Search Documents

By Riki Kobayashi, D. L. Katz

The conditions at which methane and water form solid hydrates have been extended from 4,000 to 11,200 pounds per sq. in. The curve at high pressure had been in doubt because of Villard's report of a critical decomposition temperature. The paper explains the nature of the hydrate curve at high pressure and its dependence on fluid phase behavior. Extropolation of the methane hydrate curve indicates a pressure of 40,000 pounds per sq. in. will cause hydrates to form at a temperature of 100" F. INTRODUCTION The conditions under which natural gases form hydrates when liquid water is present have been studied by several investigators up to pressures of about 4,000 pounds per sq. in. The early work by Villard on hydrates of methane and ethylene13, ethane'', and propane14 included references to critical decomposition temperatures of the hydrates. Villard's work raised the question as to whether the pressure-temperature curve for methane-water hydrate continued above a temperature of 70.7°F. or whether this was the maximum temperature at which hydrates would form. This paper extends the experimental data on methane hydrate to 11,200 pounds per sq. in. A discussion of phase relations in the hydrate region for the hydrocarbons and of the methane-water system at high pressures is included. PREVIOUS WORK Villard suspected that methane hydrates could not be formed at temperatures above 21.5°C. (70.7°F.)la irrespective of the pressure of the liquid-vapor system. However, he was limited in his studies by his apparatus in 1888 to pressures not greater than 300 atmospheres (4,410 pounds per sq. in.) so that there has remained some doubt as to the existence of such a maximum temperature in this region. For the case of ethylene hydrates, however, Villard does specifically state the occurrence of what he terms a "critical temperature for the decomposition of the hydrates"13. Fig. lB shows a pressure-temperature projection of the three-phase locus, hydrate-ethylene rich liquid-vapor (H-L2-V) terminating at the critical locus in which L² and V assume similar properties. Throughout this paper the following phase designations will be used: H for hydrate, Ll for the water rich liquid. L2 for the

Jan 1, 1949

By C. W. Binckley, F. R. Burgess, E. R. Haymaker

A method of establishing stabilized back-pressure curves for gas wells producing from formations of extremely low permeability is presented. Actual well performance under many different operating conditions is shown by the stabilized back-pressure curve. By use of the method. it is possible to conduct back-pressure tests with a critical-flow prover on wells that stabilize slowly, and save approximately 88% of the gas ordinarily vented to obtain satisfactory test data, with a great reduction in time required for testing. INTRODUCTION The reasons for establishing dependable back-pressure curves on gas wells have been pointed out by previous publications. The publication most referred to. of course, is the United States Bureau of Mines Monograph 7, titled "Back-Pressure Data on Natural Gas Wells and Their Application to Production Practices". The technique generally established therein has been accepted and used by many engineers; and, when proper tests are conducted, the results can be used for the analysis and solution of several practical problems concerning field operation and development. Even where formations of low specific permeability are encountered, the determination of a well's actual performance by the back-pressure test method permits the engineer to analyze many problems in individual well operation and also to predict necessary future field development. Such problems as the determination of the ability of a well to produce into a pipe line at a predetermined line pressure, the design of gas gathering systems and meter settings, and the determination of the time and the number of wells required to be drilled to meet future market obligations, can be solved, in part., by the use of a reliable back-~ressure curve. In addition, the computed well delivery rates determined by data from backpressure tests ordered by state regtilatory bodies, when compared with the true back-Pressure curve, permit the operator to ascertain whether such data represent unstable or relatively stabilized delivery rates for given pressure conditions of the well. The technique of back-pressure testing, as described in this report, was developed by Phillips Petroleum Company engineers from data obtained during a testing program that started in 1944 and has been continued to date. Three hundred and eleven back-pressure tests were conducted on 299 wells located in the southern part of the Hugoton Field. The gas-bearing zone is composed of several dolomitic formations of the Permian Age; the important ones are the Herington, Upper Krider, Lower Krider, and Winfield. The average bottom-hole temperature is approximately 91 °F.. and the initial wellhead shut-in pressures range from 400 to 440 psig. The spacing pattern is 640 acres per well with each well located near the center of the section. The range of back-pressure potentials on wells tested was from 500 to 23,000 Mcfd. All gas wells were acidized, and the quantity of acid used, expressed in 1574 hydrocloric acid, varied from 12,000 to 22,000 gallons per well. The quantity and concentration of each treatment depended on the stage, the formation being treated, and experience gained from previously completed wells. The gas in the Hugoton Feld is a "dry" gas. It has a gasoline content of approximately 0.25 gallons per thousand cubic feet, as determined by charcoal test, and its specific gravity averages about 0.71 as compared to air (air = 1.00 at 60°F.). Of the wells tested, 71 were completed with 7" casing, 3 with 9 5/8" casing, and 1 with 6%" casing set on top of the upper producing formation with the well bore through the gas bearing formations being open hole. Two hundred and twenty-four were completed with 5 1/2'' O.D. casing set through the gas bearing formations and perforated. For the purpose of establishing reliable back-pressure curves in the area, Phillips Petroleum Company personnel has computed data on the basis of 24-hour flows per point. Early in the program, many tests were actually permitted to flow 24 hours to obtain data for each plotting point, at great expense in man power and time. Presently, however, such tests have been replaced by tests of short duration flows which can be made to effect results that correspond to the tests obtained by flows of much longer duration. METHOD When a gas well producing from a reservoir of low permeability is opened for production through a constant size orifice, both the rate of flow and working pressure decline. first at a high rate and later at a lower rate until after several hours the decline becomes difficult to ascertain. In this paper the rae of flow and working pressure are considered to be stabilized when it becomes difficult to observe changes in working pressure during a period of three hours by the use of a deadweight pressure gage. Stalibization of pressure in the literal sense is never obtained in a producing gas well. In formations of low permeability. such as those in the Hugo-ton Field, most wellhead working pressures approach stabilization closely enough to be used satisfactorily in the determination of a back-Pressure potential curve after flow periods of 24 hours. We shall therefore describe the backpressure curve calculated from ohserved rates of flow and working pressure at the end of 24-hour flow periods.

