the name diagraphies dif- Jérées (literally, “deferred diagrams’) is applied to distinguish wireline geophysical well logs, which are made after drilling, from the drill logs (diagraphies immédiates, ie. immediate diagrams ) made during the drilling. In English no such distinction is made—the word ‘log’ is universally used, 1.2 Well logs—the necessity Many different modern wireline geophysical well logs exist. They ate records of sophisticated geophysical measurements along a borchole’ These may be measurements of spontancous phenomena, such as natural radioactivity (the gamma ray log), which re~ quires a tool consisting simply of a very sensitive radiation detector; or they may be induced, as with the formation velocity log (sonic log), in which a tool emits sound into the formation and measures the time taken for the sound to reach a receiver at a set distance along the tool (Table 1.1), Wireline geophysical well logging is necessary be- cause geological sampling during drilling (cuttings ‘Table 1.1 Classifiation'of wireline geophysical well measurements (in ‘open hole) Log type Mechanical Caliper ‘measurements Spontaneous Temperature measurements SP (sll potential) Gamma ry Indvoed Resistivity ‘measurements Induction Sonic Density Neutron Formation parameter measured Hole diameter Borehole temperature Spontancous electrical currents [Natural radioactivity Resistance to electrical current Conductivity of electrical current Velosity of sound propagation Resetion to gamma-ray bombardment Reaction to neutron bombardment —__“-. ——2 THE GEOLOGICAL INTERPRETATION OF WELL Locs sampling’) leaves a very imprecise record of the for- mations encountered. Entire formation samples can be Drought to the surface by mechanical coring, but this is both slow and expensive. The results of coring, of course, are unequivocal. Logging is precise, but ‘equivocal, in that it needs interpretation to bring a log to the level of geological or petrophysical experience, However, logs fill the gap between ‘cuttings’ and ‘cores’, and with experience, calibration and com. puters, they can almost replace cores, as they certainly ‘contain enough information (0 put outcrop reality into the subsurface, 1.3 Well logs—the making Wireline geophysical well logs are recorded when the drilling tools are no longer in the hole, “Open-hole logs*, the subject “of this book, are recorded im- mediately after drilling Logs are made using highly specialized equipment entirely separate from that used for drilling. Onshore, a motorized logging truck is used which brings its array of surface recorders, computers and a logging drum and cable to the drill site (Figure 1.2). Offshore, the same equipment is installed in a small cabin left permanently on the rig (Figure 1.2). Both truck and cabin use a variety of interchangeable logging tools, which are lowered into the well on the logging cable (Figure 1.2). Most modern logs are recorded digitally. The sampl- ing rate will normally be about once every 15cm (6 in}, although for some logs it will be as low as 3m (1.2 in). ‘An average well of say 2000 will therefore be sampled over 12.000 times for each individual log, and for a suite of 8 oF so typical logs, it will be sampled over 100000 times. This huge amount of data is stored in the computer of the surface unit. There is generally an instantancous display for quality control and a full print-out immediately the log is finished, but the raw data dre stored on magnetic tape for future processing and editing, To run logs, the hole is cleaned and stabilized and the drilling equipment extracted. The first logging tool is then attached to the logging cable (wireline) and lowered into the hole to its maximum drilled depth. Most logs are run while pulling the tool up from the bottom of the hole. The cable attached to the tool acts, both as a support for the tool and as a canal for da transmission. The outside consists of galvanized stee!, while the electrical conductors are insulated in the interior (Figure 1.3). The cable is wound around a motorized drum on to which it is guided manually during logging. The drum will pull the cable at speeds cof between 300 mh (1000 f1/h) and 1800 m/h (6000 fr/h), *Open-hole indicates thatthe formation forms the wal ofa well at ‘onpesal to ‘cased-hole'in which a tube ofmetal easing line the wel, Figure 1.2 Schematic diagram of 2 modera logging setup. The surface computer and electronic equipment are housed in a logging, truck (on aad) oresbin (offshore) Thelogging tool is winched up the hole by the logging cable which elso transmits the tool readings. The tvansnital is digital and recorded on magnetic tape. The surface ‘computer allows instant display. ie. 0.3 to 1.8km/h, depending on the tool used. As the cable is pulled in, so the depth of the working tool is checked, Logging cables have magnetic markers set at regular intervals (e.g. 100 or 50m) along their length and depths are checked mechanically, but apparen: depths must be corrected for cable tension and elasticity. Because rig time is expensive and holes must be logged immediately, modern logging tools are multi- function (Figure 14). They may be up to 18m (60ft) in Figure 1.3 Schematic diagram of a loggin cable, (Modified from Moran aod Auli, 1971)ans wae G33) bam ‘ Gant INTRODUCTION 3 {fay Mt Some tpi] modern combination logging cols Lengths areas marked diameters re many 3in (Mode om Schlumberger, is7ay length, but still have an overall diameter of only 3~4in, The Schlumberger ISF sonic tool, for example, of 3 in diameter, is 55.5ft (169m) long and gives a simul- taneous measurement of gamma ray or caliper, SP, deep resistivity (conductivity), shallow resistivity and sonic velocity. The complexity of such tools requires the use of the surface computer, not only to record but also to memorize and to depth-match the various readings. The gamma-ray sensor, for example, is not at the same depth as the resistivity sensors (Figure 1.4), so at any one instant, different formations are being sampled along the tool. The surface computer there- fore memorizes the readings, compensates for depth or time lag and gives a depth-matched output. Despite the use of the combined tools, the recording of a full set of logs still requires several different tool descents. While a quick, shallow logging job may only fake 3-4 hours, a deep-hole, full set may take 2-3 days, each tool taking perhaps 4-5 hours to complete. 1.4 Log runs When a log is made it is said to be ‘rum’. A log run is ‘typically made at the end of each drilling phase, ie. at the end of the drilling and before casing is put in the hole (Figure 1.5}. Each specific log run is numbered, being counted from the first time that the particular log is recorded. Run 2 of the ISF Sonic, for example, may ‘cover the same depth interval as a Formation Density Log Rum 1. In this case it means that over the first interval of the ISF Sonic, (ie. Run 1), there was no Formation Density log recorded (Figure 1.5). ‘Typically, through any well, more logs are run over intervals containing reservoirs or with shows, than ‘over apparently uninteresting zones. The choice of logs depends on what itis hoped to find. Logging costing 5- 10% of total well costs is expensive, so that in cheap, onshore wells, in known terrain, a minimum set is run Offshore, where everything is expensive, full sets of logs are generally run, even if hydrocarbons are not found, as each well represents hard-gained information, Cutting down on well logsis probably a false economy, but it can be forgiven when prices are considered 1.5 Log presentations A standard API (American Petroleum Institute) log, format exists (Figure 1.6) The overall log width is &25 in QLem), with three tacks of 2.5 in (64cm), tracks 1 and 2 being separated by a column of 0.75 in (1.9 em) in4 THE GEOLOGICAL INTERPRETATION OF WELL Loas Figure 1.5 Logging record. Log nuns are indicated on atypical offshore drilling eurve. Horizontal lines indicate no drilling, when logs are rin. Casing follows logging Nove log run numbers. (Tool sytabols—Schlumberger) trnck tS tc 2 twee ° vo 5 "0 15 So 3 1 . TLL 3 La 3 im rT TT Ll ° 5 og 10 “00 2000 2 3 2 . E 8 I ° s wo og 10 205 5 10 / Lt ak LU Lit we 14 Thretypcei API op orate Tricks 625i witha cena O7Sin depth coh, Crea widths 825in, Horizontal ales safe worable face ten)which the depths are printed. There are various combinations of grid. Track 1 is always linear, with ten stindard divisions of 0.25 in (0,64em). Tracks 2 and 3 may have a 4-cycle logarithmic scale, a linear scale of 20 standard divisions, or a hybrid of logarithmic scale in track 2 and linear scale in track 3 (Figure 1.6) ‘These are the classic presentations which, in the past, usually prevailed, With the advent of digitized logs, non-standard formats are becoming more common, especially on computer playbacks. On the old analog logging systems, the choice of vertical or depth scales was limited to two of 1:1000, 1:500, 1:200, 1:100, 1:40 and 1:20, From these, the most frequent scale combinations were 1:500 (lem =5m) for résumé or correlation logs and 1:200 (I om = 2m) for detailed reservoir presentation, ‘The American area was an exception, where the available scales were 1:1200, 1:600, 1:240 and 1:48. From these the commonly-chosen scales were 1:600 (1in = 100feet) for résumé and correlation logs, and 1:240 (Sin = 100feet) for detail. These scales still dominate industrial documents, buts. result of modern computer storage other scales are becoming more common. Especially useful to the” geologist are the reduced scales of 1:2000 (om myrRopucTION 5 =20m) and 1:5000 (Iem= 50m). In fact any con- venient scale can now be produced easily by the ‘computer, whereas in the past scale changescould only be made by unsatisfactory photographic methods, eo wep 8 ‘epgig t0e0d = YOm/nia Figure 1.7 Dashed log margin repreceating mivute intervals, The logging speed ean be checked from these dastes ‘Table 12 Principal uses of open-hole wireline logs > i <3 Elsl2lg g slealt 2 ESE l2 |S 1EIS I/F IE! aS == 4 | Temperature: ~ + - s|sp = ale clele 7 | Gamma ray -l- +f-|-[- + Ty lena SEF EEE : 9| Density ed we wl ow mala a [x 5 | {- . 10 | Neutron 1s el aa . _ , ie (Essentially) qualitative wse +Semi-quamtitative uses “Strictly quantitative6 THE GEOLOGICAL INTERPRETATION OF WELL LOGS TIT Aq pepuoeu| OSs BLE" ME MILI jalale) lalalo} al SCH Pee MEPTVE «PZ gh anieg nyo] S «02°90 +92 o0 XBINTOO) “aNTL098 wie Wo Wstiices — ANYeWOO oes aI EYE LATTE ona aM ‘i fee ort | AUESSEE TTR TTT |) » Lye) pean Tedtion Sood zx Ey 5 2 | ap fa 7 LOGGING DATA Gre fan fo Fe am Figure 18 A typical lop headingOne final aspect of the log grid to note is the dashed outside border on field logs (Figure 1.7), Each dash Fepresents one minute, regardless of log scale. The presentation allows a direct control of logging speed and, indirectly, log, quality. Every log grid is preceded by a comprehensive log heading. It covers all aspects which allow the proper interpretation of the log and, in addition, identification of the well, rig, logger and logging unit. Thelog heading illustrated (Figure 1.8) is but one example, each com- pany having its own format, On the log tail is found a repetition of some of the log-head data, simply for convenience, Calibration data is also added to the log tail, as are short, doubled- up or repeat sections which act as samples for empirical quality control. 1.6 The logging companies ‘The well-logging world is dominated by one, extreme- ly successful, giant international company— Schlumberger. In America a number of other compan- ies exist but in many parts of the world Schlumberger has. quasi-monopoly.The reasons for this domination are partly historical: it was Frétes Schlumberger-who created the original SPE (Société de Prospection Blectrique) in 1926, the precursor of the modern Schlumberger. The brothers, along with H. G. Doll, were the creators of the well-logging technique. ‘The international forum is becoming slightly more competitive, and in America smaller companies are active. However, three names stand out in the general logging field apart from Schlumberger: Gearhart, Dresser Atlas, and Welex. 1.7 Well-log interpretation and uses The accepted user of the well log is the petrophysicist. His interest is strictly quantitative. From the logs, a petrophysicist will calculate porosity, water satu- ration, moveable hydrocarbons, hydrocarbon density and 50 on, all the factors related to quantifying the INTRODUCTION 7 amount of hydrocarbons in a reservoir for estimates of eserves. The Society of Professional Well Log Analysis (SPWLA), the principal society of log in- terpreters, is mainly composed of petrophysicists Reservoir rocks, however, comprise perhaps only 15% of a typical well, and of this 15% only a small percentage actually contains hydrocarbons. The petro- physicist is therefore not interested in 85% or more of the well logs recorded. The exploration geologist, in contrast, should be interested in 100% of well logs, as, the amount of geological information they contain is enormous. ‘The geophysical measurements made during logging are sensitive, ‘accurate and characteristic of the for- mation logged. However, to those familiar with the aspect of rocks as seen at outcrop, the geophysical signatures of this selfsame rock in the subsurface arc impossible to imagine, To an experienced geological analyst of well logs, the reverseis true, A formation that hecan instantly identify on the logs, even to the nearest metre, he is hard put to find, even tentatively, at couterop. In the following pages it is intended to relate the ‘outcrop more closely to the wireline, geophysical well log. Logs can and should be interpreted in terms meaningful at outcrop. They contain as much infor- mation as does an outcrop, but can be stud conveniently at the desk. 1.8 This book—aims and content Table 1.2 shows the logs considered in this book, and their principal applications, which have been divided into qualitative, semi-quantitative and strictly quanti- (ative, although theres an increasing tendency to treat modern logs as simply representing a set of analytical values. A sample set of over 100 000 values for a well of 2000 m represents an enormous quantitative database. Semi-statistical, quantitative or quasi-quantitative me- thods applied to this database can bring precision to interpretation, as applied to geological problems,2 The logging environment 2.1 Introduction ‘Treated simply as an instrument of measurement, a logging tool is required to do two things: to give a true, repeatable reading, and to make the reading of a representative, undisturbed sample of the subsurface formation. For the following reasons, neither of these ideals can be realized. The first is that the undisturbed formation environ- ment is irrevocably disturbed by drilling a well. The new drill-created conditions are those in which the logging tools work. A tool can only ‘guess’ at the original states. This chapter examines what is involved in this guess, in terms of drilling pressure, drilling temperature and invasion. ‘The second reason is that the ideal conditions for a perfect geophysical measurement cannot be met in borehole logging methods. Ideal conditions would require 2 logging tool to be motionless for each individual measurement, and to have a sensor of zero dimensions measuring a point sample. Sensors have dimensions and tools move. Tool design acknowledges this, and a compromise is made between a practical and practicable measurement and one that is perfect. This chapter will also examine, in general terms, the effects of the logging method on the measurements made. The notions of depths of investigation, mini- mum bed resolution and bed-boundary definition will be discussed. 2.2 The pressure environments of borehole logging and oe The pressure environment during drilling and, in- evitably, during logging, is made up of an interplay between two elements; formation pressure and drilling-mud column pressure, The formation pressure is the pressure under which the subsurface formation fluids and gases are confined. The pressure of the drilling mud is hy- Grostatic and depends only on the depth of a well, that is the height of the mud columa, and the mud density. Maintaining the pressure exerted by the column of drilling mud at just a little above the pressure of the subsurface formations encountered is one of the noc- essities for equilibrium drilling: itis a delicate balance. The two pressure environments arc examined below. Hydrostatic pressure Fluids transmit pressure perfectly so that the pressure exerted by the column of fluid is dependent simply on the height of the fluid column and the density of the uid. The pressure in kg in a column of water can be calculated thus height of water column (m) x density (g/cm?) hea ofaer ohm Gee = pressure (kg) per sq: em Q) For a column of pure water of 2500 m (density of pure water = 1,00 g/cm?) 2500 x 1 10 250 kejem? @ In oilfield terms, the pressure of a column of fluid may be expressed by its pressure gradient. Thus pure water has a gradient of 1.00 g/cm?. That is, a column of pure ‘water will show a pressure increase of 1 kg/cm? per 10m of column (or 1 g/m? per em of column) (Figure 2.1). The term ‘column of water’ is used as applicable to wells: ‘depth’ is equally applicable and more understandable when talking about water ‘masses, stich as the oceans. ‘As water becomes more saline, its density increases (Figure 2.2). Water which has. salinity of 140000 ppm (parts per million) of solids (mainly NaC), has a density of 1,09 g/em? (at 15.5°C). A column of water of this salinity will have a gradient of 1.09 g/em? and at most olliiels orinss depth, ken (tiuld column hi 0 200 «400 «800-800 pressure, kolom? Figure 2.1 Fluid pressure gradients related to depth, or height of Med coluran.1.0 © 50100 200 250 x 109 salinity (otal solids, ppm Nach) 150 Figure22 Graph showing the increase in_water density. with inerease in salinity (NaCih. (From Piro, 1963) 2500m will exert a pressure of 2500 x 1.09 10 = 272.5kgfomn? 8) Figure 2.1 shows the various gradients for fluids of different densities and the increases with depth. All gradients are shown as Tinear. Formation pressures In most geological basins the pressure at which pore fluids are found increases from the ‘normal’ to. mo- derately overpressured. Normal pressure is defined as, hydrostatic pressure: it is due only to the weight of the fluid column above the formation. To calculate normal pressure it is sufficient to know only the depth of the formittion and the density ofthe fluids in the formation. fa formation water has the same salinity as sea water, then the Bressure at 1000 m in a formation with néritial pressure is the same as the pressure at the séa floor below’ 1000m of sea water. The graph (Figure 21) therefore shows normal pressure gradients for various salinities. Overpressure is simply defined as any pressure above the hydrostatic (or normal) for a particular depth. Thus, if the formation fluids ‘are salty with a density of 1.09g/em* and the measured formation pressure is 350kg/cm? at 250m, there is an over- pressure, calculated as follows. Normal pressure at 2500, fuid density 1.09 g/em?, from (3) 2500 x 1.09 7 272.5 kefem Measured pressure at 2500m = 350 kg/cm? Overpressure = 350 — 272.5 = 77.5 kg/em? Overpressure exists for a number of reasons, but in all cases it means that the formation fluids are being squeezed by the surrounding rocks. It is similar to the pressure regime in car brakes. When the brakes are at ‘THE LOGGING ENVIRONMENT 9 th, kn typical well profile ° 500 1000 jessure, kgfem? Figure 23 Formation ui pressure inereases with depth in atypical oiled well The pressure varies between the hydrostatic (Nuid) and the lithostatic (rock) gradients. rest, the brake fluid is at normal pressure. Putting the foot on the brake puts the fluid under overpressure: itis being squeezed by the extra pressure of the foot, Generally, most wells drilled show a typical subsur- face pressure development, Shallow formations show normal’ or hydrostatic formation pressures: there is no rock squeezing, no overpressure. Deeper into the subsurface slight overpressures arc encountered: so there is slight squeezing. As the depths increase, so the ‘overpressizre increases and the formation fluids sup- port more and more of the rock overburden pressure (Figure 23). Overpressixres: can inctease up to an empirica) ‘maximum called the lithostatic gredient. This graviicnt also: called the geostatic or overburden gradicnt, is taken as a convenient gradient representing the pro- bable maximum pressure likely to be encountered in a wellat any depth. The average gradient frequently used comes from the Gulf Coast of North America, and in American oilfield tnits is a gradient of 1 psi/ft (ie. in metric 23g/em*) and corresponds to an average rock density of 2.3 g/em? (Figure 23) (cf. Levorsen, 1967). ‘The true lithostatic gradient will in fact vary from well to well, and will depend on the densities of the formations encountered. In the example given (Figure 2.4), which is from a well in Germany, the average formation density is 2.4 g/cm? (Meyer-Giirr, 1976). The average well, therefore, encounters formation pressures somewhere between the normal hydrostatic10 THE GEOLOGICAL INTERPRETATION OF WELL LOGS rock density, g/em? 2.00 2.40 2.80 ' ° ‘Quaternary | tertiary |e — gradient =2.4g/em> 's TTiassle 2 2 00 1000 Ithostatle pressure, kglem? Figure 24 Trucrock density profile and average Bithostatic gradient from a North German well (Redrawa from Meyer-Gurr, 1976) gradient and the lithostatic gradient (Figure 2.3). In absolute terms this will give usual logging pressures of between about 150kg/cm? and 1000 kg/cm? (2000 psi 15000 psi). Most oilfield logeing tools are designed to withstand pressures up to a maximum of 1050- 1750kg/em? (15000-20000 psi), significantly above the highest pressure usually encountered Invasion—drilling pressures Under ideal conditions, the pressure exerted by the column of drilling mud will be such that when a porous formation is encountered, as the drill enters the | Figure 25 Schematic representation of dynamic filtration as a bit enters porous formation. Note the progressive mud-cake build-up. +1000 time (min atter penetration) 10 000 Figure 26 Graphic eprescatation (schematic) ofinvasionand mud cake buildup te 8 porous formation is penetrated (modified from Dewan. 1983, formation, mud will be forced into it (Figure 2.5). The porous rock will then begin to act as a filter, separating the mud into its liquid and solid constituents. The mud filtrate (the water used to mix the mud) will flow into the formation, while the solids (the mud) will form a deposit around the borehole wall once the bit has passed. In the hole just drilled, the solid deposit around the borehole wall, the mud cake, will gradually build up to form a skin over the porous interval, Initially, as the bit enters the porous formation there is complete disequilibrium and dynamic filtration takes place (Figure 2.6). That is, below and around the bit there is a continuous flow of filtrate into the formation, provided of course that the mud pressure is sufficient, Gradually, as the mud cake builds up, it creates a barrier and the movement of fluids dimin- ‘shes, until finally the mud cake becomes impermeable and filtration practically ceases (Figure 26). A cross- section through the borehole at this stage would show Figure 27) Invasion: simple representation ofthe eect of drilling on ids ina porous and permeable formation.Table 2.1 Depth of invasion (distance from borebole wall) e. pox sosity (approximate) from Miesch and Albright, 1967, Hole | T size (in) | 18 ak ! | Ratio Porosiy) Depth of invasion | invasion diameter % | | hoe diameter poe eer 1-8 |2000cm 1400em 90cm) 10 $20 | 9000 620m 430em| 5 2-30 | 25em 185m 100m] 2 Tox [Eitan sate Siam) @ mud in the hole, mud cake on the borehole wall and then the porous formation now filled almost entirely by mud filtrate, The original formation fluids have been pushed away from the hole (Figure 2.7). This is usually the situation when the open-hole well logs are run, The phenomenon of the replacement of formation fluids by drilling mud. filtrate is called invasion. Invasion affects porous and permeable formations in the immediate vicinity of a borehole. It is described by ‘depth’ or ‘diameter’ of invasion, that is the distance reached by the invading filtrate with respect to the borehole (Figure 2.7). In general, invasion is small in very porous and permneable formations, the mud cake building up rapidly to block dynamic filtration (Table 2.1). The contrary is the case in poorly perme- able zones, vugay carbonates or fractured formations, where mriud cake formation is slow and invasion may be very deep, up to several metres. Since excessive: invasion is the worst, situation for logeing and takes the real formation fluids too faraway from the borehole to be detected, chemicals.are added to the drilling mud to reduce water. loss creating a protective mud cake as quickly as possible. Products such as lignosulphonates and starch are used. Figure 28 Graph of geothermal gradients. The zone of typical oilfield gradients is indieated. THE LOGGING ENVIRONMENT 1] 23 Temperature environment of borehole logging Formation temperatures Normal sedimentary basins show a more or less regular increase in temperature with depth (Figure 28). The increase is not linear as frequently depicted; it varies according to lithology depending principally on the latter's thermal conductivity (see Figure 4.1), However, despite the irregularities there is an overall, persistent increase in temperature with depth (Figure 2.8). This increase is often expressed as a gradient, the geothermal gradient (ihe increase in temperature with depth). The metric values are usually °C per 100m or °C per km. Typical gradients for sedimentary basins are be- tween 20°C perkm and 35°C perkm (see Chapter 4 and Table 4.2). Temperatures in boreholes Just as the geopressure regime is disturbed by drilling, so is the subsurface temperature, A well drilled into a subsurface formation introduces relatively cold mud and mud filtrate into a hot formation. While drilling continues and mud is circulating, the formation is, cooled slightly and the mud heated. However, the mud remains undisturbed in the borehole when circulation ceases arid it gradually heats up to reach, or at least approach, the temperature of the surrounding for- mation. The two, however, are rarely in equilibrium. Logging temperatures taken in the mud are usually measured after only 5-10 hours of mud immobility: equilibrium is probably approached only after 5-10 days! (Temperature is considered at greater length in Chapter 4) > Typical borehole tools are generally designed to withstand temperatures up to around 200°C (400°F): this gives a guide to maxima expected during drilling. 24 Logging tool capabilities 1k was suggested earlier that logging tools should be able to sense the undisturbed formation and to make a true measurement of it. As indicated, the undisturbed formation environment is forced away from the bore- hole by drilling, to be replaced by the invaded zone, Logging tools are therefore designed either to “by-pass’ the invaded zone to reach the undisturbed formation, or to deliberately measure just the invaded zone itself. That is, they are designed with various capabilities of penetration, called the ‘depth of investigation’ (see below}. Inevitably, such demands on tool design create secondary effects. Logging is comparable to photo- graphy with its close-up lenses and long-distance lenses. Close-up logging tools give great resolution but little depth of investigation: long-distance logging12. THE GEOLOGICAL INTERPRETATION OF WELL LOGS Gistance trom borehole 100% [ on | 7 ) 100% but to total signal | SHALLOW INVESTIGATION TOOL DEEP INVESTIGATION TOOL 100% 50% ox ° 2 4 6 a distance from boraho! Figure 29 ustration ofthe notion of depth of investigition Two toolsareshown schematically slong witha graphic representation formation contribution to their overall signal E, emitter; R, receiver, tools give great depth of investigation but blurred resolution, Threc inter-related phenomena of logging and log- ging tools are examined below; depth of investigation, minimum bed resolution and bed boundary definition. Depth of investigation Most geophysical logs have an extremely shallow depth of investigation. By ‘depth of investigation’ we mean the distance away from the borehole to which the formation is having an effect on a tool reading. So- called ‘decp’ investigation is only a matter of 2-5m. away from the borehole and into the formation. The environment of logging tools is therefore from the borehole itself (shallow investigation) to a distance of, 5m from the borehole wall (deep investigation) in general, with tools that subject the formation toa Table2.2 Depth of investi- sation of the neutron tool (rao ified from Serra, 1979, alter Depth of Porosity Investigation cy (em) o 600 wo 340 0 230 0 165 90%, ofthe signal bombarding signal (Table 1.1), the depth of investi- gation of the tool depends on the separation distance between the emitter and receiver. For example, with the resistivity tools (Chapter 6), when the emitting and receiving electrodes are very close, the depth of in- vestigation is very small (Figure 2.9}. The Micro- Inverse Resistivity Tool, with electrodes 2.54em (1 in) apart, has such a shallow depth of investigation that it reads only the resistivity of the mud cake (when present), Conversely, the Induction Conductivity Tool, with emitter and receiver 1 m (40in) apart, has a depth of investigation which may reach about Sm. The Induction Tool is considered to be the most likely to ive the resistivity (in fact, conductivity) of the un- touched formation (R). The emitter-receiver separation is not the only factor affecting depth of investigation. Necessarily it varies with the character being measured. Thus for the sonic tools which measure the speed of sound waves in the formation, the waves take the quickest path from emitter to receiver: this is generally along the borehole wall (Chapter 8, For nuclear tools; the emitter— receiver separation is fixed as.a function of the average penetration of gamma rays, neutrons, ete, the field being more or less spherical around the emitter. These characteristics will be considered when each tool is described. Finally, depth of investigation also depends on the formation, whether it is susceptible to penetration or not. In thease of the neutron tools, for example, 2 non- porous bed is ‘seen’ to afar greater depth than a porous bed, due to variations in the absorbance of the signal (Table 22). In reality, depth of investigation is a very difficult term to fully understand. It is not precise; a bed is not investigated toa particular point and no further. It is a progressive character, like the radiant heat from a fire. We feel the heat near to the fire, but not at some distance away. Can we say exactly at what distance the fire has no more effect? With logging tools, the depth of investigation is ‘more realistically defined as the zone from which x% of the tool reading is derived (Figure 2.9). For instance, the neutron tool figures given above (Table 2.2) are defined on 90% of the tool signal. This is called theresistivity @ tow Ron @ ‘THE LOGGING ENVIRONMENT 13 -—— emitter-recoiver diatance short spacing value Ge, microlaterotog) ce 5, short epacing toot — = = tong spacing vatue Induction log) L, tong spacing toot Figure 2.10 The eect of minimum bed resolution oa logging-tool valucs in various scales of interbedding (1) Fine interbeds; (2) coarse interbedding; (3) single bed boundary (echematic). ‘Table 2.3. Minimum bed resolution of some common tools under best conditions (modified from Hartmann, 1975) Emiterto-receiver ‘Minimum spacing bed rezolution for rue values Tool ) (om) (em) Microlog 12 25-50 150 Microlaterolog, proximity 1 25 100 SFL. 2 305, 300 Laterolog 3 2 was «og Laterolog 8 “4 356 600 Sonic "4 0 60 Density ig $60 0 SNP-CNL 19 480 0 Lateroiog 7 2 810 150. Laterolog $ 32 810 150 Laterolog D 3 310 Bo GR 900 Induction M Isduction D 40 tooo 1200 For true’ lop reading, geometric factor, and the principle is true for all tools. Minimum bed resolution Minimum bed resolution and depth of investigation are intimately related. A tool is only capable of making true measurement of a bed if the bed is thicker than the emitter-teceiver distance of the tool (Figure 2.10). Thus, a tool with an emitter-receiver distance of 2.54cm (Lin) can resolve beds down to about 10cm, providing some idea of their true resistivity. An induction tog with an emitter-receiver distance of Lim (40 in) can resolve beds to give true tool resisti only down to 12m, and then only under ideal conditions. Table 2.3 shows some common tools, their emittes- to-receiver spacings and minimum bed resolution for true values under the best conditions. ‘A bed which is much thinner than a tool's emitter-to- receiver distance may still be identifiable. However, the value indicated on the log for this bed will only be a percentage of the real reading it should give. The tool takes a global measurement of the formation between theemitter and the receiver, the thin bed forming only a small percentage of this (Figure 2.10). The value on the log will depend on the percentage contribution that this thin bed makes to the global measurement. Aa induction log opposite a thin, resistive, limestone, bed in a shale sequence will show a subdued “blip’. On a microlog this becomes a fully déveloped peak (Figure 2.10). In reality, where lithologies vary rapidly and individual beds are thin, itis only averaged values, that appear on the log, especially the logs derived from long spacing tools. The averaged value will tend to approach that of the dominant lithology (Figure 2.10), When the mixture is 50/50, logs will even give aconstant value, butt will be somewhere between the two ‘real” values (see Hartmann, 1975). Bed boundary definition A bed, in geology, can be roughly defined as a planat unit with a homogencous composition, structure anc texture, limited by significant differences of thes: characters. The limits tend to be abrupt. Well logs ar incapable of showing these (eacures because, firstly, ¢14 THE GEOLOGICAL INTERPRETATION OF WELL LoGs 9 Movement EFFECT @) oeTEeTOR erect Figure 2.11 Logging elects on the gamma ray bed definition. (4) Sensor of zero length but a long time-constant. Distortion is due 10 ovement through the time-constant (B) Finite sie sense, long imeconstant and normal logging speed. Distortion is due 10 detector size and movement through the time constant. (Modified from Serra, 1979) point sample is not being analysed, and secondly, the logging tool is moving. The problem is well illustrated by the effect of logging speed and sensor size on the shape of the gammé ray curve opposite a sharply defined bed (Figure 2.11). The gamma ray log illustrates these effects well because the tool counts discrete events (Chapter 7). The first case illustrates essentially the effect of movement on log values; the sensor is given zero dimensions (Figure 2.114). The distortion is due to the averaging effect of tool movement. A gamma ray tool moving at 50cm/s (too fast) with a sampling count rate of once every 2 seconds will travel 1m during one sampling period. The count will therefore be the average from 1 m of formation. Over bed boundaries, half the count will be from bed A and half from bed B. ‘The average value obtained has no real formation equivalent (Figure 2.114). ‘The averaging effects of a tool reading are com- pounded by the fact that sensors have certain sensiti- vities and are ofa certain size, The sample is not a point and the sensor has its own volume, For example, the * gamma ray sensor, which may have a window 3cm ong in the direction of the tool axis, will receive radiations from a formation volume with a radius of about 30cm around the sensor (sec Chapter 7). At bed boundaries, radiations will be coming from both beds simultancously (Figure 2.11B). The actual values re- corded will again have no real formation value-eq) lent. For the gamma ray tool, averagingeffects are kept reasonable if the tool moves no more than 30cm during the sampling period (see Chapter 7) and the sensor is shielded. When interpreting logs manually, and dividing them into beds, the general tendency is to assume bed boundaries to be at the point of maximum change of vaiue or maximum slope. This may not always repre- Figure 2.12 The elect of blocking on log data. Note the reallocation, ‘of cancion’ values sent reality, but it is a good guide and is consistent, In an effort to deal scientifically with the bed- boundary problem, logs may be squared or blocked by computer. That is, the computer is programmed to eliminate the averaging that occurs on the logs between beds of different values..Log curves are resolved into zones of constant value, separated by horizontal ‘boundaries’: they become more ‘bed-like’ in ap- pearance (Figure 2.12). The usual method requires the squared log to be a true reflection of the raw log, but with transition zones or ramps eliminated, The exercise is one of reassignment of the transition zone values to re-designated blocks of real, non-transition values (Griffiths, 1982) (Figure 2.12) The algorithm applied assumes that the original log values are adequate. The method can be applied to several logs simultaneously so that they all become perfectly comparable (Serra and Abott, 1980) and the problems of comparing logs of differing depths of investigation and bed-resolving capabilities are eliminated. ‘A more complex method (Kerzner and Frost, 1984) does not assume that the raw log values are adequate, and tries to compensate for assumed tool deficiencies. For instance, it is well known that the SP will only reach its full value and real deflection in very thick beds (Chapter 5), The more complex blocking method tries to approach this full value from the raw log values using the known tool limitations, The squaring of logs certainly gives a nesrer ap proach to real formation values and formation aspects and can be a great aid to geological interpretation. However, the methods by which the squaring is done. hhave yet to become standard and accepted. Thisisia no way a criticism. 2.5 Conclusion It is suggested in this chapter that, for a proper interpretation, a logging tocl is required to make atrue, repeatable geophysical measurement of a for- mation. This was shown in fact to be impossible because of drill-created disturbances (invasion), and. because of the logging method itself. However, with a knowledge of typical formation behaviour, typical tool capabilities and log characteristics, itis possible, using the right methodology, to reconstruct the specific formation characteristics being shown on the log, A projection of this approach will be used in the ‘THE LOGGING ENVIRONMENT 15 following chapters. Each type of individual open-hole Jog, will be considered and described in terms of the corresponding logging tool's capabilities, log charac- teristics, their significance in terms of the real for- mation and interpretation in common geological terms—in short, the geological interpretation of the individual well logs. Finally, all the logs will be considered collectively and their collective interpre- tation described,3 Caliper logs Caliper tools measure hole size and shape. The sim- ple mechanical caliper measures a vertical profile of hole diameter (Figure 3.1), The more sophisticated Borehole Televiewer or Volumetric Scan gives a 360° two: or three-dimensional representation of the bore- hole walls. 3. Mechanical calipers—the tools ‘The mechanical caliper measures variations in bore- hole diameter with depth. The measurements are made by two articulated arms pushed against the borehole wall. The arms are Tinked to the cursor of @ variable resistance (Figure 32). Lateral movement of the arms is translated into movements of the cursor along the resistance, and hence variations in electrical output. The variations in output are translated into diameter variations after a simple calibration. tebite, peeey SHALE HARD LIMESTONE BED PERMEABLE “= SANDSTONE IMPERMEABLE = SANDSTONE SHALE Frequently logging tools are automatically equipped with a caliper, such as the micrologs (Chapter.6) and The density-neutron tools (Chapters 9, 10) where the caliper arm is used to apply the measuring head of the too! to the borehole wall, Sophisticated, dual-caliper fools, such as the Borehole Geometry Tool of Schhumberger, also exist specifically for measuring hole Size and volume. This tool contains two independent calipers, that is, four arms all at right angles. When the two calipers are compared itis possible to forma good idea of hole shape (see below), A gyroscope and compass included in this tool allow hole azimuth and deviation to be measured. 