Search Documents

By E. S. Machlin, S. Weing

GRAIN boundary stress relaxation has been the subject of several investigations in recent years, but as yet the phenomenon is not well understood. One of the major difficulties has been the lack of a sufficient amount of data with which to evaluate any proposed mechanism. Little information is available on the effect of small amounts of substitutional solutes on grain boundary stress relaxation. It is with this important phase of the problem that this investigation is concerned. The influence of substitutional solutes on grain boundary stress relaxation has been recently investigated in copper by Pearson: and in aluminum by Dorn et al.' Pearson studied the effect of zinc, gallium, germanium, and arsenic on the relaxation peak of OFHC copper.* The pure OFHC copper decreased the peak intensity and a second peak appeared at a higher temperature. This latter peak was termed the alloy grain boundary peak. At the lowest binary concentration studied, approximately 1 atomic pct, the primary peak was almost completely obliterated. Pearson found that the activation energy calculated from the alloy grain boundary peak for all solutes, irrespective of the concentration, was 44,000 cal per mol. Pearson concluded that the increase in activation energy from 33 to 44 kcal per mol is not continuous but abruptly changes from the lower to the higher value. As no attempt was made to investigate the intermediate range of concentration and hence calculate the possible variation of the activation energy for the initial grain boundary peak, this conclusion is hardly valid. In contrast to these results, Dorn et al. studied the effect of zinc, silver, copper, germanium, and magnesium on the grain boundary stress relaxation of aluminum. No change in the energy of activation was observed for binary alloys of the flrst four solutes, but a linear increase was found for increasing concentration of magnesium. Saturation was not found up to the maximum of 1.6 atomic pct. The energies of activation were obtained by plotting the horizontal displacement of the dynamic modulus curve vs temperature for two vibrational frequencies. As the modulus is proportional to the square of the frequency, this requires the measurement of the torsional frequency over a range of temperature. However, the technique does not lead to an unambiguous interpretation of the grain boundary

Jan 1, 1958

By L. Segle, R. Resnick

NUMEROUS experiments in the past few years have proven that void formation occurs quite generally in the diffusion zone of bimetal specimens1-4 The phenomenon has been explained in terms of the accumulation of lattice vacancies resulting from unequal diffusion of the constituents and their eventual precipitation as holes.= Attention has recently been directed to the fact that voids also form within homogeneous metals during high-temperature creep.' The idea has been put forward that these voids form by the condensation of excess lattice vacancies generated by intersecting dislocations.".' If the precipitation of lattice vacancies from solution is treated as a nucleation and growth process, the predicted vacancy concentration for homogeneous nucleation of voids at an observable rate is 100 to 1000 times the normal thermal equilibrium concentration. It appears that voids actually form during diffusion at vacancy supersaturations which are much lower than this8 According to the calculations of Machlin,' the same must be true for voids formed in homogeneous metals during creep, if these are considered to arise from the precipitation of lattice vacancies. Therefore, if the vacancy mechanism is valid, it seems that the holes which form in diffusion and creep specimens are heterogeneously nucleated by imperfections or impurities in the metal. Irregularities in the frequency and distribution of voids in diffusion couples bear out this explanation.' , It is not known at the present time what the nuclei are on which vacancies condense. Various investigators have speculated that they could be certain dislocation configurations, already existing voids produced during solidification or working, gas pockets, or inclusions in the metal. Relatedly, the precise value of the degree of vacancy supersatura-tion for void nucleation is still uncertain. An initial attempt to estimate this value from diffusion data, made by Balluffi, indicated that voids form in diffusion couples at vacancy concentrations which may