Jan 1, 1949

By Carl H. Unruh, D. L. Katz

Experimental data are presented for hydrate formation conditions for gas mixtures of carbon dioxide and methane. Equilibrium constants for carbon dioxide, defined as the mole fraction of carbon dioxide in the gas phase (dry basis) divided by the mole fraction of carhon dioxitle in the hydrate (dry basis)... were calculated usine the equilibrium constants for methane developed in an earlier paper. INTRODUCTION A series of papers have been written on gas hydrates, giving temperatures and pressures at which hydrates would form.2. 3, 4. 5, 6. 9. 12, 13, 15 Vapor-solid equilibrium constants have been developed for methane, ethane, propane and iso-butane' by which it is possible to compute for a specific gas mixture the temperature and pressure at which hydrates will form. Carbon dioxide is present in many natural gases in concentrations usually from 0.1 to 2.0 per cent. It also forms a gas hydrate with water and presumably enters the solid solution when present with a natural gas which is forming a hydrate. The equilibrium constant giving the ratio of the concentration of carbon dioxide in the gas phase to the concentration in the solid phase is required for including the effect of the carbon dioxide when computing the conditions for hydrate formation. Direct determination of the concentration of carbon dioxide in solid hydrate requires a difficult separation to be made. Therefore, this research resorted to the measurement of the temperature and pressure at which specific mixtures of methane and carbon diox- ide would form hydrates. From the equilibrium constants previously mentioned for methane, it is possible to compute the equilibrium constant for carbon dioxide, without measurement of the composition of the solid phase. Fig. 1 shows the conditions at which methane and carbon dioxide form hydrates. The methane curve FG extents to some high pressure while the carbon dioxide curve AB intersects the vapor-liquid water-liquid carbon dioxide

Jan 1, 1949

By Roscoe C. Clark

In October 1945. R. F. Farris. in an AIME paper entitled Method For Drtermining Minimum Waiting OIL Cenzent Time, presented a method for caiculating the minimum WOC* time required in oil well cementing operations. Briefly. this method consisted of shutting in a well. after the cement had been placed and noting the time required fur the wellhead pressure to reach a. maximum. This time was then multiplied by a factor of 1.5 to determine when the cement had reached sufficient strength to support the weight of the casing. As a result of this work, a number of areas have reduced their WOC time ~Wuirements with the subsequent saving of considerable rig time. The tests conducted by Farris were, for the most part, conducted under cementing temperatures in excess of 70°F. This was done because in the overwhelming majority of cases. cementing temperatures are in excess of 70°F (only in tile northern part of the country. primarily in the Rocky Mountain ares. in cementing surface casing are temperatures below 70° F. encountered, and such temperatures occur only during the winter months). The Stano-lind Research Laboratory since that time has performed a number of tests both in the laboratory and in the field to evaluate this WOC factor when cemeriting temperatures are lower than 70°F. The results of these tests indicated the need for greater WOC time in areas affeaed by abnormally low cementing temperatures. These tests indi.cated that a woc factor of 2 (Time from start to mix cement to maximum wellhead pressure x 2 = Time to drill plug) will allow sufficient time for the cement to develop adequate strength. A field test under abnormally low cementing temperatures (surface temyerature 35°F., mud discharge temperature <70°F. and cementing depth 500 ft.) in the Rangely Field, "* using the WOC factor of 2, indicated the time to drill the plug as 45 hours. Well conditions preventing getting back into the well until a 48-hour WOC time had elapsed. At this time, the plug drilled firm to hard indicating that the cement had reached its final set. It is contemplated that during next winter additional field tests will be performed to further check the results of the laboratory and field tests conducted to date,