3.2 Log presentations “The caliper log is printed out simply as a continuous value of hole diameter with depth (Figure 3.3). The HOLE DIAMETER ‘bed hole! o- ‘tight spot” rigor The caliper lp shokng hoe diameter: some ype sesponses, “Limestone, dolomi, equally applicableFigure3.2 Schematic caliper tool showing the conversion of a ‘mechanical movement to an electrical signal using a variable resistance. (Adapted from Serra, 1973) curve is traditionally a dashed line and usually plotted in track 1, The horizontal scale may be inches of diameter or, in the differential caliper, expressed as increase or decrease in hole diameter about a zero defined by the bitsize (Figure 3.3). Theordinary caliper log is accompanied by a reference line indicating bit size. The dual-caliper tools are presented in various formats, only one of which is shown (Figure 3.4) The two hole diameters measured by the two calipers indicated are with the directional elements of tool orientation (pad 1 azimuth), hole deviation and azi- CALIPER LOGS 17 ‘muth of the deviation, An integrated hole volume may be added as horizontal ticks on the depth columa giving @ continuous record of hole volume as an aid to casing cementing (not on the example} 3.3 Interpretation and uses ‘The simple caliper log records the mechanical response of formations to drilling. Where the hole is “on-gauge’, thats, has the same size as the bit which drilled it, the formation is coherent and usually quite hard. ‘On- gauge’ holes are frequent in massive limestones and calcareous shales or older, dense formations (Figure 3.1) Holes with a much larger diameter than the bit size are ‘caved’ or ‘washed out’. Thats, during deepening of the hole, the borehole walls cave in, are broken by the turning drill pipe, or are eroded away by the circulating borehole mud. This is typical of shales especially when young and unconsolidated (Figure 3.1). When two calipers are present as in the dual-caliper tools, the shape of the hole can be interpreted. If onecaliper reads much larger than the other, caving is more extensivein one particular direction and the hole is oval (Figure 3.4). Oval holes are considered to be typical of jointed formations. However, the oval shape from Jointing is not the samme as that from ordinary caving and hole wear. Caving gives a general oval shape with no particular orientation. The two calipers may show different diameters buta similar shape, or the diameter changes may be gradual: these are simply ‘washouts. Jointing or fracturing will give an oval shape with abrupt limits and a distinct orientation (Figure 3.5): BS (in) DIFFERENTIAL lis 725] CALIPER CALI Cin) o °. + ° \ q 10 10 f pom f 20m wit ea! hole size bit size Glametor=171" — diamater=19" dlameter 2 ar arrowneas) O=reterence (at arrowhead) Figure 33 Presentation ofthe caliper log: (0, in ordinary format; 2), in diferenial format.10 25 26 Figure 34 Borehole geometry log explanation), Figure 3.3 Holesize enlargement seca cx the ou-scm dualesizes (7) Meeakowu' {2}, va: houts, calipers show diferent diameter bu! snilar shops eft) or gradual diane ehagee (ight) ciTo general hole deterioration. Hu “THE GEOLOGICAL INTERPRETATION OF WELL Loos BOREHOLE GEOMETRY LOG EVI DES) Iss 709 He ABLOEG)_ bao 30 PL AZ (EQ). __ TENS (U9) tension (pounds) ote azimuth PAD fimuth | ttt craterence) presentation (soe text for 1. GREAKOUTS: these are called ‘breakouts’ (Cox, 1983). It is even suggested that the orientation of the ‘breakout’ may be relative to the orientation of the minimum principal stress (Cox, 1983). Calipers may also show a hole size smaller than the bit size. If the log has a smooth profile, a mud-cake build-up is indicated (Figure 3.68). This is an ex- tremely useful indicator of permeability. Only perme- able beds allow mud-cake build-up. The limits of the mude-eake show the limits of the potential rese: voir. Mud-cake thickness can be estimated from the caliper by dividing the decrease in hole size by two (the calip giving hole diameter), i. bit size (diam) — caliper reading (diam) Z yud-cake thickness It should be remembered that this thickness may vary between tools. A caliper on the density tool is applied harder to the formation than the caliper ofa micro-log the former probably causes a groove in the cake and ives a thinner, log-derived mud-cake thickness. Holes smaller than the bit size, but rugose, are probably sloughed (Figure 3.64). The zones of small hole will be the ‘tight spots’ during drilling, trips or logging. That is it will be at these points that tools stick or the bit gets stuck while being pulled out of the hole. A frequent cause of tight spots is abundant smectite in the clay mineral mixture, Smectite is a swelling clay which takes water from the drilling mud, expands, breaks from the formation and sloughs into the hole. 2, WasHOUTs ie cater A umoter increases right to left. (From Cox, 1983). SA } ted wellde erences in calipers dueto factsof sto CALIPER Loos 19 oxo ru ‘souoyspues aiqeoused pur snoiod seaqes: fwoose ADOTOHIN soyu) uadt1¥o. a < ehie suoms you enous > ABOTOHLNT wos w008 souou, uaditvo “asia } gy20 THE GEOLOGICAL INTERPRETATION OF WELL LOGs S80; 2400 Sa sone uo tox} osusBowoy # pasayay a 10901 uefa sede 2 g09/0 Se" Asuep pra "w 969 18 aunyo von wate) ¥ yen PBAL2p wonzauicefusuap of mOINN JP Fouoz Bun toro 8uj.e9 pv suopypUuoD 2foy Jood. 1g amd ‘Buia wou] —— oaiv3¥O 310H 7 waeThe Gulf Coast ‘gumbo’, which often causes hole problems, is smectite-rich, Finally, an extremely important use for the caliper is, in the quality control of logs. When caving is serious, the quality of all the logs is impaired. In some tools, such as the formation-density or the neutron-porosity, the caliper reading is used for an automatic hole-size correction or compensation. Caving, will demand in- ordinately large corrections and the log values will be of litle use. It is essential to verify the caliper before consulting the other logs (Figure 3.7} 3.4 Three-dimensional calipers Two- and threc-dimensional representations of the borehole wall are possible with very specialized tools. ‘The original tool was called the Borehole Televiewer. Modifications of this tool now exist, but the basic principles of measurement remain the same. The televiewer tool uses a rotating acoustic trans- ducer to produce an acoustic image of the borehole wall. The transducer is pulsed 480 times per revolution and is rotated 36 times per 30cm (I fi) of depth. Its speed of rotation is 3 revolutions per second, ic. a logging speed.of 5ft/min (Broding, 1982). For each pulse, both the amplitude of the returning signal and the travel time are detected and recorded. The original televiewer tool used the amplitude variations of the reflected signal to modulate the intensity of a korizon- tal oscilloscope sweep. A polaroid photograph was produced of the successive depth sweeps, each photo- graph covering 1.5-3m of borehole. Under ideal conditions these photographs give a good image of the borehole wall, (Taylor, 1983). However, a continuous image may be produced by digitizing the, borehole signals (Hinz and Schepers, 1983; Pasternack and CALIPER LOGS 2] Goodwill, 1983). Digitizing also allows the image to be processed and improved. A new improvement of the tcleviewer is the Volumetric Scan Well log (Broding, 1982). Instead of just one data point, at each scanning, point round the borehole, the new method now analyses a whole series. The number of data points is vastly increased— the Volumetric Scan tool takes 512 samples per-scan point-and 512 scans -per 360° revolution. It makes 36 rotations per 30cm (If. There are thus 9.4 x 10° data samples per foot as opposed to the 1.7 x 10 data samples in the televiewer tool (Broding, 1982) The Volumetric Scan tool is now able to present 2 continuous record of data using a fibre-optics cathode- ray tube, A simultaneous record is made of both the amplitude and the transit time of the returned pulses, ‘and logs of each can be presented simultancously. The grey scale of the caliper (.e. transit time) represents distance: near'is white, far is black. The grey scale of the reflectance repeats signal magnitude: whites high, low is black. With modem recording and storing tech- niques, the images from the Volumetric Scanning tool can be enhanced’ and then displayed in a. series of orientations from the horizontal slice (Plate 1) to tilted at various degrees to the vertical, The results are impressive. ‘The applications of the specialized televiewer or 3- dimensional calipers are restricted and they should only be used over selected intervals, They are ideal for casing inspection for reasons of data quality (Broding, 1984), However, under reasonable hole conditions itis found that they are extremely useful tools for mapping and identifying fractures in open holes (see Keys, 1979; Taylor 1983}. There is even the possibility of seeing dip and other‘ wiajor structural features under the right conditions. For some reason the tool has not been taken up by the larger logging companies, but it should be used more widely.‘24 THE GEOLOGICAL INTERPRETATION OF WELL LOGS: 28 PE P P GON 4 | seen \ A 4 Se “9 ~ 25 38S Figure 42 Present-day North Sea geothermal gradients. (After Cornelius, 1975: Carstens and Finstad, 198: Harper, 1971.) Taken from Comford (1984). clays (Tertiary) salt (Permian) olastice (Carboniferous) ISOTHERMS: Interval 20°C Figure 43 The theoretical distribution of isotherms around a salt dots indicated on a geological section: change in gradients shown bby isotherm spacing. (Redrawn from Evans, 1977) devised. The most frequently-used method is that of the Horner Plot (Fertl and Timko, 1972). This plot re on the concept of a straight-line relationship between BHT and thelog of At/At + ¢, where At = time in hours since circulation stopped, and = time of circulation at ‘TD before logging (Figure 4.6} At then represents the time available for an equilibrium temperature to be reached, while ¢ represents the time that the formation is exposed to cooling mud. In practice it is best to take ¢ as the time needed to drill the last metre of hole plus the circulating time at TD (Serta, 1979). Since a BHT is taken on each tool descent, several temperature read- ings are available at different times after circulation stopped (Ad). The true formation temperature is on the line through these points where it crosses the abcissa at 1 (Figure 46). This form of correction has a significant effect on measured temperatures and appears to make them more meaningful (Figure 4.7).Aeon: ‘TEMPERATURE LOGGING 25 ‘TEMPERATURE SUAVEY Tine ater +— arithmetic 8 tor reciativity a = ‘cond. and sonic Figure 619 Typical revive fog formats (1) Dual laterolog combination (2) ication, spel focused Jog combination Hoth fom Schlumberger. illiohms/m, can also be plotted. The scale is generally 0-2000 mmho. The micro-inverse and micro-normal combination of resistivities is generally plotted on track 1. The scales are identical for the two logs, generally 0-10 obm m?/m, Unwanted logging effects The resistivity and conductivity logs are especially affected by large resistivity contrasts between the logging environment and the formation. Table 6.8 gives a résumé of these effects and their importance. 6:7 Quantitative uses of the resistivity logs ve wes of fog resistivity measurements is at the heart of the whe Jog interpretation—the domain of petrophysics. Rock resistivity was the parameter depicted on the first well Jog and it was also the first parameter to be used quantitatively. The principal use of well logsis to detect oil: the principal use of the resistivity log is to quantify oil (and of course, gas). That is, resistivity logs are used to give the volume of oil ina particular reservoir, or, in petrophiysical terms, to define the water saturation, S,.RESISTIVITY AND CONDUCTIVITY LoGs 47 ‘Table 68 Factors afecting resistivity measurements apart from invasion and bed thicknees), More common effects are italicized SH, shallow; D, deep. eat Mud cake Mud Hole size Otter @ Microlog ‘ead correct Not Poor reading E Micsotaterolog applicable ‘in bad hole ye Proximity Some cretion BE Maose. Snail influence 3 SFL Mainly Readings Some errestn— _Delawate effet and F Leeretoes sh “al inked depends on etter anomalous at D investtine ratty resistivities 2k mud Curae Induction Sh Small Poor readings Nends a sand Skin eft in D ‘n condctoe of conductive beds oad When 5, is not 100% there are hydrocarbons present: 1=S.=S,. (S,. = saturation in hydrocarbons). ‘The basic equations of petrophysics Below, the fundamental equations of petrophysics appear in a specific order, followed by explanation and comment on their computation. In fact, “these equations have applications beyond resistivity measurements, but the use of these equations is not discussed here. Such information is found in logging company handbooks and specialist publications (see references). R=ER, w Overall rock resistiotty=the formation resistivity factor x resistivity of the formation fluid (sce ‘Rock resistivity’, p.37). Rock resistivity consists of two elements, the passive but constricting formation and the conductive formation fluids. As Wyllie said in 1956, “This is perhaps the most important single relationship in electric log interpretation and must be committed to memory’ (Wyllie, 1963). A Q i R (2) The resistivity index =the resistivity of a rock containing hydrocarbons divided by the resistivity of rock with 100% water. The equation introduces the notion of the ratio, in one particular reservoir, of the resistivity when entirely water-saturated, as opposed to the resistivity in the presence of hydrocarbons. The Archie Equation nF “OR ‘water saturation e where S, =saturation exponent, usually 2. PR,=R, when the formation is 100% water- saturated (See equation 1). Thus, equation (3)is usually written, R s2— Fo "OR ‘The water saturation (squared) = the rock resistivity with 100% water saturation divided by the rock resistivity with possible hydrocarbons, The equation is more commonly written, Ee « FE om This equation, due to G.E. Archie of Shell, makes use of the ratio of resistivities from equation (2). (3a) Invaded zone resistivities—movable hydrocarbons J Rug (100% nmad filtrate) FR, (with residual hydrocarbons) Flushed zone saturation =the squate root of the flushed zone resistivity in a 100% water zonedivided by the flushed zone resistivity with possible residual hydrocarbons. Residual hydrocarbon saturation, Sy =1-S,. The equation gives the saturation in unmoved or residual hydrocarbons of the invaded zone. This is the same Archie Equation as above, but here uses the resistivity ratio in the flushed zone. Comparisons of 5, and 5,, in a hydrocarbon zone is considered to give movable hydrocarbons. $,,—Sy is equal to the fraction of movable hydrocarbons in the formation. The percentage volume in terms of the reservoir is given by multiplying the term by the porosity, ic. % volume of reservoir with movable hydrocarbons = (S,_~S,) x 4 (where @ = porosity). @ Formation resistivity factor—porosity relationships 6)48 THE GEOLOGICAL INTERPRETATION OF WELL LOGS ‘Table 69 Construction oft Archie Formula Symbol Character rived from | @ Porosity Sonic fog, neutron log, deasity log cross-piots, ef -— | on OF ‘Caleulated using empirical formulae RSS Cetumbe Formula) (eg Humble Formula) and porosity # as above R, Formation-water SP oF laboratory measurements of water resitivities samples { 8 Rock redstvity RUS FRR, | Saturated 100% (€an only be calculated, ‘vith formation cannot be measured ater ‘ith Logs) x True formation Induction logs resistivity Latzologs (deep resistivities) . = ‘Water saturation S_hydrocerbons _ R, of pores Sree a where F= formation resistivity factor $= porosity m=so-called cementation factor, dependent on rock type, and probably more closely re- Practical average Archie Equation = 0.62 x Ry PIT © Iated to pore geometry than to cementation @=a constant ‘The equation indicates that the formation resistivity factor is a function of porosity and rock type (m)— Archic discovered this relationship between F and porosity (see Figure 6.5) and equation (5) is the result. ‘Subsequent rescarch and empirical correlations show that the global relationship varies; average figures used for the relationship are: on in most sandstones a) 0.62 pe —this is the Humble Formula (best average for sandstones) (5b) r= compact formations has 5) F a where m= variable (usually 18-3) (Sd) most frequen! used formula is (58) whieh is appiicable to sendsiones. Ya the porosity relationships are quise wilt This is the general equation for finding the water saturation, and values for the unknowns may be obtained as shown in Table 69. 68 Qualitative uses General indications for resistivity log interpretation ‘To interpret the geological significance of resistivity logs itis essential to realize that the same porous bed can have a multitude of resistivity responses, depend ing on fluid content (Figure 6.20). In petrophysical terms, F will remain constant while R, varies (ee Figure 6.6). No porous bed can be said to havea typical resistivity—this is a general principle for qualitative geological work. General notions of depth of investigation and bed resolution as previously described (sce ‘Log charact: istics?) must also be considered. The indications for interpretation are: gross indications—deep logs; texture— intermediate logs; structure—micro-logs. Textures “The resistivity of a rock is intimately related to texture; quantity’ F being the mathematical expression of this. The simplest relationship between resistivity and fectuce is tha: demonstrated by an increase in re-ose RESISTIVITY AND CONDUCTIVITY LoGs 49 Figure 6.20 Schematicillstration ofthe behaviour of resistivity logs over the same reservoir bed but wth diferent Muids and, i the last eae, no porosity sistivity as porosity decreases, other factors being equal (Figure 621), This is the basis for the porosity resistivity (Hingle or Pickett) cross-plots. Moreover, tight limestone or sandstone will have high resistivities, and all curves, both deep and shallow, will give the same reading (Figure 6.20), As the porosity increases, the resistivity will decrease logarithmically, and the shallow and deep devices will show separation (Figure 6.20), ‘The resistivity-texture relationship can in fact dis- play considerable subtlety, as the following example shows, RESIETINTY 86 ohm mo & to Figuee 6:21 The close relationship between resistivity and porosity inva water-bearing sandstone. Resistivity rom sphesiealy focused log, SFL: porosity is log-derived. 10 100 3000 resistivity Ro/R iy bimodal porosity system 2 unimodal porosity system Figure 622. Textural pattems indicated on the resistivity log inthe Rodessa Limestone, East Texas. Porosity is density-log derived: resistivity is from the shallow lterolog (Prom Keith and Pitman, 1983) ‘The Rodessa limestone of the East Texas Basin shows different porosity characteristics in different sub-facies, depending on grain type. Ooid limestones tend to have bimodal porosity, skeletal limestones tend to have unimodal porosity (Keith and Pittman, 1983). The effect of each of these facies on the resistivity log is quite distinct. Thus, for the same porosity value, the unimodal porosity facies (skeletal) shows a higher resistivity than the bimodal (ooid). This is brought out by plotting density log porosities against resistivity from the shallow laterolog corrected for Rye (Figure 6.22), The textural difference between the facies is more distinctly shown by their behaviour on invasion—plotting corrected invasion-zone resisti- vities against R, (un-inyaded formation) clearly se- parates the facies. The authors found that resistivity, ra i TEE50. THE GEOLOGICAL INTERPRETATION OF WELL LOGS especially from the invaded zone, was better discrimi nator of facies than porosity. ‘The example of the Rodessa limestone simply illus- trates the sensitivity of rock resistivity to texture, Inasmuch asa certain facies produces certain texture, then resistivity is sensitive to facies, and F may be facies-related. Gross lithology Itis the domination of shale as a subsurface lithology which enables the resistivity logs to help indicate gross lithology. The only commonly-found conductor of currents, besides formation waters, is shale, and for- mations with moderate resistivity tend to be shales. However, there are no characteristic resistivity limits LitHoLoey 20 30 440 cdapth (nm 60 60 470 20 Shile intervals shown on the resistivity logs. In most Figure 64 sand -shale saquenees, shales tend {9 have a constant, 1ypial value forshale. The values depend on many variables, such as compaction, porosity, mineral content, quartz content, etc. But inany restricted zone, shale characteristics will tend to be constant, so that the resistivity log, although iit does not allow a first indication of lithology, can subsequently confirm it (Figure 6.23) The same may be said lor certain distinct lithologies ‘whose resistivity is typical. The most common are sal, anhydrite, gypsum and coal, all of which have very high resistivities (Figure 6.24, Table 6.10). High resistivities will alo be associated with tight limestones, sandstones, or dolomites. Unusually low resistivities may also be indicative. A low resistivity can be associated with electronic (metal- lic) conductivity as opposed to ionic conductivity. This is the case for mineral concentrations. An example comes from the basalts drilled on DSDP leg 26, where low resistivities were attributed to secondary hae- matite (Hyndman and Ade-Hall, 1972} As the re~ sistivity decreased in the basalts, so the amount of secondary haematite was considered to increase (Figure 625). In oil exploration, the most frequen’ mineralizations are due to pyrite Subtle lithological variations Although the resistivity logs do not allow the direct identification of lithology, they are nonetheless very sensitive lithology indicators. This is illustrated by @ series of tests made on a core from recent sediments in the Mississippi Delta (Chmelik ef al., 1969), It was found by the authors that physical parameters (such as vane shear), grain size, chemical content, density and water content all gave expression on the measured resistivity (Figure 626). The authors concluded that there were subtle differences easily visible on the resistivity log which were not initially identified on photographs or on X-ray photographs. ‘The sensitivity of the resistivity logs was used with effect in the exploration of reefs in central Alberta (McCrossan, 1961). As reefs are approached, there are subtle increases in carbonate content and bedding, characteristics. These subtle changes can be picked out by careful average resistivity mapping, higher aver resistivities surrounding the srefs (Figure 6.2 map of resistivity values: map. In order to explain tse trends, were plotted against labors 3 of carbonate content, porosity and siexsity, aod aise against depth. The interrelationships were complex, but the author considered that the most important influence on the resistivity was that of porosity (McCrossan, op. cit) (Figure 6.28). This latter, in turn, was shown to have a straight-line relationship with carbonate content, the carbonate grains probably directly filling up the pore spaces. Thus, the increase in carbonate content as the reefs are approached isSHALE SANDY SHALE-SILT RESISTIVITY AND CONDUCTIVITY LoGs SI Resistivity Sete: tans m®/mC 2) SHALE very variable CALCAREOUS SHALE TG LIMESTONE 180-6000 POROUS UMEsTONE (SALT WATER) con. sar 4 210 090 -» anevoRte 210 000 GYPSUM RAR RA RA RA anal 1000 SHALE Figure 624 Responses ona deep resistivity log of some minerals and sone typical di the following uid values: pure, fresh water 267°C) ~ ce, saltesaturated water (267°C) — ‘Table 610 Some typical (diagnostic) resistivity values (mainly feom Serra, 19721 lnetve lithologies. To theve mineral values shovld be added (0320, methane = a. ithology/ Resistivity Resistivity Range Mineral ‘ohm nym Shale Moderate Exremely variable sco 0 . (03000-10000), : Limestone Generally Variable— depends on aw Fee high poresity and formation . Dolomite “ . * Sandstone Mederate-low Salk Very 10000-iainty "208 be 0 Bich fo secondary ematte-—— Sah iaoneiy 2 in seconsry nasmatite High 1000 ‘ete eume sare High (variably 10-1 000000 Figure 6:25 Low resistivity in oceanic basalts due t0 secondary _ haematite. The amount of haematite can be judged by the Cure very low 0.000101 lemperatore; the higher the temperature the more haematite, From Hyndman and Ade Hall, 1972)52. THE GEOLOGICAL INTERPRETATION OF WELL LOGS. tem = Ge do 7s jag [7s 0 os spies ood a7” ard [a0 a0] coum ta | Sats Ty end *7) 179, Resist eaten] [soar aenry —| PBR © 626 Resistivity log ofa Recent sediment cove from East Bay, Mississippi Delta, When compared to subsequent analyses on core material, the fine detail that the log contain i evident, (From Chmelk eo, 1969) accompanied by a decrease in porosity and hence an increase in resistivity. The resistivity logs show them- selves to be as sensitive, ifnot more so, than laboratory analyses of carbonate content. Correlation ‘The sensitivity of the resistivity logs to subtle lithelogi- ccal changes is the basis for their use in correlation. Ideally, logs which correlate well are those which are more Sensitive to vertical changes than to lateral variations. Within a limited geographical extent, thisis often the case with the resistivity logs, especially in shale or silt intervals. Distinctive shapes, trends or peaks over shale zones are related to subtle com- positional changes reflecting original patterns of sedi mentation, and as such can be correlated. The best log for this purpose is usually the deep induction loz (Figure 6.29) Despite its frequent (and successful) use for cor relation, mainly as a result of its availability, the resistivity Jog has drawbacks for this task. It is influenced by changes in formation pressure and interstitial water salinity which are non-stratigraphic, post-depositional elements that tend to obliterate the original depositional features. Facies From the reef mapping example shown previously, itis clear that facies and facies changes can be followed on the fesistivity logs. Indeed, there is some overlap between a facies change and 2 subtle lithological change. One of the principal uses of the resistivity log in facies analysis is its ability to register changes in quartz (cand)-shale mixtures. This is especially so in the fine- grained rocks, shales and silts, and more so,than in sandstones themselves. The example (Figure 6.30) shows small-scale deltaic cycles 15 m-20m thick, pic- ked out by resistivity trends. The increase in resistivity corresponds to an increase in the silt (quartz) content. Even slight, subcyclic events are visible on the logs. Within sands themselves, it is suggested that in hydrocarbon-bearing zones, different resistivity values may be correlated with differences in grain size. A coarser-grained sand will generally have a low irreduc- ible water saturation and hence higher resistivity, the saturation hydrocarbons being higher (Figure 6.31). A fine-grained sand with higher irreducible water will show lower resistivity. A clean, fining-upwards sand- stone filled with hydrocarbons should show a re- gular upwards decrease in resistivity. Compaction, shale porosity and overpressure ‘The normal compaction of shale seen along a borehole shows up in a plot of shale resistivity against depth: as ‘compaction increases so the resistivity increases (in homogeneous shale) (Figure 6.32). This trend is es- pecially apparent in conductivities and a plot of shale conductivity (deep induction) on a log scale against depth shows a near-lincar distribution (Macgregor, 1965) corresponding to persistent, normal compaction, “The reason for this trend seems to be a relationship between conductivity and shale porosity. The same relationship to shale porosity is shown by the sonic log, which also gives persistent trends with shale com- paction (Chapter 8). In some wells a reversal in shale conductivity with depth is encountered: in such cases overpressure is diagnosed. When a zone of overpressure is encoun- tered, shale conductivity increases abruptly and con- siderably (Figure 6.33) although possibly taking on a(A\ edmonton al C LLB root WBIBL> 20 orm m RESISTIVITY AND CONDUCTIVITY LoGs 53 Inter-reet low 15-20 ohm m <1 ohm Figure 6.27 A resistivity map of the middle and lower Treton. Devonian ref complex, Canada. The cefsare surrounded by resistivity gradient IRedrawn from MeCrosan, 19615) new diminishing trend below. The abrupt increase is probably related to and caused by an increase in shale Porosity which occurs as the overpressured zone is entered (Schmidt, 1973). Plotting shale conductivity with depth therefore brings out normal pressure and compaction trends,and abnormally pressured zones. Care must be taken with such plots to ensure that the changes are not due to variations in shale. If enough data are available in a particular region, tables can be constructed to give quantitative estimates of over- pressure from resistivity values (Ichara and Avbovbo. 1985) “The use of normal conductivity depth trends may be used simply to indicate geological compaction. In rapidly-deposited or stratigraphically contiguous zones, compaction will be persistent and regular. When there is a break, either in sedimentation or, more importantly, an unconformity, compaction trends will54 THE GEOLOGICAL INTERPRETATION OF WELL LOGS Figure 628 Resistivity versus porosity as 2 function of carbonate Content (oy weight) for the Devonian tecf complex shown in Figure 6.27, (Redrawn from MeCrossan, 1961.) RESISTIVITY (DEEP INDUCTION) be interrupted and hence so will resistivity trends. Breaks in resistivity trends can therefore be used to diagnose geological breaks (see Chapters 12, 13) Source rock identification Organic matter is generally highly resistive to electric current and when present in shale can ctuse an increase in the shale resistivity (Figure 634). When organic matter matures, free oil is liberated and may be found. in the pores and voids of the shale: this again increases the resistivity. The resistivity may increase by a factor of ten or more (Meyer and Nederlof, 1984). Thus, source rocks in shales are associated with high resistivities. ‘A source rock cannot be identified on a resistivity log alone, because many factors cause high resistivity values. However, analysis of source rocks from around the world shows that resistivity log values, cross- plotted with either sonic log or density log values, are sufficiently diagnostic to be used as a source-rock identification tool (Meyer and Nederlof, 1984). Presumably the cross-plotting technique eliminates the lithological effects and leaves that due to the organic matter (Figure 6.35). For these calculations, ‘hen mm ohm wm ° = 10 fz ao [oz oh 7m 210 “40100 1000TNDUGTION GFF 167 NORMAL Wo oben m/e 100 LITHOLOGY ° BreET 28 Baal aE Ps depth Figure 6.30 Resistivity logs showing small-scale delta eycles. The resisuvity varies with changes in the sond-shale percentages, the resistivity values must be standardized to 24°C (75°F) using the Arps formula, They can then be put into formulae derived by the above authors with either RESISTIVITY AND CONDUCTIVITY Logs 55 SHALE RESISTIVITY ohn m2/m 10 20 30 opts the igure 6.32 Shale resistivity weads with depth, The example shove normal compaction treads from the Gulf Coast, 1, Oligocene Miocene: 2.3, Miocene, Louisiana, (Redrawn fcom Magara, 1978, alter Hottman and Johoson, 1985) sonic or density values. For example, for the sonicjresi- tivity combination D=6.906 + 3.186 log, At +0487 log, oR a positive D value indicates source rock (Meyer and Nederlof, 1984). Although any type of resistivity log may be used in these calculations, when source beds are thin the shallow reading devices are preferred (even micro- logs), although care must be taken to ensure that the latter are not over-affected by borehole conditions. 1.0 scree 190 aoe tut ul 1 4 {> @ | . Figure 631 Grain steesctivty relationship in an il zone. The sesistivity variations are related to grain-size changes (oe text)56 THE GEOLOGICAL INTERPRETATION OF WELL LOGS ome sone | conoENITY 4 see f fe 2000 & _ a_i / 18000 he % POROSITY % Figure 633 Indication of overpressure on a conductivity plo. ‘Condustivity changes corselate with changesin shale porosity (Data source, Schmidt, 1973.) ology nessTIiTy SFL depth (md 50) Figure 634 A source-rock interval fadicared by an ines = resistivity, 759° (ohm m) Figwe 635 The identification of soure-ock intervals on crost-lo of resistivity against soni anit ime, The oblique ine = =. om Seriminant snalysis using points of known souree-rock potential. (From Meyer and Nederot, 1984)7 The gamma ray and spectral gamma ray logs TA Generalities The log The gamma ray log is a record of a formation’s radioactivity. The radiation emanates from naturally- occurring uranium, thorium and potassium (see be- low}. The simple gamma ray log gives the radioactivity of the three elements combined, while the spectr gamma ray log shows the amount of each individual element contributing to this radioactivity. ‘The geological significance of radioactivity ies in the distribution of these three elements. Most rocks are radioactive to some degree, igneous and metamorphic SHALE compact sanosrone “7 LUMESTONE ARKOSE Micaccous SANDSTONE SILT/FINE SAND KAOUNITE — pune Mowr— CLAY MORiLLonire MINER nuTe ALS BITUMINOUS: SHALE COAL FINING UP SHALY. SANDSTONE CLEAN SHALE Figure 7-1 The gamma ray log some typical responses, the gamma ray log shows natura radioactivity. rocks more so than sediments. However, amongst the sediments, shales have by far the strangest radiation. It is for this reason that the simple gamma ray log has been called the ‘shale log’, although modern thinking shows that itis quite insufficient to equate gamma ray emission with shale occurrence. Not all shales are radioactive, and all that isradioactive is not necessarily shale (Figure 7.1)—see section 76, Principal uses ‘The gamma ray log is still principally used quanti- tatiyely to calculate shale volume, Qualitatively, in its GAMA AAY LOG (natural radioactivity? glauconite o heavy minerat bands \) SHALE aAse ‘LINE: spar. M = mica,* = gJaucente sume58 THE GEOLOGICAL INTERPRETATION OF WELL LOGS ‘Fabie 7.tlo) Principal wees of the gamma ray low Dieeipline Used for Kaowing ‘Quantitative Petrophysies Shale volume Gamma ray (max) ) ‘Gamma ray (min) Qualitative Geology shale Gamma say (2%) istaliness) Gamma £33 (min Litholosy “Typical Bbiosetiviey values Mineral Miser idenifeation radioactivity Sedimentology Faces ‘Chayjera size relationships Swatigrapity Corellon = Uncoaformity = identification “Table 74h) Principal ses ofthe spesrah eatin T39 TOR Ditcipline Used for Knowing ‘Quantitative Peirophysies Shale volume Th (max), Th (ein) ‘Wa or pure shale Radionctive Yn CT, ‘mineral "kim, Kirin) volume shale Semiquantitative Geolony Chay mineral Fh, K, U content and ‘dentifieation of individual clay cualitaive minerals Decsital Radioactive content ‘ineral ‘f detrital minerals Mdentfication Sedimentology Depositional “Thy content of ‘enviroment “hate depositional Reservoie Fracture ‘Uranium contribution ‘peolosy idestifiestion fo radioactivity Geochemistry Souree rock ‘Uranium content valuation ‘of organie matter simple form, it can be used to correlate, to. suggest facies, and of course to identify lithology (shaliness). In its complex form, the spectral gamma ray log, ican be ‘eed to calculate radioactive mineral volumes, identify ‘lays and detrital minerals, indicate depositional en- Vironment, and suggest fracture zones and souree-rock intervals (Table 71a, b). 7.2 Natural gamma radiation ‘Natural radiation in rocks comes essentially from only three elemental sources: the radioactive clements of the thoriam family, of the uranjum-radium family and of Oa the radioactive isotope of potassium “°K (Adams and Weaver, 1958) ‘itatively, potassium is by far the most abun dant of the three elements (Table 7.2) but its contri- tution to the overall radioactivity in relation to ts tieight is small. In realty, the contribution to the pyerall radioactivity ofthe thre clementsis of the same Oger of magnitude, the abundance seeming to be the Javerse of the contribution in energy: a small quantity ingranium has a large elect on the cadioactivity, 2 farge quantity of potassium a scl effect Bich of the three sources erait: eAmme Fays SPO taneously. Thatis,theyernity *9tous with womass Bh‘Table 72. Abundance and relative radiation activity of the naturat sadioactive elements K Th U 2pm ~3ppm unit weight 1 1300 3600 "serra (1979), Seara ex af (1980) ‘adams and Weaver (1958) ino charge but great energy (this being the definition of a gamma ray). The energy in the case of uranium, thorium and potassium emissions occurs in the spec- cum from 0-3 MeV (million electron volts), ‘The radiation from *°K. is distinct, with a single energy value of 146 MeV (Figure 7.2). Both thorium and uranium emit radiations with a whole range of energies, but with certain peak frequencies. These peaks are especially distinct at the higher energy levels of 2.62MeV for thorium and 1.7MeV for uranium (Figure 7.2). The spectra and the energy levels illustrated are those at the point of emission, One of the characteris- tics of gamma rays is that when they pass through any material their energy is progressively absorbed. The effect is known as Compton scattering, and is due to the collision between gamma rays and electrons which produces a degrading (lowering) of energy (Figure 7.3). ‘The higher the common density through which the gamma-rays pass, the more rapid the degradation or loss of energy (in reality it depends on the material's Potassium Figure 7.2 The gemma ray emission spectra of naturally radioactive sinerals. The principal peaks used to identify each source are indicated (After Titman eral, 1965, redrawn from Schlumbetges wr) THE GAMMA RAY AND SPECTRAL GAMMA RAY LOGS 59. Figure 7.3 Schematic drawing ofthe Compton scattering of gan ays. The effects mote marked in denser matics (@ Lavenda, 1985) electron density, which is very similar to common density) In borehole logging, when radiations are observed by the tool, they have already passed through the formation and probably also the drilling mud, both of which cause Compton scattering. Thus, the discrete energy levels at which gamma rays are emitted become degraded, and a continuous spectrum of values is observed (Figure 7.4). When each of the radioactive minerals is present, their radiations become mixed and the resulting spectrum is very complex. However, a glance at the original spectra (Figure 7.2) will show hat the final complex, mixed spectrum, even after Compton scattering, will still contain diagnostic peaks in the 1-3 MeV region. The original distinct peaks of potassium at L46MeV, uranium at 1.76MeV and thorium at 2.62 MeV still exist and can be used to identify the original source of radiations. This is the principle used in the spectral gamma ray tool. 73 Tools Gamma ray tools consist essentially of a sensitive gamma ray detector, generally a scintillation counter made with a large sodium iodide crystal. When « gamma ray penetrates the erystal it produces a ash of light, which is then converted to an electric pulse by a photoelectric cell. The tool literally ‘counts’ the gamma rays. In the spectral gamma ray tool the energy level of the ware ween Figure 7-4 Complex spectrum observed from a radioactive source containing poiassium, thorium, and uranium, after Compton seat. tering. (Alte Hassan etal, 1976),60. THE GEOLOGICAL INTERPRETATION OF WELL LOGS ‘Table 73 Ratios of radiodetive to non-radioactive material in rormal elemental mixtures (Serra et ol, 1980) Ty OY SFradiosetive isotopes in norma mixtures 00199 199 99.27 072 0.0057 wR aay may “Table 7.4 Typical medecn gamma ray tools ‘Name Symbol Company Gamma ray log. GR all Spectralog Dresser Atlas [Natural garoma ray NGT Schlumberger spectrometry gamma ray entering the scintillation counter is iden- tified by the intensity of the flash it produces. This energy level allows the radioactive source to be iden- tified as described above. The count is therefore separated into the relative contributions from the radioisotopes of uranium, thorium and potassium. The ‘overall amount of each of the three elements is then calculated from the known ratios of radioactive to non- (A) GAMMA RAY Los radioactive isotopes in naturally-occurring mixtures (Table 73}. Typical modern gamma ray tools are listed in Table 74 Log presentation: seales and units ‘The accepted unit for radioactivity logging is the API (American Petroleum Institute). The API unit is defi- ned in a reference well in the grounds of the University of Houston, Texas. The well contains spectally-mixed high-radioactivity concrete surrounded by equal- ly special, low-activity concrete, An API unit is 1200 of the difference between the two radioactivities, If a particular gamma ray tool is tested, the APT unit is 1/200 of the deflection between the low and high values for that tool. Thus not only does the Houston pit serve asastandard for the API unit, italso serves to calibrate 4 tool, This is also true for the spectral tool, where the values for the individual elements are found empiri- cally by calibration. ‘The simple gamma ray log is usually recorded in track 1 along with the caliper. Seales are chosen locally, but 0-100 or 0-150 API are common (Figure 7.5). The tool is small and can be combined with practically any other tool, be it a resistivity or porosity device. ‘The spectral gamma ray log is run alone (Dres- ser) or combined if required with the FDC-CNL ther toos 2 otner togs (se, caliper ote) |S Gesiatvity, sonle, density neutron ete.) GAMMA RAY 2 capa) 3 1 T 1400 gamma-ray log reading about 70 API) (©) SPECTRAL GAMM/ sone é THe ea pore ‘gamma-ray: thot thorium |, | uranium ppm) potassium |s mt = Uranium oT 7 aa wa Figwe 75 Typical yrmma ray and spectral gamma cay log headings.(Schlumberger), The log format depends on the logging company, but generally a combined, recon- stituted log is given in track 1, while tracks 3 and 4 give three logs, potassium % uranium ppm and thorium ppm (Figure 7.5). 