Jan 1, 1958

By B. C. Raynes

IN the early part of 1952, under the auspices of the U. S. Atomic Energy Commission, Horizons Inc. undertook an investigation dealing with the preparation of high purity thorium metal in order to develop an economically competitive production process for this metal based on a recovery of the metal by a fused salt electrolysis, since a survey of the literature of thorium chemistry indicated that electrolytic methods had not received much attention, despite the consideration that electrochemical production methods often offer attractive commercial potentialities. Thorium metal has been prepared by methods which fall into four general categories: 1) reduction of halides or double halides; 2) reduction of the oxide; 3) thermal decomposition of halides; and 4) electrolytic methods. Some of the important results obtained by various investigators since about 1900 in recorded attempts to prepare pure thorium metal are summarized in Table I. A bibliography of pertinent literature is included in this paper. Thorium is strongly electropositive and possesses a great affinity for oxygen and nitrogen. As a result the prospects for the production of high purity thorium from aqueous media or oxygen bearing compounds are remote. The most promising and profitable area of investigation for production of thorium metal lies in fused salt electrolysis. Thorium has been produced previously by this technique from halide salts in fused alkali chloride eutectic mixtures. Until recently, however, the lack of known large uses for thorium has inhibited the investigation of these methods. The basic problem in an investigation into fused salt production methods for thorium is to develop a suitable fused salt system for the economic preparation of thorium. The first portion of the work in these laboratories was a small scale investigation of the electrolysis of potassium thorium fluoride, KThF5, under a variety of conditions. High purity thorium was prepared in this manner, but the process did not appear to offer the simplest and most attractive continuous commercial operation. The work reported here relates to an all-chloride system employing thorium chloride as the cell feed. ThCL4 was produced as an anhydrous material stabilized in molten sodium chloride. This product, further diluted in NaCl, was subjected to electrolysis. The electrolytic process has been studied in cells capable of holding up to 50 lb total salt charge. Ductile, pure thorium metal has been reproducibly obtained in this operation. Considerable experimentation was carried out in the preparation of ThCL as a cell feed free from moisture, oxides, and other impurities. Thorium was cathodically deposited as a coarsely crystalline powder interspersed with residual bath salts. Aqueous procedures for separating and recovering the metal were developed. The thorium metal powder obtained has been evaluated with respect to its hardness, chemical purity, and certain other physical properties. Preparation of ThC1,—Thorium chloride, and the chlorides of some other metals (aluminum, beryllium, tantalum, zirconium, etc.) cannot be obtained from aqueous solution by direct heating and evaporation. Evaporation and heating of a solution of thorium chloride yields a hydrated salt or partially hydrated salt, and finally Tho,. Anhydrous thorium chloride is obtained by forming a hydrated ammonium chloride complex salt in acidified water solution, dehydrating this compound, and finally decomposing the double salt and subliming the ammonium chloride component. Various methods for the preparation of anhydrous thorium chloride, in accordance with this concept, were investigated as indicated in Table I1 from hydrated thorium chloride (ThCl, . 3H2O), and from thorium oxycarbonate (ThOCO,). In the method found most satisfactory for this work, thorium oxycarbonate was dissolved in an excess of concentrated HC1 to give a clear yellow solution. At least 2 mol NH,Cl were added to the solution per mol of contained thorium. Ammonium thorium chloride hydrate was crystallized from solution by cooling and saturation with HC1. The resultant complex, (NH,Cl),ThCl;xH,O, was dehydrated by evaporation at 130° to 150°C and thoroughly dried at about 250°C. The dried ammanium thorium chloride was then mixed with sodium chloride in proportions to form NaThCL or to produce a ThCL4-NaCl mixture with a given thorium content. The ammonium chloride was sublimed off at temperatures above 500°C in an inert atmosphere furnace to give an anhydrous chloride source electrolyte, essentially free of Tho, and of NKC1.

Jan 1, 1958

By A. B. Dyes, P. H. Braun, B. E. Cole

This paper presents a reservoir engineering analysis of the performance data for a waterflooding project in the Johnson Lease, South Ward field. This type of engineering analysis adds to the knowledge of the mechanics of water flooding. This study reveals that in general (a) high ultimate areal sweep-out efficiencies (95 to 100 per cent) can be expected, and (b) cross-flow among strata of different permeabilities during the fill-up phase can have an important influence on the behavior and should be dealt with in studying reservoirs for water flooding. The analysis further showed that the recovery predictions based on what was considered a consistent interpretation of core data and logs were low because little contribution was assumed from the low permeability, oil-bearing sand sections which have apparently yielded considerable waterflood oil. This project has yielded in excess of 2 1/2 times as much oil by water flooding as by primary depletion. These very favorable results were achieved without the use of any special injection well completion techniques. However, all of the low permeability material war included in the section open to injection despite its poor estimated primary yield. LEASE PERFORMANCE HISTORY The South Ward field lies in the southeast corner of Ward County near the southern tip of the tremendous Yates trend. Deposition of sand bodies along the seashore reefs during the latter part of the Permian period established what have turned out to be ideal conditions to form a great number of independent productive sands. Probably, these reservoirs in Ward and Winkler counties are providing operators with some of the nation's richest sources of secondary oil reserves through the use of waterflooding techniques. In South Ward field water has been very successfully used to flood oil from the Bennett sand occurring at a depth of about 2,300 to 2,400 ft. About 150 ft above the Bennett sand reservoir, the Grand Falls sand produced gas from a large cap extending over most of the field. The Atlantic Refining Co.'s 320-acre W. D. Johnson lease (Fig. 1) is on the northwest edge of the structure with only gas cap apparently present in the Grand Falls sand. The Bennett sand under this lease is oil productive, however, having a gross sand thickness averaging about 80 ft with a net of 18 ft of fairly clean, highly permeable sand, and about 36 ft of oil-bearing sand with low permeability and high water saturation. The lower permeability sand is interbedded with numerous layers of lime, shale, and anhydrite. Radioactivity logs from 12 wells and cores from two, examples of which are shown in Fig. 2, reveal that the total Bennett sand pore volume is about 71,000 bbl/acre. The estimated initial water saturation averages 41 per cent in the 54 ft of sand considered to contain oil; thus, also considering an estimated 18 per cent oil shrinkage, the lease originally contained about 34,000 bbl of oil per acre. Original pressure was probably about 1,400 psig. PVT correlations indicate that the reservoir fluid was originally saturated with gas. After discovery of the field in 1930, Atlantic completed the first well on the Johnson lease in March,