Jan 1, 1949

By Owen Thornton

Recently equations have been presented by Rose and Bruce&apos; and by Rose², showing how the relative permeability of a reservoir rock may be determined from the capillary character of the rock. In particular, equations were developed to show the relationship between capillary pressure and the effective permeability to the wetting phase in a poly-phase flow system. The equations are as follows (nomenclature the same as in the Rose and Bruce paper) Rose and Bruce assume that the average length of path (L,), which the wetting fluid follows in flowing through a porous media is independent of the saturation to the wetting phase, so that 1 is a constant dependent only on the k rock, and the permeability to the wetting phase is a function of the saturation and of the capillnry pressure: In measuring the flow of ekctric current through partially saturated core samples, it has been observed that the electrical resistance usually is not a simple linear function of saturation. This fact indicates that the average length of path in the flow of electric current is not independent of saturation, but rather that the tortuosity of the path depends upon the average saturation to the conducting fluid as well as upon the characteristics of the rock. itself. Inasmuch as the flow of fluid and the flow of electric current in many respects are analogous, it may be indicated further that the length of path for fluid also may not be independent of saturation. The following relationship can be derived to express the resistance to flow of electric current through a core sample partially saturated with conductive wetting phase, the non-wetting phase being non-conductive: where L is the average length of path followed by the current in flowing through a length L of partially saturated core, where L is the length of current path when the core is fully saturated, and where Rs and R, are the specific resistivities of the partially saturated and saturated core sample, respectively. It will be noted that R., the specific resistivity of the core sample when S, = 1.0, is equal to the "formation resistivity factor"&apos; multiplied by the specific resistivity of the saturating fluid. If it be assumed that the average lengths of path for fluid flow and for the flow of electric current are the same, then equation (5) may be combined with equations (l), (2) and (3) to give the following relationship: The above equation suggests including a correction for tortuosity when calculating the relative permeability of a reservoir rock to a wetting phase. The correction factor may be obtained from the resistivity-saturation relationship. For instance, the following calcu- lations are obtained for the unconsoli-dated sands described by Leverett Kw Kw Sw R°/Rs Pt/Pc¹ (Calc.) (Obs.) 0.30 0.08 0.88 0.02 0.0 0.40 0.16 0.94 0.06 0.04 0.50 0.27 0.97 0.14 0.11 0.60 0.40 0.98 0.26 0.21 0.70 0.53 0.98 0.39 0.35 0.80 0.68 0.99 0.57 0.54 0.90 0.84 0.99 0.77 0.76 1.00 1.00 1.00 1.00 1.00 It will be noted that the agreement between the calculated points and the observed data is rather good. Further investigation of the above method for obtaining relative permeabilities may be merited. For many sands within specified ranges of saturations the following relationship has been found to hold approximately&apos;: — =SW"......(7) Rs The exponent n in the above equation has been found to equal two for many sands so that equation (6) reduces to the following: The writer acknowledges the permission of The Texas Co. to submit this note for publication. 1. Walter Rose and W. A. Bruce—Jnl. of Pet. Tech., Vol. 1, No. 5, p. 127 (1949). 2. Walter Rose—Jnl. of Petr. Tech., Vol. 1, No. 5, p. 111 (1949). 3. G. E. Archie— -Trans. AIME, 146, 54 (1942). 4. M. C. Leverett— Trans. AIME, 142, 152 (1941). 5. M. C. Leverett—Trans. AIME, 132, 149 (1939).

Jan 1, 1949

By Walter Rose

In a recent technical note, Owen Thornton&apos; suggests that wetting liquid relative permeability may be derived from the relationship: where Pd/Pc is the ratio of displacement pressure to capillary pressure at the wetting liquid satdration Sw, and I is the resistivity index at this saturation. Thornton shows that this expression gives values for relative permeability in good agreement with those experimentally determined by Leverett. We have recently developed another expression for krw which seems preferable to Equation 1 since it requires fewer experimental data for its verification. This equation may be derived in the following manner. Making use of the analogy between mass transfer of fluid in a porous medium and electrical conductivity through the fluid in the same medium, we can postulate a hydraulic formation resistivity factor analogous to an electrical formation resistivity factor. From Poiseuille&apos;s Law the resistivity to flow in a tube of radius R is of the form 8u/R², where is the viscosity of the fluid. Similarly7 the resistivity to flow in a porous medium of the same dimensions as the tube is given by D&apos;Arcy&apos;s Law as where k is the permeability of the medium. The hydraulic formation resistivity factor, Ff, is thus the quotient of these two resistivities, or: But permeability is defined by the KO-zeny equation as: where: r is the average pore radius of the porous medium, L. is the average tortuous length of the average pore, + is the porosity, and L is the bed length. This gives Fr as a function of the (La/L) .-.-.tortuosity ratio and porosity, or in explicit form: where: T = (La/L)&apos;, and N is the number of pores of radius, r, which will be found in any cross sectional area, x R&apos;. That is, N = R²/r². It will be clear from the foregoing considerations that the hydraulic formation fat. tor should be dependent on the value of R which in every case must be arbitrarily selected. By analogy, the hydrauiic formation factor, Fcf, characterizing the porous medium at Sw<1,will be given by: R² NT. where ke is the effective wetting liquid permeability obtaining at Sw<l, and Te is the square of the (Las/L) ratio defined by Thornton. Therefore, the hydraulic resistivity index, If, is: However, Thrnton gives I as: and it follows that: Equation 2, when checked against the data of wyckoff and Botset² and some of the Morse et al data³, gives comparisons as shown at the bottom of this page. Computations based on Leverett&apos;s data, quoted by Thorntaon, gives computed krw values somewhat lower than the experimental values. Moreover, other instances can be cited where we find that neither Equation 1 nor Equa tion 2 is checked by experimental data. The work of Botset&apos; on the Nichols Buff sandstone and Morse et a1 on oil wetted Bradford sandstone are such instances. A reason for these discrepancies may lie in the fact that if both Equations 1 and 2 are presumed always to give accurate values for krw, it follows that PC = Pd/Sw". Such an expression, it is well known, is too simple to describe accurately the capillary pressure behavior of all porous media. However, it is believed that a particular advantage resides on the use of Equation 2, since it is not dependent for its utility on a knowledge of the capillary pressures obtaining in dynamic flow svsterns. No practical technique for measuring the capillary pressures characterizing the fluid distributions in dynamic flow systems has yet been proposed. Acknowledgment is given to Dr. Paul D. Foote, executive vice-president of the Gulf Research and Development Company, for permission to publish this note. 1. Owen F. Thornton, "A Note on the Valuation of Relative Permeability," J. Pet. Tech., 1:7, Section 1, July, 1949. 2. R. D. Wyckoff arid H. G. Botset, "The Flow of Gas-Liquid Mixtures through Porous Media," Physics, 7:325-345, 1936. 3. R. A. Morse, P: L. Terwilliger, .and S. T. Yuster, "Relative Permeability Measurements of Small Core Samples," Oil and Gas J., 46:109-125, Aug. 23, 1947. 4. H. G. Botset, "Flow of Gas-Liquid Mixtures Through Consolidated Sand," Trans. AIME, 136:91, 1940.