7.4 Log characteristics Depth of investigation As discussed previously, gamma rays are subject 10 Compton scattering. Knowing the gross energy char- acteristics of natural radiation and the usual range of sedimentary densities, the approximate volume of rock contributing radiations measured by a logging tool can be calculated, Figure 7.6 shows that 50% of the signal will come from within a radius of 18m of the borehole and 75% will come from within 30em (Desbrandes, 1968), These are approximate values, since they will vary according to the density of the formation, Two rocks with identical quantities of radioactive elements but different densities will show different gamma ray counts; the less dense will be more radioactive. However, in general, the principal contribution to the radioactivity detected by the gamma ray tool will come from within 30cm of the tool detector, both horizon- tally and vertically. Logging speed Because gamma radiations are discrete events and, as described, are measured in the gamma ray tools by ‘counting’, there are restrictions on logging speeds. Radiations are ‘counted’ bya tool over a fixed periad of time, say two seconds, called the time-constant. But as radiations are to some extent random, the actual count in one time-constant varies, perhaps between 185 and Figure 7.6 The depth of investigation of the gamma ray tool. The ‘raph shows thatthe formation at 10cm from the borehole wal contributing 25% ofthe signal, at 40 em it contributes 57 or ls, Figures are spproximate. (Redrawn fiom Desbrandes, 1968) THE GAMMA RAY AND SPECTRAL GAMMA RAY LoGs 61 aman & ° : 2 x t * To inches 20 ee Figure 7.7 Statistical variation inthe gamma ray log. Two rans, GR and GR2, made with the same tol and separated by about 10 220, with an average of 200: this is the statistical variation, The effects of this on a typical log are illustrated in Figure 77. The comparison is from the same tool making a repeat run over the same section, the second run being made 10 minutes after the first. To avoid large statistical variations the ‘count’ should be as large as possible, which means it should be made ‘over a long period of time. However, since a borehole tool is constantly moving, too longa time-constant will blur bed boundaries and mix several lithologies (Figure 215). With a rapidly-moving tool, the rock being ‘counted’ at the beginning of a long time- Constant is not the same as the rock being ‘counted” at the end of the time constant (for a discussion of this see “Bed boundary definition’, Chapter 2) In practical terms, the compromise is that the ordinary gamma ray tool should not travel more than 30cm in the time-constant. Table 7.5 sets out the limitations of time-constants and logging speeds. ‘A gamma ray tool that is pulled too fast up the hole62. THE GEOLOGICAL INTERPRETATION OF WELL LOGS Table 75 Logging speed & time-constant imple gamma ray too! “Fime-constant Logging speed Formation logged (eevonds} caib) in Gime-constant (em) 1 1080) 300 2 350, 305 4 25 305 6 140 310 “Table 76 Spectral too! logging—time constants Logging ‘Fim. logged ‘Time-constant speed in time jeconds} (enh) constant (em) Company 4 25 305 ‘Sehivmberger 305-406 oa 183 Dresser Atlas shows two defects: the shape of beds is distorted and the full count value is not achieved (see Chapter 2). ‘The spectral tool is also sensitive to logging speeds, in fact more so, The time-constant for the Schlumberger tool is 4 seconds and the recommended logging speed 275 m/h, that is, 30.5cm is logged in the time-constant (Serra et al, 1980). For Dresser Atlas tools the logging speed is slightly lower (Fertl, 1979) (Table 7.6) Unwanted borehole effects ‘The amma ray log is relatively unaffected by small- scale borehole irregularities, but is markedly affected by large caves (Table 7.7). The effect is due to the increased volume of drilling mud between the for- mation and the gamma ray detector which causes increased Compton scattering and a consequent dimi- nution in the gamma ray log value. ‘A quite different effect is caused by the use of radioactive mud additives such as KCl. The potassium radioactivity of the KCl is detected by the gamma ray tool and the usual result is a marked increase in absolute values. However, since the mud volume through the hole is relatively constant for any one hole size, the relative sensitivity of the log values remains the same: it is the base line that is usually increased (Figure 7.8) although this is not always the case, ‘Table 77 Unwanted environmental eflects— gamma ray log Factor fect on log Seventy Caving ‘Diminution of values Common ‘elated to size of cave Mud additive ‘Increased absolute values, Present sensitivity unchanged (2) 2 When the effect makes the log values unusable, Ratings: requent, csommon, present, r2r8 GAMMA RAY API Figure 78 The effec on KCI in the driling med on gamma ray Fiore Mell Le with ordinary mud, well 2 with KCI mud, The Tomttion values should be the same A, thedilference ereated by the el content, The wells are 3km apart 7.5 Geochemical behaviour of potassium, thorium and uranium and natural radioactivity {The old tenet that the gamma ray log is a “shale log" was based on its use as a black box, not understanding, what was inside. In modern interpretation an under- standing of the mineralogy and geochemistry leading to radiation is used, Described below are the natural occurrences of the radioactive minerals and their ‘geological significance. Potassium Potassium is both chemically active and volumetri- cally common in naturally occurring rocks. Because of its chemical activity it is generally chemically com- bined. In the clay minerals, for example, it (and invariably its radioactive isotope) occurs in the clay silicate structure. In evaporites it occurs chemically as & salt, and in rock-forming minerals, such as the feld- Spats. itis again chemically combined in the silicate‘Table 78 Potassium ia clay minerals: chemical content, From Serra (0979), Dresser Atlas (1385) Potassiven content Minerat lite Ye by weight Average 351-831 520 Construction K. Al, Sitcate Ghvconite 320-580 450 ‘state Kaolinite 000-149 063 Al, Silicate Smestie 000-060 022 Ca, Na, Mg, Fe Silene Chorte 0 ° Mg, Fe, Al, Siliene average shale = 2.7% potassium structure, The behaviour of potassium can therefore be considezed in terms of chemical composition, as can its contribution to radioactivity. ‘The potassium content of the clay minerals varies considerably. Ilites contain by far the greatest amount, while kaolinite has very little or none (Table 7.8). The consequence of this is that clay mixtures with a high kaolinite or high smectite content will have lower potassium radioactivity than clays made up essentially of illite (mica), However, since most clays are mixtures of several clay minerals, the differences discussed above are muted, and natural shale has a relatively constant, potassium content of about 2.7% (Table 7.8). Potassium is present in many rock-forming minerals besides the micas, considered above as clay minerals ‘The most important of these are the feldspars Microcline contains approximately 16% by weight of potassium, and orthoclase approximately 14%; such percentages render the feldspars highly radioactive in geological terms (see Table 7.15). Feldspathic sedi- ments may therefore be detected by their radioactivity. inally, potassium occurs in some of the less com- monly occurring evaporites but in sufficient quantities to have an important effect on the radioactivity (Table 7.9), In these salts thereiis between 10% and 50% potassium by weight. When it is considered that the average shale contains only 2.7% potassium, the very Table 79 Potassium content of evapories Thal % Potasivm gamma 3 Species Formula Ste” SRT ART Shite KCl sas 500 Chmalite KAMECKQO' 14d 200 ovate K,S0MASO, "129 90 ‘Serra etal, 1980 "Ser, 1979 THE GAMMA RAY AND SPECTRAL GAMMA RAY Los 62 strong radioactivity of these potassium evaporites is understandable (Table 7.9, Figure 7.19). Uraniun Original uranium is generally associated with acid igneous rocks which contain on average 4.65 ppm of uranium. However, the mineral forms soluble salts, especially the uranyl ion (UO,?*) and in this form is present in river and sea water. The salts are unstable and pass easily out of solution. From sea or river water, uranium can be ‘fixed’ and pass into sediments in three geologically important ways (Serra, 1979): 1, chemical precipitation in acid (pH 25-40) or reducing (pH 0-0.4) environments: 2. adsorption by organic matter, plants, plankton, shells or by animal skeletons in certain environments: 3, absorp- tion of uranium by phosphates. ‘The extremely acid, reducing conditions required for the direct chemical precipitation of uranium indicated above are found in few natural environments. An environment of stagnant water and a relatively slow rate of deposition (Adams and Weaver, 1958) is re- quired 10 produce black shales and this is geologi- cally the most important. The high gamma-radiation values of the North Sea Jurassic “hot shales’, typical black shales, come mainly from a high uranium content (Figure 7.9, Table 7.10) (Bidrlykke et al, 1975) High organic matter values are often associated with high gamma ray values (eg. Schmoker and Hester, 1983) (Figure 7.26), This is due, certainly in part, to the uranium adsorbed by the organic matter as suggested in (2) above (Koezy, 1956). However, the exact re- lationship between organic matter and total uranium content is not easy to establish, since high organic matter content is not always related to high uranium content (Figure 7.10). Nonetheless, the principal pre- sence of organic matter in shales (Table 7.11) suggests that some part atleast of all radioactivity comes from the uranium contained in this organic matter. In general, uranium behaves as an independent constituent: it is not chemically combined in the principal molecules of rocks like potassium, but is loosely associated with secondary components. For this reason it has a very heterogeneous, original, sedimentary distribution, Moreover, its continued solubility even in the subsurface, which is a function of its loose attachments, makes it susceptible to leaching, and redeposition an¢ its distribution is therefore even more irregular. ‘Typically, on the logs, uranium is shown by irre- gular, high peaks corresponding to its uneven distri- bution, Due to the unusual requirements of its original deposition, these peaks are associated with unusual environments such as are found in condensed se- quences or at unconformities (ef, Figure 13.1).64 THE GROLOGICAL INTERPRETATION OF WELL LOGS 5 a GAMMA RAY (TOTAL) API 2 (Weorow | upanium POTASSIUM % meh OTE A ago |B fs these 2k ees a . = joaging speed s0om/m 5° stand-otls af 25 = = a ra : % j & g 3 60 ed I Figure 19 “Dak shale radioactive. Aspsstral enmma ray lop over the Upper Jurassic lac shales ofthe Noth Sea showing she ighucaniumy contribution ‘Table7.10 Black Kimmeridgian shale radioac- tivity, Averages from well 2/11-1, North Ses, Norway, rom Gigslytke of ai, 1975), “Table 711 Average weight (,) of ovganic mutter in sediments (from Shas 1980) Sediment ‘Average weighs Brenent Conent % of pamma ray 3 — lve™ Shales 290 Uranivm Sa7ppm 61 Corbonates 029 Thorium osepom 33 = —___— Potassium 1H 6 Sendstones 005 SContbtion in % (© ie wot gamma ray valve Thorium ° ° Like uranium, thorium has its origins mainly in the 3 gt acid and intermediate igneous rocks. However, it is 2 extremely stable and, unlike uranium, will not ge- y ° nerally pass into solution. For this reason it is found in : ° bauxite: (residual soils). Thorium and its minerals find 4A their way into sediments principally as detrital grains, 128 ie * They are typically heavy minerals such as zircon, ot emses 3 thorite, monazite, epidote and sphene (Table 7.12) Figure 70 Organic carbon conten there is wide dispetsion. (Soure: of a waniuon content ‘Weaver, 19SK| They are all very stable. yhorium minerals may be found as silt-sized par-Table 7.12 Thorium-bearing heavy minerals (Serra etal, 1980) Composition "TO, content ( Thorite 25-63 Monazie on Ziecon Less than | Uranium pra Thocium ppm Zircon 300-3000 SSphene 190-700 Epidote 20 Apatite 5-150 ‘Table 7.43 Thorium abundance in lay minerals [From Hassan eval, 1976; Dresser Atlas. 1983) THE GAMMA RAY AND SPECTRAL GAMMA RAY LOGS 65 ‘Thorium prow Mineral fapproximete average) Bauite 3-132 44) Kaolinite 18-26 More ice—museovite 62 Siectte 10-24 Mote Glaveonite 2s marine particles in placer concentrations (see “sandstone radio activity’, below) but occur more generally distributed throughout most shales where the thorium seems to become fixed by adsorption (Serra, 1979). Moreover, because of its detrital nature and consequent transport by currents, thorium shows an affinity for terrestrial minerals and, amongst the clay minerals fos example is more abundant in kaolinites (of terrestrial origin) than in glauconites (of marine origin) (Hassan et al. 1976) (Table 7.13) 1.6 Radioactivity of shales and clays In petroleum borehole logging, the gredtest amount of natural radioactivity (by volume) is found in shales. A high gamma ray value frequently means shale. A \ypical shale analysed by a spectral gamma ray tool shows that each of the three elements (U, Th and K) is contributing to shale radioactivity (Figure 7.11). An analysis of shales in general shows the relative contri- bution of each to overall shale radioactivity (Table 7.14), A discussion of shale radioactivity is pertinent to the utilization of the gamma ray as a ‘shale indicator’, that isthe gamma ray used to indicate the volume of shale in 4 rock both quantitatively and qualitatively. For the spectral log this requires an analysis of the relationship of the individual radioactive elements to the clay matrix, The use of the simple gamma ray is secondary to this discussion. Potassium, as discussed above, is part of the clay Figore 7.11 typical shale interval analysed by a spectral gamma ray tool. The log shows the individual contributions of thorlum, potassium and uranium to the overall radiosetvity mineral structure and will have a fairly constant distribution through most shales. As such it is a good ‘shale indicator’. However, potassium also occurs in detrital minerals and in sand-sbale mixtures may ‘occur in both the shales and in the sands. Uranium is shown distributed irregularly because of its affinity with secondary components and not the rock-forming minerals, Thus, in the average shale it may contribute only 1074-20% of the overall radioac- tivity (Table 7.14) but in certain cases this can increase dramatically (e.g. Table 7.10, Figure 7.9). This distri- bution is not related to clay volume, and consequently uranium is a poor ‘shale indicator’. Indeed, for this reason, on some spectral gamma ray logs uranium is subtracted from the total gamma ray contribution to give a better clay volume estimate (Figure 7.5). The behaviour of thorium in shales is not fully understood, Experience, however, shows that despite its varying content in clay mineral species (Table 7.13), ithasaconstant value in almost all naturally-occurring shales. The average value is about 12 ppm (range 5 to 20 ppm) for a typical shale. This constant value is said to contribute between 40% to 50% of overall shale Table 7.44 Average radiotctive mineral conte bution fo shale radiouetivity. (From Dypvik and Eriksen, 1983) ~ Contebation to radioactivity eo Postassiam 3s-45 Urasiore 10-20 ‘Thorium66. THE GROLOGICAL INTERPRETATION OF WELL LOGS radioactivity (Table 7.14 and ref), Considering there: fore the constant average value and the high contri- bution to the overall radioactivity, thorium is a very ‘good ‘shale indicator’, In mixtures of sand and shale, thorium will occur only in the shale fraction (except in rare occurrences) ‘Asa ‘shale indicator’ then, thorium may be used in ‘most cases, potassium may be used in many cases but uranium should not be used at all. This obviously also has implications for the simple gamma ray log: where uranium content is variable the log is a poor ‘shale indicator a oe e |B a a J; ol) 5 (| SB . ow) P| |B Figure 7.12 Sand line and shale line defined on a gamma ray lo, ‘These baselines are for the quanlitative use of the Top, and may be reasonably constant in any one zone 7:7 Quantitative use of the simple gamma ray log, The gamma ray log may aften be used quantitatively, and although the gamma say value for shales varies enormously, in any onearea or well, the values for pure shale tend (0 be constant (Figure 7.12) Thus, if one considers the maximum average gamma ray log value to be pure 100% shale ie. shale line, Figure 7.12) and the lowest value (o indicate no shale ata (Le. sand line, Figure 7.12}, a scale from 0-100% shale can be constructed, Since this scale will be linear, any value (GR) ofthe gamma ray log will give the volume of shale from the simple calculation GR value (og) — GR (min) volume of shale % of shale“ =—“GR (imax) — GR (min) a) GR (max (00%, shale, GR (min) = 0% shale, i ‘clean formation, Generally the value is not very accurate and tends to give an upper limit to the volume of shale (Vay OF Va): ‘A modification of the simple linear relationship used above has been proposed as a result of empirical correlation (Dresser Atlas, 1982). The relationship ‘changes between younger (unconsolidated) rocks and older (consolidated) rocks (Figure 7.13) for pre-Tertiary (consolidated) rocks, Vy, = 0.33 2"— 1) e for Tertiary (unconsolidated) rocks, Vis, = 0.083(27"™ — 1) 3) where Vj,=shale volume from these formulae (see Figure 713) and GR-GR(min) ¥& = GRmaa) — GR (min) as shown previously in (1) Figure 7.13 Graphical representsion of the relationship between Felative gamma ray deflection and shale volume. (From Dresser ‘Alas, 1982)7.8 Quantitative use of the spectral gamma ray log The spectral gamma ray log, like the simple gamma ray, is used to calculate shale volume. it can also be used to calculate the volume of radioactive minerals, Shate volume In the description of shale radioactivity given above, it was shown that the three elements are not distributed equally in shales. Some spectral logs are therefore plotted with a computed potassium + thorium radio- activity curve as a beter shale indicator (Figure 7.5). However, as indicated previously, potassium can occur in detrital minerals such as micas and feldspars so that thorium can be considered as the best shale indicator (Fertl, 1979; Schenewerk et al, 1980) The shale volume calculated from the spectral gamma ray log therefore may be based entirely on the thorium values, The mathematical relationship between thorium value (in ppm) and shale volume is taken as linear, the same relationship as between the simple gamma ray and shale volume. The equation becomes Th (log value) — Th (min) alt) Th (max) — Th (min) (4) ‘Th (min) =thorium value in clean formation (ppm) 1, NON-RADIOACTIVE ELEMENTS. 100 otrtal grains ‘auaetz o 7.0 mm t 0063 0.125, gat 025 ne median grain size Me ese 90 composition s » © THE GAMMA RAY AND SPECTRAL GAMMA RAY LOGS 67 Th (max) = thorium value in pure shile (ppm) V(t) shale volume from thorium values As with the simple gamma ray, an empirical, expon- ental relationship to clay volume may be used instead of the simple linear one shown above (Fertl, 1979), i. Vig=0.33 (2""— 1.0) for consolidated and Mesozoic rocks 6) 0.8327"! — 1.0) for Tertiary clastics (6) where Vy, hale volume. Radlioactive mineral volume Attempts to quantily the presence of radioactive minerals such as feidspars or mica are based on lwo. assumptions: ()all thorium radioactivity ig from shale, and (2) radioactive detrital minerals show only pot assium radioactivity. For the quantification, the potassium values are normalized for shale volume using the maximum and minimum method as for thorium, The normalized potassium value will give shale volume + radioactive minerals volume. Subtracting the shale volume derived, from the thorium log will leave the volume of radioac- tive minerals (Schenewerk et al., 1980), 2, RADIOACTIVE ELEMENTS radioactive detaial ‘grains lay clas S eowoleaniclclasts feldspar YC 0 = = oo) 008s ads «OBO 70 mm sur WINE ein Me ose. median grain size Figure 7.14 Radioactive elements in detrital rocks. Typical sandstone composition shown against grain size for the Reindeer Forcation, Mackenzie Delta, Beaufort Sea, Radioactive elements vary wih grin sie Ite 8 deltaic sand of Lower Tertiary ape. {Redrawn from Nentwick and Yole, 1982)68 THE GEOLOGICAL INTERPRETATION OF WELL LOGS Volume of radioactive minerals K(log value) ~ K(mnin) — MiyC(K(max) ~ (roin)] 0 K(min) = potassium (max) a =empirical factor for the formation concerned 4, in clean formation jotassium % in pure shale The quantitative use of the gamma ray spectral log is in the experimental stage and several methodologies exist for shale and radioactive mineral volume caleu- laion besides those detailed above (e.g. Quirein et al., 1982), All the proposed methodologies need empirical and theoretical confirmation (see also ‘Clay mineral identification’, below), 7.9 Qualitative use of the simple gamma ray log Lithology As a first indicator of lithology, the gamma ray log. is extremely useful as ic suggests where shale may be expected (Figure 7.1), Moreover, as shown above, the higher the gamma ray value, the higher the percentage fof shale (Figure 7.13), But the log is only a first Table 715 Potossium content of some common detital mtnerals (from Serra, 1979: Edmundson etal, 1979; Diesiee Alls, 1983; Sehlumberper, 1985), ypotassivm by Average Gamma ray value Mineral species Weight x carn) Glauconite™ ase 45 751-90" Muscovite 39-98 38 a-270 Biotie 62101 89 e715 Microctine 109-46 16 220-280" Orthociase hieta 4 20-280" “Deira! or auihigenic "For Sin hole, 1.2 gm? mud, 3Zin Nal scintillator PETROLOGY averages) MEMBERS: TAGE 8 GAMMA BAY API zi lo 8000150] & 28 = 50 ° so 100150 re 7.18 Radicactivesand, the mica sands’ ofthe Nec Sea 15-30% mica, mainly muscovite, which causes the °asiout wily quartz 65% feldspar 4% clay 15% ‘mesium sand (8804) our son [8 | g|8 pyrite 5% ale vane’ gena toon |B 5 clay 19% 5 ‘coarse sand (1000n) COWER. JORASSIC. uni sssesc They refine-prained shallow marine sandstones with perhaps 20%elayindicator. The radioactivity of some typical lithologies other than shale is now considered. This shows that any lithology indicated by the simple gamma ray log must be confirmed by other logs. Radioactivity of sandstones and other arenaceous racks Quartz, the principal component of the coarse-grained devrital rocks, shows no radioactivity, Sandstones consequently usually show low gamma ray values (Figure 7.1) However, associated detrital minerals are radioactive. The most common of these are feldspars, micas, heavy minerals and lithic fragments (Figure 7.14}, The first two groups contain potassium (Table 7.15}, the third thorium (Table 7.12) and the last contains shale, These all cause sandstones with high to moderate gamma ray values. There are many examples of radioactive sandstones that may be quoted. The fine-grained mica sands of the North Sea (Nyberg et al, 1978) are a typical, well- Table 7.16 Radioactivity in sandstones Radioactive Species Mineral element Mica sand Muscovite/biotte “°K Glauconitic sand ——Glaueonite 2K sArkose Potassiefeldspars 9K Placer sit Heavy minecals Th GAMMA RAY API 300" SANDS AND SILTS (with heavy miner placers) 7500" 7600" THE GAMMA RAY AND SPECTRAL GAMMA RAY LOGS 69 F GAMMATRAY API 30 fess : ‘ | asi {ere 50 f| Toe see 12 Figure 7.16 Glauconite causing radioactivity ia 2 sandstone in terval Sity sands envelop this matineglauconite-rich sand giving the sands higher gamma ray log values than the shales. An oil flow confirms the reservoir characterises. DST=Drill Stem. Test, *Glavconite, known example (Figure 715). Some marine sands contain glauconite and, if the concentrations are sufficiently high, render the sands radioactive (Figure 7.16). In fact radioactive sandstones are far transgressive horizon tidal channel 7a ; BEACH-MARGIN COMPLEX BEACH - SHALLOW MARINE COMPLEX radioactive placer deposits Figure 7.17. Heavy mineral concentrations placer deposits} causing spiky gamma ray log. Shales have lower gimma ray values than the hay ‘mineral deposits (Nigeria) (Redrawa from ra, 1978)JO THE GEOLOGICAL INTERPRETATION OF WELL LOGS more common than realized. Arkoses are radioactive by definition (Table 7.16). Thorium, as previously described, is present in heavy-mineral suites. Placer silts (concentrations of heavy minerals) are frequently radioactive, producing a spiky aspect to the gamma ray log (Figure 7.17) However, this is the only case, and in general detrital grain radioactivity is caused by potassium (Table 7.16) For sandstone reservoir studies, identifying clay as opposed to non-clay radioactive elements is important, Neglecting radioactive sandstone intervals as being shales means missing essential reservoir, The fact that only potassium should be causing detrital mineral vadioxetivity in sandstones (e.g. Table 7.15) is used in the interpretation of the spectral gamma ray log to separate. shale radioactivity from detrital grain radioactivity (see above, ‘Quantitative uses’) Radioactivity in carbonates Carbonates in their pure state are not radioactive and this aids their identification (Figure 7.1). Nonetheless, in certain facies, carbonates contain organic matter and this is frequently radioactive due to uranium. This is certainly the ease in the example given (Figure 7.18) and itis proposed (Hassan, 1973) that pure carbonate radioactivity is due only to uranium. Shaly carbonates will show the presence of potassium and thorium. Radioactivity in evaporites The most common evaporites, such as salt and anhyd- rite, give extremely and abnormally low values on the 3 [tose at wn tacersee a —] Sine tn ey agate on Figure 7.18 Radioactivity of Ypresian (Eocene) Limestones, Tunisia, related (0 uranium concentrations, The ucanium is 35 sociated with eatly diagenesis, orgaaic matter and phosphine Concentrations. (Redeawn from Hassan. 1973) 3 e DOLOMITE (argitlaceous) Zs sede = es i . ANHYORITE - with dolomite bands 8 combined loge sa in, Nort Sesgamma ray log (Figure 7.1). However, the high radioactivity in some evaporites caused by potassium content hus already been mentioned (Table 7.9). The log example shows & typical aspect of this evaporite radioactivity. Frequently there are extreme contrasts between the potassium and non-potassium- bearing zones (Figure 7.19}, and those zones without pot assium should be the more voluminous. Icis considered that in logging potassium salts the percentage of K,O can be estimated from the gamma ray response. Thus, for a 6.25-inch, liquid-filled hole, Edwards et al, (1967) found a correlation of 126 API units per 1% KO. Obviously, the logs must always be calibrated before making generalizations of this kind, Igneous and voleanic rock radioactivity lgneous rocks are not volumetrically important in petzoleum wells, but occur sufficiently frequently to be a nevessary element in the lithologic vocabulary, Both uranium and thorium originate in the acid-to-in- termediate igneous rocks, but their distribution is very irregular singe they are associated with secondary minerals such as apatite. Potassium is. present, es- pecially in the acid igneous rocks, principally in the alkali (potassic) feldspars, The net result is that basic igneous rocks have low radioactivity, while the in termediate and acid types show progressively higher values (Keys, 1979; Sanyal et al,, 1980) (Table 7.17) The example shows a typical basalt which may be confused with sand (Figure 7.20). Mineral identification Certain minerals can be identified on the gamma ray log by their abnormal contribution to the radioac- ‘ivity. This is certainly the case for the potassium-rich evaporites described previously (see ‘Radioactivity in evaporites’) which give high peaks. It is often the case for coals, which give excessively low values (Figure 7.1) However, these are only general indications Table 7.17 Radioactive elemenis in igneous and voleanie rocks (Grom Serra, Baldwin eral. 1980, figures approximate) eck type Taippm) Ulppm) —K,0% Typical sadioacivity Acid intrusive 1-25 1-8 4.11200 AT Acid extrusive 9-25 2-57 2.00-600 Basie imrusive 05-5 03-2 090-220 peck —— Low Basie exirusive 0510 02-4 140-250 a Very Ultabasic = — 00001 _1.60 Low 093 THE GAMMA RAY AND SPECTRAL GAMMA RAY LOGS TI GAMMA RAY APT 090, 60, 90 120, 150 sie composition basic le. pyroxene ane 25 plagioclase: 50 cept Gm) 7s 100 125 re7.20 Low gimma ray values through @ basic sll It may be infeed with & snadstone interval, Unconformivies Unusually high gamma ray values often occur as narrow, isolated peaks. Considering the geochemistry of the radioactive minerals, these peaks are generally associated with uranium concentrations. As discussed (ee*Uranium’ above) uranium concentrations indicate extreme conditions of deposition. Experience has shown that these conditions frequently occur around72 THE GEOLOGICAL INTERPRETATION OF WELL LOGS GAMMA RAY API €= 0 00 a0] F223 8 3 3 ° ° | 20m osn| T A. LOGS. {clay volume) = 100% e E “rand sit clay coarse fine LITHOLOGY/GRAIN SIZE 8, GRAPHIC RELATIONSHIP (schematic) igure 7.21, Facies com the gamma ray log (A) The changes in sandstone grain size are reflected in changes i the gamma ray value, This allows 2 facies to be supsested.(B) Graphic representation of the unconformities where a long passage of time is repre- sented by little deposition. The minerals associated may be uranium-enriched phosphates or uranium- enriched organic matter (see also Chapter 13, and Figure 13.1) Facies An interesting and fairly comprehensive scheme for facies identification in detrital sediments (sand-shale} has been developed using gamma ray log shapes. The basis for the scheme is the relationship between grain size and shale content. It is shale content that the gamma ray log indicates, but it is interpreted in terms of grain size. For example, a coarse-grained sand will havea very low shale content, 2 medium-grained sand some shale, and a fine-grained sand may be very shaly. “Thechanges in grain size will be followed by changes in gamma ray value (Figure 7.21), This method of indicating facies with the gamma ray log, however, is not straightforward. The relationship between grain size and shale content is very varisble,as is the relationship between shale volume and gamma ray value (see ‘Shale volume’). Empirically, if the shation ol grain size with gamma cay value. Heveitsexpressed as astraight line bu he relivonship i very vanable, It should parallel che clay volume change. gamma ray log shows a typical shape it can be taken as indicating grain-size changes. A lack of shape is not evidence for lack of grain-size change sinceit cannot be interpreted, ‘The facies scheme derived from the gamma ray log is fully described in Chapter 12. Correlation ‘The gamma ray log is one of the most frequently-used logs for correlation. It has ‘character’ is repeatable, is not affected by depth, it gives some indication of lithology and is simple (Figure 7.22). Moreover, it is almost always run and the sensitivity scales are always relatively similar. Generally, because it is used for correlation, it is reproduced on the well completion log, the document used to reassemble the essential drilling and geological data at the end of a well (see Chapter 11) Besides its availability, the gamma ray log has inherent advantages for correlation, especially when this concerns shales. The gamma ray value of shale formations is often variable, depending on the various amounts of clay minerals, carbonate and organicsw BN=IT _BN-10 Logs comRecrea to 1v0 Figure 7.22 Correlation using the gamma ray og mutter present, Horizontally, at the same stratigraphic level, these various elements tend to show only slight variability in the complex mix, ie. the depositional environment which controls the mix is. laterally persistent. The complexity does not persist through lime, as most abrupt changes are vertical. There arc changes, amongst others, in source and age, Thus, the gamma ray log value in shales remains constant laterally but changes vertically. These are ideal charae- teristics for correlation, In sandstones, gamma ray log shapes are often used to correlate. However, the shape isa facies characteris- tic and often leads to false corrclations (c.g, Figure 13.8) Log shapes in carbonates are generally related to shale distribution and as such are more reliable for correlation. However, the shapes must be sufficiently consistent to ensure that they are not related to uranium concentrations, as discussed above (sce ‘Carbonate radioactivity’) Although it has many advantages for correlation, the gamma ray log also has disadvantages. The fine detail on the logs is merely statistical variation. A comparison between any log and a repeat section shows to what extent this has an effect (Figure 7.7. Fine peaks therefore cannot be used for correlation, ‘The second disadvantage is that the gamma ray cannot Baronia field, Sarawak. (From Scherer, 1980) be calibrated. Although absolute values are given on the logs they are relative both to hole size and tool, the former because of Compton scattering by hole mud and the latter due to detection capabilities. Logs, to be entirely comparable, must be ‘normalized’ (sec Chapter 11). 7.10 Qualitative uses of the spectral garmma ray log Shale and clay minerals amount of literature exists on the possibility of identifying individual clay minerals using the spect- ral gamma ray log. As was shown previously see (Geochemical behaviour’) the potassium content of the clay minerals varies considerably between species but is moderately constant within species (Table 7.8) ‘Thorium, too, varies but with slightly less consistency with each species (Table 7.13). The intent is to find if these variations enable the individual species to be identified qualitatively, and eventually quantitatively. ‘The interval of the Muddy ‘I’ formation of Eastern Wyoming has been studied by Donovan and Hilchie (1981). They found a fairly good correlation between potassium radioactivity and illite content. However, they also found that while there was no correlation4. Tue GROLOGICAL INTERPRETATION OF WELL LOGS possible 20 12 Th loom) monimeritonite, tite ‘ca 100% kaolinite . —— 70% tite feldspar ine Figure 723 Geaph othe heretical disribtin oelay minerals, easy minerals and evaporite in terms potassium and shorn sone, (Redeaven from Quirein era, 1982), between clay mineral content and total gamma ra- diation, there was a strong correlation between total counts and uranium content. The essential radiation was therefore coming from uranium. The evidence suggested that the uranium source was principally smectite, its presence being caused by the exchange of the uranyl ion from the formation waters. Uranium radioactivity was therefore related to the presence of smectite ‘Almost extetly the opposite was found in the analysis of shales around the North Sea (Dypvik and Eriksen, 1983), The authors found that potassium and thorium were the dominant contributors to gamma ray activity with uranium being of minor importance (cl, Table 7.14) ‘A complex quantitative approach to clay-mineral identification has been proposed (Quirein er a, 1982). The authors suggest that clay mineral species, along with feldspar and evaporites, can all be identified relatively simply by their Th/K ratios (Figure 7.23). There is certainly a tendency for this behaviour (ct. Tables 78, 7.13) and it is the basis for using just thorium as a shale indicator (see ‘Quantitative uses’) However, it is not certain that individual clay minerals fall into such 2 simple classification. Such 2 classifi- cation demands a striet chemical control for the distribution of the elements. As was indicated, po- tassium is chemically involved in the clay lattice, but the exact behaviour of thorium in terms ofclay-mineral composition is not clear. This method needs both empirical and theoretical justification. Local variations, complexity of elay-mineral mix- tures and many other contributory variables allow no convincingly clear picture as yet for clay-mineral identification. The use of the spectral gamma ray log for this purpose is still in the future. Detrital minerals Using the spectral log to identify detrital minerals such as feldspar and mica has already been discussed (see ‘Quantitative use of the spectral gamma ray log’ above). Qualitatively, where sand (and porosity and permeability) are suspected from other logs, high potassium radiation values on the spectral log will indicate feldspars or micas. High thorium values will indicate placer deposits. The spectral log is therefore ‘more specific than the simple gamma ray log. Depositional environment In as much as thorium, potassium and uranium are environment indicators, the gamma ray spectral logs can be used for environment identification The principal use of the three elements is in the identification of the depositional environments of shales, The affinity of uranium for shales of marine origin has been documented (Koczy, 1956), as has the affinity of thorium for terrestrial sediments (Hassan te al, 1976) Consequently, it has been proposed to contrast the content of uranium in shales to that of thorium to give an index of the amount of marine influence on the environment of deposition (Adams ‘and Weaver, 1958) (Figure 7.24). Marine shales should have a low Th/U ratio, with the converse the case in continental shales. Published figures give some sup- port for this idea (Table 7.18).pr THE GAMMA RAY AND SPECTRAL GAMMA RAY LOGS 75 ‘CONTINENTAL | ENVIRONMENT Figure 7.26 Schematic representation ofthe use of the Th/U ratio to indicate environment of deposition (Source of data, Adams:and Weaver sg) However, the make-up ofa shalein any environment is extremely complex, and simple systems of environ- ment identification will lead to simple errors, The chemical composition of any shale and the chemical conditions of the depositional environment are very complex. In addition there is frequent reworking and redeposition, Cody (1971) examined the use of boron as salinity indicator, and commented ‘the complexities, of natural environments make definite conclusions (about reliability) extremely difficult’. The same com- rents can be made about the radioactive elements as environment indicators. ‘COUNTS PER MINUTE [See TTT TTT IT FT] pT = al aC ‘goa | = = Figure 7.25 High values of uranium activity identiied on the spactrallog corvelation with fractures. Completion on zone 4 alone gave 3 BOD. while the addition of zone B gave 4 total 18 BOD. (Modified from Fert, 1973.)16 THE GEOLOGICAL INTERPRETATION OF WELL LOGS ( 2 ? — == Tain aa Figure 7.26 High organic carbon values und the total gamma ray giving good corelation, in his case due Lo Uranium associated with organic ‘Table 7.18 Uranium and thorium content of shale from Adams and ‘Weaver, 1958) “Thora top) Shale ype “Uranium (opm) Grey-green American Russian Platforen Black shales ‘Average shale (est) ning average valce Figutes with too wide a spreadto be really representative, added by the author Fracture localization ‘The mobility of uranium and its presence in formation waters is considered to be the cause of high uranium radioactivity in fractures and faults (Fert, 1979; Fert] and Rieke, 1980), Using the spectral gamma ray log, zones of high uranium radiation can be detected (Figure 7.25), and other logs may be used to confirm the hypothesis. Source-rack evaluation ‘The relationship between organic matter and uranium is the basis for being able to identify source rocks using the spectral gamma ray log. The theory has already been discussed (see ‘Uranium’ sbove) and illustrated (Tables 7.10, 7.11; Figures 79, 7.10). High uranium values may well indicate high organic content (Figure 7.26), but not necessarily so. If uranium can be confidently used for the identification of paths of formation-water flowage (see Fracture localization’) it cannot also be used to identify source rocks since it is far too mobile. Moreover, investigation show that in lacustrine environments organic matter is not enriched in uranium and that organic-rich shales can show normal uranium values (Meyer and Nederlof, 1983). The use of high uranium values in shales as a means of identifying organic matter is unreliable, Such values should be used only as an indication, The use of uranium content as a quantitative estimate fer organic matter content is not advisable.8 Sonic or acoustic logs 8.1 General The log The sonic log shows a formation’s internal eransit time, designated Az (delta, the reciprocal of sonie velocity) Ic is a measure of a formation’s capacity to ransmi¢ sound waves. Geologically, this capacity varies with lithology and with rock texture, notably porosity (Figure 81) Principal uses Quantitatively, the sonic log is used to evaluate porosity in liquid-filled holes, As an aid to seismic interpretation it can be used to give interval velocities and velocity profiles, and can be calibrated with the SHALE compact SANDSTONE, compact LIMESTONE compact povomiTe: Less ‘COMPACT SHALE compact * porous oa SANDSTONE WATER Sons SHALE coat SALT ANHYORITE. SHALE i Transit Time, At 0 = 10°VAt onic velocity, fice, seismic section, Cross-multiplied with the density, the sonic is used to produce the acoustic impedance log and it is the first step in making & synthetic seismic trace. Qualitatively, for the geologist, the sonic log can indicate lithology, may help to identify source rocks, overpressure and to some extent fractures. It is fre- quently used in correlation (Table 8.1) 8.2 Principles of measurement ‘The sonic tool simply measures the time it takes for a sound pulse to travel from the emitter at one end of the logging tool to the receivers at the other end of the tool. ‘The sound measured is that carried by'P’ or compress- ional waves (Figure 82). Between the emitter and the receiver the *P” waves travel through the formation, SONIC Los + soate: miorosecondett (at) =aTnih 352 pitt 344 itt values vary considerably 80"170 wtt 3 SS re 8.1 The sonic log: some typieal sesponses. The soni log shows a formations ability co wansmit sound waves, fis expressed as Interval78. THE GEOLOGICAL INTERPRETATION OF WELL LOGS ‘Table 8.1 The principal uses of the sonic log Knowing Discipline Used for Quantiative Petroplysies Maurie velocity uid velocity Interval Titegrated craved selecity ume Seismic markers ‘Check shots beatin ‘Acoustic Direct use of| - impedance sonic lor Qualeative Geoloey Litholoey Mate and minerai and velocities quantiasive Conetation - “Texture Fracture Density log iceniieation porosities ‘Normal compaction Geochemistry ‘The sound emissions from the tool generally have a frequency between 20-40 kHz (kilohertz) or 20000~ 40000 cycles per second (Figure 8.3). The tool will register about five complete measurements per second (borehole compensated) which at normal logging eT Figure 82 Schematic representation of 2 typical “PY or com= pressional (sound) wave. In the logging environment they are the most rapid Figare 83 Sonic tol emitter patterns (schematic) Typically a pulse lasting 200 microveconds is emitted cvery 50 milliseconds, ie. 20 times a second, Four pulses are needed for 2 complete log measure ricnt, (Redrawn from Serra, 1979) Compaction and ‘verpeeseore trends Source rock Resistivity Jog ‘evaloation values speeds of about 1500m/h (5000fi/h}, produces one reading every 8cm (3in) of hole, (These figures do not apply to thenewer long-spacing sonic tools: sce below.) 