Jan 1, 1958

By C. R. Garr, A. R. Troiano

FLAKING or hair-line crack formation has been a major problem confronting the producer of large alloy steel forgings.' Today it is generally conceded that hydrogen in one or more forms in alloy steel sections is a requisite for flaking. Hence an obvious remedy is a thermal cycle that reduces the hydrogen content to some low level. In practice, this generally means holding the material at some subcritical temperature for a period of time depending on the size of the ingot or forging. Another method that will suffice in some cases is to cool large forgings slowly, i.e., bury in ashes. In practice, this hydrogen bake is sometimes accemplished inadvertently by a series

Jan 1, 1958

By Dilip K. Das, Douglas T. Pitman

ONE of the major uses of barium in metallic form is as a getter material in vacuum tubes. Because of the high chemical reactivity of the metal, Ba-Al alloys are extensively used. Numerous methods for the preparation of Ba-Al alloys have been published, a few of which1-4 are cited here. Most of these methods were found to be quite elaborate, involving the reduction of BaO, and not too well adapted for the close control of the final composition of small amounts of alloys prepared for laboratory use. A simple laboratory method for the preparation of Ba-A1 alloys in small batches starting from pure metals was devised, so that it was possible to control the desired compositions to within 1 pct. The pertinent features of the alloy system Ba-Al5 are 1) an intermediate compound BaAl, with the melting point of 1050°C, and 2) a eutectic between aluminum and BaAl4 at 98 pct Al. The accompanying sketch shows the experimental arrangement for the preparation of the alloys. Weighed amounts of aluminum and barium were placed in an alumina and a stainless steel crucible, respectively. According to the supplier's specification, the purity of the metals used in the alloys is as follows: a) aluminum rods—99.9 pct Al, and b) barium rods—99.5 pct Ba. The stainless steel crucible, tapered at the bottom and having a 1/16 in. diam hole, rested on top of the alumina crucible. The assembly was placed inside a graphite sleeve which rested on a refractory platform. The platform moved the assembly up and down through the field of a radio frequency coil. A glass bell jar was placed between the crucible assembly and the radio frequency coil to maintain a steady flow of helium around the melt. A small window was cut out on the wall of the alumina crucible to observe the progress of the re- action and to record the temperature with an optical pyrometer. The platform was first raised high enough to move the barium out of the radio frequency coil field in order to allow only the aluminum to melt. The assembly was then lowered so that the barium began to melt and flow out through the small orifice into the molten aluminum. In order to keep the violence of the exothermic reaction under control, the rate of flow of barium was carefully regulated by raising or lowering the crucible assembly. All the samples prepared by this technique were examined by a Norelco X-ray diffractometer using CuKa radiation, The diffraction specimens were prepared by placing the finely powdered samples in flat specimen holders. The Ba-Al alloys prepared with a high barium content were found to consist mainly of BaAl4. The structure of BaAl4 has previously been reported by Alberti and Andress.8 They found that BaAl4 was body-centered-tetragonal with an a0 = b0 = 4.530Å and c0 = 11.14Å. An alloy whose composition was found by chemical analysis to be almost 100 pct BaA1, was used to determine the relative intensities. The d-spacings were obtained from the same alloy to which a small amount of tungsten had been added as a calibrating material. Accurate values for a, and c, were calculated according to the method proposed by Taylor and Floyd.' The calculated values are: a, = b, = 4.566Å and c0 = 11.250Å. The measured d-values for BaAl4 are shown in Table I along with relative peak intensities above background and hkl indices. Acknowledgment The authors are grateful to L. J. Cronin, the head of the Techniques Dept., for suggesting the problem and for his constant interest. References 1Froges and Camargue: German Patent No. 809107, 1951. French Patent No. 935324, 1949. 2E. Bonnier: Annales de physique, 1953, vol. 8, pp. 259-312. 3M. Orman and E. Zemhela: Prac Institute of Metals; 1952, vol. 4, pp. 437-445. 4E. Fujita and H. Yokomizo: Reports Gov. Chemical Industrial Research Institute, Tokyo, 1952, vol. 47, pp. 291-297. 5E. Alberti: Ztsch. fur Metallkunde, 1934, vol. 26, p. 6. 6E Alberti and K. R. Andress: Ztsch. fur Metallkunde, 1935, vol. 27, p. 126. 7 A. Tnslor and R. W. Floyd: Acta Crystallographica, 1950, vol. 3, p. 285.