Jan 1, 1949

Jan 1, 1949

By Leonard D. Jaffe

In recent years, the advantages of tempered martensite as a microstructure for steel parts have been well established. For parts that must not fracture brittlely when loaded at high rates, at low temperatures, or under conditions of combined stress, tempered martensite is especially desirable. One of the factors that determine whether a steel part will have a martensitic structure after heat treatment is the hardenability of the steel. Despite the large amount of recent research, there are still many unsettled basic questions concerning hardenability and its relation to chemical composition, and, indeed, considerable disagreement among investigators in the field. Much of the uncertainty that beclouds the existing knowledge of hardenability arises from the fact that hardenability is still a wholly empirical concept, and has not been quantitatively related to the more fundamental knowledge of isothermal decomposition of austenite. Hardenability is, essentiaily, a measure of the time required for small percentages of austenite to decompose, during continuous cooling, to pearlite, bainite, proeutectoid ferrite, or proeutec-toid carbide. In the practical heat-treat- ment of steel, also, the decomposition of austenite is usually not isothermal but rather during cooling, or, in general, is not at a single constant temperature. Thus, both hardenability and practical heat treatment involve anisothermal decomposition of austenite. In a previous publication1 the author and his collaborators outlined the principles that govern the relations between isothermal and anisothermal decomposition of austenite, and reviewed the experimental information available in 1945. To avoid repetition, familiarity with that article (Paper I) will be assumed, and attention given primarily to work carried out or published since 1945. The method generally employed was explained in Paper I: anisothermal transformation is studied through investigation of the effect of holding austenite at one temperature upon its subsequent decomposition at another temperature. Both temperatures are below the stability range of the austenite. This paper is confined to studies in which only a small amount of transformation occurred at the first temperature. This is the only case of importance in practical heat treatment when an essentially martensitic structure is sought. Materials and Procedure It was necessary to use different steels for the studies involving proeutectoid ferrite and for those involving proeutectoid carbide, beeause the two constituents have not been found to form in the same steel. Several steels were tried and rejected

Jan 1, 1949

By R. Woodward Moore

A graphical method of analyzing the data obtained from shallow earth-resistivity depth tests is presented. The method is based upon empirical results and has no theoretical basis. The usual apparent resistivity-electrode spacing curve is used together with a cumulative resistivity-electrode spacing curve plotted on the same sheet. The greatly reduced scale required for plotting the cumulative values of resistivity together with the effect of the summation of the individual resistivity values serves to minimize the effect of purely local surface anomalies and inadvertent errors of measurement. The point of intersection of tangents or straight lines drawn to intersect at zones of maximum curvature in the cumulative curve indicates the depth to the underlying material. Numerous figures are presented in which data from published reports and from recent field studies are analyzed and the results compared with actual depths established by borings or with depth values obtained by the use of theoretical methods of analysis. Smoothly rounded curves of apparent resistivity such as are often obtained in the field, and which have been a serious drawback to attempts to analyze the data empirically heretofore, appear to be susceptible to rather accurate analysis by the method described. The method is best suited to analyses involving shallow two-layer formations. It has been applied successfully, however, in analyzing the data obtained from tests made over shallow three-layer formations. As with most empirical methods, its chief advantage is its simplicity and ease of application. There have been published many papers that discuss the interpretation of data obtained from earth-resistivity tests when using the four-terminal method of electrode spacing developed by Wenner.&apos; The majority of these have dealt with theoretical analyses for two-layer and three-layer formations. Some sets of "master curves" have been presented for use in analyzing field data to determine the depth to the first and possibly the second horizon below the earth&apos;s surface. Although practically all of these theoretical methods of analysis have appeared to have particular merit and some have been used successfully in practice, they have been found to be of little value where the local conditions surrounding the test failed to conform to those assumed in the theory. In certain fields, particularly in civil engineering, relatively shallow explorations are often involved and geophysical methods of test must compete with the direct methods of exploration ordinarily used. Only when it can be demonstrated that geophysical methods of test can materially reduce the time and cost of a given exploration project will the civil engineer abandon direct methods in favor of the interpretations of geophysical exploration data. Empirical methods of analyzing earth-resistivity data have been used in many instances in the past. Such methods have been employed for a number of years by the Public Roads Administration in research work relative to the application of