83 Tools Modem sonic tools consist of a double array of sonic pulse emitters (transducers) and receivers, each array consisting of one emitter and two receivers. This arrangement compensates for borehole effects. In the conventional borehole compensation (BHC) arrange. ment the emitter-receiver arrays are inverted, while in the newer long-spaced arrangement the arrays are parallel but pulsed sequentially (Figure 8.4). ‘The tools are mostly run hole-centred so that the sonic pulse radiates symmetrically about the tool and measurements come from all sides of the hole simultaneously. ‘Table 82. The principal sonic fools Name ‘Symbol ‘Company Borehole Compensated BUC Schlumberger Sonic Long Spacing Sonic Lss Schtumberger Borehole Compensated BHC Dresser Atlas ‘Acoustibog Acouatilog Long-Spaced BHC Long Spacing ‘Acoustog Acoustilog Acawatc Velocity Lo = Welex Borehole Compensiied BCS Gearhart ‘Sonictle SONIC OR ACOUSTIC Locs 79 Figure 84 Sonic tools. Representations ofa borehole compensated sonic tool which gives instantancous readings with an inverted reve tuansmitier array and the Long Spacing Sonic Tool (Schlumberger) which gives long and short-spaced readings using time Ge. postion} delay system: pastions (1) and (2) ate both felat've {0 the same measure point, (Modified from Thomas, 1977 and Purdy, 1982), Log presentation, scales and units Sonic log values are given in microseconds (xs)per foot (1 microsecond = 1 x 10~® seconds). The value is cal- Jed the interoal transit time and is symbolized as At (Figure 8.5). The most common interval transit times fall between 40s and 140s: this is the arithmetic sensitivity scale usually chosen for the log (Figure 8.54). The velocity i the reciprocal of the sonic transit time, ie. velocity ft/s = 1/4 yl. Even on logs with a metric depth scale, the transit time is mostly still given in js/ft. The necessary conversions must be made to extract the metric velocity, thus: ‘At = 40 ys from the sonic log. Velocity = 25000 ft/s = 7620 m/s. 40x 10° When @ sonic tool is run on its own it is presented in full-width track 2 and 3 Figure 8.5a).If, as is often the case, the sonic log is combined with other tools, the log appears only on track 3, often with the sensitivity scale (of 40 js-140 ys maintained (Figure 8.56), An integrated travel time (or TTI) is recorded simultaneously with most sonic logs. It represents the average velocity for the formation logged in milli seconds (ms=1 x 10"? seconds) (Figure 85}, each ‘millisecond appearing on the inside depth column as a bar. Each 10ms isa longer bar (Figure 8.5), Adding the milliseconds and dividing by the thickness of the interval covered gives the velocity, The TTI milli- seconds may be added together to cosrespond to the travel times on the seismic section: seismic sections are usually in two-way time, that is TT] x 2, “The sonic tool is frequently run in combination with, the resistivity logs (e.g. Schlumberger ISF-Sonic tool; Dresser Atlas Acoustilog-Resistivity tool). Itis best run hhole-centred, although modern tools may be excen- red, especially in large holes.30. THE GEOLOGICAL INTERPRETATION OF WELL LOGS (a) BOREHOLE COMPENSATED SONIC LOG 3 INTERVAL TRANSIT TIME | Pht ry | | Jtii tt gt integrated | Ts | travel ime—7pee ELL LY B [LE] + mntsecond es | | | \i | Litt * Lt | oe PTTTTT rt (©) LONG SPACING SONIC LOS == 5 Lae] 3 ‘tes [799° able“tension rere) 1098 om T i | | travel tine “H 3 | T i [or oot “P ; a im (i i Figure 85 Typical sonic log headings. (2) BHC tool; (b) ong 8.4 Log characteristics Depth of investigation ‘The path of sonic waves measured by borehiole tools is essentially along the borchole wall with little penet- ration. The penetration in fact seems to depend on the wavelength of the sensed waves. The theoretical depth of investigation is between 12cm and 1 metre and will depend on the velocity for the formation; the higher the spacing tool (on the ISF-conic combination of Schlumberger} velocity, the shallower the penetration (Serra, 1979). Bed resolution TThe vertical resolution ofthe sonic isthe span between receivers for the borehole compensated tools and should be similar for the long-spacing tools (Figure 8.4) This is frequently two feet (61 cm), Beds of less than 60-crn thickness willbe registered on the sonic log, but a true velocity will not be recorded.SONIC OR ACOUSTIC LOGS 81 . one ABERRATION (= & al a4 o at 1 cycle skipping a (b) o E e.. il LSS ABERRATIONS (c, a) ce) [NT TRANS TE (ay) [NT TRANSIT Tine j2s0 psitt a ia eit 4a 10'| Separation 25m 25m Papaired spurious peaks Figure 8.6 Unwanted environmental effects on the sonic log. (a) BHC tool, eye skipping; (b) BH too, noise spikes: (e¢ long-spacing tool, paired aberrations: DTS ~short-spsced son, DT L—long spaced sonic. The paits are separated for the most pact by Of, the Feceiveetsansmitter distance,82. THE GFOLOGICAL INTERPRETATION OF WELL LOGS ‘Table 83. Unoanted environmental efleets—sonie fog 8.5 Quantitative uses Factor Efe 09 Jor Severity” The sonic log can be used to calculate porosities, coving “Cycle shipping Camaen although its usually inferior 19 neutron or density-log, Diminished ar tough 19 ‘aleulated values. ae ae “Fo use the Jog it is necessary to propose that when a poired anomalous formation has, on average, a uniform distribution of peaks (LSS) mall pores and is subjected (o a heavy confining, Tote rugosity Noise wiggering commen Increased Ar spikes (BHC) High or low oF altemate " paired anomalous peaks (E85) TWhen the effet makes the log feuding umusable. Ratings: frequent, comman, present, Fe BHC — Borehole Compensated Sonic. LS Long-Spaced Sonic Unwanted logging effects ‘The most common unwanted logging effects on the sonic log are caused by poor holes, either rugose or raeate caved. Thece may be an increase or 2 decrease in mente X interval transit time (Figure 86 and Table 83) votesiy (Thomas, 1977; Purdy, 1982). Sonic logs registered we oon after logging are the most reliable, Prolonged Fawe 87 Disurammais Wren) of the path of P waves seogsure to driling muds eauses deterioration, es ey cosh showing te rntonship between time spent inthe A og (7) eng The as fo pecially in shales (Blakeman, 1982} mat a) a om sone loc sso é cone ronowiry Se Fe geutatyscone ronosire > <7 Fonone chain oensites se reEW 224 ANO 288 DOLOMTEL esr Pehaika 0.2 REEF WELLS ad eal ss as as a is as ee DELTA ment Figure 8 Ines! ans sons compared to measured porosity ina dolomite (From MeFadaeun, 1973)pressure, there isa simple relationship between velocity and porosity (Wyllie et al, 1956), : a a which can be written, replacing Ar for V, as AL= GAG + (I= b)At gs 2 vere Y= toob-meatured wlocty ¥, = velocity of the interstitial Nuid Vg, = velocity of the matrix material = porosity At= tools measured interval transit time ‘Ar, = transit time of interstitial uid Atgs = transit time of mairix material Equation (2) simply states that the transit time mea- sured by the tool is the sum of the time spent in the solid ‘matrix and the time in the fluid: itis called time average relationship (Wyllic et al,, 1956). This ‘time’ is a func tion of the matrix velocity and volume (distance) (Figure 8.7). The relationship is best translated into graphic form, where it becomes obvious that the ‘measured interval transit time has a linear relationship with porosity (Figure 88), The relationship will vary depending on the velocity of the matrix material (see equation 2). Some of the more common matrix velo- cities are shown in Table 8.4 Although the quantitative method works for ave: rage conditions in compacted formations and es- pecially in carbonates (Sarmiento, 1961), where At, and At, are known, there are many variables besides, SONIC OR ACOUSTIC LOGS 83 ‘Table 8.4 Some typical sonic matrix velocities fee sho Figure 41) (rom Schluenberger, 1972; Serra, 1979; Gearhart, 1983), Arti) Mims vats Sandstones 555-51 490-5950 18000-19500 compacted Quant 334 ss10 ra1s0 Fimesiones 53-476 S8o0-7000 19000-23000 Calene 465 8555) 21500 Dolomites 45-98 6770-7925. 32200-26000 Dolomite 40 1620 25000 1600-5000 5000-16000 Shale 167-625 matrix and porosity which affect velocities (Figure 8.9) The effects which cause the most deviation from the simple law in porous formations are compaction (. external pressure) especially in sandstones, and the presence of gas. In unconsolidated sandstones, for example those buried at less than 1500 m~2000 m, travel times are far too long, and the calculated porosities may be over 40% (impossible theoretically). Most logging compan- ies provide compaction correction coefficients, based on cross-plotting sonic porosities and density- or neutron- log porosities. It is, however, best to avoid using the sonic log to calculate porosity in unconsolidated formations (cf. Sarmiento, 1961). ‘When gas replaces liquid in the formation, the time- average graph no longer applies (ie. V, is replaced by ¥, and ¥,,.). Even though the sonic pulse does not penetrate deeply into the formation, there is often sufficient gas in the sitallow invaded zone to affect the (ait, prass.= cons.) Figure 8.9 A nuinber of the factors which affect sock velocities. From Sherif, 1980, after Hilterman, 1977)84 -ZOLOGICAL INTERPRETATION OF WELL LOGS g INTERVAL TRANSIT TIME: Tan y2Ay 90 40] i Figure 4.10 The effect of gas om the sonic log, The sone velocity in Urs pita uandstone is lowered by about 8% velocity. Indeed, the effect can be used to identify gasevus hydrocarbons when a gas/water contact is present (Figure 8.10). To estimate the real porosity in the presence of gas, the porosity calculated from the raw log should be multiplied by about 0.7, but this is aly an estimate, To calculate porosity in the presence of shale the sonic Joy must be corrected for a shale volume derived from other logs. 8.6 Qualitative uses Lithology identification The velocity of the common sedimentary rock types is rarely diagnostic: there is too much variation within gach type and too much overlap between types (Table 44, Figures 8.11, 812). However, such is the natural occurrence that high velocities are more fikeiy to be associated with carbonates, middle velocities Figure 8.11 The average velocity ranges of common lithologies Compared. The considerable amount of overlap indicates that “cloety alone i seldom diagnostic lithology (values are for depths {pica of il exploration wells). See Table 88 for matrix velocitiss hd Table 83 for mineral velocities. with sands and shales and low velocities with shales (Figare 8.10) Velocity, nonetheless, may be diagnostic and is certainly indicative of certain rock types which are often in the near pure state in mature. Such is the case for halite (rock salt), gypsum, anhydrite and coal (Table 8:5) Correlation—sonie tog character ‘The sonic log is a sensitive recorder of a formation’s lithology. Although the precise lithology may not be identified, the sonic velocity of a particular formation is apparently very typical. The log shows even the slightest changes. It is rather like the colour of a formation: it is not diagnostic of a particular lithology, ‘but in some formations the colour is both very typical ‘Table 85 Some disgnostc (mineral) velocities (rom Serra, 1979; Gearhart, 1985; Schlumberger, 1985). ‘rity Velocity® (as) Velocity* ds Water (line) ‘189-200 H6IO-1525—$290-"H0 Hale ear 850 ‘Anbydiie 30 6100 zowwe Gypsum ses _5860 a AnthraciteCALIPER inches SONIC OR ACOUSTIC LOGS 85 INTERVAL TRANSIT TIME (a) |S GaMGAR RAY API 100 z 140, muisecones 40 2000 ee — Pelee _ 23 aor [a a Li oe bit size! 3 — b i 7 Figure 812 Sonic log in sand-shale sequences (a. The sands have a lower sonic Velocity (about 3385m/s) than the sles (3900s) (b) The reverse, where the sands have higher velocity (about 4350./s} than the shales (3300 m/s Sonic velocities are therefore not diagnti cf ithology and at the same time indicates subtle changes. The sensitivity of the sonic log is especially evident in fine- grained sediments or beds without porosity, The sequence illustrated (Figure 8.13) is entirely shaly: cuttings and side-wall cores find only shale, The sonic log, however, picks out small variations, probably in texture, in carbonate and in quartz content, to show a very distinct stratigraphic interval, despite depth differ- ences. This characteristic sensitivity makes the sonic Jog excellent for correlation Texture The inherent capacity of the sonic log to indicate texture is recognized ints use to calculate porosity (see above). The way in which sound travels through a formation is intimately associated with, apart from porosity, ‘matrix and matrix materials, grain size distribution and shape, and cementation’ (Wyllie et al, 1956), in other words texture (Figure 8.7), Theexample shows the use of the sonic log to indicate grain-size86 THe GFOLOGICAL INTERPRETATION OF WELL LOGS 90 us | = Lo 700+ 5 5 & 5 8 200: 200 Figure $8.18 Compaction in a shale sequence shown by a regular esresse in interval trait time with depth. The velocity dereases from approximately 160 yf to 140 yf over 500m, representing a shale porosity of 62%, Back-plotting the present surface values from a well {0 the 200-ys origin gives the amount of erosion (Magara, 1978). High-pressure identification Acoustic velocity ean be used to identify overpressure. Other things remaining constant, an increase in pore- pressure or overpressure is indicated bya drop in sonic velocity. A plot of shale interval transit times through an overpressured zone shows 2 distinct break in the average compaction line (Figure 8.18). The principal reason for this drop is probably the increase in shale Figure 8.16 The lationship berween sudstone po ingerval transi time io Miocene mudstones, Japan 1968)88 THE GEOLOGICAL INTERPRETATION OF WELL LOGS INTERVAL TRANSIT TIME && too Bf 3 8 om { shift to tind} | "eon wai 2 100m wa] ea compaction | 3000 | & snatod | Figure 817 Uplift and erosion indieated by shale sonic interval Tennst tines, Well has been uplfed and shows shale transi¢ times (compaction) consisten with burial 1100m deeper than at preset. Weil? defines the normal” shale trend without upli porosity, although several factors arc probably com- pounded (Figure 89). It is considered possible to cal- culate the amount of overpressure from the extent of deviation of the sonic velocity from the normal com- paction trend (Table 86) (Hottman and Johnson, 1965). Overpressure may also be calculated by an equivalesit depth method, the simplest of which gives the following formula (Magara, 1978): Pp. where P = formation fluid pressure at depth D (psi) 6, =formation-water gradient (psifft) 6, = lithostatic gradient (psi) D = depth of calculation point (ft) D, = equivalent depth ({t) with same sonic tran- sit time (see below). Sy x D)+5{D-D) Disa point in the section at normal pressure which INTERVAL TRANSIT “ren 90 200199 $0.9 imi sooo] hate At = : \ ;NomMat! TREND 2000 imal pressured ‘ror OVERPRESSURE seco ve Zone 2 4000 Figure 8.18 Overpressute indicated bys plot ofshale ntervad transit Tinks against depits A deerease fom the normal compaction trend wate overpressure (D and D, are for overpressure calculations fee text) has the same interval transit time as the point being measured. An example of D and D, equivalence is marked on the sonic-log depth plot (Figure 8.18). The above calculation suggests that the pressure at D is the sum of the hydrostatic pressure to D, and the litho- static pressure from D, to D. ‘Although the sonic log can be used to identify overpressure, it can only do so ence drilling and ‘Table 8.6 Overpressure estimates (after Hottman and Johnson, 1965), ‘At decrease from ‘average trends) er ee Reservoir Mid pressure eradient(giem) 1or Le 216 2.37 Gradients) 0.300 0365 5SONIC OR ACOUSTIC Locs 89 NEPUBLIQUE PRANGaIsE, MINISTERE DU COMMERCE ET DE L'INDUSTRIE. DIRECTION DE LA PROPRIETE INDUSTRIELLE. BREVET D'INVENTION. Gr, 8. — GL 4. N° 786.863 Procédé ot appareiliage pour la reconnaissance des terrains traverséa par un sondage. Socdé dite: SOCTETE DE PROSPEGTION BUSCTRIQUE (Precdés SCHLUMBERGER) résident eu France (Seine). Demandé le 1 juin 1934, & 46 heures, & Parts, Délivré le +7 juin 4935. — Pablié le +» septembre 1935, [Breve divest dont Id dienes af ajrne en exdeutioa de Tar. 14 7 de la eid 6 juile 1868 f ‘modifie parla loi da jal vgos. , te a ren Bae Brever Schlumberger du log sonique Figure 8.19 The breve: 'inwestion for the sonic log deposited by Schlumberger in Paris, Jane 1964. (From Allaud and Mastin, 1976,,90. THE GFOLOGICAL INTERPRETATION OF WELL LOGS logging are completed, by which time it may be too late! Source-rock identification By itself, the sonic log cannot be used to indicate source-rock potential. However, the presence of or- ganic matter, especially in shales, lowers sonic velo- cities, apparently in direct relationship.to abundance. If the sonic-log values are cross-plotted against ano- ther diagnostic log, such as the resistivity, then organic rich zones may be identified (Figure 6.35) (Meyer and Nederlof, 1984). “The principal logs for semi-quantitative source rock identification are the density and resistivity logs (see Chapters 6, 9), but the density log is allected in poor hole or when grain density varies, whereas the sonic log jg relatively unaffected and the sonic-resistivity plot may be more reliable (Meyer and Nederlof, 1984). 8.7 Seismic applications of the sonic log Acoustic velocity is the essence of the seismic section and the sonic log. Indeed, the sonic log was originally invented as an aid to seismic prospection as is seen by Schlumberger's brever invention registered in Paris in June 1934 (Figure 8.19) (Allaud and Martin, 1976). Alter its invention the sonic log became a tool for petrophysicists and geologists, but today its reverting, to its origins and is increasingly becoming a sup- plementary 100! in seismic prospection. Seismic v, sonic velocities “The frequency of the sound pulse used in sonic logging is in the range 20-40 kHz; the equivalent pulse in seismic work is 5-50 Hz The sonic tool can resolve beds down to about 50cm or even thinner, The seismic wave can resolve, typically, down to about 10m in shallow section but only 50m in deeper section; it depends on velocity and wavelength. Seismic re- solution, then, is approximately 1/100 that of the sonic og (Sheriff, 1980). The difference is well iMustrated when seismic and sonic traces are directly compared (Figure 8.20), Sonic log data, iit is to be compared to seismic data, must be brought up to the same scale and must be averaged. Interval velocities “The results of sonic logging may bepresented in several ways so that they may be used in seismic interpretation. ‘Two presentations, which are complementary, are the and the time-depth curve. To find interval velocities, the sonic velocity is averaged over important stratigraphic intervals, oF Figure £20 Thecontrastng requency content ofthesonictogend a Seiemic trace, (Redeaw rom Shes, 1980) U Np it eral ieee | / vo Figure 8.21 The presentation of sonic velocity data to match the Selle ofscismie data:the time-depth curve and the interval velocity rapit-The two hocizoptal scales are independent the depth scale Sammon 10 Both curves,IAAT A A Ne A conde) two-way tima ( SONIC OR ACOUSTIC LoGs 91 sonie velocity avs ‘time seate sonic log’ é 2 : 20 Figure 8.22 A sonic log re-played on a time sea urmished with lithology and stratigraphy, gives an accurate visual geology to the seismic intervals likely to be indicated on the seismic section (Figure 821) The velocity is found by counting the integrated travel-time marks (Figure 8.5) over the interval concerned, and then dividing by the depth covered, For instance, if 200 marks are counted (ie, 200 milliseconds) between 240m and 3400m (thick- ness 100m), the interval velocity is 1000/200 x 10° m/s = 5000 m/s, Interval velocities are usually presented in histogram form against depth (Figure 8.21) The time-depth curve is made by accumulating the interval velocities. That is, the accumulated milli- seconds are plotted against depth (Figure 8.21). For example, the interval discussed above would be plotted as a straight line from a point with coordinates of 800 milliseconds (x axis) and 2400 metres (y axis) to 1000 milliseconds and 3400 metres (i.¢ 200 ms for 100m). The interval above, 790 metres covering 175.5 milli seconds (ic. interval velocity of 4500 m/s) is represented ‘on the scale for depth from 1610 m-2400m and on the time scale from 625-800 milliseconds (Figure 8.21). ‘The presentation on the time axis then becomes similar to the seismic section. A normal time-depth curve is, taken from zero time and zero depth (ie. corrected from well KB to surface datum) to the well TD. From this can be read the average time to any particular depth or stratigraphic horizon, and this value can then be used to convert seismic time maps (isochron maps) to depth maps (isobath maps). In practice, when @ well is completed a series of ‘check shots’ is run to calibrate the sonic log. That is, a geophone is lowered into the well and a shot is fired at the surface. The time taken by the sound pulse from the surface to reach the geophone is recorded. The precise depth of the geophone is known and therefore also the precise time to this depth. The shots are made through- ut the well with the geophone at strategic stratig- raphic and scismic levels (just above a probable reflector). A time-depth curve can be made from the check shots which is independent of the sonic-derived time-depth curve. Alternatively, the check-shot depths may be used by relating them to the soniclog, the latter then being squeezed or stretched from shot to shot so that the average velocities between check shots cor- respond to the average velocities on the soniclog, With the achieved precision, the sonic log may now be re plotted by the computer on a linear time scale similar to that of a seismic section (say 10em=1 second) rather than a linear depth scale as in a well (Figure 8.22). A geological and stratigraphic repre- sentation on a time-scale log isa powerful tool for both geophysicist and geologist. The seismic section takes on a direct geological significance (Figure 8.22). Synthetic seismic logs A synthetic seismic log is a presentation of the data contained in a sonic log in the form ofa seismic trace. In92. THE GEOLOGICAL INTERPRETATION OF WELL LOGS (aan SMe Mee . — ! + ve! Figore 8.23 Diagrammatic representation of the construction of @ synthetic seismic trace from the sonic log (From Thomas, 977) a computer-derived calculation, the high frequency data of the sonic log is replayed at the low frequency of seismic data. [A seismic section is the result of acoustic reflections from subsurface strata. The reflections depend on the contrasts of the acoustic impedances (1c. velocity x density) of the adjacent layers, that is the reflection coefficient (R): acoustic impedance below acoustic impedance above ‘aooustic impedance above + acoustic impedance below DV: Diy DV, + DV, ie When both a sonic log and a density log are run in a ‘well, the acoustic impedances of the layers logged can bbe calculated (Figure £23). The acoustic impedance log represents the logged section as it would be sensed by the seismic pulse. With the aid ofa computer,asynthetic seismic signal is formulated and passed through the acoustic imped- ‘ance log. The seismic signal is distorted just as it would be if it were going through these layers in the subsur~ face. Recording the signal distortions, the computer constructs a synthetic seismic response (Figure 8.23) ‘The original sonic data have been converted into @ seismic trace, The synthetic scismic log is invaluable for ‘tying’ wells to the seismic log, and demonstrating the effective resolution on the section9 The density log 9.1 Generalities The log The density log is a continuous record of a formation's bul density (Figure 9.1). This is the overall density ofa rock including solid matrix and the fuid enclosed in the pores. Geologically, bulk density isa function of the density of the minerals forming a rock (ie. matrix) and the volume of free fluids which it encloses (ie. porosity) For example, a sandstone with no porosity will have a quartzite 0 SANDSTONE 10% 80 LIMESTONE 610% oo DoLomite: 310% SHALE SANDSTONE 9 20% poorly compacted SHALE compact COAL ORGANIC SHALE SALT SILL(IGNEOUS) SHALE bulk density of 2.65 g/cm?, the density of pure quartz. At 100% porosity the bulk density is only 2.49 g/em?, being the sum of 90% quartz grains (density 2.65 g/cm?) and 10% water (density 1.0 g/cm), Principal uses Quantitatively, the density log is used to calculate porosity and indirectly, hydrocarbon density. Its also used to calculate acoustic impedance. Qualitatively, it is a useful lithology indicator, can be used to identify DENSITY Log 1719 24,28 25 27 29 2.85 g/em> 2.71 gfem? +(2.54 g/em3) 2.87 grem? 2.68 prem) f Cons ettect. 2.03 gems 2.08 gvems| Figure 9.1 The density log: some typical responses, The density log shows bulk density. ‘Density and porosity with fresh formation.water Aensty £0 fee (f. Figure 10.1, which ison a compatible scale of neutron porosity)94 THE GEOLOGICAL INTERPRETATION OF WELL LOGS ‘Table 9.1 The princinal uses of the density log Discipline Used for Knowing Quantitative Pewophysies — Porasiy Matrix density Flvid density Seismic Acoustic (Use a fog) impedance ‘Semi-quantitative Geology Lithology (With newtron)| and —— awvaltative Shale ‘Average tends textural changes Mineral Mineral identfestion densities Reservoir Overpressure Average trends ‘eeolony Identification Fracture Sonic ecognition porority Geochemistry Sourse-rock Density-OM. cevalvation calibration ‘Table 9.2 Density and electron density of some commo, Alas, 1983; Gearhart, 1983). ‘Actual density Compound Formula py (em?) Quartz Sid, 2654 Calcite cco, 2H Dolomite CaCO, MgCO, 287 28-299) Anhydrite caso, 295 (289-305) Sylvite Kql 198 Halite Nac 216 Gypsum CaS0,2H,0 232 (23-235) Anthracite coal 16 (132-18) Bituminous coal 138 (iss Fresh water 1,0 19 Salt water 200600 ppm 1.146 “oir 40° API CH) oss Methaue cH. ‘0.000677 Gas’ (average) CuK. 09007726 ‘Schlumberger 1985 only, ‘Gearhart 1983 only. certain minerals, can help to assess organic matter content and may help to identify overpressure and fracture porosity (Table 9.1). 9.2 Principles of measurement The logging technique of the density tool '5 +> subject the formation to a bombardment of medivm-energy (0.2-2.0 MeY) collimated (focused) gamma rays and to ‘measure their attenuation between the tool source and, sn compounds from Schlumberger, 1972; Dresser Tool-derived Too! density (gem?) illerence (electron density) (gem) 264-266 +0006 2n ero 235-288 +002 10 +001 2.89205 4001 to 0.02 1.86-1.99 $012 to +006 203-208, F013 tw +012 23-240 = 001 to 005 132-1.80 (147) +003 10 ~004 LIS-117 (7124 +002 t0 0.05 Logit) Zero 1130.24) +0016 085-097 010-012 ‘0.00076 ‘o.000886 detectors. Such is the physical relationship that the attenuation (Compton scattering—see section 7.2) is a function of the number of electrons that the formation contains—its electron density (electronsjem?)—which a cam is very closely related to its common density (g/em*) (Table 9.2). In dense formations, Compton scattering attenuation is extreme and few detectable gamma rays reach the tool's detectors, while in a lesser density the number is much higher. The change in counts with change in density is exponential over theCOUNTS per SECOND ioaane sts DENSITY gem? Figure 9.2 Correlation between the density-tool radiation count (counts per second) and bulk density. A high density gives @ low ‘count, (Redeawn from Desbrandes, 1948) Figure 9.3 Mluctration of the energy bands of density and litho- density (photo-eletrc) logging and the dominant types of gamina ay interacdos, (Data sources Titman and) Wabl, 1965; Schiuraberger, 1982) ‘THE DENSITY Loc 95 Toot HEAD Gehematic) aa | | Figure 9.4 A density tool (Deosiog from Dresser Atlas) and a tool head (echematic) (Modified from Dresser Atlas, 1982) 5 =e] 2 =o CALIPER BULK DENSITY (HOLE DIAM IN INCHES crawsicn® Seamer] be 2 ad FTTH E i \ fe % i I | 1 Pa pecctggar z 14 ea | 3 in e a . lel £comeenon ea TP erent, t = | “ees Wy scat I S J L E| i 34 Figure 95 Typieal log heading ofa density log. The density logis over tracks 2and 3:the scales gfem. The automatically applied correction is shown in log form (dashed) Tais isan ‘old style’ heading ic. before about 1980.96 THE GEOLOGICAL INTERPRETATION OF WELL Loos “Table 9.3 Modern density tools Name Symbol Company Formation Density FDC Schlumberger ‘Compensated Litho-Density Tool LDT Schlumberger Compensated Densilog CDL_— Dresser Alas, Weles, ‘Geathart geological applications to be tested and will not be considered further here 9.3 Tools ‘The usual modern density tools have a collimated gamma ray source (such as radiocobalt or radio- ‘Table 9.4 Unsanted environmental effects—density log Factor Caved oF roveh hole Bifect on Fog ‘Decrease in format Severity” Yo approach a driling mad density value ‘Automatically correced i the Barite in the drilling mod tool—when mud-cake is thick, gives the cake density ‘When the effect makes the log reading unusable. Ratings frequent, common, present, rare, average logging density range from about 20-30 g/em® (Figure 9.2). Detector counts in modern tools are converted directly to bulk density for the log print- out (Figure 9.5), ‘A recent advance in density logging has scen the introduction of a tool using the photoelectric effect (the Schlumberger Litho-Density Tool). At energy levels below Compton scattering, gamma rays become so attenuated (low in energy) that they are ‘captured” and absorbed. The effect is dependent on both the medium’s electron density and its atomic number (Figure 9.3). The log records the formation’s photo- electric absorption index (Schlumberger, 1982). The Jog has not been in use for sufficient time for its J FACTOR 96 SIGNAL, DISTANCE from BOREHOLE WALL Figare 9.6 The depth of investigation of the density tool is very shallow. The graplis show experimental results for a 35% porosity, ‘waterfilled sind. (Redrawn from Sherman and Locke, 1975) caesium) and wo detectors (near and far) which compensate for borehole effects when their readings are compared and combined in calculated ratios. Source and detectors are mounted on a plough-shaped pad which is pressed hard against the borehole wall during logging (Figure 9.4). Densityog readings therefore refer to only one point on the borehole wall. Log presentation, scales and units ‘The density log is normally plotted on a lincar scale of | bulk density (Figure 9.5). The log is run across tracks 2 and 3, most often with a scale between 20 and 3.0 xem, The main log is accompanied by a curve indicating the borehole and mud-cake corrections that have been automatically applied. A record of cable tension may also be included, as the density tool tends to stick in poor holes. The tool is run typically as a density-neutron combination along with a gamma ray tool and a caliper—e.g. CDL-GR-N (Dresser Atlas), FDC-CNL- GR (Schlumberger). The caliper is an essential accom- paniment to the density log for reasons of quality control 9.4 Log characteristics Depth of investigation and bed resolution Research into the density tool's characteristics shows that its depth of investigation is very shallow. Figure 9.6 shows that 90% of the Schlumberger FDC. response probably originates from 13cm (15 in) or less from the tool, These are experimental results using 2 sand with 35% porosity (Sherman and Locke, 1975). In normal logging, the investigation cepth will probably bbe even less, around 10cm (4 in) for average densities. Consequently the density tool is likely to be much affected by hole conditions. Moreover, in porouszonesTOOL MEASURED Figure 9.7 Tool measured bulk density nda visualization of the derivatio porosity where the tool has its principal petrophysical use, it will be measuring the invaded zone. Theres little chance of it detecting Muids, notably liquid hydrocarbons, in place, While the depth of investigation of the density tool is ‘small, the bed resolution is good. At average logging speeds (about 400 m/n, 1300 ft/h), true densities can be read in beds down to about I m (3ft) At lower spoeds (about 250 m/h, 800 fy/:), thinner beds may be resolved down to 50cm (18 in). Partial reaction from the density too! may be caused by very thin beds, especially ifthey have a very high or very low density. Calcareous nodules 510m thick, for example, are seen as peaks on the density log. Good bed resolution renders the density log useful for drawing bed boundaries. Unwanted logging effects The most frequently-encountered unwanted logging effects are shown in Table 9.4. The shallow depth of investigation of the density tool makes it very suscep- tible to hole conditions, despite compensation and automatic corrections. The density log should be interpreted along with its corresponding caliper log. 9.5 Quantitative uses Porosity calculation The density log is used to calculate porosity and it may also, with difficulty, be used to calculate hydrocarbon density. To calculate porosity from tog-derived bulk density it is necessary to know the density of all the individual materials involved. The density tool sees global (bulk) density, the density both of the grains forming the rock and of the fluids enclosed in the interstitial pores (Figure 9.7). As an example, ifthe tool measures a bulk density of 2.5 g/cm? in a salt-water- bearing formation (uid density 1.1 g/cm?) we can interpret any of the following Grain density Lithology Porosity 26Sgfem? ‘Sandstone " 271 gfem? Limestone 287 gfe" Dolemite THE DENSITY Loc 97 COMPONENTS: e PORE FLUID Density 14 fem? Al __warans (arains) Density 20055/em? 4-8 8 othe porositycomponent. The figuresare fara sandstone with 10% Of course, if we know the grain (matrix) density and the uid density we can solve the equation that gives porosity from the summation of fluid and matrix components (Figure 9,7), For example, bulk density (a, = porosity (p) x fluid density (p,) + (= @) x matrix density (Pq) When solved for porosity this equation becomes i Poa — Po Pores oe = matrix (or grain) density fluid density ul: density (as measured by the tool and hence includes porosity and grain density). ‘The relationship between the bulk density (as mea- sured by the too!) and porosity can be extremely close when the grain density remains constant (Patchett and Coalson, 1979}. The example shows a reservoir of orthoquarvzite composition and a reasonably constant Brain density of 2.68 g/em? (Figure 9.8). The porosity derived from the bulk density log in this example corresponds well to the core porosity when a matrix density of 2.68 g/cm? and fluid density of 1.1 gem? are applied. If constant grain-density figures are applied to a formation and the grain density is not constant, the porosity calculated is inaccurate. This is the case with the North Sea Jurassic sands, where up to 30% mica ‘can increase the average grain density to 2.84 g/cm? (mica density is about 2.76-3.1 g/em?), When too low a grain density is used, the porosity is underestimated by the density log (Figure 9.9). Erroneous porosities may also be'calculated when the fluid density changes. This s the case when a rock is saturated with gaseous hydrocarbons. As shown above, the porosity equation is furnished with a grain density and a fluid density. The latter is 1.0 g/cm? for fresh water and 1.1 g/cm? for salt water (but may vary with temperature). In the presence of gas (typical density 0.0007 g/cm*) the fluid density drops dramati- cally. As the example shows, the density log gives too high a porosity (Figure 9.10) Ifthe porosity (and water saturation) can be calculated by other means, theQ8 THE GEOLOGICAL INTERPRETATION OF WELL LOGS BULK DENSITY gem" fue tt 68 g/m (quartz) Figure 9.8 Close correspondence between the bul density log end ‘core measured porosity in an orthoquartzite. The blk density can De converted to porosity using a mairix density of 26Bg/em", 25 Tndicated by the last histogram. density log can be used to calculate the hydrocarbon density, ‘When oil is present the porosity given by the density log is essentially correct. This is because the density tool investigates the flushed zone (see ‘Depth of investigation’, section 9.4) where only a small volume ofoil remains. Moreover, the density of oilis quiteclose ot that of water (0.7 gfer? v, 1.0 g/cm). Gas, however, is more mobile and frequently occurs in the flushed sno where, because of the large density difference with ‘er, it has the effect of diminishing the bulk density as described above. 