Jan 1, 1958

OFFICERS AND DIRECTORS President and Director AUGUSTUS B KINZEL, '60, New York City Past President and Director GROVER J HOLT, '59, Ishpeming, Michigan President-elect and Director HOWARD C PYLE, '61, Los Angeles, Calif

Jan 1, 1958

By Oren C. Baptist, Eliot J. White

Samples of four reservoir sands having different degrees of water sensitivity were subjected to several laboratory tests to determine differences in capillary behavior attributable to clay-mineral effects. The degree of water sensitivity was estimated by the difference between gas and water permeability. Unsaturated pore volumes determined with mercury injection were comparable to irreducible water saturations determined with the semipermeable bar-rier method only for the non-sensitive sands that contained small amounts of non-swelling clay minerals. Irreducible saturations were independent of salinity of the water, which suggests that most residual water is adsorbed 0.7 a film over solid surfaces and that internal retention by clay minerals is relatively insignificant. The water-sensitive sands imbibed water to higher saturations than did non-sensitive sands, and this result is believed to be a combination of the effects of clay minerals and preferential wettability; however, the relative importance of each of these effects is riot determinable. Thus it is concluded that the semipermeable barrier method is the best of the laborafory methods used for estimating formntion water .saturations in the sands studied. INTRODUCTION This investigation is a continuation of studies of the role of clay minerals in petroleum production in the Rocky Mountain area. Previous work has shown large differences in gas and water permeability in numerous reservoir sands; these water-sensitivity effects observed in the laboratory have been correlated with well behavior in the field, and the degree of water sensitivity imparted to reservoirs by the various clay minerals has been determined.'.' The effect of interstitial clay minerals on formation water3 saturations is not so well known. It has been reported that formation water saturations probably are higher in clayey than clean sands,' but this effect has not been evaluated in the laboratory, and the relative effect of the various types of clay minerals on formation water saturation has not been determined. The clay-mineral groups of most importance to this study and to petroleum production in general are kaolinite, illite, and montmorillonite. The minute layers of montmorillonite are loosely bound to one another so that water enters between them, causing great expansion of the particles. Kaolinite and illite do not exhibit interlayer hydration, but they do immobilize considerable amounts of water owing to their large surface areas, which are of the order of 15 to 97 m2/gm of clay.5 The four Wyoming sands used in this study were selected to represent a wide range of water sensitivities. These sands, with the number of wells and fields from which the samples were obtained, are: Second Frontier, five wells in two fields; Newcastle, six wells in one field; Tensleep, two wells in two fields; and Lower Muddy, five wells in one field. EXPERIMENTAL METHODS AND PROCEDURES Water-displacement data were obtained by the semipermeable barrier method," using both fresh water and brines having either 16,500 ppm sodium chloride or calcium chloride. A displacement pressure of 35 psi was found to be sufficient to obtain the practical irreducible water or brine saturation (IWS, IBS) in most samples. Mercury-injection data were ob-tained with equipment and pro-cedures similar to those described by Purcell.7 Exceedingly high pressures are required to attain complete mercury saturation by this method; however, since saturations in most samples change very little with large pressure increases above 1,000 psia, the maximum pressure used was 1,400 psia. To facilitate comparison of mercury-injection and water-displacement results, it is convenient to express mercury-injection data in terms of unsaturated pore volume. The abbreviation, (100 - SHg)min, is used to denote the pore volume unsaturated by mercury at 1,400 psia injection pressure. Gas permeabilities, ki, were deter-mined by the method proposed by Klinkenberg.8 Methods of determining types and amounts of clay minerals by X-ray diffraction analyses, as well as methods of determining water permeability, have been described previously.' The imbibition characteristics of