Jan 1, 1946

By Morris Muskat

TEe method of R. W. Moore&apos; for determining subsurface interfacial depths by means of integrated curves of apparent resistivity has been analyzed theoretically. It is found that the only unique tangents that can be drawn to such curves are the asymptotes at infinite electrode spacing and the tangents through the origin at vanishing electrode spacing. Explicit expressions have been derived for the relationship between the electrode spacing at the points of intersection of these tangents and the thickness of the surface strata as a function of the conductivity parameters for the two-layer and three-layer earths. It is found that in all cases the electrode spacing at the points of intersection will exceed % of the thickness of the surface layer, and may even become indefinitely large as the resistivity of the deepest layers increases as compared with that of the surface layer. These results do not agree with the empirical findings of Moore that the intersection of the tangent lines fall at an electrode spacing very approxirnately equal to the thickness of the surface layer. Introduction In a recent paper, R. W. Moore1 proposed a new method for the analysis and interpretation of earth-resistivity measurements. This consists essentially in plotting against the electrode separation of a Gish-Rooney system the integral of the apparent resistivity with respect to the electrode spacing, and observing the breaks in the integral curves. Specifically, tangents are drawn to the various segments of the integral curves, which show appreciably different slopes, and the intersections of these tangents are used as indications of the depths of the various underlying strata. By numerous examples Moore has shown that the electrode separation at the first intersection so found agrees very closely with the depth of the bottom of the surface layer. Moreover, in a number of instances the intersections for greater electrode spacings seemed to correspond to the depths of deeper beds. From a strictly empirical point of view, the procedure proposed and demonstrated by Moore appears to be particularly interesting because of its simplicity as well as its apparent accuracy, Moreover, such a simple method would be especially valuable in view of the well-known difliculties heretofore encountered in the quantitative interpretation of resistivity data. It therefore seems appropriate to attempt a mathematical analysis of the problem to see whether the integral method of Moore has a sound basis. Such an analysis is presented here. Even on cursory consideration, it is clear that the remarkable successes observed by Moore in the application of his method are indeed surprising. For it depends upon the determination and choice of the slopes of smooth and continuous curves, which by their nature must show even more gradual- variations than the original apparent-resistivity plots that are already so difficult to interpret quantitatively. It is well established, of course,

Jan 1, 1946

By C. E. Van Orstrand

An intrusive body is a mass of igneous rock that has migrated upward, presumably from great depths. Great variations in form, composition and depth of burial occur. It is not proposed in this paper to go into all of the details of the problem. Instead, it mill be assumed here that the intrusive has the form and characteristics represented in Fig. I. According to W. 11. Emmons minerals are not deposited in appreciable quantities below the "dead line." Important amounts of ores of virtually all metals may be found around summit cupolas, but only gold, copper and zinc (in limited quantities) are usually associated with trough cupolas. Most students of ore deposits agree that temperature changes are largely instrumental in precipitating many ores in the "hood" and that the difference in temperature between the intrusive and that of the country rock may be the chief factor in determining the mineralogy of a deposit. A problem then of some interest is to correlate the principal physical features of the intrusive with the phenomena of heat conduction as evidenced by the cooling curves and the isothermal surfaces. Mathematical Theory In order to show the possible application of heat conduction to the interpretation of the processes of metalliferous deposition associated with intrusives, it will be as- sumed that the intrusive is a large par-allelopiped of dimensions, 2l, 2m, 2n. The equation that represents the flow of heat from the parallelopiped into country rock (v = o, 1 = o), which has the same thermal properties as the intrusive, is2,3 l-x v I FT 2 [2 2 — l-x m — y — m — y n — z In this equation, the origin, x = o, y = o, z = o, is at the center of the parallelopiped. v = temperature at point x, y, z, time 1. v, = initial temperature throughout the intrusive, 1 = o. k = coefficient of thermal diffusivity. a, ß, y are variables of integration. Put n = m in Eq. I, and we have v I [2/5 [2I-x 2 — l-x m — y 2 2 2 kt 2 20xt e-a2da ebrdß

Jan 1, 1946

By R. C. Hyslop

The object of this experiment was to obtain a usable set of field curves for determining the susceptibility of rocks with the vertical magnetometer. The need often arises for determining the susceptibility of rocks in the field in order to interpret magnetic anomalies obtained. This is particularly true when diamond drilling has been done and the relation to the surface rocks of the samples taken is desired. Many methods of measuring rock susceptibility have been developed, but in general they require careful shaping of the sample or pulverizing it in addition to delicate laboratory measuring instruments designed for the purpose. The requirements for a practical method for the determination of susceptibilities are as follows: 1. There should be as little alteration as possible of the specimen. 2. A minimum of time should be required in the preparation of the specimen and the calibration of the equipment. 3. Theequipment used should be portable. 4. An accuracy within the limits of the survey at hand should be obtainable. 5. A means of measuring the remnant magnetism should be available. Laboratory Procedure With these ideas in mind, the problem was approached in the manner in which Koenigsbergerl developed his method using the principle of the electrical theory of images of Lord Kelvin and applying it to the magnetic case. This solution of the problem states that the apparent change in the magnetic field intensity AZ due to the proximity of a specimen is given by: mk AZ = 2a2 Where k is the susceptibility of the specimen, m the strength of the pole of the needle, and a its distance from the surface of the specimen, as shown in Fig. r.2 This relation is true only for an infinite body and corrections are necessary for a finite body. Koenigsberger used in his work a variometer of special design. He suggested that a Schmidt vertical balance or horizontal magnetometer, as well as some other instrument, might be used. The Colorado School of Mines horizontal torsion magnetometer operates on this principle. In the