14 germ (salt water) Acoustic impedance “The density log is used in conjunction with the sonic log to calculate acoustic impedance. The subject is brielly described in Chapter 8 (sce “Seismic appli- cations of the sonic log’) 9.6 Qualitative uses ‘The density tool gives a continuous log of the formation’s bulk density and it needs no interpretation asthe character is given directly. The qualitative use of this log therefore depends on the geological signific~ ance of the deasity of « formation, Lithology identification The densities of the more common lithologies are rarely diagnostic since there is too much overlap and too much spread caused by differences in composition and texture. Shales, for example, may have densities ranging from 1.8 g/cm? to 2.7 g/cm? the density differ- cence between a plastic clay and a compacted shale. Overall, oilfield densities generally measure between 2.0 g/cm? and 30 g/em’, the common lithologies span- ring the whole of this range (Table 9.5). Although the density logis itself'a poor indicator of lithology, combined with the neutron log it becomes excellent. In fact the neutron—density log combination is probably the best qualitative indicator of general lithology. The subject is described in Chapter 10 (see ‘Neutron-density combination). The density log in shales: compaction, age and composition Although it is impossible to recognize shales solely by their density, the density and. variations in density ‘become diagnostic once a shale has been identified. ‘Shale compaction “The compéction of shales with burial is a well-known, phenomenon and it can be followed on the density log. Shale compaction involves a series of textural and ‘compositional changes, resulting in a progressive in- crease in density (eg. Burst, 1969). For example shallow, uncompacted clays have densities around ‘Table95 Densities of common lithologies Lithotogy Range (gem) ‘ (Clays-shales 18-275 Sandstones 19-265 Limestones 2.271 Dolomites 23-287THE DENSITY LOG 99 L 1 L M L i so te © SANDSTONE no mica © MICACEOUS SANDSTONE x > 24 | = g g 2 | | S 2 wi § 0 | g ° T T T T T T T 20 2s 29 15 10 5 ° SANDSTONE POROSITY (LOG)% a a a a 22 cS 24 25 26 27 BULK DENSITY gem? Figure 99 The effect of mica on porosity values derived from the bulk density log. For the graph a matrix density of 2.65 fem? was used (giving the disgonel line) For the micabeous sands, core-measured porosities are consistently higher than those given by the log, because the grain density is too low at 265 pfem?, Mica has densities up to 3.10 p/em>, (Redrawa from Hodson, 1975) 2.0 gjem?, while at depth, ths figure commonly rises to 2.6 gicem* ‘Changes due to compaction are gradual and, when seen in one well, occur over a considerable thickness of sediment (Figure 9.11). To sec clay compaction changes, unless the shale series is very homogencous, average shale values should be read off the density log and re-plotted at 2 small vertical scale (say 1:5000). ‘This method allows clay compaction to be examined even in shale-sand or shale-Time sequences. Shale age Although itis by no means diagnostic, shale density is often indicative of age. In general, older shales are more dense, Palaeozoic clays are rare, as are Tertiary shales. The increase in shale density during compaction, although essentially due to.a decrease in porosity (Figure 9.12), is accompanied by irreversible diagenetic changes (Shaw, 1980). Compaction trends therefore become fossilized’. This means that in the subsurface, ‘change in compaction trends will indicate a change in age, in other words an unconformity (Figure 9.13). Beyond this, if general compaction curves for a region can be established, the maximum depth of burial of a formation can be estimated. The methodology is similar to that described using the sonic log (see ‘Compaction’, Chapter 8). However, for compaction studies the density log must be used carefully. Itis very responsive to local lithological variations and a usable average is often hard to obtain. Shale composition Shale density changes due to compaction are gradual, while small-order, local variations are more likely due to changes in shale composition. For example, an increase in carbonate content is generally accom panied by an increase in shale density. The increase in density is even more marked when iron carbonate (Giderite) is involved (density when pure, 3.89 g/em*). ‘When organic matter is present, the reverse occurs and the density diminishes, organic matter having a very low deasity (around 1.2.g/em?; Figure 9.14). This re- lationship may be quantified (see “Source rock eva- uation’ below), ‘The density lag in sandstones—composition and diagenesis Bulk density variations in sandstone generally indicate porosity changes. However, as explained above, this is100 THE GEOLOGICAL INTERPRETATION OF WELL Loss, BULK DENSITY ao LOG DERIVED SANDSTONE POROSITY zg GAS WATER 2 Figare 9.10 The effect of gas on the density log. In this example the sas zone reads about 35% porosity: it should read 27% porosity Bit Crea Figure 9:12. Porcsity dectine,as a function ofageofclaystone, shale, slate, (Modified feem Burst, 1969, and Manger, 1963), AVERAGE SHALE DeNSmes rod Depth im so 210 g/cm? E a 2:22 gyom? |_| s0c0 222 g/m? yo 225 g/em? I ; 4 LOWER CRETACEOUS 2-50 g/em* ‘Figure 9.11 Shale compaction with depth shown on a compressed vertical scale, bulk density log. not true when there are changes in grain density. Pure quartz sands are considered to have a grain density of 2.65 g/em?, but in reality such sands are rare. Overall grain density will change depending on the non-quartz constituents. Sands are commonly mixed with ‘eld- spars (density 252-263g/em>), micas (2.65~ 3.1 g/em) lignite fragments (0.5-1.8 g/cm?) and rock Fragments (variable density)—see Figure 7.14. Heavy minerals may als” be a constituent (2.7-5.0g/em?). The3 BULK DENSITY g z g/cm? BE le ee pie es we me 5 SHALE E a vunconronmny| i BS L a z SILTY 5 ian Figure 9.13 Smoothed bulk density log values showing an abrupt, r0ss density change across a major unconformity BULK DENSITY gfom? raters re (come ica) eae 3%) Figure 9.15 The effect of muscovite (grain density 2.76-3.1gfem?) ‘on the bulk density log in micaccous sands. The increase in density below 15m is due to the mica content. The pereentage of mica indicated is based on thin-section analysis of cove materia, THE DENSITY Loc 10 a ee eS oe Figure 9.14 The effect of organic matter on shale bulk: density values. When organic matter content is high shale deosity is considerably lowered (note density scale is reversed from loz) T0.C, =Total organic eatbon, well known mica sands of the North Sea Jurassic reservoirs (as already discussed) contain up to 30% muscovite (Figure 9.9). The density of muscovite (2.76~ 3.10g/em’) increases the average grain density from 2.65 g/em? to c. 2.82 g/cm? and it varies with the mica content (Figure 9.15). In sands without shale, therefore, grain density can give some idea of sand composition. Changes in grain density in sands are generally gradual and of a moderate order. Abrupt changes, especially in otherwise homogencous beds, often in- dicate diagenetic or secondary changes. The example shows a sand with a zone of secondary carbonate ceinent (Figure 9.16) In cores these zones are shown to have very abrupt limits. A similar phenomenon may also occur with secondary pyrite cement. Mineral identification Density becomes a criterion for lithological identii- cation when itis ether abnormally high or abnormally ! | Figure 9.16 Secondary, diagenetic cement causing an abrupt lop petk in a sandstone itervasore 102 THE GEOLOGICAL INTERPRETATION OF WELL LOGS, ee 3 BULK DENSITY gem? \ 7 i ARTE Sha i = : sa tlh ' eee) EE Sclow f f edded PYRE nig pak Figure 9.17 The ientifcation of cost, with low density and pyrite, with high density, on the bulk density log Lithology fom core analysis ow (the average for sedimentary rocks in oil wells being about 2.5 g/cm"). Coals, for example, are iden tified by very low densitics, between 1.2g/em* and 1.8g/m?, and pyrite, conversely, by very high den- sities, between 4.8 gjem® and 5.17 g/om?, The extreme & [ouncoensiry | NevTRONS g ‘fem % & ipo “25 aduo — eyo 7g || : 2 ea £ erator] Le sat pisith; 7204 ovens eet eater 4 ene \ f75m Figure 9.18 Bulk density log overs sat-shale series. Thedensty log lover the evaporite intervals tends to give constant values. The neutron log aessts in the kdemification of the evaporite intervals, (GN ak=— 3) 25m § § S & fom 5 i 3 2 ‘Table 9.6 Diagnostic migeral and ftholog cal densities rom Serra, 1972, 1979; Gearhart, 1983; Dresser Atlas, 1983). tem) Low ignite Coal ‘Aathracite Organic shale values for these minerals may not be reached under natural conditions, but abnormally high and abnor- ‘mally low peaks are still easily visible (Figure 9.17). The more common extreme and diagnostic densities are shown below (Table 9.6) Evaporite identification Chemical deposits, because oftheir purity, may at least bbe suspected if not positively identified by their den- sities (Table 9.7). Care must be taken, as ‘may be impure and densities will be altered." most evaporites tend to give intervals of density with very little variation, When this occurs, along with densities near the pure mineral values, evaporites are probable (Figure 9.18) Overpressure identification “The general increase in shale density with depth of burial was described under the heading of compaction. The principal cause for this gradual increase is @Table 9.7 Evaporite densities (typical va les as seen by the density tool) THE DENSITY LOG 103, Table 9.8 Some shale porosity values (13,4 fom te 2) ‘Approximate shale porosity Evaporites Density (jem) = Area 500m 100m 200m — 5000 m salt 204 Anhydrite 298 LGulfCoat 45% 28 Gypsum 235 a _ Carmallite 137 72. Nagaoka Sylvite 185 Plain 4% ym — Polykalite 29 Japan 3 Venenuels 30% = 4 Oklahoma 23% diminution in shale porosity with increasing over- burden, Mudstone porosities may be as high 2s 50% near the surface, diminishing rapidly to below 20% from about 600m downwards (Figure 9.19) (Magara, 1978). The actual figures and gradients vary from one region to another (Table 9.8), although the normal trend of a progressive porosity loss is universal, However, porosity may increase with depth and when it occurs there is overpressure. The general decrease in shale porosity is accompanied by an expulsion of both pore-water and interstitial water (Burst, 1969). The fluids are gradually squeezed out during burial Ifthe fluids cannot escape, once trapped AE ras = g |__| =m a Foci MUOSTONE POROSITY 96 Figure 9.39 Diminution of mudstone porocity wth depth, Neogene, Japan. (Redrawn from Magara, 1968) 1. Dickingoa (1953), 2 Magara (1968) 3. Hedberg (1963) 4. Athy (1930), they inevitably become overpressured: they begin to support some of the overburden pressure (see Chapter 2), This has the effect of preserving porosity. It is this preservation which causes a break in the compaction trend which is registered by the density log. The density break therefore identifies zones of abnormal pressure (c.g. Fert, 1980) (Figure 9.20), Fracture recognition ‘Numerous methods have been proposed for the iden fication of fractures (Schafer, 1980). One of these involves the comparision of density-log porosity with sonic-log porosity, The density tool records bulk density, and as such will include both intergranular porosity and fracture porosity. For the sonic measure- ment, however, the sound waves will take the quickest path from emitter to seceiver. This path should avoid the fractures. The sonic velocity will therefore give only intergranular porosity. When the density-derived po- rosity is much less.than the sonic porosity, the differ- ‘ence is due to the fracture porosity (Schafer, 1980). In practical terms, the two logs should be norma- lized to permit this comparison. This may bedone with the logs themselves (Figure 921) or by cross-plotting, core-verified values to define ‘fracture fields’. ‘The method is by no means everywhere simple and successful, Cross-plotting log density against log ve- locity on core-verified valuesfor the Grove Formation, Kansas, showed distinct fields for fractured and non- fractured formation (Etnyre, 1981). However, the fields showed exactly the reverse relationship to the one expected :for a given velocity, fractured zones showed a high density, and for non-fractured zones, a low density (Figure 9.22). The problem seemed to be one of li- thology, with the fractured zones occurring over a particular lithology. The cross-plot may be showing lithological differences rather than the presence of fractures.104 THE GEOLOGICAL INTERPRETATION OF WELL LoGs sowie BULK DENSITY ite fem PRessune PRESSURE = son f0 arafon Figure 920 Overpeessure shown by a diminutton of bulk density Source rock evaluation ‘The presence of organic matter in shales lowers their density. The normal average matrix density of 2 mixture of clay minerals is about 2.7 g/em* (Table 99), while organic matter has densities between 0.70 g/em?— 1.80 g/om* (Table 9.6). The presence of organic matter therefore has a marked effect on shale bulk density (Figure 9.14), The organic-matter effect on the density log can be quantified (Tixier and Alger, 1967) and the log used to Gerive the amount of organic matter in a shale, To do Con Vest Energy Corp N01. John B. Lupe Figure 9.21 Identification of fractures in the Austin Chalk, Texas by plotting the bulk density and soni logs on 8 compatible porosity Seale. The fractured zones give 4 ower density and higher porosity than the sonic. (Modified from Schafer, 1980) nytt acoustie 27 26 2s, 24 2a pensiryg/em? ; FRACTURED ZONE ©, NON FRACTURED, Figue 922 Fracture identifeation by cross-plot, in the Grove Formation, Kansas. Fractured zones. show higher than normal 20 8 0 5 z 16 2530 noutron porosity units % Figure 10.2 Logs (A) end cxoss-plt (B) showing the efet of gas on rose plot, The neutron shows too low a porosity in gas zones hydrogen index compared to water. The presence of gas makes the neutron log give oo low a porosity (Figure 10.12). Corrections for gas content can be made (Gaymard and Poupon, 1968) but the best use of this phenomenon is qualitative. Moreover, on the neutron-density combination (sce below) gas stands ‘out very distinctly, giving a large negative separa- tion (Figure 10.12). Shale effects on neutron porosity Since the neutron log is sensitive to all hydrogen nucle, itis sensitive to both free and bound water, The former is formation water, the latter occurs in clays ‘Giher within the molecule or adsorbed between day mineral layers (sec ‘Neutron log in shales’ section 10.6) (CNL Schlumberger) the neutron (and density) fogs. The nutton values from A are used inthe Slight admixtures of shale with reservoir matrix ‘material therefore disrupt neutron porosity values, and the true porosity cannot be calculated without cor- rections. The example (Figure 10.13) shows that the neutron porosity stays constant while the true porosity varies considerably. ‘A study of shaly sandstones showed that in quartz~ clay mixtures the hydrogen indices of wet clay and formation water are very similar (Heslop, 1974). In other words, the neutron is incapable of separating wet clay from water. Cross-plotting gamma ray values (as clay indicator) against neutron log values illustrates this. The gamma ray log shows diminishing clay volume and the neutron maintains a constant value (Figure 10.14). The neutron-derived porosity is there- fore erroneous and the neatron cannot be used. to derive a clay volume. In shaly sandstones, therefore, the neutron poresity value is best not used,10.6 Qualitative uses Lithology identification The use of the neutron log to identify lithologies depends on an understanding of the distribution of the hydrogen index in natural materials. ‘The hydrogen detected by the neutron tool occurs in two principal chemical combinations, one between hydrogen and carbon (the hydrocarbons), and one between hydrogen and oxygen (simply water) Hydrocarbons occur as gases (methane, etc), as liquids (cil, bitumen, etc.) or as solids (coal, organic matter), Water occurs as free water (in pores), as absorbed ions (asin clay interlayer zones), as water crystallization (as in evaporites), or as combined water (as in igneous rocks) The lithologies in which these various forms of combined hydrogen are found have hydrogen indexes which cover the entire scale between 1 and 0 (Table 10.5), Probably only pure water can be re- cognized categorically by its hydrogen index, which is However, the neutron log gives an extremely sen- sitive reflection of lithological characteristics and changes, and combined with other log responses the THE NEUTRON Log 113 sandstono porosity Mine = CNL cone PoROSITY a 3 tow NEUTRON POROSITY LOG VALUES Figure 10.13 The effect om stale on neutron porosity values. The neutron log registers shale 2s poresity (water), The true porosities should fal along the diagonal line (compare Figure 1014) Figure 10.14 Neutron log values cross-ploted with gamma ray values in 2 shaly sand formation. The neutron log shows the same porosity in shaly sands, with gamuna ray values between about 50-70 API, as inthe clear sands, with gumma ray values below about 50 APL This is aot the case; they are lower in the shaly sands compare Figure 1013} (From Heslop, 1974)114 THE GEOLOGICAL INTERPRETATION OF WELL LOGS Table 10.5 Neutron log values of some common lithologies (rom Serra, 1979; Edmundson and Raymer, 1979) Limestone Tiydrogen poresity units index! NL ‘Water, fresh 100 00 ‘Water salt oe Quan -2 Seidstones™ 21025 Cake a — Limestones* =i 1030 Dolomite 1 o Dolomites* 110 30 ‘Shales 251075 209 to Coal, ignite 2 066 Coal anthracite 38 040 Methane (2910 59) 049, ‘Approximate ranges up (0 30% porosity "200° F, 7000 pai. hydrogen index becomes diagnostic. This is examined below. Neutron log in shales: compaction ‘The neutron log shows abnormally high ‘porosities’ in shale or clay intervals. Values vary between 75% @N and 25% 6N, buta typical shale has values around 40 50%4N (Figure 10.15). The neutron log will therefore indicate probable shales (40-50% N) as opposed to sandstones (030% @N) or limestone (0-35% oN). The values in brackets are only ‘typical’, not exact. A shale “Table 10.6 Combined water in clays Chay ype “water? Hydrogen Neutron ro) index” porosity value (exuy ite 8 00 30 Kaolinite 3B an 37 Chiorive 4 03 Pa Smectte 122 a7 4 Weaver er al (1973) Serra (1979) with a neutron porosity value of 50% does not have a real porosity of this value, The hydrogen index is bigh because of the presence of both free and bound water. ‘When a clay is deposited, up to 70% or more of its volume may be water. This diminishes very rapidly, and over shallow geological depths typical of oil wells, is generally between 10% and 25% depending on the degree of compaction (of Figure 9.11), Clay water is divided into free pore-water, adsorbed ‘water clinging to the clay but also in the pores, and lattice-water which forms part of the clay mineral structure. Clays with no lattice-water show a gradual elimination of both pore and adsorbed water by compaction. A residuum of about 10% usually re- ‘mains. Interstitial water is an important element of the smectites (Table 10.6) and complicates compaction since it is more or less stable up to an abrupt NEUTRON POROSITY UNITS % som Figure 10.15 Typical nevtcon log response ina sand~chale sequence. Shales givehigh valtes @N 40—45%, sands give lower values $V 28-30% (ON in shales ig due 10 free, adsorbed and inteslayer Water, and does nol give real porosity.TE og CE agony om F005] Figure 10.16 Indication of clay.water types on a density neutron cross-plot. The outside (heavy) line has a slope of SNP 27554 1.14 GFCD, (Redrawn from Mapara, 1982) doeoe U1 100R0—— anaow oe ‘Sisal ‘og valet) [Figure 10.17 The effets of shale compaction on neutron log val in a sand-shale sequence. Histograms’of the log values show the diminution in sand porosicy, but only a seal change inthe shales, THE NEUTRON LoG 115 dehydration point. This point is largely temperature- controlled, but in oilfield work is often related to depth. A clay rich in smectite above the dehydration point may contain up to 20-25% interlayer water: below itis rapidly eliminated. In depth terms this can be any- where between 1500m and 5000m but is usually at about 2000m (Shaw, 1980), Anattempt has been made to study the behaviour of the various clay waters using neutron-density cross- plot techniques (Honda and Magara, 1982), Adsorbed and free water will be detected as porosity by both the density and the neutron tools. Interlayer water will, however, only be detected by the neutron. The line of equal density-neutrosi porosity indicates the limit of the interlayer water (Figure 10.16). This theoretical behaviour of water in clay for- mations (Figure 10.16) is difficult to see on the neutron logs. In a typical well of 3000m there is a gradual diminution of the average neutron shale value (Figure 10.17). This is presumably the compaction elect yet itis surprisingly small when the theoretical diminution of water content with depth is considered. Frequently, when compaction is indicated on other logs (for example the sonic) the neutron-log value remains constant Figure 10.18) A verified explanation of these phenomena has yet to be found, but they suggest that the bulk water content of clay, as seen by the neutron, remains constant, while its distribution between pore-water, adsorbed water and interlayer water changes. Pore-water should diminish with compaction, RANSIT TIME NEUTRON POROSITY UNITS % z eitt E eso. 40 soo 4s 309-15) 00m 600m oom 700m 200m 800m 300m 000 s000m 1000m Figure 10.18 Compaction in shale sequence shown by a decrease in interval trent time (increase in velocity} while the neutzoa log shows constant values,116 THE GEOLOGICAL INTERPRETATION OF WELL LOGS war SEDIMENTARY ‘organic shale | |LOG fae Gama ar] | NEUTRON POROSITY UMTS % | THIN secTION 3 consstruents a ro 2 leo “5 30 on MSIE 0% rs | 5 s FE : 2 2,2 = a £ 5 Fie 1019 Changs inthe wrong corel with hangs in arts dint The ihe the quae cone the lower the neutron SEDIMENTARY LOG (core) organic mai. thin section % NEUTRON POROSITY UNITS % loo 45 30 16 i : rn ae . 40m -* _ organic matter peak aoe Ne, eo a $F 20 S Sze al Flewe 10.20 The elles of ogaie mater o he eton oy, Organi trata ih onc index cause x norte ees in the seuteon log valves,THE NEUTRON LOG 117 Table 107 Nevtzon tog values of some evaportes (rom [Toa an | NEUTRON Schlumberger 1985) z Ponosiy % - 5 fo roopoz0 10 0-19) Minecat Composition Nevizon - Jom tog Lteten hae [rete Carnailie | KCIMgC, 64,0 oot | retete Gypum | CaS0,2H,0 oo ca SAT Wee Kaine MgSO,KCLIH,O | 60" j beet Poiykatie | K,SO.MgSO, eer 26256 2,0 cannature [tate | Seth + wo say [Ses £ Hale Rect s 2 shay room Aanydrte | C250, = g - Fae a a | carmaume Pele CNL —apparent limestone porosity c cc Neutron log in shales: composition ple sat [siete ; gees 200m Eyen though there are differences in the amounts of ae combined water between clay mineral species state (Table 10,6), variation in the proportions of these in [etete natural shales probably has very little effect on the aes neutron log, Variations in non-clay minerals are much more noticeable and they dominate the neutron response, Rapid or short-amplitude changes of the neutron- og values in shales mostly result from two causes: changes in admixed quartz content and changes in organic-matter content. ‘Typical quartz-clay changes are seen on the neutron Jog occur in cyclic deltaic sediments Figure 10.19). The neutron is sensitive to the addition of quartz, with a hydrogen index of 0.01, replacing clays with a high hydrogen index (Heslop, 1974). The same result is obtained on the admixture of any material with a very low hydrogen index with clays. This may be quartz as above, limestone, dolomite or certain evaporites: the lower the clay volume the lower the neutron log value. Entirely the reverse effect is caused by the admixture of organic matter with clays: they cause an increase in the hydrogen index. This is because organic matter has ‘a higher hydrogen abundance per unit volume (hy- drogen index around 0.66) than clay (Table 10.6) (Figure 10.20). The increase in neutron values with organic matter is notable, but not as marked as the attendant decrease in bulk density (see Figure 9.18). Evaporites The neutron log can be used to distinguish between evaporites on the basis of water of crystallization (Table 10.7). Figure 10.21 The hydrated evaporite camallite (KCIMgCI,6H,0) identified on the neutzon log. The halite which accompanies tbe carmalite has 2 nevtron value near 20. Of the evaporites with water of crystallization, gypsum (CaSO,,2H,0)is the mostcommon. However, carnallite, polyhalite and kainite also contain the water radical (Table 10.7), All these minerals have a high neutron-log value which differentiates them from other evaporites such as salt (NaCl) or anhydrite (CaSO,), which contain no water and hence have a log value of zero (Figure 10.21) Idemtification of hydrated minerals Hydrated minerals (excluding evaporites) are not com- mon. However, hydrated iron compounds may be sufficiently abundant to affect the logs. Thus a zone of| chamosite (2Si0,Al,0,3FeOnH;0) in the Liassic shales of the North Sea is sufficiently rch to be seen on the neutron log (Figure 10.22). The zone is characte- rized by an increase in the neutron log value associated with a corresponding increase in the bulk density (see also section 19,7, ‘Neutron—density combination Volcanic and intrusive rocks ‘The neutron log is especially useful in the recognition of intrusive and volcanic rocks: most of these show118 THE GEOLOGICAL INTERPRETATION OF WELL LOGS 2.0 BUCK BENGITY a7en9 GAUMIR BAY APT TEGTRON POROSITY UNITS % rn 6 rom) 20m peaks related to mineral hydration igure 1022 Charnosit, a hydrated icon mera of formula 5i0,Al,0;3FeO nll, causing high neutron vals, Note the high density of the chamosite zone. Liasie, Nerth Sea, 5 WEUTRON LOG counts/second 8 e500 2000 15001090500 2 ‘SOUND WATER wt % EP + 2 ‘3 4 DIABASE 400. bound water aq donth (m) 950 Figure 1023, Nevtron response te volcanicrocks with bevnd wate the example is ofa diabase with 1-27 of sulphide minera=sion. 1 has no porosity, “Bound, water values trora analyses of 108% (3m) ‘composites on pulp, (Redfawa from Nelson and Glenn, 1975) hhigh neutron-log values associated with high densities. The high neutron-log values in igncous rocks are due to their high content of chemically-bound water. The bound water may be original or associated with alteration products, mainly clay minerals, which re~ sult from the reaction of hydrothermal fluids with the original intruded rock. ‘An example is shown of an altered diabase with no porosity (Figure 10.23). The neutron log is seen to be responding entirely o the bound water (between 1-47% by weight). The alteration products in this example are diotite, sericite, kaolinite, montmorillonite and chlor- ite, all of which themselves have varying amounts of water (Table 10.6) (Nelson and Glenn, 1975). ‘Similar high neutron values with high densities can be seen ia a series of basalt flows encountered in the subsurface (Figure 10.24) The flows were subaerial, ‘and show weathering and soil development on their surfaces. Alteration products, mainly chlorite, are more abundant at the tops of the flows than at the base. The neutron log values reflect this. 10.7 Neutron—density combination: lithology identification Clean fornsations By themselves, beth the neutron and the density log are difficult tu use for gross lithology identification[GaN AY AT utHoLosy 50 e 25m 50m B= i Po | ie 750 Figure 10.24 Neutron log response to basalt ows. The Nows show 2 Jow gamma ray value corresponding toa high neutron log vale ‘The lithology consists of basalt lows with interbeds of iron-rich lay soil However, once combined, they become probably the best available indicator for the reasons given below. Both the neutron log and the density log should be showing the same formation parameter—porosity. Plotted on compatible 1/1 scales, they should give identical values and it should be possible to super- impose the two logs (Figure 10.25), In practice, this is usually not the case, but clean, water-filled limestones give almost perfectly superimposable logs, as shown in Figure 10.25. ‘The explanations can be taken in two stages. Firstly, the scales of the two logs are made compatible (nor- mally) on a clean-limestone scale. A nentron-log value of zero (no porosity,100% matrix) corresponds to a bulk density of 2.70 g/cm? (the density of pure calcite is 2.11 g/em*), and so on to a neutron value of 100 (100%, Mluid) and a density of 1.0 g/cm* (the density of fresh water) (Figure 10.26). A cross-plot of density-log values against neutron-log values will show a straight-line relationship, a point on the line corresponding to a particular porosity (Figure 10.26). This is the ‘clean- limestone’ line. ‘The second stage of the explanation is that the THE NEUTRON Los 119 NEGTAON POROSITY © is T Nee mene ai wil oan act Se tre q ies : Le Figure 10.26 The density-neutron cross-plot The plot ie necessary ‘ofind real clean formation porosities because ofthe dillering effects ‘of matrix type onthe two logs seetext) Example: density 228 nem, ‘eal porosity 25%, lithology’ clean limestone-ef Figure 1025} (Cross-pot for Schlumberger FDC-CNL logs in fresh waterfiled sandstone, Plot from Schlumberger, 1979) straight-line relationship only holds good for clean limestones because matrix material has variable effects con both logs. A sandstone is seen differently from a limestone by the density log because of a different matrix density (see Chapter 9) and by the neutron log because of the different hydrogen index of the matrix (Table 10.5). On the cross-plot of density-log values against neutron-log values, the clean-sandstone line plots as shown on Figure 10.26. Again, a point on this, line corresponds to a clean sandstone with a particular porosity. In the same way a ‘clean-dolomite line’ may also be constructed (Figure 10,26). For logs plotted on compatible scales, the variations in matrix ar¢ trinslated into a separation of the curves and it is this that is used for lithology identification. A clean limestone shows no separation, while for a clean sandstone the separation is slightly negative and for a clean dolomite moderately positive (Figure 10.27), For a constant matrix the absolute values will change with porosity, but the separation will remain more or less constant (eg, Figure 10.25) If the formation is clean, the type and degree of separation therefore indicates matrix type ie. lithology (Figure 10.27), Shale and shaly formations Clean formations and the ideal reactions described above form the minority of cases. Shale is usually present, Pure shale is recognized on the neutron—density combination when the neutron value is high relative to120 THE GEOLOGICAL INTERPRETATION OF WELL LOGS NEUTRON POROSITY Gimestone matrix) BULK DENSITY jensity| shale low porosities) porosity pimestone) 25m (righ porosities) 50m Figure 10.25 Neutcon poresty Jog and bulk density log plotted on compatible scales. The neutron porosity is displayed with a scale for Timestone matric the density fora mattix of 2:70 glem® (—2er0 porosity) The two logs fellow each otber closely oves the limestone interval Example: g = 25% Bulk density = 228 jem, See Figure 10.26 for eross-plot position. 5. NEUTRON POROSITY HOEK % =a 4296.20. 24 1812-8 0-8 =D Ay St BR ht E SS a1 22 SEO O SS 27 20 Ve" SapSraTOA| shale SANDSTONE 18%, 1 [te separation LUMESTONE s188 DOLOMITE 15% shale Figure 1027 eulzod newtyoo-deusty log combi water filled porssity Jo tesponses. The Ngure shows clean sandstone, limestone a dolerite, all with 152%the density value, It gives a large positive separation to the logs, the neutron well to the left of the density. This separation is typical and diagnostic (Figure 10.22)and is due to the high hydrogen index of shale matrix material (gee ‘Neutron log in shales’, above). If shale becomes diluted by matrix grains such as quartz or calcite with low hydrogen indexes {Table 10.5), the neutron-log value decreases rapidly. Such a change is not seen so markedly on the density logsince the matrix density of shales (2.65-2.7 g/em*)is similar to that of quartz and calcite (2.65-2.71 g/cm?) On the log combination, the result is a decrease in the ncutron-log value and a decrease in the log separation. ‘The decreases continue until clean formation values are reached (Figure 10.28). Ideally, the changes from pure shale to clean for- mation are progressive on both logs as the volume of shale decreases. The relationship can be considered as roughly linear. A 50% shale mixture should thus show 50% of the change from pure shale to clean formation. Qualitatively large or small separations can be con- sidered to indicate more or less shale (Figure 10.28). In practice, small separations ic. slightly shaly for- mations, tend to be related to low neutron values, while pure shales show large positive separation and high neutron values Used properly, the neutron—density combination is the best log indicator of shale. t allows a more reliable HEUTRON 4, POROSITY; gtimestone)o] THE NEUTRON LOG 121 indication than the gamma ray log and, at least qualitatively, can be used to evaluate the degree of shaliness (Figure 10.28), Used thus, asa shale indicator and with typical known separations in clean for- mations, the neutron—density combination can give 00d idea of lithology in almost all normal formations. Unusual lithologies and minerals Although knowing the log separations expected on the neutron-density combination from clean formations and shales allows the construction of a gross lithology, there will be cases where the separation is anomalous and cannot be explained by simple, typical lithologies. For example, the two logs generally move together, a higher density corresponding to a lower neutron-log value. When a very high density corresponds to an even higher neutron-log value, a simple lithological expla- nation is not possible, Such a case is that of the chamosites in the North Sea Liassic (Figure 10.22). The increase in neutron-log value is caused by the chemically-bound water in the chamosite, while the deasity increase is because the mineral is dense (6.03 g/cm"). The large separation and extreme values are very characteristic. Some similar unusual separations which may be diagnostic are shown graphically in Figure 10.29. feo Figure 1028 Changes in the neutron-density combination separation dv to changes in quartz-clay admisture. tn the two coarsening-up, ‘tise sequenors shown, the quartz content increases upwards relive (0 the sale, The separation changes are due mainly tothe changes in neutron value (compare Figure 10.19,122 THE GEOLOGICAL INTERPRETATION OF WELL LOCS waren 19 7 # 77 % “ye 3 “ge 20 fons 4% Onsite ye Vf APPROX. CML POROSETY _(eandstone) 180% APPROXIMATE DENSITY RESPONSE £ 5.0 (pyne i ED haematite Figure 10.29 Neweeon-desity plot with tentative locations of some saree oaity non-quarte materials. (From Ranson, 1977) Evaporites Evaporites are also distinctive. Since their densities see Section 9.6, ‘Mineral identification’) and also their hhydrogen indices (see above, “Evaporites’) may be diagnostic, evaporites become very distinct by the combination of both logs. The absolute values of both jogs and the log separation (Figure 10.30) are both indicative. Conclusions The neutron-density combination is the best lithology indicator for most formations. Shales and shaliness and evaporites can be identified, clean formations and even matrix type can be suggested, and unusual minerals located with the possibility of identification. ‘Neutron and density values can also be used quanti- tatively for lithology identification. The method is described in Chapter 11 (see ‘Cross-plotting com- patible logs’, section 11.6) . NEUTROW POROSITY UNITS =a & [Se oo so ee 18 2 8 Oe 8 eri 2 a eae ia oS fies BULK DENSITY gfeme 55 | z |z0 2 Ope 2 26 27 2 wostone 6=0 FREE] rats SALT = HALITE 1 ANHYDRITE: 1 1 CO heey aucoot REA) evesu aaa C 1 Ieee ’ tte S cl _SAMMtiTE 1.87 gien® eoP oP yp) POLYHALITE, limestone d =0 Eaimundson and Raver. 19793 re 10.30 Idealize! neutran-density jog Combination responses in series of pure evaporites (For FDC-CNL Schlumberger, Figures (ror11 Lithology reconstruction from logs LL.1 Introduction ‘There are two independent sources of lithology data available from oil wells, one set of data coming directly from the drilling and one set from the wireline logging, ‘The drilling data consist of cuttings, cores and all the recorded drilling parameters. The logging data consist of the wireline log suite and sidewall cores, For a reliable lithological reconstruction, the two sets of data are essential. As a result of the great sophistication of wireline logs the drilling data are often forgotten. This should never be the case since the only continuous sample of formation lithology comes from drill cuttings, ‘This chapter describes the methods for interpreting lithology by the manual method using log and drill data, semi-automatic methods using mainly logs and fully automatic methods using only a computer and logs. 112 Lithology from drill data—the mud log ‘The mud log and the way in which it is made is described briefly so that the data it represents can be | RETRO used properly in log interpretation. Drill-derived data and log-derived data often appear to be in’ conflict. Which can be believed? This book describes how logs can be used and this section describes how drilling data can be used, which are reliable, and some pitfalls. ‘The mud log is the geologist’ record of the drilling of a well. Before logging was invented, it was the only record that existed. On this log is recorded the lithology, the drilling rate, bit changes, gas record, calcimetry, dates and events (Figure 11.1). The li- thologyis based on an examination of cuttings—small chips broken off the formation as the drill advances. They are washed away in the stream of drilling mud and brought to the surface. They do not usually ‘float’ in the mad, but are pushed upwards by friction and drag. At the surface, the drilling mud is passed through. a large metal sieve, the ‘shale shaker’, and the cuttings are recovered. It is from the shale shaker that the geological cutting sample is taken. On the mud log, lithology is usually recorded in percentage of cutting types in a particular sample, say 10% sand and 90% brown shale, New samples are taken every 2 to 20 metres, depending on the rate of drilling, Fast drilling rates of perhaps 1 minute per ‘SLE UO] TOTAL GAS [EALEMETAY MOD pescrrnos sha,” x ors, 2 1 6 18] a0coe090 [easing SST gy me smateix carb. por ‘SH as above scr as SLST th ay. cat le, non-esle SST a, ab, ors [calcinotry onaivela ‘SH ayn, ben. tem 10% dolomite subtle. ol sity DOL. tees. mas. yet Figure 11.1 A typical mud log. The logis the well-site record of lithology (eutings) and some drilling parameters, J this example eutings samples were taken every 2-1,124 THE GEOLOGICAL INTERPRETATION OF WELL LOGS metreat the top of the hole allow only one sample to be taken every 20-25 m drilled (ie. 1 sample per 20-25 minutes). At greater depths, rates of 30 minutes to drill 1 metre of formation are common, and a sample can be taken every 2 metres (je, 1 sample per hour) oF even. closer. To reconstruct the mud log, the time it takes @ sample to seach the surface after being cut, the ag time, ‘must be calculated. A sample drilled at 3000 m will take perhaps I hour and 10 minutes to reach the surface. It will be travelling at 43 metres per minute up the hole (about 2.6 km/h), the exact rate depending on the rate at which the mud is ‘circulated’, that is pumped through the mud system. The drilling rate is used as the basic curve for the mud log (Figure 11.1) and is presented as real depth. A metre drilled is instantly recorded, but the ‘cuttings sample is tied to the drilling-rate log, so the lithology corresponding to the depth 3000 m-3002m will only be recovered at the surface 1 hour and 10 minutes after it has been drilled. ‘The same principle of lag time is applied to gas readings and to shows as to cuttings. ‘On most mud logs the geologist has recorded not only an analysis of the percentages of the cutting types present, but also what this means in terms of real subsurface lithology. For instance, if a sand-shale— coal sequence is being drilled, all these lithologies become ‘smeared’ while travelling to the surface. In the cuttings sample will be seen 50% shale, 40% sandstone and 10% coal. The geologist will then makea ‘guess’ at the real lithological column using the drilling rate, Knowing that shales will drill slowly, sandstones faster (Figure 11.1) There are occasions when the drilling rate can be compared to a sonic log or a gamma ray log (Figure 11.2) and a good interpretation of bed boun- daries can be made from it. However, in general, the drilling rate involves too many variables, such as (GAMMA RAY API Ss BeOS RATE nT 2400 samme ray, \ 2are api rate * 2ans| Se = fig data as inFig. 114 @=t0e oma Figuee 11.2 Comparison between rig-derived lithology interpretation and subsequent log interpretation, Note he fiet o es atc Only "good guess Iivhologs: on the eutings percentages. The rg bed Dou: ‘emeared’ bed boundaries on cuttings log end “delayed! arrival ‘of new lithology Lora new‘weight on the bit, bit wear, pump ratesand so on, for it to be an accurate boundary indicator. The mud-log interpretation of lithology should not be used to interpret bed boundaries on the well logs. It is important, therefore, to use the cuttings per centages in the right manner. As described previously, 4 rock cutting from 3000m takes over one hour to reach the surface. During this time it becomes mixed with other cuttings taken at shallower depths and moving morestowly up the hole. It also becomes mixed with chips washed from the well walls, higher up the hole, so-called ‘cavings’. By the time they reach the surface, samples are therefore considerably mixed and heterogeneous. An experienced rig geologist will usu~ ally recognize cavings and eliminate them from his count. The mixture of cuttings, however, from the various lithologies is the reason why percentages are recorded. All lithological boundaries have become very blurred, When interpreting the cuttings logs, it is the arrival ofa new lithology which is significant. During drilling, froma thick shale into a thick sandstone, when the bed is actually penetrated only a small percentage of the cuttings will be sandstone, This increases, but there will be 80% sandstone only several metres lower (Figure 11.2), The drilling rate will correlate with major lithological changes—the so-called “drilling break’. Gas levels are also likely to change. Some pitfalls Certain lithologies are notorious for appearing on mud Jogs, but not in the cuttings samples. Suchis the case for loose sands, silts and soluble evaporites (eg. salt) ‘The mesh of the sieving ‘shale shakers’ is such that loose grains of sand or silt, even coarse sand, pass through the mesh. If this is suspected, the mud may be diverted through de-sanders, where all small grains are extracted. However, de-sanding is not routine and there are many cases where unconsolidated sand reservoirs have been drilled and shale has been re- corded on the mud log! Salt is a very similar case, Unless the drilling mud is salt-saturated, no cuttings will be found on surface. Shale, mostly cavings, will be recorded. An experienced rig geologist, however, will note mud salinity changes along with drilling-rate indications which suggest the presence of salt. ‘The exact opposite exists where lithologies which seem to appear on the mud log do not actually occur in the formation. A typical cause is the use of lignosul- phonate, a mud additive which reduces water loss. It resembles lignite and has very often been interpreted as such on logs, but this interpretation is suspect if it implies the presence of coals in pure shale intervals: realcoals mostly occur in zones of both sand and shale, Despite these various anomalies, the mud log is essential to the lithological interpretation of wireline LITHOLOGY RECONSTRUCTION FROM Logs 125 logs. As previously indicated, it represents the only continuous record of real lithologies (except where cores exist: see below). However, because of the way in which the samples are collected, lithologies have become mixed and bed-boundaries smeared, Wireline logs can be used to separate the mixed elements and define the bed boundaries. 