Jan 1, 1958

By John C. Shyne, Eric R. Morgan

BORON as a hardenability agent of commercial importance has been the subject of extensive study in recent years. It has been suggested in the past that boron increases hardenability by combining with nitrogen, thus rendering it innocuous.&apos; More recently it has been proposed that boron increases hardenability by reducing the free energy of sites of ferrite and pearlite nucleation.&apos; The segregation of boron to grain boundaries, the primary site of nucleation, would account for the large hardenability effect of minute additions of boron in steel. This grain boundary theory has found general acceptance while the concept of boron as a nitrogen scavenger has been disregarded. However, the interaction of boron and nitrogen in steel is considered to be important to hardenability because boron may be made ineffective by combination with nitrogen.". * The present work was undertaken to examine some of the effects of boron and nitrogen arid their interaction in steel. A base composition of moderate hardenability was chosen. This was a 0.35 pct C steel containing 2.50 pct Ni and 0.30 Mo. This composition was used because it contained no strong carbide, nitride, or Table I. Composition of Steels by Analysis, Wt Pct Steel C Ni Mo 0&apos; B N* Base 0.35 2.50 0.30 0.0012 nil <0.0001 Boron 0.35 2.50 0.32 0.0012 0.0022 <0.0001 Nitrogen 0.33 2.60 0.35 0.0015 nll 0.0052 Boron plus nitrogen 0.35 2.48 0.30 0.0006 0.0020 0.0040 * Obtained by vacuum fusion technique. boride-forming elements. The other alloys examined were modifications of the base composition and contained nitrogen, boron, or nitrogen plus boron. The compositions of the four alloys used are listed in Table I. The alloys were vacuum melted from electrolytic iron, electrolytic nickel, ferromolybdenum, and fer-roboron. When required, nitrogen was added by admitting a partial pressure of nitrogen over the melt after vacuum melting. The alloys were cast into 21h-in. diam ingots and hot rolled to %-in. sq bars. No ingot pattern was discernible on the polished and etched cross sections of the hot-rolled bars. The alloys were normalized at 1650°F, then machined into standard %-in. diam end-quench hardenability test bars. Four different austenitizing treatments were used: 60 rnin at 1550°F, 45 rnin at 1800°F, 30 rnin at 2000°F, and 20 rnin at 2000°F followed immediately by 30 rnin at 1550°F. The 1550" and 2000°F treatments were carried out in duplicate for each of the four steels. In order to prevent surface oxidation the bars were austenitized while buried under charcoal in an atmosphere of argon. After quenching the bars in a conventional end-quench fixture, the hardness surveys were made in the usual fashion on parallel flats ground along opposite sides of each bar. These were ground 0.050 in. deep rather than the conventional 0.015 in., to avoid decarburized or deboronized regions.5 The austenite grain sizes which resulted from each heat treatment were observed metallographically using Vilella&apos;s etch for martensite. The criterion used as a measure of hardenability was the distance from the quenched ends of the bars at which a hardness of RC 35 was observed. This hardness represented the inflection point on the hardness vs distance plot. Fig. 1 shows the hardenability of each steel after the several heat treatments. The duplicate tests demonstrated excellent reproducibility; the distance to Rc 35 was reproduced within 1/16 in. for duplicate test bars. All four alloys had the same grain-coarsening characteristics. Neither boron nor nitrogen had any observable effect on grain size. The grain sizes resulting from the various austenitizing treatments were ASTM 7 at 1550°F, ASTM 4 at 1800°F, and ASTM 2 at 2000°F. The small amount of boron in the boron steel greatly enhanced hardenability when the samples were austenitized at 1550°F. The low hardenability of the boron plus nitrogen steel showed that nitrogen eliminated the boron contribution to hardenability. Both steels containing nitrogen, with and without boron, exhibited lower hardenability than the base composition when quenched from 1550°F.

Jan 1, 1958

By Goro Shimaoka, Kichizo Niwa

IN the past, the crystalline states of graphite in cast iron have been studied by many investigators. Recently several authors 1-4 have published interesting information on the crystal texture of the individual graphite grain, especially spherulites in nodular cast iron. In regard to the crystal structure of graphite, Lipson and Stokes0 howed by powder X-ray analysis that most of the crystals contain about 14 pct rhombohedra1 structure (ß-form) which differs from the hexagonal structure (a-form) in the sequence of layer stacking. Hoerni6 onfirmed the existence of the ß-form in a single crystal with electron diffraction by Kossel-Moellenstedt&apos;s method. Furthermore, Hoerni and Weigle7 found extra reflections in electron diffraction patterns, which are given by a unit cell twice as large as the usual one in the basal plane. Recently, Lukeshs observed graphite twinning and satellite reflections in his X-ray work and considered them as evidence for the existence of an orthorhombic superstructure. Owen and Street" and Matuyama" studied the structures of graphite crystals in several types of cast iron, using the X-ray powder method, and they showed that the graphite crystal contains a certain amount of the ß-form, an amount which varies with the sources of the cast iron used. The present study was made in order to secure structural information on graphitization in cast iron. Electron diffraction studies on the structure of single graphite flakes segregated from molten carbon or silicon-rich iron at 1200" to 1500°C are reported. Experimental Honkeiko pig iron (4.05 pct C, 1.59 pct Si, and 0.65 pct Mn) was used as the solvent of carbon or silicon. The pig iron was melted in an open furnace and a small quantity of metallic silicon was added to the melt at 1300°C. The melt was heated to 150O°C, and then retort black carbon or metallic silicon was added little by little. The components added to the bath contained approximately 5.0 pct C and 2.0 pct Si in the carbon-rich material, and 3.8 pct C and 5.4 pct Si in the silicon-rich iron. These two types were held at 1500°C and then cooled without graphitizing by cutting the heat sup-

Jan 1, 1958

Jan 1, 1958

By Earl A. Gulbransen, Kenneth F. Andrew

DRY oxidation of zirconium has been studied by several groups.&apos;" The present work extends our early study1 to the high-temperature studies of Cubicciotti2 and Belle and Mallett.8 ulbransen and Andrew1 showed that the parabolic rate law fitted the 2-hr experiments between 200" and 425°C after an initial deviation. An energy of activation of 18,200 cal per mol was calculated. Cubicciotti,&apos; in a study between 593" and 880°C, showed that the parabolic rate law fitted the data and an energy of activation of 32,000 cal per mol was calculated. In contrast to these studies on sheet specimens, Belle and Mallett3 studied the oxidation reaction on rod specimens and found a cubic rate law to fit the data.