Jan 1, 1946

By R. G. Patterson

The usual procedure in determining the magnetic susceptibilities of rocks and formations has been to take samples in the field and measure their volume susceptibilities in the laboratory, using one of numerous methods available. In none of these methods is the measuring instrument suited for field use. Usually the sample is pulverized; sometimes it is machined to a desired sue and shape, and in one method1 a cast of the sample is made and the susceptibility of the rock compared with that of a material of known susceptibility of the same size and shape, made from the cast of the original sample. Some writers assert that the difference in susceptibility of a rock sample in powdered and solid forms is negligible, and while doubtless this is true in many cases there are others in which the difference becomes appreciable. Further, the susceptibility of a small specimen may differ from that of the rock from which it is taken, owing to the stresses produced by the hammering or other forces used to obtain the sample and shape it for testing. Wilson2 found that an individual specimen of granite had about the same susceptibility in the solid as in the powdered form, but that pieces from different parts of the rock varied as much as 750 per cent because of unequal distribution of magnetite. The most accurate value of susceptibility would, obviously, be obtained from averag- ing a number of tests on the undisturbed rock instead of from one or two samples. No such procedures have been published, to the writer&apos;s knowledge. It is the purpose of this paper to describe a method by which susceptibilities of rocks may be determined in place by means of a portable instrument. Method The principle involved in this new method is that there is a change in inductance of a coil due to the presence in its magnetic field of a paramagnetic or dia-magnetic medium. In this case, however, the windings of the coil do not enclose any part of the rock under test. Instead, the coil is circular, being small in cross section as compared to its radius, and consists of several hundred turns of fine wire. It is placed flat on the rock of which the susceptibility is to be measured. Thus only part of the magnetic field from the coil penetrates the rock under test. The relation between the change in field strength or inductance of the coil and the susceptibility of the medium will be developed later. To measure the change in the inductance of the coil due to the presence of a magnetic material an alternating-current inductance bridge was chosen. Description of Apparatus Of many alternating-current bridge methods applicable,3 the one shown in Fig. 1 (Owen&apos;s bridge&apos;) was adopted for its simplicity and sensitivity. In Fig. I, the test coil is indicated by PlL1. It is connected to the bridge by about

Jan 1, 1946

By Parke A. Dickey

Potential differences between strata of shale and sandstone have been recognized for about 15 years, and they form the basis of the electrical logging of oil wells. Hitherto these potentials have been ascribed solely to electro-¦ chemical reactions caused by the disturbing effects of the water in the drill hole through which they are generally measured. Observations by the author in wells empty of water and in a mine shaft suggest that potential differences between sandstones and shales are natural and specific properties of the rocks, although they are moditied in measurement by the disturbing effects of the water in the drill hole. The cause of these potentials probably is to be sought in the relative polar adsorptive capacities of quartz and clay particles. Introduction The existence of potential differences in the surface of the earth has been known for some time, and their measurement has facilitated the discovery of geological discontinuities such as ore bodies. The technique of measurement has been well established.1-3 In measurements in oil wells, it has been found that the potential of shales with respect to some arbitrary reference is very nearly constant in any one well, at least below the zone to which meteoric water has gained access. Sandstones and some other types of sedimentary rocks have a difference of potential with respect to shale that is negative arid may attain a value of 200 millivolts or more. The difference of potential, as usually measured commercially with water in the hole, is distributed over about one foot vertically at the top and bottom contacts of the sandstones. The sharpness of the curve at the formation contacts can be improved by using a short electrode that is close to the wall of the hole&apos; These potentials generally have been ascribed to the disturbing effects of the water in the drill hole. Potentials have been found, however, in wells in which no water was present. In the course of experiments in electrical logging in pennsylvania, carried out by the pennsylvania Geological Survey, observations were made of several phenomena that hitherto have not been described, to the author&apos;s knowledge, ill the published literature. The potential differences across the contacts of sandstone and shale are called ill this paper "natural potentials," as distinguished from the potentials caused by electroendosmosis (electrofiltration) and concentration differences, which will be discussed subsequently. Potentials in Cored WeLl In the spring of 1942, the author measured the potentials in an exploratory well (Day 13, near Grand Valley, Pa.) which had been cored With diamond tools. The potentials were first measured with fresh water in the hole. * null balance potentiometer was used in this and all the other measurements described in this paper. One nonpolarizing porous pot electrode, containing copper sulphate solution and

Jan 1, 1946

By Richard H. Zinszer

Application Of electric logs has been used in correlation of subsurface structure to determine the size and shape of the oil reservoir. Such a knowledge is hardly complete until saturation and productivity are determined for the various horizons. It is difficult to use electric logs successfully for this purpose because of the many factors that influence the curves. In this work an effort was made to establish an empirical correlation between formation resistivity for a series of electric-log curves and water saturation and oil productivity of sand. Electric logs were obtained by recording with 10 different electrode spacings through the same interval of a well. In the analysis of these curves all factors except two, a dependent and independent variable, were held constant. The interval studied had been cored and a large number of samples were analyzed in the core laboratory in order to compare the values of permeability and saturation with the measured values of formation resistivity. Results of this analysis indicate: (I) the depth to which mud filtrate enters a sand body, (2) a correlation between permeability and sand resistivity for shallow penetration depths, and (3) that the water saturation for an oil sand can be determined from the measured sand resistivity after making suitable corrections for sand thickness. The procedure in the last method depends on a correction for thickness factor and a theoretical curve. The accuracy of the theoretical curve is now being tested in the laboratory Introduction In an effort to place the analysis of electric logs on a quantitative basis, there is need for a better understanding of the many factors that influence an electric-log curve. These factors may be large differences in sand thickness, permeability of the sand, salinity of the interstitial water, mud resistivity, formation temperature, and other sources. A number of examples were presented by Archie,&apos; which treat the effect of sand thickness on measured resistivity. He compared the true resistivity of a sand with the observed resistivity for sands 8 it., 16 ft., and 24 ft. thick, and found the observed resistivity of the 8-ft. sand to be approximately