11.3 Lithology from cores—direct physical sampling Cores may be cut during drilling, when a continuous, cylindrical sample of the formation is recovered, or they may be taken after drilling, when small, punctural samples may be taken from the borehole wall. Cores cut during drilling During drilling and before logging, when a complete record of lithology is required (for example in reservoir) a continuous samaple is taken by coring. The rill bit is replaced by a core barrel. The retrieved core, depending on the preceding hole size, will bea cylinder of rock 5-15 centimetres in diameter and up to 20me- tres long. Itis a direct physical sample of the formation. Being a real physical sample of the formation lithology, a core appears to need no interpretation, In fact the teverse is true; cores should be used as a reference to calibrate the logs from a lithological point of view and to compare with the lithology interpreted from the wireline logs. Indeed, such com- parisons are used frequently through this book: they are essential to a proper understanding of the capabi- lites of wireline logs. Cores provide the geologist with the only record of real subsurface lithology. Cores do in fact need interpretation and processing before they can be compared to logs. The principal problem is one of depth. Cores are cut during drilling, so that their depth limits are calculated by adding all the lengths of drill string together. Mistakes often occur, and frequently these depths do not agree with thedepths shown on the well logs. The logs re taken as the reference: for detail the reference may be just one log, frequently the sonic or the density log. The drill depths for a core must therefore be adjusted to log, depths. The changes are usually about + 5m but may bbe as much as 15m. Moreover, because of recovery problems, the depth changes between contiguous cores are frequently different. (Minor miscalibrations within acoreare also sometimes seen because of cable stretch, but in most cases these can be ignored.) Even with the core depths perfectly matched to the log depth, inconsistencies between the two may arise. It is at this point that the capabilites of the logs emerge. ‘The problems of bed definition (see Chapter 2) become. clear, as do those of depth of investigation (also Chapter 2)and what this implies in terms ofthe volume ‘of rock sampled by a log, In extremely heterogencous126 THE GEOLOGICAL INTERPRETATION OF WELL LOGS “GAMMA RAY APL ‘CORE DATA rain se By DETAILED RECORD a) 1:20, COMPLETION Log 100 seamen cates eH wile ding, The natural dtl of acre geeraliel 8 the sedimentologial Jog at 1.200 scale. The sedimentological detail is generalized to lithology only at the 1.500 scale SIDEWALL SAMPLE SIDEWALL SIDEWALL SAMPLER, CORE 2.5em grey-brown claystone formations, the directional tools such as the density- log tool may be difficult to calibrate because ofchanges actoss the small distance between the core and the borehole wall. These particular problems are discussed in the chapters on individual tools. ‘The level of detail at which corescan becompared to logs is an important aspect of core utilization, A core. derived sedimentological log should already bring the natural detail of a core to a manageable level for comparison with the logs, at say a scale of 1:200 (Figure 11.3), Reservoir, sedimentological and calib- Zation studies can all -be kept at this scale (see Chapter 12), However, for general stratigraphical work and the completion log, (sce below) much less { ‘, Fay = [2 Si $ sample depth ums, Fiwe I Sideval core sample. Schemati stration of «sidewall spl. sidewall cor and its record on @ lithological top. Logs etait is desirable: 20m of core is represented by only em on the log, so that considerable lithological generalization is wsually necessary (Figure 11.3) Generalization is a skill learned through experience and it must give clear and concise, but accurate, results ‘The natural lithological detail of a core is not ‘mimic. ked" by a sedimentological log. The latter should be a stylized representation of a core, an implied interpre- tation having been made. This sedimentological detail remains at a scale of 1:200 even though there has been 2 200% decrease in scale, For general work this detail is usually lost, and at a working 1:500 seale the only detail eft over a cored interval isa slightly more precise and accurate lithology (Figure 11.3),Cores cut after drilling Several methods are available for core sampling once 2 hole has been drilled and logged. All of them involve cuttiig into the borehole wall. The most frequent method is sidewall coring. A sidewall ‘gun’ is lowered into the hole on the logging cable: it consists of a series of hollow cylindrical LITHOLOGY RECONSTRUCTION FROM LOGS 127 ‘bullets’ 18cm in diameter and 2.0-3.0em long (Figure 11.4). The ‘bullets’ are arranged in series along, the sidewall tool, The tool is run to total depth, depth- calibrated with a gamma ray tool, and then pulled up the hole, The sampling points are decided in advance and are based on an inspection of the logs already run ‘When a sampling point is reached, the sidewall tool is stopped exactly at the depth chosen and a ‘bullet’ is [;NoueTION orn BULK DENSITY gfe GAMMA RAY CALIPER * xz 24 ae pe es E [srt ‘ola ’” | sontcastt APL 20010 20 24 2001240 40) PHI NEUTRON % P.O, 8 7 ] i 1800 k | 1850 I. | eT s900 [I ee E Figwre 115 Wellogcomposite Allogsrun over he stme inervalarereplowted together. The compas forms the bse documentfora manual interpretation ofithobgy. (GAMMA RAY API feo. _ NEUTRON ® 109 0 ‘cuTTINGS CALIPER (aid ‘BULK DENSITY 0% tod es ees 2400m % => suits @ Figure 11.6 Horizontal routine, manual method. The stage fo building up lithology: I, rig data; 2, simple wiceline log; 3, all logs and ws ‘amples (only neutron-densty shown) ‘The lithology s built up from I to 2 to 3, but with reiteration. Nete the inereasing detail and preciso128 THE GEOLOGICAL INTERPRETATION OF WELL. LOGS fired, A smal directional charge shoots the hollow steel sidewall bullet into the formation wall which it penet- rates by force. The sample it recovers is, as indicated, [bout 18cm in diameter and up to 3om long, This sample is a reliable indicator of lithology (depending fon the accuracy of the calibration), but because of the Sampling method the rock is frequently shattered, so that the petrological characteristics are destroyed Porosity measurements, for example, should not be made on these cores, as the grains are usually shattered, ‘A new tool (the Hard Rock Coring Tool of Gearhart) ives a similar size of sample, 24m in diameter and 44em long (if x 13 in), but has a revolving sampler fand so avoids shattering, Samples from this tool ean presumably be used for porosity measurements ‘Sidewall coring as a method of lithology sampling should be used essentially for verification. As the sample is so small, interpretation problems can arise, and sidewall core results should be used with care. In sands with shale laminae, for example, a sidewall may fall ina shale lamina and it will not be representative of the zone as a whole. For this reason, in reservoirs, & Closely-set series of samples is taken. The obvious advantage of a sidewall core is that its depth is known dand it ean be taken in 2 specific, chosen lithology. 11.4 Lithology interpretation from wireline logs— manual method “The manual interpretation of lithology from well logs GAMMA RAY APL 0 “00 TEUTTON FOROS, should be undertaken only using all the logs registered, Now that logs are digitally recorded, all the runs from a well can be se-plotied by computer to give one eSmposite log (Figure 11.5). Thisis an extremely useful Gocument, The final lithological interpretation may ‘appear on this composite log or, €0 avoid over: chutering, may be transferred to a document with only the logs usually used for correlation. This is often the tgamma-ray (or SP) and a resistivity Tog, oF the gamma Tayand a soniclog (Figure 11.8) The original litholog\- tal interpretation, however, is made on the composite document showing all the logs. Horizontal routine There are no simple rules for the quick manual interpretation of lithology from logs—if there were, this account would be superfluous. This book outlines the capabilities and characteristics of each of the open; hhole tools. To interpret thology, these capabilities and ‘characteristics must be known and used, A systematic proach is best. Thus, the eross lithology is suggested by the mod log, this can then be corroborated aid compared at the same depth, horizontally, to a simp\e jor such as the gamma ray or the SP (Figure 11.6.2) ‘The. interpretation is then continued, again horizontally, through the other logs—resistivity, so- mnie and density-neutron. If all corroborate the sam jnterpretation, the lithology can be noted and then INDUCTION ohm fy 24 a CAL oy sig 8] ‘BULK DENSITY gene, he ‘a2. 4 Cc se. cal, SHALE SAND sl. shaly SAND clean aL SAN shaty ‘SAND cloan ssp clean SAND >| on igure 11.7 Vertical ror ‘ 1 rama mathod, Base ne and curve separains ate wd for she vet anai inthis end shale saocompared to sidewall cores or other samples (Figure 11.6,3), If the lithology is not corroborated, then there must bea feedback’ from one log to the next. ‘The first aspect to check is that of log quality. The hole may be very caved, one or more of the logs may be badly recorded, and hence the readings are anomalous. These aspects are considered in the chapters on the individual tools: However, the anomaly may be within the lithology itself, For example, ina sand-shalesequence, there may be 40% sand and 60% shale marked on the mud log. The gamma ray log may read persistently high, so that only shale is suspected. The resistivity log and the sonic loz LITHOLOGY RECONSTRUCTION FROM LoGs 129 are not diagnostic, but the density-neutron com. bination shows obvious shale with either sandstone, limestone or dolomite: sandstone is indicated from the mud log, The sandstones are then marked on the thology log and compared to the mud log or sidewall samples. A check with the SP shows that the neutron~ density indicated sand intervals correspond to per- meable zones, and that in turn these have mud-cake indicated by the caliper. The anomalous log, therefore, is the gamma-ray. From the interpretation it can be concluded that the sandstones have a high gamma ray count because of included feldspars, micas or other non-shale radioactive elements. The manual interpre- ‘Table 11.1 Logging-tool response values. All values are shown so that they cover the range of values found in the various sources, noe of the sources has values outside those ised above. From Serra (1972) Serra (1979), Dresser ‘Aas (1985), Gearhart (1983), Schlumberger (1985) Density Neston Resistivity Gamma ray tg to porosity Material {ohm mtn) API sity? tafe?) (pa) ite” ‘Sandstones upto 1009 18-160 33-100 Deze Oa Limestones 80.6510? 16-100 16-53 266-27 0-30 Dolomites 17x10? 12-100 383-45 -28299 0-30 ‘Shales 05-100 41000170 265-27 25-75 Quan iio? 512-56 264-266 Caeite 10-10% 455-3 20 Doloaite 1710 35455-2808 . Tite 250-300 22.30 30 ‘ nE— Chlorte 130-250, 232 82 EE Kaoliite 80-130, toe 7 OE Siectce 150-200 20-30 a ¢— Glauconite 1390 2228 38 & Nussonte 1oN102 M28 236-31 20 2 Bowe 110!" 90-275 508-31 Dest & Microcine moi 8 zat 3 2 Orthoctse mor 2326-3 = Anthracite ws 0 90-120 eis 38 g fumious coal 10-19 0-18, ooo LIST Lignite 4x10? ou io oS 32 Gas (es) « ° 00088 3 Methane = ° 026 “oon Sonar APH or—t0# —oD-04 238 a « 3 Water °F) 3. pare « ° 189-207 .00(1419 100 & alt (3000ppm) ——03t a0 180 he « 2 Pyrite 10-10 382-39 aes 3 E224 Serie lot—1000 0 a 30-339 2 3 Halite interoods of NOSTORE, ‘l “CLAY, grey con! BATHONIAN TP aS Bata SAND FORMATION [BAYOSTOW negiumt-eosrso OF arained, mic. ora} arey-beowl [Poteet ab one enone igure 1-8 The completion log. An example ofthe log with Sater thon, siraphy, ydrocarbs: shows. tests and dil data. Kise svlogis’s “basic eoard” of well, The lithology comes from sn interpretation of the log composite cores and drill Gata,ragquency « Figure 11.9 Histogram of log values. This histograsm fy of gamma ray values over 100 m of formation, The sand and shale populations are indioated. Between them is the zone of shaly sands and sandy shales, 11.5 Computer aids to lithology interpretation A welHog curve represents a series of quantitative values, each value derived from the formation. The formation, and hence the lithology, can be quantified. For quantification to be an aid to or the tool for interpretation, however, there must be a grouping. That is, particular set of values must be shown to represent a particular lithology. The lithology will be defined numerically, rather than by its subjective appearance to the geologist through cores and cuttings. ‘This quantification may be made at several levels of sophistication: one-log, two-log or multi-og. The most, sophisticated multilog quantifications can give an entirely computer-derived lithological interpretation Histograms—one-log quantification. The simplest way of grouping well-log values is by using @ histogram, where the log value is plotted against frequency (Figure 11.9). The histogram bas various uses. It can be used to define populations or average values. For example, the ‘shale’ and ‘sand’ values of a gamma ray log may be presented in this mannet (Figure 119). A second example shows a gamma ray histogram ina series of voleanic rocks with values forming distinct populations (Sanyal et al, 1980) (Figure 11.10). Used in this way, the histogram helps to define the log limits of lithology and average log values (Walters, 1968). A second use is in the normalization of particular logs. This is done by selecting a consistent stratigraphic interval and comparing the log responses by compar- ing histograms of the log values (Figure 11.11) (Kowalchuk et al, 1974) This is done as much for petrophysical as for LITHOLOGY RECONSTRUCTION FROM LoGs 131 Figure 11.10 Histograms of gamma ray log values. Gunma ray values from a volcanic sequence showing distinct papwlations for various lithologies. (Redrawn from Sanyal et al, 1980) gcological reasons. In this same manner, histograms of Jog valucs may be compared to histograms of labo- ratory values when the same character is being mea- sured (see Chapter 9} In the petrophysical usage, logs and laboratory values are made consistent for reservoir calculations. In the geological usage, slight lithological changes may indicate facies trends (see Chapter 12) Finally, histograms have been used to make stratig- raphic correlations when curve comparisons were inconclusive (Walters, 1968). The method seems dif- ficult to apply and to envisage (Figure 11.12). However, itdoes illustrate that a particular formation may give & certain shape of histogram, the advantage being that depth differences are eliminated. Cross-plots—two-log quantification When any two values are cross-plotted, the resulting series of points may be used either to define the relationship between the two variables, or to define 1 I 7 ts veanaon | 5 & 20 i | = 10 5 AM, ul oof o2 03 o4 os oe Figure 11.11 Comparison to 2 master histogram, Newton and density log value histograms of one well compared to master Listograms (contiouous Tine) of six wells trom the First White Speckled Shale. Upper Cretaceous, Alberta, Canada. (Redraw from Kowalehuk er al, 1974)132 THE GEOLOGICAL INTERPRETATION OF WELL LOGS, il Manse —— @=— Walters, 1968) fields, using both x and y axis values, giving the upper and lower limits of both variables. Roth the above outputs are used in cross-plotting well logs. ‘Three types of well-log cross-plot exist: (1) Cross-plots of compatible logs, that is those measuring the same parameter, for example porosity logs Cross-plots of incompatible logs, for example a plot of gamma ray p, neutron values—the logs do not measure the same parameter ‘Cross-plots of laboratory or sample values against Jog values, for example porosity core values against neutron porosity values. 2 eB) Cross-plotting compatible logs ‘Typical and illustrative of this type is the neutron— eo Figure 11.12 Stratigraphic ws of histograms, Swatjgraphie zones character ad by histograms of log values a8 an aid to correlation, (From density cross-plot: the plot of neutron porosity values, against density porosity values. The method has been explained previously (see Chapter 10, ‘Neutron density combination’), It was shown that cross- plotting neutron and density values can be used to identify pure matrix and/or the related porosity. This is impossible using only the value from one of the logs. For example, a neutron log value of 25% may be a dolomite with 17% porosity, a limestone with 25% porosity or a sandstone with 29% porosity (Figure 10.26), but when associated with 2 density of 2.28 gfom! it has a unique attribution, aJimestone with 25% porosity. This cross-plot uses a straightline relationship between the two variables to quantify the desired characteristic and to identify lithology: It is the simple case with only two variables, in this example porosity and matrix (lithology). When shale, for example, is DENSITY ofen® 20 NEUTRON Figure 11.13 Shale point dined ona esizon density roepot. Vales ‘point’ to represent shal=. 2 ™ POROSITY som 200 m ofa sand-shale formation, Note the arbitrary choice of neLITHOLOGY RECONSTRUCTION FROM LoGs 133 BULK DENSITY fen? NEUTRON POROSITY Figure 11.14 Lithologial fields defined on a neutron-density crossplot, 1 is based on the values from a 500 m interval in one well added to this system, a more usual combination in nature, a solution for three variables must be developed. On the cross-plot being used as an example, when three variables are present a plotted point will not give a solution—the grid is defined only in terms of two variables. Thus the point uscd above, density 2.28 g/em* and neutron 25% can only be interpreted as was shown, in terms of pure end-members. Ifshaleis added, then a shale end-member with neutron and density values must be defined, For this plot, this is frequently done; a neutron— ity ‘shale point” is defined empirically. A great many Wihologially ‘unidentified points are plowed chosen to correspond to the extreme value but within the supposed shale field (Figure 11.13), Once the ‘shale point’ is defined, the cross-plot can be divided into a compositional triangle, the end-members being shale 100% (shale point) porosity 100% (uid point, ie, neutron porosity 100%, bulk density 1.0 g/cm*) and matrix 100% (matrix point, ie, for limestone, neutron porosity 0%, bulk density 27 g/cm"). Any point on the plot now has a precise value of the three variables. This ‘shale point’ cross-plot has many drawbacks. Firstly, only one matrix can be considered at a time. A ‘zone will be interpreted as only shaly sandstone or only shaly limestone—never both. But more importantly, it mixes definable with undefinable values. Shale is inevitably very variable, and the shale point therefore very imprecise (Figure 11.13) yet the matrix and liquid points are both quite precise. ‘A more realistic approach is to define fields of values on this plot in which a particular lithology is likely to be plotted. The approach is empirical and the log limits ofeach lithological field wil vary from well to well, and even within one well with depth (Figure 11.14). However, the approach docs help towards a more rigorous, numerical definition of lithology on the logs, especially in zones of rapid alternations of lithology. ‘A sophistication of the cross-plot method of li- thological quantification is to cross-plot cross-plots. This is the basis of the Schlumberger M-N plot (Burke c Figure 1115 Crose-plot of cross-plots. The M and plat from ‘Schlumberger, (From Burke etal. 1989,134. THE GEOLOGICAL INTERPRETATION OF WELL LOGS, oni = logyy resistivity gamma ray API units Figwe 11.16 Cross-plt of incompatible logs. Gamma ray and resistivity values eros potted to define lithology fields. (From Heslop, 1974) et al,, 1969).N is defined by a density-neutron cross- plot, M is defined by a density-soniccross plot, and M is plotted against N (Figure 11.15). Theoretically, mi- nerals and shale become separated into fields, and porosity scliminated. However, the geological vitlue of the logs is lost and a mineral identification more reliably done using mud log and simple cross-plots. In fact there isa tendency to rather obscure cross-plotting in the vain hope of finding a unique ‘shale point’ or “mineral point’. These points rarely exist in nature, except for ‘chemically-precipitated deposits. In this respect, certain minor minerals which usually oceur in the pure state can be identified by their end-points Generally, these minerals tend to the extremes. Cross-plotting incompatible logs Thecross-plotting of incompatible logs is usually done to quantify lithology. Incompatible logs are those which do not, in the first instance, measure or indicate the same parameter. Resistivity and gamma ray logs areincompatible, one gives the resistivity, the other na~ tural radioactivity and, by inference, shale volume. However on cross-plotting, compatibility will become evident if there is any), The resistivity logs, for instance, will show a consistent set of values in shales, as will the gamma ray log; this will become evident on cross- plotting (Figure 11.16}. In fact in almost all cross-plots, it is rare not to find a consistent relationship of some sort (cf. Heslop, 1974). Often empirical relationships become evident which otherwise would remain hidden. Plotting the gamma ray log values against the neutron log values, for example (Figure 11.17) brings out séveral relationships (Rider et al, 1979). There is a consistent, straight-line relationship between the two where both the gamma ray and the neutron logs are reacting to a shale-sandstone mixture. Each log is showing the volume of shale in its own-way. In cleaner sandstones, the gamma ray logcontinues to react to the shale but the neutron log is now Teacting also to porosity, and the cross-plot relationship changes. The shale field can therefore be defined by the break between the cross-plot relationships. Alone, the neut- ron log is unable to differentiate between a shaly sand with pore-water and a wet shale with enclosed water. ‘The gamma ray log alone cannot identify porosity. When cross-plotted, the limits become clear (Heslop, 1974)Figure 11.17 Trendlines roma gamma ray-nevicon porosityceoss- plot. The onset of porosity i visible, and slso the influence of Abundant organic matter. (From Rider et ol, 1979.) On this same plot (Figure 11.17), at higher gamma ray and neutron values there is also a relationship break. It is due to organic matter, Since the neutron, tool reacts to all hydrogen present (Chapter 10), it reacts to the hydrogen combined with oxygen in organic matter (a solid hydrocarbon), In the example, the organic matter is mostly coal and lignite grains with low radioactivity. Thus, while the gamma ray values diminish as the organic matter replaces the shale, the neutron values increase or remain high (Rider et al, 1979). This same gamma ray/neutron plot can also be used to identify anomalous shale zones. A straight-line relationship between the two log values showsa typical shale-quartz. mixture. Deviations indicate a new ele- ment. When high uranium values increase the radioac- tivity of a shale, for example, the gamma ray log will react, but not the neutron log. There will bea deviation from the straight-line relationship. Cross-plotting log values against sample values The first use of this technique was to verily log calculated values of porosity against those from the laboratory (Figure 11.18). This, obviously, can only be done over cored intervals. The cored zone is then used to calibrate the logs, and the normalized log values applied more confidently to zones without cores. This technique is essential to log interpretation. However, when it is applied it must always be remembered that measured values and log values are not a priori identical. Core porosities are measured in a small plug about 10cm? in volume, logs usuall; give an average of about 1000-15 00cm? of formation (ie. over 1000 times the sample volume). Core porosities are mea- sured under atmospheric conditions, log porosities under reservoir conditions, notably of pressure and temperature. A persistent difference between log and core porosities of | or 2% will often indicate a difference LITHOLOGY RECONSTRUCTION FROM Logs 135 ux DENSITY (405) sien? Figure 1118 Crosr-plot of log values (density porosity) against sample values (core porosity). The plo shows the limits of accuracy to be expected from log values in physical conditions, rather than badly-calibrated logs. Sample calibration may also extend to lithologies. For example, the validity of the gamma ray log as a shale indicator can be checked against laboratory measurements of clay percentages (Heslop, 1975) (Figure 11.19). The amount of ash in a coal can be compared to its bulk density as measured by logs (Lavers and Smits, 1977) (Figure 9.24). This technique of calibration for lithology is not used enough for geological purposes. Cross-plotting log against laboratory values means that the position of the various points used is lost. This is a statistical comparison; an individual laboratory value is not compared to an individual log value, and represents the real advantage of the cross-plot: it ‘compares averages, In order to re-introduce individual readings defined and identified points from a core can be re-plotted on a cross-plot of log values. For example, in order to identify lithological fields on a neutron-density cross-plot, lithologies defined from a core.can be plotted and labelled on this cross-plot grid. Thisisa process once removed from plotting the values ona depth scale and comparing with the log. 11.6 Multi-tog quantification of lithology ‘Two typical methods for the multi-log treatment of logs will be briefly described below. Many methods exist, so that mention here is only by way of illus- tration. The first method described is used essentially by the petrophysicist: it is designed to indicate hydrocarbons, and lithology is a secondary consi-136 THE GEOLOGICAL INTERPRETATION OF WELL LOGS 20: 0 CLAY VOLUME FROM X RAY a a) 3 CLAY VOLUME FROM LoGs a0 Figure 11.19 Crose plot of laboratory values (ofelay volume) against log values (of clay volume). The plot is a partial verification ofthe log Setivation of shale volume. (From Heslop, 1975) deration. The second method is principally designed to indicate lithology. Petrophysical multi-log analysis Any multi-log quantification of lithology or, for that matter, hydrocarbons, must begin with a definition of what is being sought. Thus, if we are searching the log values for oil, oil should be quantified on each log, In fact, in the oil search itis not possible to give log values, to oil, so this is done for water, its antithesis. Any calculation for oil can only be done using defined values of formation-water resistivity, formation-water density, formation-water neutron value, liquid re- sponse to the sonic log, and so on, In reality, character definition is progressive and built up from one log to another. For example, porosity may be first defined using the porosity logs. To define porosity the matrix characteristics must be defined. The porosity thus defined is then applied to a resistivity log which cannot define porosity itself, but ‘once porosity is defined, can give the water saturation, For this the water resistivity must be known. The system is reiterative, and ‘best fit’ results are found (Figure 11.20). This form of methodology is the basis for many of the logging-company systems, It relies essentially on the definition of certain absolutes such as ‘shale point’, ‘matrix type, and formation water resistivity. Even so, the number of variables that can be manipulated is perhaps three or at most five. A typical system, for example, will define clay, quartz and porosity volumes, over a certain interval (Figure 11.22). A more complex system can define clay, silt, sand and porosity volumes, or clay, feldspar, mica, quartz and porosity volumes. The output, however, is always in volume percentage of the defined elements. The log is usually called a CPI (Computer Processed Interpretation). This sort of output can be criticized from a geologi- cal point of view as being dependent on artificially defined absolutes which have little relation to lithology in the usual sense. A sandstone is not defined by its quartz percentage: it has a compositional and textural definition. The output of these computer-defined litho logies' in percentage of constituents does not, therefore, represent geological lithologies. Statistical mutti-log ai An entirely diferent approach is to use statistical methods to cope with the enormous amount of data generated by logs. The first method considered is that of discriminant analysis (Broggi et al, 1983). Each depth point is represented by a certain number of variables (i. log values). The whole set of depth points Ge. the well) then represents the population to be examined. All typical lithologies are identified man- ually at reference levels in a well, preferably whereaa iweuT WELL PARAMETERS [CORRECT LOG ATA] [DETERMINE Parenox FROM N-D PLOT] NO Rr FROM OIL OR OLL eres eicinaca Ranta DETERMINE Ru 8 Amt FROM Rwo & Rita CURVES PRE-INTER- PRETATION ‘SELECT INTERVALS BY LITHOLOGIC UNITS AND. HYDROCARBON TYPES EVALUATE INTERVALS SELECT ay caosseuors PARA- Litlogy Poesy Type, METERS iyeacarben Etveaecay Params END z MODEL [SELECT LITHOLOGY oPTIOR| Losic FoR INTERVAL EVALUATE SH ALINESS AND correct Py, Pb, At OBTAIN Ry SONPUTE|_ {CORRECT FOR LiGHT] Syo Shr | “L_HYOROCARBONS ; LEVEL BY [compute Su] [COMPUTE $ 8 Ano} LEVEL INTERPRE- rest ‘convencence TATION [GOMPUTE Fea (ty eons] I [COMPUTE Atmpa (oleae) | soute,® SEC.POROS. INDEX compute $ (i-Sui8 Pet-ss0} For Moved Ot! (ROVERSE-HOLE LOGIE EVALUATE RESULTS] STATISTICALLY EvALvaTION PRESENT RESULTS os FILM or LISTING: Pore, Flude, Litalogy RESULTS ‘igure 11.20 Logic forthe computer iatecpretation using the Com- plex Lithology method of Schlumberger (Poupon etal, 1970) there are cores, and the characteristics are supplied to the program, Bayes’ discriminant analysis is then used for characterizing and sorting the va‘ious depth points into populations which represent the type lithologies. LITHOLOGY RECONSTRUCTION FROM Loos 137 Bayes’ system is apparently capable of selecting the most useful variables to define a population and rejecting those which have no discriminant power (Broggi et al, 1983). ‘A somewhat similar approach is used in the Schlumberger ‘Faciolog’ (Wolff and _Pelissier- Combescure, 1982). In the method used, a certain amount of preparation of the logs precedes the final statistical analysis. First of all, the logs are rendered compatible by choosing scales of equal magnitude. For example, the neutron-log values range from 0-80 while those for resistivity range from 0.1 to 2000, Rescaling using a log’s standard deviation reduces the variations to the same order. Before passing to the next phase, shoulder and thin-bed values are eliminated by comparing them to the values of preceding and following depths. Significant differences cause a value to be eliminated since they are considered to be unstable. Once normalized in this manner, new reference axes are defined for the log-value populations using prin- cipal component analysis. ‘The first principle com- ponent axis is the most important and is in the direction of maximum variation. A principal com- ponent space is thus defined. Using the principal component analysis, data are grouped into small clusters or local modes by a combination of algor- ithms. The reduction in data at this stage is of the order of 10 to 1 Figure 11.23) Crossplots and squared logs of local mode values indicate whether the statistically- derived data are representative of the original data NEUTRON DENSITY, [25 originat ~15]106 original 2.95} las squared -15|1.05 squared 2.95| Figure 11.22 Squared logs made from statistical clustering, The ‘example shows the elect of plotting clustered values alongside the original values for a neutron log and a density lop. The result should bbe compared to the ctose plot of Figure 11.25, (Redravn from, ‘Schlumberger, 1982)138. THE GEOLOGICAL INTERPRETATION OF WELL LOGS (Figure 11.21). The local modes, numbering say 100- 150, are then grouped by linking those that are closest, with a result that between 4 and 20 clusters are obtained. This grouping can be automatic or manual if geological data is available, as it should be from the mad log. It is the final groupings that become ‘elec- trofacies’. In most cases these are in fact geophysical lithofacies or electrolithofacies: they are of the same The advantage of statistical methods is that natural variability is accounted for. The geological recognition of a lithology can then be reduced to the classification ofa series of geophysical numbers: it is conceptually a simple operation (albeit complex mathematically). A geologist’ lithology, formerly only a concept, becomes numbers, more easy to manipulate and more consistent order as the lithologies used by geologists (Figure 11.24), permit uTHOLOGIC POROSITY AND FLUIDS FORMATION cnaractenisrics | FLUIDANALYSIS SeBULK VOLUME seBULK VOLUME | SHALE % BULK VOLUME | WATER SATURATION. [HYDROGARBON| WATER az |Poros| 0 100) :00 % Q PERMEABILITY INDEX. | HYDROCARBON VOLUME S ao PS~ 025/60. 6 % A lo _ HS" fp F HYDROCARBON WEIGHT 925] faydrocarbon isciostion. (fom Dreseer Atl, 1982) (er processed interpretation (CPI) €* wat The lihalogy #wc monm one ne Rtn ARRREARR IIRL + nem nerghenntes weRaRSERER 139 8 : i 7 ‘. . : So: ; : : : : : : g : : : : : : : BO Po : : : : é 2 § ~ bas § : : pig : git d fs ee ff Baa 8 DOSE De bg : git, 3 eo : : 3 g : : BB lemecnnt Fo cst bs Pat A : : i : oe a : : : ger : : : f: : : 1a maar 13 va Figure 11.23 Effects of clustering data for Schlumberger Faciologon the neutron-density cross plot, Tep:unzoned data. Bottom:zoned data. ‘The zoned data show a large reduction ip scatter, (From Serra and Abott, 1980.)140 THE GEOLOGICAL INTERPRETATION OF WELL LOGS FACIOLOG FROM GST DEPTH SHALE 1 SHALE 2. SHALE 1 SAND 3 sano 3 ann 3 SAND 1. sano 3 SHALE 4. SHALE 2+ sano 3 SHALE 4, sana 3 SAND 2+ SAND 1, SHALE 4. SAND 2, sano 3 sanp 3 sanay SAND 2. CALCAREOUS LInesToNE 1 SAND 2, CALCAREOUS SAND 2, CALCAREOUS SHALY very sanpy stuty LIMESTONE 1 very sanpy caLcarcous sHaLy SAND 2, “cALcARECUS VERY SANDY cALcarcous SHALE 2. SILTY SHALE 2, SILTY SHALE 2. SANDY 500 Figure 11.24 Lithology derived from a statistical computer analysis of log data, Faciolog from Sehiurmberger. (From Wolf etal, 1982)12 Facies and depositional environments from logs 12.1 Introduction ‘The use of well-log analysis in geological disciplines is not sufficiently well developed. Traditionally, logs are used to correlate: one well is compared to another and lines drawn between the wo. This is a primitive approach, and logs have a far greater potential. ‘The previous chapter describes the use of logs to construct lithology. This chapter takes geology a step further and describes how logs can be used for facies and sedimentological analyses. Modern subsurface ‘geological analysis can and should employ a thorough and sophisticated analysis of well-log data, 12.2 Facies Log shapes in sand bodies: geometry A basic scheme to classify sand bodies in the Gulf Coast area of the USA, apparently developed by Shell (Cf. Serra and Sulpice, 1975) was based on the shape of the SP log (Figure 12.1). The scheme was intended to give. classification of log shapes and types in order to aid correlation: it was essentially a geometrical approach. ‘Smooth Serrated é BELL cevunoen Loc) FUNNEL ‘Figure 12.1 Log shape classification. The basic geometrical shapes and description used to analyse SP log shapes. ager CHARACTERISTICS, Syren cpt-gonraer Figure 122 Sand-body log shape. The shape is described by purely geoftetcoal characterises, Several independent elements can be ted besides overall shape. Purely on geometrical considerations, a log through a sand body can be characterized by a number of factors (Figure 12.2) In a simple scheme, shape, curve characteristics, the nature of the lower contact and the nature of the upper contact can all be used for classification, At least eleven variables can be found to describe the geometry. Even in this simple scheme, a classification with eleven variables becomes com- plicated; moreover, it has no geological explanation, Rather than classifying shapes, an attempt should be made to understand the origin of the shapes, Log shape: lithology and sedimentolog y To understand a log shape, the log itself must be understood. Although the SP log was at the origin of the interest in log shapes, it is the gamma ray log that is best used for examining shapes; the curve gives greater variety of shapes, shows greater detail and has more ‘character’. Many authors, in fact, regard the gamma ray log as characteristic’ when illustrating shapes. ‘Therefore, as the gamma ray log is frequently an indicator of clay (shale) content (but by no means always—sce Chapter 7), an explanation of gamma ray logshapescan be related toclay content. A bell-shaped log (Figure 12.1), where the gamma ray value increases regularly upwards from a low value, should indicate increasing clay content: a funnel-shape, with the log142 THE GEOLOGICAL INTERPRETATION OF WELL LOGS teat — ene crasaernce “SSE “TRIASSIC FLUVIATILE SAND BODY ~ POINT BAR Figue 12.3 Sedimentology of bell shape, A core cut through & sind body with atypical bell shape onthe gamma ray log shows 11:0 bea flaviatle channel. Note the close correspondence between the ‘gamma ray (giving clay content) and grin size. Triassic, Sahara {Redrawa from Serra and Sulpice, 1975,) 3 Gawwa ray Sane 3 et g | anak DISTRIBUTARY CHANNEL, SAND BAR, sit lamioae DISTRIOUTARY CHANNEL igure 12.4 Grain size~gamma ray correlation: & very close re lationship is possible between the two, (Modified tron: Simon Beyz00, 1980) ‘value decreasing regularly upwards, should show the reverse, a decrease in clay content. ‘Acote cut through Triassic luviatile sand body in the Sahara (Figure 12.3) shows a typical bell-shape on the gamma ray log response (Serra and Sulpice, 1975). The increase in gamma radiation corresponds to an inerease in clay content regularly upwards, The in- ‘crease in clay content is correlated to a decrease in the sand-grain size. A sedimentological analysis ofthe core shows a set of sedimentary structures typical of fhu- viatile point bar deposits (Serra and Sulpice, 1975). The bell-shape, therefore, can be interpreted as indicative of a fining up, luviatile point bar sandstone (in this case) This example shows the close relationship possible between the gamma ray log and sandstone geain size Shapes on the gamma ray log can be interpreted as grain-size trends and, by sedimentological association, aseycles, A decrease in gamma ray values will indicate fn inerease in grain size: small grain sizes will cor- respond to higher gamma ray values (Figure 12.4), The sedimentological implication of this relationship leads to adirect correlation between facies and log shape, not just for the bell-shape as described above but for a ‘whole variety of shapes. ‘Numerous publications show the log shapes expec- ted or found in various facies (Krueger, 1968; Galloway, 1968; Fisher, 1969; Fons, 1969; Pirson, 1970; Goetz et al, 1977; Coleman and Prior, 1982, Galloway and Hobday, 1983). They all depend on the relationship between log shape and grain-size trends in sandstone bodies. A bell-shape indicates a fining-up sequence which may be an alluvialluvial channel but also a Uansgressive sand. A funnel-shape is a coarsening-up sequence which may be a deltaic pro- gradation or a shallow marine _progradation (Figure 12.5), The analogies may even be extended to deep-sea deposits. In these cases the log shapes are those of overall sequences rather than individual bodies (Parker, 1977). The shapes come from the diminution in bed thickness associated with dimi- nution in grain size, rather than the direct change in ‘grain size itself, However, the principles are similar (Figure 12.5). To add another dimension to log shapes, a simple, theoretical scheme has been proposed (Selley, 1976) to differentiate between the various environments which ‘may give similar log shapes, based on the presence or absence of glauconite and carbonaceous matter (Figure 12.6), Glauconite is the marine indicator, car- bonaceous the non-marine, However, there is no replacement fora serious and complete sedimentologi- cal study of core material, when it is present (e.g. Budding and Inglin, 1981), to confirm the origin of log shapes (Figure 127). Moreover, the careful compi- lation of sedimentological data shows that interpreting ‘gamma ray log shapes in terms of sodimentology is far too simplistic, 1FACIES AND DEPOSITIONAL ENVIRONMENTS FROM LOGS 143 DELTAIC-FLUVIAL ENVIRONMENTS ‘A. CHANNEL-POINT BAR, alluvial or fluvial 26m] ga or 5 B. DELTA BORDER PROGRADATION Ga dr sp. MARINE ENVIRONMENTS ©. TRANSGRESSIVE MARINE SHELF D. PROGRADING MARINE SHELF ae: Ga or SP GR or SP 250 or dsp DEEP SEA ENVIRONMENTS, ] PROXIMAL DISTAL E,SLOPE CHANNEL —F. INNER FAN 6. MOLE FAN H. SUPRA-FAN I. BASIN ¢ CHANNEL ‘CHANNEL DEPOSITIONAL PLAIN a. os, Figure 125 Facies indications from gamma ray (or SP) log shapes, These are idealized examples both of log shape and sedimentologic facies. (Modified from Serra, 1972; Parker, 1977; Galloway snd Llobday, 1983) The gamma ray-grain-size relationship—a critical examination Neither the gamma ray nor the SP log varies because of changes in grain size: they vary (often) because of changes in clay content. The sedimentological in- terpretation of gamma ray log shapes depends on the variations of logs with grain size. By inference, the sedimentological interpretation of log shapes will in turn depend on the extent of the relationship between, clay content and grain size, There is no doubt that a grain-size/clay-content’ A on oF tog ae Ses, ° 400 ° 100 ore a © site c i | | 1 1 submarine gage of submarine Movie or deltaic fan (urbidte) marine bar deltaic bar sand channel Sistibutary ge channel —_ distal slope or crevasse splay | proximal GL = glauconite present C= carbonaceous matter present cal mineral emia flog shapes A simple system icrpetng the environment of and bodies using a combination of og shape and detrital mineral components. (From Selle, 1976) relationship exists (e.g. Figure 12.3), However, itis by no means a constant one. The textural analysis (Figure 12.8) shows an alluvial molasse sandstone with consistent relationship, grain size to clay content, while the well-winnowed marine sandstone shows no relationship (Pettijohn et af, 1972). A similar variation is expressed by the statistical analysis of quartz content in various grain-size fractions from different enviroa- ments (cf. Davies and Ethridge, 1975) (Figure 12.9), ‘This latter analysis shows two contrasting aspects. The 30 8 % CLAY «20n) ©, 8040 60 70 100 26 GRAIN size og median) Flgwe 12-8 Testural analysis of the layers size relationship. SEY Senza! compared to grain size (or an allivial sandstone (dg Ron-quartz content of sandstones (ic. clay matrix, rock fragments and feldspars) may vary linearly within one sand body or even one environment, but, in contrast, the variations of these same elements between different environments show widely different patterns, In the Auviatile environment, for example, the non-quartz elements show a second-order polynomial variation. In fact, such are the differences in the variations between the quartz and non-quartz content of sand stones that the differences themselves may be used to ‘identify environments (Davies and Ethridge, 1975} That grain size shows a variable and complex ‘lationship to clay content is amply demonstrated by the correlation of core-measured grain size to cor. responding gamma ray log values (Figure 12.10), A relatively good correspondence is found in the deltaic (QUARTZ CONTENT % ‘QUART? WEAN Size Figus :2.5 Yasistions in sxedstone vssnpociion with grain size and 8 GERRY AF SOCAN TELTRGHS ese los ashes. pensitY gem? 24 : Dy, catser donsty = om “ie BREAK 2 TRANS 20 RESIN Z oon) - . GARI RAY AI SON TI Jo. ——RETIRONS: #4 ob ome 8 eda ogienv'giegs 2a Z density 10: - GR Bt aneak gE a rican 20 al 3} < FORMITY 20 se 2 § peony eg sone 55 En Fila SNEHTT og art bit GR is up—Y— SFL 104 sizo 3 4 GE S j paeaK F FAULT 7 E g consity eo cal, 3 § 20 Figure 12.17 Examples of abrupt breaks, Lithological related breaks: 4 breaks:C, unconformity; D, fall Frequently an enormous amount of data isavailable which is not exploited. The entire suite, lithology interpretation, electrofacies analysis and sample ana- lysis data, can then be subjected to a sequential analysis—from the base of the well upwards. All the data are used (Figure 12.13) channel erosion: B, cycle top transgression, Non-lithologicaly related In this inal step of the interpretation, the intent is to reconstruct a vertical sequence of individual depo- sitional environments. The evidence for the interpre- tation comes from each preceding step. A particular lithology, such as coal for example, indicates swamp or marsh; electrofacies analysis indicates a zone of clay tojayeBor agBnoug ayv Biep 2 P fares puT 5 wonboepoiaioo Y gI2I amy osu yoray vo wean peevun een) io ‘suNaHHOD “an ve aurenz0 33409 aaa Ts 152 THE GEOLOGICAL INTERPRETATION OF WELL LoGssand progradation; a core analysis indicates a sand to be fluvial; faunal analysis suggests a shale was de- posited in a neritic environment, and so on, The resulting interpretation should bea sedimentologically logical sequence of environments. The breaks re- cognized from the logs and from faunal analysis should bbe added to the environmental interpretation and the principal sequences extracted (Figure 12.18), The en- vironmental logic should only continue between the major breaks. This document, as well asthe detailed set of environments, may now show thick or distinctive FACIES AND DEPOSITIONAL ENVIRONMENTS FROM LOGS 153 depositional sequences which have a stratigraphic importance (Figure 12.18) (see ‘Depositional sequences and stratigraphy, Chapter 13). Itis impossible to give the details of all the reasoning, used to construct this final document, Many geological disciplines are involved. This section simply describes, the system of sequential analysis, and how the body of data is built up to form the interpretation tool. To use the tool one needs not just the data from this book, but training and experience in all fields of oil geology.13 Stratigraphy and logs 13.1 Introduction ‘As an aid to stratigraphic analysis, wireline logs are invaluable, As a means of correlation their use is obvious. However, they may also be used for more sophisticated forms of stratigraphic analysis, This chapter describes the use of logs to define lithostratie- raphy, various aspects of the use of logs for correlation, ‘and stratigraphic breaks and stratigraphic sequences seen on logs. The latter are the common factor between logs and seismic sections, and between sequential analysis and seismic stratigraphy. 