Jan 1, 1958

By M. I. Sherman, J. D. H. Strickland

USE of a hydrometallurgical approach to the oxidation of sulfide ores and extraction of metals therefrom may have advantages over the more common smelting techniques when a low grade deposit is difficult to concentrate or the subsequent separation of metals, coexisting in the ore, is laborious by any known smelting operation. For economic reasons, the most promising oxidants are either atmospheric oxygen or electric power. The use of oxygen, or air under pressure, has recently been revised. Pyrrhotite has been converted to iron oxide and elementary sulfur&apos; and a variety of sulfides have been treated by Forward and co-workers.2-4 Generally sulfate is the end form of the sulfur but with galena in an acid medium, elementary sulfur can be formed." For economic reasons chlorine and ferric iron salts are about the only possible alternatives to the atmosphere as oxidizing agents for base metal sulfides. If aqueous solutions of chlorine or ferric iron are employed, the reduction products can be oxidized electrolytically in situ and used again, thus acting as catalysts for electric power as oxidant. The use of ferric salts for this purpose is established hydrometallurgical practicea but, although chlorine gas has been employed in the dry state at an elevated temperature, its use in aqueous solution at or near room temperature has not found favor. The reaction of chlorine water with the soluble sulfide ion has been studied by several workers,7-9 and both sulfate and elemental sulfur are found as end products, the latter being favored by the presence of a low concentration of oxidant relative to that of sulfide in solutions of about pH 9 to 10. Of direct bearing on the work in hand are an early American patent" and a recent Austrian patent." The former advocates stirring powdered ore with an aqueous solution of ferric chloride chlorine oxides and chlorine. In the latter it is claimed that both metal and sulfur can be obtained by electrolysis, in a diaphragm cell, of a metal ore slurry in brine. Details in these patents are scant and no data or explanation is given for the mechanism of the reaction which, in the Austrian work, is attributed to the (unlikely) action of nascent chlorine at the anode surface. No mention is made of possible differences in behaviour between various ores. Apparatus A complication encountered when working with chlorine water is that a serious loss of chlorine occurs by gas partitioning unless an enclosed system is used and any air space in the apparatus is kept very small and constant. Arrangements were made, therefore, to take out samples for analysis without letting air into the system to replace the liquid removed. For convenience in studying a heterogeneous reaction the apparatus was so designed that a reproducible controlled stirring rate could be maintained and the ratio of surface area of ore to volume of solution was approximately constant throughout any experiment. The apparatus used is shown in Fig. 1. The ground ore was placed in the horizontal cylindrical vessel, A, of about 1 liter capacity, heated by a constant temperature circulating bath pumping water through the concentric jacket, B. By adding chro-mate to this water, an ultraviolet radiation filter effectively surrounded the reaction vessel, greatly reducing any possible photochemical decomposition of chlorine solutions. Stirring was effected by glass paddles, C, attached by an axle to a magnet which was rotated by another powerful Alnico magnet, D, outside the glass end, this magnet being itself rotated by an electric motor electronically controlled to constant speed. Speed could be varied from about 150 to 900 rpm and was measured and held to within 1 pct of a given value. The end of the reaction vessel remote from the stirring magnet was closed by another one-ended glass cylinder, E, connected by thin polyethylene bellows, F, clamped by screw clamps and watertight rubber gaskets to the main vessel. Through E, a glass electrode and calomel electrode projected into the solution and a hypodermic syringe pierced a small bung and allowed acid or alkaline to be added to maintain a constant pH. By pushing the fully extended bellows until the two cylinders touched, from 50 to 100 ml of solution could be forced out through a sintered disk into the three-way tap system, G, either to waste (for flushing purposes) or up into a 10 ml burette where the solution could subsequently be measured out for analysis. The ore samples were introduced at H, the tube being stoppered by a thermometer of —1 to +52ºC range, graduated to 0.1°C intervals. To prevent ore from being ground in the end bearings of the stirrer these bearings were pro-

Jan 1, 1958

By A. R. Troiano, W. J. Barnett

IN recent years the demands of space limitations and increased loads, particularly in the aircraft industry, have accelerated the trend toward utilization of ultra-high strength steels. The increased application of such steels has focused attention on the phenomenon of delayed failure or static fatigue. Parts subjected to a static load, after successfully sustaining the load for an extended period of time, may fail in a brittle manner. Subsequent examination of such failed parts by conventional laboratory tests in many cases reveals no apparent signs of embrittle-ment. That is, the material exhibits good ductility and impact resistance. Furthermore, consideration of the service conditions (environment and temperature) under which this brittle failure occurred eliminates its classification as a stress-corrosion or creep type of failure. Recently this unique time dependent fracture process has been the subject of numerous investigations. The most significant result of these studies was the recognition of hydrogen as the source of embrittlement. While the effect of numerous variables on the occurrence of static fatigue failure has been defined, an adequate explanation of these behaviors is lacking. The occurrence of this phenomenon in steels generally has been limited to strength levels in excess of 180,000 lb per sq in. Although steels at high hardness levels have exhibited delayed failure in the absence of hydrogen introduction after heat treat-ment,&apos;.&apos; this behavior was most prevalent when the material in processing had been subjected to an environment conducive to the absorption of hydrogen. Delayed failures have been produced in high strength steels by the introduction of hydrogen