Jan 1, 1946

By H. G. Doll

&apos;This paper discusses a method and apparatus for determining the dip of formations traversed by a drill hole, by means of electrical measurements in the hole. The process consists in recording the Spontaneous Potentials present along three oriented generatrices in the hole. By comparison of the three curves thus obtained, after application of a correction factor to compensate for any deviations of the hole, the dip angle and the dip direction are determined. Excellent accuracy is obtained for dip angles over 10°. Because the method is simp1e and the apparatus rugged, the measurements are rapid and reliable; consequently, they can be made at all the levels that are of interest at any time after the well has been drilled. Mechanical cores are not needed to evaluate the dips. A convenient, time-saving procedure is to perform the dipmeter survey at the time the electrical log is taken, using for both operations the same cable and truck. Actual examples of results obtained in wells of the Gulf Coast area are given. Introduction The problem of determining the angle and direction of the dip of the strata traversed by a drill hole has practical interest, and sometimes vital importance, which need not be emphasized to the oil industry. The need for that type of information has existed ever since wells have been drilled. Certainly it has not subsided, in spite of the numerous inventions and improvements made in other fields of the art of drill-hole exploration. The few examples given in this paper are but illustrative of this fact. In the past, methods have been devised, and used, to solve this problem, but generally they have met with limited success, possibly because they failed more or less to answer the requirements of a modern age for fast yet accurate measurements. If accurate, the results were obtained at the cost of very tedious and lengthy operations. The new S. P.* dipmeter service offered to the oil industry is the result of exhaustive and in this particular field. In 1032 a dipmeter and electromagnetic teleclinom-eter service was made available. The apparatus employed then was of a rather delicate and costly construction, and the measurements had to be made with precautions sometimes difficult to secure on the rig. Although the information given by the electromagnetic dipmeter was limited to the direction of the dip, very interesting results were obtained in California, Trinidad, Venezuela, Rumania and, most notably, in the Netherlands East Indies.+ Apparatus The S. P. dipmeter surveys present none of the drawbacks mentioned above. The apparatus is rugged and free from constructional complications, therefore well adapted to field use. This, combined with the new, basically simple approach to the problem, explains the rapidity of execution of the measurements. The results are remarkably accurate and they are complete, giving the

Jan 1, 1946

By W. M. Rust, W. D. Mounce

The almost universal acceptance of electrical logging by the petroleum industry calls for a critical examination of the physical bases of the common methods. This is particularly needed for the natural potential log. The conventional explanations are based on the phenomena of electrofiltration and electro-osmosis. However, experimental data show that they fail to give a basis that is quantitatively sound. Moreover, they lead to interpretations of details of electrical logs that are erroneous. It was found experimenlally that when a plug of shale is placed in a chain of dissimilar electrolytes, a current flow results. Experiments on oil-well cores and fluids similar to those existing in wells show that potentials are set up which are in reasonable agreement with those encountered in electrical logging. Doubtless other sources of potential are important but this explanation not only accounts for much of the potential but also clarifies some fairly familiar examples of misinterpretations. Introduction The extraordinary value of electrical well logging as a tool for petroleum engineers and geologists is now so universally recognized that it is not necessary to point out the ways in which it can be useful. It is sufficient to mention only the saving that has resulted from the elimination of much coring that would have been necessary under former practices and the considerable additions to reserves resulting from the discovery of sands missed in coring. Paradoxically, the very remarkable success that has resulted from the application of electrical logging to correlation and completion problems has cost some of its most consistent users dearly.&apos; The fact that some operators have placed unwarranted faith in erroneous interpretatioils of electrical logs does not reduce the value of the logs. It should, however, make one pause long enough to ask why experienced users should be misled. The answer to this question is not hard to find. In nearly every case, the fault has been oversimplification. This is illustrated very aptly by the familiar interpretation and theory of the natural potential log. Recording and Interpreting Logs The natural potential log is extremely simple to record precisely enough for all present-day uses. One needs only an electrode, a cable to lower the electrode through the drilling mud, a recording meter and a surface ground connection. The electrode can be elaborate but need not be. A simple brass rod 2 in. in diameter and 2 or 3 in. long is quite satisfactory. The cable need not be particularly well insulated, nor need it have a very low resist-