13.2 Lithostratigraphy Type subsurface lithostratigraphy The simplest use of well logs for stratigraphic analysis is in the identification of lithostratigraphic units, Such units allow a well to be divided up into intervals which may be identifiable in other wells or even on the surface. The best use of the lithostratigraphic unit in the subsurface is to define type sections. Just as type stratigraphic sections exist at outcrop, so type wellscan be designated in the subsurface for particular litho- stratigraphic formations. Well logs form the basis for these definitions. The various lithological units of the North Sea, for example, have been defined by govern ‘ment agencies of the countries concerned (Rhys, 1975; Deegan and Scull, 1977; NAM/RGD, 1980), Each lithological unit has been designated a type well which can be used as a reference for both lithological and log, characteristics (Figure 13.1) Ona different scale, the stratigraphy of a particular field may be described using a type well and well logs illustrating each lithostratigraphie interval (€.g. Jamison et al,, 1980). Names in fields are frequently specific to the operating company. The illustration of the named intervals using well logs enables them to be identified outside the field and by other companies. It avoids annoying name confusion. 133 Some aspects of correlation Correlatable time-constant horizons Time-constant horizons are rare. They are, by nec- essity, more or less instantaneous in terms of geological time, and independent of environment. To be useful, moreover. they must be of wide geographical extent. ‘The most common horizons which fulfil these crite- ria are those made of voleanic tull. Thei¢ origin, as the products of voleanie explosion, is independent of the final depositional environment. The dispersion of the clements of the explosions can be very large. The Lower Tertiary tulls of the North Sea and Norwegian Sea are a fine example (Jaequé and Thouvenin, 1975; Knox and Morton, 1983). They are found over a huge area, a minimum of 2000km from north to south, from the Lofoten Islands in Northern Norway to the coasts of Holland. The tuff horizons can only be recognized with difficulty in cuttings, except with a microscope, but are casily seen on the well logs as they show a characteristic low radioactivity, high density and high velocity (Figure 13.2), Their age is everywhere Upper Palacocene (Thanetian) and the main phase seems to have lasted less than Imillion years (Knox and Morton, 1983) ‘A second type of time horizon may be caused by chemical changes in water masses due to changes in ‘ocean water-cisculation patterns. These changes may create distinet horizons in sediments, especially those from deeper-water environments. The effects are rapid, of large extent and essentially independent of environment Marker horizons ‘The simplest and most evident form of log correlation is by using log markers, This marker may be a distinctive peak, a distinctive shape or a distinctive lithology. The Jurassic of the Paris Basin, for example, shows the behaviour of marker shapes over a stable, intra-eratonic basin (Figure 13.3). Correlating these log shapes gives a reliable chrono-stratigraphy (Serra, 1972), False correlations — facies In sand-shale sequences, correlations oiten become extremely complex. It is exceedingly casy to corrciaic two similarlooking sand bodies between owo we only to find that the liquids that they contain arc incompatible for the structure, and that, in reality, they are not related. The correlation was of like facies. It is essential to take account of the facies being correlated. Insand-shale sequences. although sand bodies are the reservoirs and therefore important, they should, nonetheless, not form the basis for correlation, In contrast, in the example shown in Figure 13.4, ‘om sedimentoiogical considerations, facies were si, bysand progradation; a core analysis indicates a sand 10 be fluvial; faunal analysis suggests 2 shale was de- posited in a neritic environment, and so on. The resulting interpretation should bea sedimentologically logical sequence of environments. The breaks re- ognized from the logs and from faunal analysis should be added to the environmental interpretation and the principal sequences extracted (Figure 12.18) The en- vironmental logic should only continue between the major breaks. This document, as well as the detailed set of environments, may now show thick or distinctive FACIES AND DEPOSITIONAL ENVIRONMENTS FROM LOGS 153 depositional sequences which have a stratigraphic importance (Figure 12.18) (see ‘Depositional sequences and stratigraphy’, Chapter 13). Iis impossible to give the details of all the reasoning, used to construct this final document. Many geological disciplines are involved. This section simply describes the system of sequential analysis, and how the body of data is built up to form the interpretation tool. To use the tool one needs not just the data from this book, but taining and experience in all fields of oil geology.TYPE SECTION Coordinator: 62 16" 347 ‘TOP. OF RED CHALK TO BASE OF CRETACEOUS STRATIGRAPHY AND Locs 155 BURMAH 48/22-2 WELL Rae. 09 tt (25m) AMS.L. Or 22" 30" rilog tia November 1007 ~ 260% Jonwary 1068 « nites 419 3 3 @ & x a [oamma RAY] SONIC pitt = APL 5 lo ‘Go sof [eso _s90 40 aPleres RED CHALK FORMATION oa [<206m) Mudstone, calcareous to slightly ~ Calearnovs, ra brown wit i aa tein geyrmtiing tn tener ES a SHEEP tay ronUATION fzc0m Mdotoce grading to sale wilh Gooth, ahty calearoous Coce fonaly Voy catearecta ht brown to ory becoming orosominenty ‘revs otve gn wh dope sees Sr'glevcente, come shall regents 200 Thal, gy ealearoove Coccentoaty very calearaous ove fs voy to grey blacks mocorstoty E Tndurated, hell regents abundant = | n'naces, sport bande of ity so00'8 3 | sendstons,vory testo, abundant 3 2 | Srosphate ance, brown ack é & | Sear bese wih atin sittone, G cam genes # 1100 a 3 g g 1200 Pa 2 3 br200 & & ssa" SpICSOY SANDSTONE FORURTON fat eo Sancs\ones very fine to madam, 3 Satcnie'ts tepals anaoraat~ Le eaters Se cemented band, EB seer gule a neler ia ‘aar’ BRGEams a Saleateon to Bae fel 2 ‘green, firm to flsallo, (435m) JURASSIC Kimmeridge Clay Figure 13.1 Type section of an offshore formation ustrated by wel logs. (Redrawn from Rhys, 1975.)t 156 Tit GEOLOGICAL INTERPRETATION OF WELL LOGS ‘GARR FAT mareY saa] 3 x Figure 132 Time-constant horiven. The main North Sea Tuff zone~an effective ‘ime line’ which ean be correlated over huge aveas. Peessentally many thio, graded beds of tu, given an expected continuity and value for correlation Thick coals represent a relatively long period of time and are usually persistent: they havea high correlation value, The same can be said of log patterns showing a progradational upward increase in quartz: these too are good for correlation, However, the sandstone bodies, as proved by Muid incompatibilities, cannot casily be correlated. This is an area of active deltaic deposition: reservoir correlation is exceedingly complex. The rule of correlating only certain facies is uni- versal, In general, the finer-grained the facies, the better the correlation value. The deeper the environ- ment of deposition, the more continuous the cor- ‘elation (with the exception of turbidites!). Palacontologically-controlled correlation Itis generally safer not to correlate well logs without the corroboration of palscontological datings. Adding palacontological datings gives to a correlation, a time stratigraphic value. It is a correlation which can be interpreted in terms of regional and perhaps global time-indexed events. The exaraple (Figure 13.5) shows the careful log correlation and dating of a series of Middle Cretaceous rocks which cover a small pod-like horst (the Buchan Horst) in the central North Sea (Burnhill and Ramsay, 1981), The lithostratigraphy shows some continuity but also inexplicable discontinuities. The datings, not fine enough for correlation, do however add sufficient data to explain the discontinuities and to corroborate the continuous log correlations. The com- bined log and palacontological correlations show a mid-Cretaceous, early Turonian unconformity over the horst itself, combined with a later Turonian onlap, By correlating the unconformity with a regional Turonian regressive phase which occurred throughout north-west Europe and North America, the authors show that the erosion was not the result of local movement on the horst, but the result of a global change in a sea level (Burnbill and Ramsay, 1981), The succeeding onlap had a similar cause. Obviously, this sort of reasoning is not possible with undated log correlations. Correlation and depositional topography ‘The topography of the present-day surfaces of sed ment deposition must be similar to those that existed in the past. The environments of shelf, slope and basin cxisted in past geological time, as now. Long-distance correlations should reflect these topographic elements. The Pennsylvanian to Permian deposits of West Texas were apparently laid down close to the shelf edze (Van Siclen, 1958). Correlations over 10km in these beds across the palaco-shelf edge show the depo- sitional topography as it then was (Figure 13.6). The topography is associated with changes in lithology and, in this case, limestone and reef development. Limestones develop on the platform, reefs on the platform edge and shales in the basin off the shelf slope. To show depositional topography the choice of datum is critical. In the example cited, the arca has not suffered too severe a tilting or differential subsidence. The present-day surface used asa datum still allows the palaeotopography to be seen. This is not always the case, and a chosen horizontal datum must correspond toa facies which was deposited on a horizontal surface. For example coal beds were originally quite flat: unconformities were not ;Sc ; Tati ww 3 ALA WA fverhay ; CLIC een "havo NvUA-OTY158 THE GEOLOGICAL INTERPRETATION OF WELL LOGS nya oucsinstogajsanag = Sf 280% “and w 49 S13 9PSINg ¥9S WON HOH pied UEYoNg ay Sosoe sneaoelNE-pH NU) 10 pjjonuos Creer eosaey, ong wos) susr019 siuoWonag FaponaUNet}ge]D = AD “vor HruanenIU mBiogeayt ~ gAY“avore!g puowLEnD Diody>IoY = ingjorsia= Z.4 4 I00| J90 spuozesosolu Fa aos “MUA [gy Kes PUN “49 ojoluozeteg SE] am fe i] wa sappy | 2 << sto nvtong ——____» Ei | I eves unoions ett CEA am | a Te ; i f e ge i x FI fa eoreecf i 1 ct 2. a3 oe bo : r| wacmuen |S 2 LIT J b ! | I | I z ar Fag oF gone sa} once Oo a omg nee El cums [san eos up one ‘ oe —db. STRATIGRAPHY AND LOGS 159 coal correlates sand bodies 0 NOT corre! shale cycle te correlates, Figwe 13.4 Correlations ina detaic complex Incompatible us show thatthe sand bodies are separated Persistent coreations are bared on ENT snd the fine-grained facies, MC marker coal, M = marker, CYC= marker cycle, Wto Z is 35km. Correlation and seismic sections Using well logs to identify mapped scismic horizons is common (see Chapter §) but their use should 0 beyond this. Log correlation and seismic mapping should be undertaken together: the approach should be iterative. Horizons correlated using wireline logs should be identified on seismic. Mapped «seismic horizons should be identified on the logs. Seismic ties and log correlations should obviously be parallel. The cxamination of a seismic section before completing & log correlation will indicate whether such 2 correlation is possible. ‘The log tied to the seismic section in Figure 13.L0, for example, shows a prominent progradational in- terval surmounting a series of parallel sequences. Nearby wells may be correlated to the lower parallel sequences (supercycles Ta-Tb—Te) but not at all to the progradational interval (especially super cycle Ta). Such simple concepts need to be demon- strated by the simultaneous use of well logs and seismic sections sw PLACID MONNELL i= . eo eee @ @ mune © snare e ‘mane S13Rng* urscoup er aL. sanugonte E Lyasoe Figure 13.6 Correlation and paleotopography in the Permian of West Texas. Logs—SP 0 the left electrical survey tothe right (Redrawa from ‘Van Sielen, 1958)160 THE GEOLOGICAL INTERPRETATION OF WELL LOGS. ee sone \ A 7 NS we Ee = 20m aE = SieeS ~~ op 40m \ é ee som = a Lower CRETACEOUS, Figure 13.7 Anomalous gemma ray peak a an unconformity. The peak corresponds to gap of 15 Ma: itsstratigraphie importance isevident, The radioactivity is probably due (0 uranium concentrated in phosphatic nodules, 13.4 Stratigraphic breaks and stratigraphic sequences Unconformittes and sequence limits Modern stratigraphy is paying more and more atten- tion to the breaks in the sedimentary and stratigraphy records. Such breaks can vary in their duration in time and in their geographic extent and can be made up of non-deposition or erosion (or both) (Sloss, 1984). Channel downeutting forms an erosional break, short in time and of very limited extent. A eustatic sea-level drop creates non-deposition over huge areas and may last several million years. The terminology for such breaks as yet lacks definition, but it is clear that angular unconformities such as that seen by Hutton at Siccar Point, is but onc type of stratigraphic break and probably the least common, Major stratigraphic breaks are usually easily re- cognizable on wireline logs: there are dip changes, @ distinct faunal gap, a change of compaction, and there may be anomalous log peaks. This sort of break corresponds to an angular unconformity (cl. Figure 13.7), But many important breaks are far more subtle, Current theory suggests that eustatic sea-leval chan- ges are frequent and that drops are rapid while rises are gradual (Vail et al, 1977). The effect of this on the sedimentary record is not yet clear. What is clear, however, is that any persistent sedimentary cycle, however thick, must have been deposited between any possible abrupt sea-level changes. A break in a cycle is not proof of a change: a continuous cycle is, however, proof of no abrupt change. Depositional cycles are often clearly marked on the well logs and their limit~ become especially clear following sequential analysis, (Figure 12.18). Correlation of sequentially-analysed ‘well logs, both within basins and between basins, will demonstrate the continuity and importance of the log- defined sequence limits. The common practice of tying imprecise palaeontological limits to some distinct log, break may have 2 sound scientific basis: such breaks can be related (0 eustatic events (Figure 13.8). This is a useful concept but not infallible, as is shown below. Palaeontological limits and lithostrarigraphic limits A palacontological boundary tied to a lithostratig- raphic limit assumes that an obvious lithological change has time significance. This is frequently true, as discussed above, especially when the lithological change represents a time gap. However, there are cases, where. very distinctivelog breaks does not correspond to the major time-defined break. A well-known example illustrates the point. ‘The widespread ‘Late Cimmerian Unconformity’, identified over much of the North Sea, corresponds to the limit between the Jurassic and the Cretaccous. It is easily identified on logs as the top of a black organic- ich shale with very high gamma ray log values (Figure 13.9), The same ‘unconformity’ has been iden- fified on the basis of logs in 600-700 wells. It now appears to be a lithostratigraphic limit. Very careful datings ofthe top of the radioactive shale show that he unconformity is not at the top ef the shale but within 2 carbonate-rich lithology above (Rawson and Riley, 1982). There is no distinctive marker on the logs at this level (Figure 13.9), Because the lithological change was confused with the unconformity, the former was con-Stas ++ STRATIGRAPHY AND LoGs 161 WELL RECORD ol |e 8\./F 8)2]& |onma | 7 E\S|E | mr |S] some | E Ey TE] an (8) we é Bla uool Eleso alt SEA LEVEL CHANGE | AGES | Ma z[g/. £ Se bar AE 2 a 2\2 alg |é B18 |e a 38 8 7 ee ae ee Bl. [ez = | men Ole [547] = 100 é = rT = non-deposition “ls APTIAN i = i MW Ta Ue egg aaa TO 2 eae g\2 5 180 ole |S : SATION ale |g I Sacer a8 |6 2 ALENT dl ae 8 | z zie ala g\é Figure 13.8 Enstatc ea level changes related to log breaks: Cortelation of Jurassic and Cretaceous stratigraphy as sen in the well record, with lobal sea level changes (cea level from Vail etal, 1977}, PM =Plenus Marl; GM =Glauconite Marl; Kell= Kellaways Beds; Corn ‘= Combrash; FM orest Marble, sidered to be diachronous. However, the diachronous termination for such a distinet facies as is represented. by the black shales is difficult to accept, since the ‘environmental conditions under which they were de- posited were extreme. The careful dating work, in fact, seems to show that the end of the black shale facies was indeed isochronous over a large area but that it does not always correspond to the unconformity (Rawson and Riley, 1982). This example illustrates that a lithological break is not always coincident with structural events, and vice versa, Seismic stratigraphic sequences sequences and log-defined “The correspondence between sedimentary and stratig- raphic units has been highlighted by the methods of seismic stratigraphy. Seismic sections may now be presented interpreted into time-stratigraphic units162. THE GEOLOGICAL INTERPRETATION OF WELL LOGS 1980 1075 STRATIGRAPHY — LITHOSTRATIGRAPHY Sanat é from DINOCYSTS: pee GAMMA RA E | avo awwonrtes | STRATIGRAPHY =e Turonian ‘uronian ae Coviacian Contacian sores Albian Aibian= Aptian = Early Aptian fea eed pO JRE | Early Barcomian Barremian “Hauteriviaa PRINGIPAL UNCONFORMITY | > MAJOR LOG earty Vatangintan | Late Ryazinian h cHANGE black organic shales LATE \| tate vototon | ciMMERIAN lunconroRMiry wigaie vos Iicmeridgian Kimmeridgian a - tate Oxtoraia Figue 139 Careil dating around the-Late Cimmerian Unconformity’ ofthe North Sea shows that the depositional change, indicated y the Toes doce not precisely coincide with the stratigraphic break {cedrawn from Rawson and Riley, 1982) (Gheriff, 1980) which are in fact ‘seismic sequences’. A time-stratigraphic unit is ‘a three-dimensional set of facies deposited contemporaneously as parts of the same (sedimentary) system, genetically linked by de- positional processes and environinents” (Sheriff, 1980). For a seismic sequence to be equivalent to the time- stratigraphic unit, its limits, like that of the time unit, must be isochronous. In seismic stratigraphy, these isochronous limits are sea-level falls, rises being con- sidered as gradual, falls as essentially instantaneous (Vail et al, 1977). It is the rises especially which give rise to recognizable seismic sequences which are boun- ded by important seismic horizons created during the falls. For the scismic section to show an important reflection, thereis an important changein lithology. An important change of lithology is inevitably marked strongly on the well logs. The replay of logs, especially sonic and density logs, on a time-constant (instead of ‘depth-constant) scale (Chapter 8), allows logs to be placed over a seismic section and compared directly (Figure 13.10} The sequential analysis of logs, de- scribed previously (Chapter 12), allows sedimentary stratigraphic units to be identified. These units are on the same scale and should be identical to the time stratigraphic units of the seismic section. Figure 13.10 shows @ mega-sequence identified on the well logs,STRATIGRAPHY AND LOGS 163 puto sor 24 uo 109 pur 80} 21 yeos-oul) ® Uo pauysp s2ouanba oud 0) spuodsa1soss80} 24) uo el aE Sai (7010506) ton mont164 THE GEOLOGICAL INTERPRETATION OF WELL. LOGS, projected on to the seismic record and showing large seale progradational forests, Sequential tog analysis and seismic stratigraphy should be complementary. 13.5 Conclusions ‘The brief survey of this chapter covers only the fundamentals of the ways in which well logs may be used in stratigraphic analysis. The simple lithastratig- raphic unit is identifiable by well logs but has only limited value. Logs have always been used for cor- relation, but this chapter shows that they are capable of considerable refinement in this task. Itis in a more complete stratigraphic analysis that the use of logs is capable of the most refinement. The recognition of sedimentary sequences and depositional and stractural breaks from the logs was previously described by the use of sequential analysis. This chapter shows that this lor-derived information can be turned into a powerful stratigraphic tool. Sequential analysis leaves the log interpretation at a scale com parable to a stratigraphic analysis of seismic data. The two can be integrated. Combined with seismic data, the single data point represented by a well can be greatly expanded in both geometry and geography: the seismic data becomes annotated with geological meaning.14 Concluding remarks 14.1 The geologist’s problem Fully ball of this book is taken up with linking geophysical responses to geologically recognizable parameters, for instance, what natural radioactivity ‘means in terms of shale content; tow sonic velocity can be interpreted in terms of shale compaction; what bulk density means in terms of mineral composition; and so on, In other words, geophysics is translated into geological terms. This is the central problem for the geologist using well logs. There is a gap between what can be defined gcophysically and what is significant geologically. The following section illustrates an aspect of this, problem by discussing lithology, both geophysically and geologically defined. 14.2 The lithology problem In Chapter 1 it was suggested that the geologist using Togs can quickly identify a subsurface formation from its log characteristics, but when faced with an outerop of this same formation, is at a loss to recognize it. The reverse of course is also true; the field geologist cannot identify a formation using well logs. The two data sets are in quite different terms. Chapter 11 discussed computer-derived lithology from log interpretation. It was pointed out that an output from this interpre- tation, in terms of quartz, shale and porosity per- centages, does not suggest lithology to the geologist. Indeed, such percentages do not even define lithology: a greywacke may contain 50% shale and 50% quartz but it is still a sandstone: a silt may have an 80% quartz content but it is not a sandstone. The geo- logical definition of lithology is textural, composi- tional and structural. A fine-grained, bioturbated orthoquartzite is comprehensible to a geologist, he needs to know how such aspects can be defined using, geophysical logs. ‘There are two approaches to the problem. One is empirical—to confront the geophysical data set di- rectly with the geological data set; the other is ‘mathematical —to convert the geophysical data into a uasi-geological format. ‘Throughout this book the geophysical data, the well logs, have been shown compared to core data. This is ‘one method of defining the geological data in geophysi- cal terms: by a direct comparison, However, cores are rare and generally found only in reservoir zones. A far more comprehensive comparison may be achieved by geophysically logging outcrops. Here is where the 2 | sunrace £8] cama nay g fe ors, BBG iw ae gag 58 ai lssoo* g See 5 Figue 141 A gamma ray curve constructed from measurements made at the outcrop. Mississippian, Chainman Shale and Penasylvasian, Ely Limestone, Northern White Pine Range, Nevada. Dasted line on lithology columa indicates rubble oF covered. outcrop. CPS =counts per second, (Redraim from Chamteriain, 1984) HEIGHT trom base geological data cannot be disputed, and here is where the geophysical monitoring can be regulated for re- quirements, The example shown here (Figure 14.1) is of gross natural radioactivity measurements, but the results are promising (Chamberlain, 1984). There is no reason why this approach cannot be extended. Spectral gamma ray, induction and sonie-velocity tools can be adapted for outcrop work. The outcrop can then be characterized in terms of geophysical response. It provides an empirical solution to the lithology pro- blem, and must surely be one of the promising directions for the future, The second approach to the lithology problem is to refine the geophysical data set with geological prin- ciples in mind. The approach was briefly described in Chapter 11. The logs are considered within themselves and, de facto, grouped into populations with numeri- cal similarity. The grouping commences with each individual log but then brings them together into one complex. The final groupings are therefore defined in terms of all the available logs. The groups are entirely geophysical and, as such, have numerical limits, but166 THE GEOLOGICAL INTERPRETATION OF WELL LOGS they can be made to approach the level of homogeneity used in lithological description. The geological validity of these groupings is not yet fully understood. This numerical approach to the lithology problem is obviously attractive since it can be applied rigorously by the computer. It takes into account the natural limitations of the geophysical measurements and uses all the available data in the form of an interactive complex. It is necessarily the tool for the future, but it must be adapted to geologically meaningful con- straints. 14.3 The dipmeter problem This book is about the geological uses of well logs and yo the one log that directly gives 2 geological charac- ter, the dipmeter, is not discussed. The omission is deliberate but regrettable; an explanation follows. The dipmeter tool produces an apparent dip by comparing detailed micro-resistivity curves from op- posing sides of the borehole (eg. Serra, 1979), The dip, therefore, is that shown across the borehole at particular depth: the log is a plot of dip value and azimuth against depth. In the older tools, before 1967, three curves were measured and compared. Later generations of tool used four or six curves, while the present, new generation uses eight curves. Data points, that is dips, from the older tools were 3~4 every 10m, and these were increased in the four-curve tools toup to three per metre. With the newer tools, however, dips can be given accurately every 12.5 cm, that is 8 dips per metre: an enormous mass of data (Figure 14.2). This, exponential multiplication of data has caused a break down in dipmeter interpretation, From the outset, dipmeter interpretation has sul- fered by being subjected to alLembracing interpre- tation systems based on theoretical models. Originally it was considered that dipmeter logs should be in- terpreted in terms of three patterns; constant dip, upwards-increasing dip and upwards-decreasing dip, the so-called green, blue and red patterns 3 CURVE | 4 cuaVE | 8 CURVE DipMeTER | DIPMETER | DIPMETER op—~fop— lor om - om 5 x | s s | g ey x g 08 8 x 8 g x x s 2m = 2m dates: pro-i867 1907-1084 1004-7 Figure 14.2 An illustration of the increase in data density of the ipmeter duc to tool improvements, A 2-m interval is shown; the increase in data is exponential (Schlumberger, 1970). Refinements have been made to this simple concept, but usually again in terms of turnkey interpretation systems to be applied uni- versally. As data quantity has increased, so these systems have begun to show their inadequacy for geological uses, The problem with the dipmeter log, strangely, is the same as for the other wireline logs. The geophysical language is not the same as the geological one. The geologist has no field model of dip changes every 1cm vertically, although he should have. The only way in which the dipmeter may be reliably interpreted is by theuse offield examples. Thisis why the dipmeteris not included in this book: there is not enough field evidence for its interpretation. This is in no way = criticism of the tool, only of the way in whieb it is used. 14.4 Not a conclusion—an approach In Chapter 1 the various wireline logging tools were listed with their principal geological uses (Table 1.2}. ‘The uses were specific to each tool. This approach, through the individual tools, was made in order to introduce the contents of the book. Chapters 11-13 took a different approach, that of a common problem to which all the logs contributed a part of the solution, ‘This approach was used in the discussion on lithology above (this chapter) and is certainly the way in which logs should be considered as an aid to geology. To illustrate this, the common geological elements which can be interpreted from the well logs are put into five major groupings. They form part of the table (Table 14.1) in which the relative contribution of each geophysical tool to the interpretation ofthe geological elements is given a subjective percentage value. In reality, the table indicates the extent to which each tool isinfluenced bya particular geological element. But the degree of influence indicates the degree to which the tool may be used for its interpretation. The resistivity tools, for example, are strongly influenced by texture (P) and of course fluid content: they are used accord- ingly for the interpretation of these (Table 14.1). The five geological groupings used for the table are detailed below (ef. Serra and Abbott, 1980). Composition This grouping covers the chemical and mineralogical make-up of rocks. It may be given by a simple chemical formula, as for the evaporites, but more frequently it has no chemical limits and is a complex mixture of various minerals, asis the case for a typical rock such as sandstone. The term, naturally, covers shaliness. {also ‘covers the variations in secondary elements such as organic matter and detrital minerals. Geological composition (as indicated earlier in thisCONCLUDING REMARKS 167 ‘Table 14.1 Relative contribution % of geophysical outpat parameters to information on geological purtmeter groups (6. rock resistivity contributes 13% of the information op textures, when the contribution rom all the logeing Cools listed i considered) Geophysical Geologie! parameter groupeys . ‘ouipue — - Chapter Too! pacemeters Composition Texture Siructure Strata ‘Fluids 3 Caliper Hole diameter 3 6 6 3 ° 4 Temperature Temperature é 2 0 3 3 5 SP Spontaneous 0 6 6 ‘ B potential 6 Resistivity Whole rock 10 8 6 10 30 resistivity 7 Gamma Nata 3 6 6 13 ° ray radiation 2 Spectral Spectra of 0 ‘ 6 83 ° gamma ‘natural ‘adiation 8 Sonic 6 3 10 13 6 velocity 9 Density Bulk density B B ‘ 10 0 10 Neutron Neutton porosity 10 B 6 5 17 Other tools ete 0) ya «en chapter), is only one of the terms in the definition of lithology. Texture ‘This grouping covers such elements as porosity, per~ meability, grain size, sorting and matrix distribution. It concerns all the ways in which grains, minerals or chemicals are distributed in a rock. In well logging this, also concerns vertical variations in texture, such as compaction or overpressure, Texture is also intimately concemed with diagenesis. Structure Bedding is the principal element of this term. This may be seen normally in laminations, lenses or beds but also in the constructed geometries of sedimentary struc- tures. On a larger scale, folding and faulting are structural elements: on a smaller scale, so arefractures, stylolites and veins. Strata This is a bulk term intended to concern the more complex geological aspects of strata arrangement. It covers facies, log trends, correlation and stratigraphic analysis. Alsoimplied is the parallel use of wireline logs with seismic data. The term has been given meaning principally for the construction of the table. Fluids Free hydrocarbons come into this group. Also in- cluded are free formation water and as an extension, chemically or physically-bound waters, such as in water of crystallization. For the construction of the table, the suggested degree to which a particular geological parameter will be explained by the geophysical output of a wireline logging tool, s entirely subjective. It simply illustrates which tools are best used to solve a particular geologi- cal problem. However, italso illustrates (and this isits object) that geological-information comes from a complex of wireline geophysical logs: that geophysical infor- mation should be pooled and used as an interactive set. ‘This aspect was illustrated previously in the discussion on computer-derived lithologies (this chapter). It is without doubt the approach that will bring the biggest advances in the geological interpretation of well logs. In the future new tools will come into service, new uses will be found for old tools, or old tools will be significantly refined (eg. Fertl, 1984). However, what will remain unchanged is the huge amount of geologi- cal information, in numerical form, that is contained in wireline geophysical well logs. It is there for the interpretation,References ‘Adams, LA. and Weaver, CE. (1958) Thorium-wraniuny ratios as indicators of sedimentary processes: example of concept of geochemical focies. Bull, Ans. Assoc. Petrol. Geol. 42{2), 387-40 Allaud, L. and Martin, M. (1976) Schlunberger, histoire Pune Acchnique, Berger-Leviault, Pacis Archie, CF. (1950) Intraductian to pecrophysics of reservoir cocks, Bull, Am, &ss0e. Petrol. Geol 3434 943-961, Asquith, G. and Gibson, C. ((982)' Basic wel log analysis for geolopisis. Am, Assoc. Petrol. Geol, Tulsa, Methods ia Exploration Ser. 216. Belknap, W.B., Dewan, IT. Kirkpatrick, CV, Mott, WE. Pearsoa, ‘Adand Rabson, W.R, (1959) Calibration facility for nuclear logs Drill and Prod. Prac. APE, 289-311 Benoit. WR. Sethi. DK. Fert, WH. and Mathews, M. 11980) ‘Geothermal wellloganalysisat Desert Peak, Nevads, SPWLA2Ist dn, Sip, Trams, paper AA, I= Bjpclykke, K., Dypvik, Hand Finstad,K.G. (1975) The Kimmesidge shales composition and radioactivity, NFP Juratcie Norther 1. Sea Symp. Stavanger, Paper 12, 1-2. Blakemsn, ER. (1982) A esse study of the effect ofshaealteration on sonic transit times. SPWEA 23rd Aun, Synip. Trans, paper Il, rn Broding. R.A, (1982) Volumetric scan well logging. Lag Analyst XXIM() 14-19. Broding, RA. (1984) Application ofthe sonic volumetric sean log to een cratuation, SPWLA 280 Aan Symp. Trans e373, r Brogsi, G. and Gerdil-Nevillet,F. (1943) Cluster method and log data can determine lithology. Oil & Gos J, July 18, 129-140 Budding, M.G., snd Ingln, HL. (1981) A reservior geological model ofthe Brent Sands in Southern Cormorant. In: Petroleum Geology of the Continental Shelf of NIV. Europe, eds. ling, LV. and Hobson, G.B., Heyden, London, 226-334 Burke, JA, Carapbell, RL and Schmidt, A.W, (1969) The ltho= porosity cross plot. SPHLA 10ch Ann. Symp. Trane, 1-29. Burahil, TJ. and Rarasay, W.V. (1981) Mid-Cretaccous palacon ‘ology and stratigraphy, central North Sea In: Petroleurt Geology of the Continental Shelf of N.W. Europe, eds ing, LLV. anc Hobson, G.D. Heyden, Landon, 245-254 Burst. LF (1969) Diagenesis of Gull coast dlayey sediments and its ‘possible relation to petroleum migration, Bll. Am. Assoc. Petrol Geol, 53) 73-93, Bussian, AE. (1982) A generalized Archie equation. SPWLA 2rd ‘Ann. Srmp. Trans. paper B, 1-12. Bussian, AE. (1983) A compatison of shaly sand models. SPWLA 24th Jinn. Srnp. Trans, paper E. 1-16, ‘Chamberiain, AK. (1984) Surface gamma-ray logs: a correlation twol for frontier areas. Bull Am Assoc. Petrol Geel. 68(8) 1040 1033, Comelik, FB, Bouma, AH, and Rezak, R. (1969) Comparison of ‘lecircal logs and physical parameters of marine sediment cores, Gulf Coast Assoc. Geol Sos. Trans. XIX, 63-70. Clavier, C, Coates, G, and Durpanoir, J (1977) The theoretical and ‘experimental bases for the ‘dual water’ model for the intepre ‘ation of shaly sands. 52nd Ann. Conf. SPE of AIME, 1-18, SPE paper 6859. Cody.R.D, (1971) Adsorption and the reliability of trace elements as environment indicators for shales. J Sed, Pet. 41(2) 461-471 CCokmsn, 1M, and Prior, D.B. (1982) Deltaie environments. In: Sanistone Depasiional’ Enoironmenis, eds. Scholle, PA. and Spearing, Dy, Am. Assoe. Petrol Geol. Mem. 31, 139-178, Cooper. BS, Coleman, S.H., Barnard, P.C. and Butterworth, 5S (1975) Paleotemperatures in the northern North Sea busin. in Petroleum and the Continental Shelf of NW. Europe, cd. Woodland, A.W., Applied Sei, Publ, Barking, 487-492. CComford, C. (1984) Source rocks and hydrocarbons of the North Sea. In: Iniraduetton tothe Petroleun Geology ofthe North Sea, el Glennie, K.W.,ch, 9, Blackwell, Oxford, 171-204 Cox J.W.(1983)Longais orientation inclongated boreholesandits Corfelation wath rockstessdata. SPW LA 2ath Aen Spmp. Trans, prperd, 1-7 Davies, DIK. and Elbe, FG. (1975) Sandstone competion aod depositional envionment. Bull Am Asoc. Pett Gel 893) 239-264 Death, NG. and Schuyleman, SF (1981) The geology ofthe Maghusoltfiel fo Petroleu Geoiogy ofthe Continental Shelf of NA Earopeeds. hing, L¥-and Hobuon,G.D. Heyden London, 542-351 Deegan, CE and Seu, BJ.(1977) A proposed standard theses ‘aphicnomencltuce forthe entra and northern North Seu. ep. inst Geol Set Le, 7725, and Bul Nor, Pet. Dir 138. Desbrandes. R96) Théorie et inerprtation des dingrophis “Fetip, Para a Desbrandes, R (1982) Diopraptes dons ls sondage, Technip, Paris, p37 Dewan, 17. (1983) Essentials of Modern Open Hole Log Inter retain. Benet Tote Dickey, PA (1969) Increasing concentration of subsurface brines with depth, Chem. Geol 4361-390. Donovan, WS. ad Hicks, DW. (961) Natural gamma ray ‘emissions i the Muddy J formation in eastern Wyoming, Loy ‘ays XU (2) 17-29 e-pudication}, Dresser Alas (1902) Wel logging onl incerpretatin echnigues. The Course for tome sul. Oresat Nias Publication. Dresser Atlas (1983) Log interpretation charts Dresser Ads Publication Dyprik, Hand Eriksen. D.0. (1983) Natural adioactiviy oft ‘eliments and the contributions of J, Thad K Pet, Geol St) ‘09-416 Edmunéson, HIN, and Raymer, LL (1979) Table of ruosctive Ioaaing parsmetes of comiaon minerale.SPWLA 6th Europea Symp, Trans, Lin. Paper BB, 1-6 edwards. M_ Ovinger NHC and Haske, RE. (1967) Nukes og ealuation of potash deposits SPWWLA Bik Ana Symp. Tae, Paper b 1-12 Exnyte, L981) Fractuse idcatiaton in the Panoma Field ‘Coute Grove formation. Lag Analyst XXING) 3-6 Evang, TR, (1977) Theroal properizs of North Sea rocks, Lag ‘Analyst XVI) 3-12. Fel, WH. (€8) (1996) Abnormal Formation Pressures— ‘plications 19 Exploration, Dritingand Production of Oiland Gas Resources. Elsevier, Amstedam. Fert, W.lL (1977 Shale density studies and their aplication. In: ‘Deelopnens In” Petroleum. Geology. al. Hobson, GD. Applied Science, Barking 293-327, Feal, W.H. (1979) Gamma ray spectral data assists in complex formation evaluation SPWLA 6th European Simp, Trans. Le, Paper Q, 1-32. Ferth W.ML (1586) Advances in well logging intenvetaion. OW & Gos J. Apri 16h, 35-91 g Ferd, WH (985) Logs ep approximate reservoir temperate oi! {Gas J. Apil 2th, 89-98 Fert, Wit and Rieke, HLH. (1980) Gamma ray spectral evaluation Techniques idenilyTractired shale reservoirs and soure rock characteristics J. Pet Teck November, 2053-2002. Festh WAHT and Tinke, DJ. (1972) How downhole temperature pressures allt dling, World Oi 73-8210 par. June 1972 Mares 1993 Ferth WL, and Wichmann, PA. (1972) How to detecmine static ‘BHT from well data, World Oil Iasuary é Fisher, W.L- (196) Facies charateieation of Gulf Coast asin eta ‘ystems, with sme Helocne analogues Gulf Coat Assoc Geol Sect. ans, XIX.239-26 Fons, L (1968) Geologicst applications of well ogs SPH”LA 10k “an Sip. Trans. Pape: Ah, m4 Galloway. WE 1968) Depositions systems inthe lower Wileoxgroup, North-Central Gulf Coast basin. Gif Coast Assoc. Geo Socs. Trans XVIM1,275-289, Galloway, WE. and Hobday, D.K. (1983) Terrigenous Clastic Deposttionai Systems. Applications to. Petroleum, Coal and Uranium Exploration. Springer Veriag, New York, Gaymard, R. and Poupoa, A. (1968) Response of neviron and formation density togs in hydrocarbon bearing formations, Log Analyst 18(5) 3-12. Gearhart, Ine (1983) Formation Evaluation Data Handbook. Gearhart Inc. Publication. Gieath, JA. (1966) Electric log characteristics of diapiric shale. In: Diepirism and Diapirs, eds, Braunscin J ad O'Brien, G.D. Am,