Jan 1, 1958

By J. L. Huitt, IM. L. Slusser, E. E. Glenn, M. L. Slusser

This paper is concerned with: (1) an analysis and interpretation of the filtration characteristics of drilling muds on filter paper, and (2) an interpretation of early stage "filtration" on consolidated synthetic porous media, herein referred to as the "surge loss" period. It is shown that water-loss data can be plotted so as to disclose the existence of a surge period, a non-uni-form cake thickness period, and a constant pressure filtration period. The same periods occur during exposure of consolidated porous media to muds. The data on consolidated porous media indicate that: (1) mud particles bridge pores inside porous media during the surge period, and (2) "filter cake" build-up inside porous media occurs during subsequent filtration. The data also show that the surge loss increases with sample permeability and with pressure differential. INTRODUCTION The phenomenon of mud filtrate invasion into porous sands of a drilling well has been recognized for many years.1,2,3,4 Part of API Code 29 is concerned with the filtration testing of a drilling mud to yield data on this subject. In a number of investigations "plastering properties" of muds were studied,5,6,7,8 "plastering" apparently meaning cake forming properties. In such studies the characteristics (thickness, permeability, and toughness) of the cake and the rate at which the liquid phase of the mud was separated from the solids (filtration rate through the cake, commonly referred to as water loss) were measured or at least considered. From the standpoint of its effect on well productivity, filtrate invasion into "clean" water-wet pay sands (i.e., those containing no clay minerals) may not necessarily be a serious problem. An increase in water saturation in the zone immediately around the wellbore would cause loss of oil permeability. However, in such a system, much of this water may be removed by oil production. If filtration occurs into "dirty" sands (i.e., those containing appreciable clay minerals), clay swelling or other mechanisms may cause losses in oil permeability which cannot be overcome by oil production.&apos;.&apos; The phenomenon of mud particle invasion into the pay sand of a drilling well was considered to be a possible serious factor in the early days of rotary drilling. Opinions varied as to whether drilling fluids "mudded up" and "sealed" oil sands. In 1932 Rubel,9 Gill,10 and Parsons," in separate papers, presented differences of opinion together with some laboratory and field data concerning loss of oil production caused by drilling fluid invasion into pay zones. These papers, and the discussions, furnish a large background on this subject and indicate the very clear thinking of many men at a time when petroleum production technology was in its infancy. Jones and Babson,6 and Silent,&apos; experimentally found that the depth to which mud particles invaded water-saturated sand packs was small (no greater than 1/16 in). Arn-quist12 in 1937 observed loss of water permeability due to invasion of mud particles into water-saturated, consolidated, sandstone samples. Williams and Cannon5 studied the filtra-tion properties of drilling muds and mentioned "sludge" which penetrated a short distance into the water-saturated filtering medium (sand pack). Byck5 concluded that formation permeability had no effect on the plas-

Jan 1, 1958

By W. W. Eckles

This paper is presented as a suniniary report of the use of well gas composition correlations obtained from mass spectrometer recordings as a means of identification and determination of reservoir continuity. Conventional methods for detecting composition differences are en-pensive, elaborate, and difficult to obtain. This excludes the use of extensive composition data for most applications. During recent years the mass spec-trometer has come into general use as an analytical tool in petroleum refineries. The use of mass spectrometer composition patterns ir~ characterizing or "finger-printing" the produced gas from a reservoir, presents a novel method for correlatitlg gas samples from well to well. The mass spectrometer provides a trace similar to an electric log. having peaks which represent the abuntlnnce of certaitz hydrocarbons in the well gas sample. Without going further into the detailed analysis the idea has been advanced that these traces or patterns could be used as a means of identifying a particular natural gas. This theory has proven to be essentially correct. The mass spectrometer pattern method is simple and cheap as COi?7pared to other standard methods. It greatly facilitates the solution of reservoir and geological problems in which correlation of well gas Com-positions is a factor. Specific field applications have been made. This paper concerns the results obtained in 465 individual gas analyses from 35 fields and 77 res- ervoirs. In a number of cases it has been found that such data have been extremely valuable in the determination of reservoir continuity. In at least one case, the method was a valuable contribution in tracing a reservoir from sand to sand in a coinplex fnulted field involving n11rnerous gas reservoirs. Field applications are presented to illustrate the possibilities of the method at the present stage of developnient and to stimulate the ernployrnent of this new approach by geologists and petroleum engineers in the industry. INTRODUCTION Identification of producing horizons and the determination of reservoir continuity are often a problem in those areas where dome structures and highly faulted sands are encountered. To complicate the picture further, there may be numerous sands, one on top of the other which dip and diverge in different directions. Even though it may be possible to develop some solutions to the preceding problems on the basis of the geological and reservoir data on hand. it is readily recognized that substantiating data based on independent methods would he extremely valuable. It has been found through field studies that correiation of well gas composition can be used to advantage in the geological and engineering study of a complex reservoir identification problem. METHOD FOR COMPARISON AND CORRELATION OF WELL GAS SAMPLES Since a large number of well gas samples are required in a gas identification or reservoir con- tinuity problem, it is necessary that a method of analysis be employed which can detect differences in composition readily and inexpensively. Although detection would be possible by means of low temperature fractional distillation (POD) analpsis, the method requires a relatively large sample and is comparatively slow and expensive to run. The mass spectrometer affords an inexpensive and precise analysis of small well gas samples taken at the surface which are about 1/300 as large as a POD sample. These samples can be obtained by regular field personnei and shipped to the spectrometer for analysis. It is, therefore, a practical approach to the problem. A typical record from the mass spectrometer is illustrated in Fig. I. The peaks on the record represent the abundance of ions produced from the different hydrocarbon molecules making up the gas sam-ple. In well gas comparisons only five peaks are employed, since the other peaks are formed from the same gas molecules and furnish no additional information. The five peaks used represent the abundance of methane, propane, ethane, butane and heavier, and oxygen. They are referred to respectively as the 16, 29, 30, 43, and 32 peaks. The oxygen peak is used to correct for air content in