Jan 1, 1946

By Warren J. Jackson, John L. P. Campbell

Automatic recording of the radioactivity of the earth&apos;s formations provides a log of relative intensities that, if properly interpreted, can be applied to oil-field engineering. Production, engineering, and geological departments regard the radioactivity log as a forward step in the securing of more conclusive information for successful well completions. Explanations of the technique, together with some of the problems to which radioactivity logging have been applied, are presented in this paper. Introduction Those responsible for the completion of new oil and gas wells, or with planning workover operations on old wells, have followed closely the development of radioactivity well logging because they are interested in reducing the number of unknowns that usually are present in such work. Many operators have made use of the log, or are acquainted with at least one of the applications, but it is unlikely that any one operator is familiar with all of the many applications that are possible. The purpose of this paper is to present a representative example of each of the many applications, wherever the information has been released by an operator, and to illustrate by hypothetical examples other applications, releases for which have not been given. In this way, it is hoped that a better understanding of the scope of this new engineering tool may be obtained. Athough the literature contains many references that describe in detail the theory, development, and application of radioactivity well logging, a brief discussion of the technique may be helpful to those who are investigating the subject for the first time, All rocks contain radioactive material in varying concentrations. In general, shales contain relatively more radioactive material than sandstones or limestones. These radioactive materials disintegrate with time into other materials of lower atomic weight, and in that process of disintegration the rocks are emitting many rays, the most penetrating of which is the gamma ray. The intensity of emission of the gamma ray is proportional to the quantities of radioactive materials present. A higher rate of emission would be observed, therefore, opposite a shale than opposite a sandstone or limestone. Since measurable intensities of gamma rays are capable of penetrating as many as five concentric strings of casing, and as the gamma-ray intensity is relative to the formations, the measurements of variations in gamma-ray intensities offer a means of identifying cased-off formations in their proper stratigraphic sequence. TEe Gamma-ray Curve The measurement of gamma-ray intensi ties is accomplished by means of an ioniza-tion chamber that consists of a heavy cylinder, about 3 ft. long, that contains two insulated electrodes and is Med with an inert gas under high pressure. Under normal conditions no current will flow through the gas when an electrical potential is set up between the electrodes, but when

Jan 1, 1946

By G. E. Archie

Data given in this paper indicate that certain relations exist between different characteristics of sandstone. These relations are not rigid, for all sandstones are more or less heterogeneous. Not every increment of a given horizon follows exactly the relationship; however, studying a reservoir as a whole from a large number of samples, definite trends are found to exist. Apparently nature&apos;s processes of sedimentation and diagenesis are such as to give some degree of uniformity. Average relationships are valuable because electrical measurements in a borehole comprise a considerable volume of rock near the hole. Also, it is desirable that a given producing reservoir be studied as a unit. In a permeable reservoir whose pores are filled with a saline solution, it was found that the total porosity and the type of sandstone influence the relationship between its electrical resistivity and the RO and the resistivity of the brine R,. The relation can be expressed by the following empirical equation, Ro = R10O— [i] where 0 is the porosity expressed as a fraction and the numerical value of the exponent lies between 1.8 to 2.0 for consolidated sandstones, and is about 1.3 for clean, unconsolidated packs. This equation indicates that an appropriate electrical log can be used to estimate the porosity of a saline water-bearing reservoir when Rw and the type of rock are known. The fraction R0rw, is called the "formation resistivity factor." It reflects the effect of the nonconducting rock structure on the resistivity of the salt water that fills the tortuous paths through which the electricity must pass. The formation resistivity factor F is found to be related to permeability k. In this case the type of rock plays an important role, as shown by the relation F = ak-b [2] where the values of a and b must be determined experimentally for each horizon. Therefore Eq. I rather than Eq. 2 is used when it is desired to determine the value of F, because small changes in rock structure particularly in the pore size are not critical. Another useful relation, R = R0S- [3] where R = resistivity of rock when partially saturated with brine and partly with oil or gas, S = fraction of the pore space filled with brine, and n equals approximately 2 for both unconsolidated and consolidated sandstone, shows that the connate water of an oil or gas reservoir can be estimated in situ from a resistivity log. Equation 3 shows that in order to estimate the water content of a reservoir a knowledge of R and Ro is essential. The first value can be obtained from the electrical log when all factors influencing the resistivity curve can be properly taken into account. The latter may be obtained when a log is available on the same horizon where it is entirely water-bearing and has

Jan 1, 1946

By William M. Newton, John E. Sherborne

The value of the electric log as a means of interpreting underground structures has been increasingly demonstrated by its almost universal present-day use. It becomes important, therefore, to be able to evaluate the factors that affect the values obtained. By far the greatest number of electric logs are measured today in boreholes filled with rotary mud, the properties of which may exert an important influence on the electric log. In an effort to establish what the order of magnitude of such effects might be, a series of investigations was made in the laboratory and the field. Apparatus suitable for either field or laboratory use was designed and a procedure established for determining the factors that affect the resistivity of drilling muds over a wide variation in mud condition. Five muds representative of the types of clay-base muds employed in California were tested and the following observations were made: 1. The effect of raising the temperature from 8o° to i8o°F. is to decrease the resistivit~ of the mud or filtrate approximately 50 per cent. 2. The resistivity of the mud in most cases closely approximates that of its filtrate. 3. The change in the resistivities of muds caused by the addition of chemical is not the same function of the amount of chemical added for each mud. 4. The effect produced by increased sodium chloride content is to reduce markedly the resistivity of the mud and its filtrate. 5. Weighting materials, such as baroid and limestone, tend to increase the resistivity of drilling mud. 6. Cement and counteracting reagents reduce mud resistivity. The results of the investigation indicate that the mud resistivity is materially affected by the addition of electrolytes whether these are added for the purpose of conditioning the mud or enter the mud when salt beds or highly saline structures are penetrated. It should be observed, however, that the magnitude of the effect produced upon the electric log by changes in mud resistivity is also dependent upon the relative salinity of the water en. countered in the formation being logged. The laboratory data are supplemented by field data, which indicate that both the self-potential and resistivity curves are influenced by mud resistivity.

Jan 1, 1946