S-ar putea să vă placă și

• Operational Procedures: 7.04 Stuck Pipe

  

  Document7 pagini

  Operational Procedures: 7.04 Stuck Pipe

  ciucalata88

  Încă nu există evaluări
• Operational Procedures: M-I/SWACO Norge AS

  

  Document5 pagini

  Operational Procedures: M-I/SWACO Norge AS

  ciucalata88

  Încă nu există evaluări
• Operational Procedures: 7.05 Wellbore Stability

  

  Document4 pagini

  Operational Procedures: 7.05 Wellbore Stability

  ciucalata88

  Încă nu există evaluări
• • Reviste
  • Podcast-uri
  • Partitură
• Despre Fluide - Utile

  

  Document60 pagini

  Despre Fluide - Utile

  ciucalata88

  Încă nu există evaluări
• Distance Learning Drilling Calculations Part 3

  

  Document13 pagini

  Distance Learning Drilling Calculations Part 3

  ciucalata88

  Încă nu există evaluări
• 7.03-7-Lost Circulation

  

  Document13 pagini

  7.03-7-Lost Circulation

  ciucalata88

  Încă nu există evaluări
• MECHANISMS, MEASUREMENT AND MITIGATION OF BARITE SAG

  

  Document14 pagini

  MECHANISMS, MEASUREMENT AND MITIGATION OF BARITE SAG

  D. Nathan Meehan

  Încă nu există evaluări
• Multilateral Wells

  

  Document133 pagini

  Multilateral Wells

  Bogdan Stanciu

  Încă nu există evaluări
• BARITE SAG CLEANING SOLUTIONS

  

  Document66 pagini

  BARITE SAG CLEANING SOLUTIONS

  ciucalata88

  Încă nu există evaluări
• Cavings Analysis: A Component of Real Time Wellbore Stability Monitoring

  

  Document11 pagini

  Cavings Analysis: A Component of Real Time Wellbore Stability Monitoring

  ciucalata88

  Încă nu există evaluări
• Cavings Analysis:: A Key To Managing Wellbore Instability in Real Time

  

  Document1 pagină

  Cavings Analysis:: A Key To Managing Wellbore Instability in Real Time

  Andrey Kazak

  Încă nu există evaluări
• Optimize Fishing Operations with Decision Trees and Calculators

  

  Document53 pagini

  Optimize Fishing Operations with Decision Trees and Calculators

  ciucalata88

  100% (1)
• L 'Ffi-,.-: Bffii

  

  Document118 pagini

  L 'Ffi-,.-: Bffii

  ciucalata88

  Încă nu există evaluări
• Distance Learning Drilling Calculations Part 3 PDF

  

  Document143 pagini

  Distance Learning Drilling Calculations Part 3 PDF

  António Pedro Tavares

  Încă nu există evaluări
• ISCWSA Error Model Rev4

  

  Document57 pagini

  ISCWSA Error Model Rev4

  ciucalata88

  Încă nu există evaluări
• Limite Functii

  

  Document4 pagini

  Limite Functii

  Diana Margiana Negru

  Încă nu există evaluări
• TC1625 Catalog Pgs30-40

  

  Document11 pagini

  TC1625 Catalog Pgs30-40

  Vanessa Rocha

  100% (1)
• Rig Selections 296

  

  Document9 pagini

  Rig Selections 296

  ciucalata88

  Încă nu există evaluări
• Designing & Managing Drilling Fluid

  

  Document11 pagini

  Designing & Managing Drilling Fluid

  fisco4ril

  Încă nu există evaluări
• Model for casing design optimization

  

  Document112 pagini

  Model for casing design optimization

  ciucalata88

  Încă nu există evaluări
• Gs Lift Design Oourse

  

  Document200 pagini

  Gs Lift Design Oourse

  Raul Dolo Quinones

  100% (1)
• Halliburton Packer Information PDF

  

  Document76 pagini

  Halliburton Packer Information PDF

  rlk2806

  100% (4)
• 477 - 7in TOL Inflow Test Program

  

  Document4 pagini

  477 - 7in TOL Inflow Test Program

  ciucalata88

  100% (1)
• Drill String Design

  

  Document30 pagini

  Drill String Design

  Victor Reyes

  Încă nu există evaluări
• Hazards and Precautions

  

  Document8 pagini

  Hazards and Precautions

  ciucalata88

  Încă nu există evaluări
• Error Ellipse Survey

  

  Document4 pagini

  Error Ellipse Survey

  ciucalata88

  Încă nu există evaluări
• Drilling Motors

  

  Document20 pagini

  Drilling Motors

  ciucalata88

  Încă nu există evaluări
• OMV Petrom: A3 Muntenia Vest 2868 Slatioarele 2868 Slatioarele G1 2868 ST G3 Rev E1

  

  Document15 pagini

  OMV Petrom: A3 Muntenia Vest 2868 Slatioarele 2868 Slatioarele G1 2868 ST G3 Rev E1

  ciucalata88

  Încă nu există evaluări
• The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  

  De la Everand

  The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  Mark Manson

  Evaluare: 4 din 5 stele

  4/5 (5783)
• The Yellow House: A Memoir (2019 National Book Award Winner)

  

  De la Everand

  The Yellow House: A Memoir (2019 National Book Award Winner)

  Sarah M. Broom

  Evaluare: 4 din 5 stele

  4/5 (98)
• Never Split the Difference: Negotiating As If Your Life Depended On It

  

  De la Everand

  Never Split the Difference: Negotiating As If Your Life Depended On It

  Chris Voss

  Evaluare: 4.5 din 5 stele

  4.5/5 (838)
• Shoe Dog: A Memoir by the Creator of Nike

  

  De la Everand

  Shoe Dog: A Memoir by the Creator of Nike

  Phil Knight

  Evaluare: 4.5 din 5 stele

  4.5/5 (537)
• The Emperor of All Maladies: A Biography of Cancer

  

  De la Everand

  The Emperor of All Maladies: A Biography of Cancer

  Siddhartha Mukherjee

  Evaluare: 4.5 din 5 stele

  4.5/5 (271)
• Fear: Trump in the White House

  

  De la Everand

  Fear: Trump in the White House

  Bob Woodward

  Evaluare: 3.5 din 5 stele

  3.5/5 (738)
• Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  

  De la Everand

  Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  Margot Lee Shetterly

  Evaluare: 4 din 5 stele

  4/5 (890)
• The Little Book of Hygge: Danish Secrets to Happy Living

  

  De la Everand

  The Little Book of Hygge: Danish Secrets to Happy Living

  Meik Wiking

  Evaluare: 3.5 din 5 stele

  3.5/5 (399)
• Team of Rivals: The Political Genius of Abraham Lincoln

  

  De la Everand

  Team of Rivals: The Political Genius of Abraham Lincoln

  Doris Kearns Goodwin

  Evaluare: 4.5 din 5 stele

  4.5/5 (234)
• Yes Please

  

  De la Everand

  Yes Please

  Amy Poehler

  Evaluare: 4 din 5 stele

  4/5 (1888)
• Grit: The Power of Passion and Perseverance

  

  De la Everand

  Grit: The Power of Passion and Perseverance

  Angela Duckworth

  Evaluare: 4 din 5 stele

  4/5 (587)
• Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  

  De la Everand

  Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  Gilbert King

  Evaluare: 4.5 din 5 stele

  4.5/5 (265)
• A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  

  De la Everand

  A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  Dave Eggers

  Evaluare: 3.5 din 5 stele

  3.5/5 (231)
• On Fire: The (Burning) Case for a Green New Deal

  

  De la Everand

  On Fire: The (Burning) Case for a Green New Deal

  Naomi Klein

  Evaluare: 4 din 5 stele

  4/5 (72)
• Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  

  De la Everand

  Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  Ashlee Vance

  Evaluare: 4.5 din 5 stele

  4.5/5 (474)
• Principles: Life and Work

  

  De la Everand

  Principles: Life and Work

  Ray Dalio

  Evaluare: 4 din 5 stele

  4/5 (599)
• Rise of ISIS: A Threat We Can't Ignore

  

  De la Everand

  Rise of ISIS: A Threat We Can't Ignore

  Jay Sekulow

  Evaluare: 3.5 din 5 stele

  3.5/5 (137)
• The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  

  De la Everand

  The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  Ben Horowitz

  Evaluare: 4.5 din 5 stele

  4.5/5 (344)
• The Unwinding: An Inner History of the New America

  

  De la Everand

  The Unwinding: An Inner History of the New America

  George Packer

  Evaluare: 4 din 5 stele

  4/5 (45)
• Steve Jobs

  

  De la Everand

  Steve Jobs

  Walter Isaacson

  Evaluare: 4.5 din 5 stele

  4.5/5 (806)
• The World Is Flat 3.0: A Brief History of the Twenty-first Century

  

  De la Everand

  The World Is Flat 3.0: A Brief History of the Twenty-first Century

  Thomas L. Friedman

  Evaluare: 3.5 din 5 stele

  3.5/5 (2219)
• Angela's Ashes: A Memoir

  

  De la Everand

  Angela's Ashes: A Memoir

  Frank McCourt

  Evaluare: 4.5 din 5 stele

  4.5/5 (440)
• The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  

  De la Everand

  The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  Brené Brown

  Evaluare: 4 din 5 stele

  4/5 (1090)
• John Adams

  

  De la Everand

  John Adams

  David McCullough

  Evaluare: 4.5 din 5 stele

  4.5/5 (2409)
• Bad Feminist: Essays

  

  De la Everand

  Bad Feminist: Essays

  Roxane Gay

  Evaluare: 4 din 5 stele

  4/5 (1015)
• The Glass Castle: A Memoir

  

  De la Everand

  The Glass Castle: A Memoir

  Jeannette Walls

  Evaluare: 4.5 din 5 stele

  4.5/5 (1711)
• The Outsider: A Novel

  

  De la Everand

  The Outsider: A Novel

  Stephen King

  Evaluare: 4 din 5 stele

  4/5 (1800)
• The Woman in Cabin 10

  

  De la Everand

  The Woman in Cabin 10

  Ruth Ware

  Evaluare: 3.5 din 5 stele

  3.5/5 (2322)
• A Man Called Ove: A Novel

  

  De la Everand

  A Man Called Ove: A Novel

  Fredrik Backman

  Evaluare: 4.5 din 5 stele

  4.5/5 (4609)
• The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  

  De la Everand

  The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  Viet Thanh Nguyen

  Evaluare: 4.5 din 5 stele

  4.5/5 (119)
• The Light Between Oceans: A Novel

  

  De la Everand

  The Light Between Oceans: A Novel

  M.L. Stedman

  Evaluare: 4.5 din 5 stele

  4.5/5 (789)
• Brooklyn: A Novel

  

  De la Everand

  Brooklyn: A Novel

  Colm Tóibín

  Evaluare: 3.5 din 5 stele

  3.5/5 (1937)
• The Perks of Being a Wallflower

  

  De la Everand

  The Perks of Being a Wallflower

  Stephen Chbosky

  Evaluare: 4.5 din 5 stele

  4.5/5 (2099)
• Sing, Unburied, Sing: A Novel

  

  De la Everand

  Sing, Unburied, Sing: A Novel

  Jesmyn Ward

  Evaluare: 4 din 5 stele

  4/5 (1103)
• Wolf Hall: A Novel

  

  De la Everand

  Wolf Hall: A Novel

  Hilary Mantel

  Evaluare: 4 din 5 stele

  4/5 (3811)
• Manhattan Beach: A Novel

  

  De la Everand

  Manhattan Beach: A Novel

  Jennifer Egan

  Evaluare: 3.5 din 5 stele

  3.5/5 (791)
• A Tree Grows in Brooklyn

  

  De la Everand

  A Tree Grows in Brooklyn

  Betty Smith

  Evaluare: 4.5 din 5 stele

  4.5/5 (1929)
• Little Women

  

  De la Everand

  Little Women

  Louisa May Alcott

  Evaluare: 4 din 5 stele

  4/5 (104)
• The Art of Racing in the Rain: A Novel

  

  De la Everand

  The Art of Racing in the Rain: A Novel

  Garth Stein

  Evaluare: 4 din 5 stele

  4/5 (4193)
• Her Body and Other Parties: Stories

  

  De la Everand

  Her Body and Other Parties: Stories

  Carmen Maria Machado

  Evaluare: 4 din 5 stele

  4/5 (821)
• The Constant Gardener: A Novel

  

  De la Everand

  The Constant Gardener: A Novel

  John le Carré

  Evaluare: 3.5 din 5 stele

  3.5/5 (104)