Jan 1, 1958

By L. F. Stutzman, George Thodos

A new graphical method, which is a modification of that proposed by Buckley and Leverett&apos;, is presented for the determination of the displacing fluid saturation at breakthrough for frontal drive mechanisms. This method is applicable to all types of fluids, which may be either soluble or insoluble in the oil at reservoir and injection conditions. This adaptation has unique advantages in handling problems in which either liquids or gases are the displacing fluids. INTRODUCTION In the displacement of oil from a reservoir with water or gas injected into the formation or by the expansion of a gas cap, the determination of breakthrough or flood advance is important to define the initial and subordinate phases of the operation. Thus, the determination of the displacing fluid saturation at breakthrough is fundamental to the establishment of over-all reservoir production. Methods for analyzing this saturation and for calculating production rates have been previously reported1,&apos;*".". All of these methods utilize the rate equation for linear flow developed by Buckley and Leverett1. xs = QT - Qn/A? (?fu/?Su) .....(1) Sp where f, represents the volumetric fraction of displacing fluid in the total flowing fluid within the reservoir as is defined by: fo = 1/1+ ku/kd µn/µn .......(2) Since the time, ?, necessary for the flood front to travel distance ?X is unknown, Eq. 1 cannot be solved directly for the determination of the breakthrough point. Therefore, it becomes necessary to solve Eq. 1 simultaneously with an equation representing a fluid bal-ance on the reservoir behind the flood front. To develop such a fluid balance equation, it has been assumed that the fluid composition is uniform in the reservoir behind the flood front. Based upon this assumption, the follow-ing equation can be derived from a material balance on the displacing fluid: x = (?X) s Dx ? x=0 [ Snx - Sn1 ] dx = QT - Qn/A? It is readily observed that the displacing fluid satura-tion behind the flood front, Snx, may be determined through the simultaneous application of Eqs. 1 and 3. Although the approach may vary, this objective re-mains the same for all of the previously mentioned graphical or analytical methods. After Sox is determined, it is then possible to calculate the gas-oil and/or water-oil ratios and the rates during the initial and subordinate phases of production, and thereby define the production history of the reservoir. In this presentation only the de-termination of Snx is considered, and the reader is re-ferred to the several texts in reservoir engineering2,4,5 for a complete discussion on the methods for establish-ing the production history of a reservoir. PROPOSED METHOD FOR DETERMINING Snx Since at any time the quantity, QT - Qn/A? is a constant, Eq. 1 may be expressed in differential form as follows: dxsn = QT - Qn/A? d (? fn/? Sn) sn .....(4) Eq. 4 may be substituted directly into Eq. 3 to obtain: ? fn/? Sn ? [ Snz - Sn] QT - Qn/A? d ( ? fn/? Sn) = QT - Qn/A? ...............(5) Eq. 5 simplifies to the expression: ? fn/? Sn ? [ Snz - Sn] QT - Qn/A? d ( ? fn/? Sn) = 1 ....(6)

Jan 1, 1958

Jan 1, 1958

By M. King Hubbert, David G. Willis

A theoretical examination of the fracturing of rocks by means of pressure applied in boreholes leads to the conclusion that, regardless of whether the fracturing fluid be of the penetrating or non-penetrating type, the fractures produced should be approximately perpendicular to the axis of least stress. The general state of stress underground is that in which the three principal stresses are unequal. For tectonically relaxed areas characterized by normal faulting, the least stress should be horizontal; the fractures produced should be vertical with the injection pressure less than that of the overburden. In areas of active tectonic compression, the least stress should be vertical and equal to the pressure of the overburden; the fractures should be horizontal with injection pressures equal to or greater than the pressure of the overburden. Horizontal fractures cannot be produced by hydraulic pressures less than the total pressure of the overburden. These conclusions are compatible with field experience in fracturing and with the results of laboratory experimentation. INTRODUCTION The hydraulic-fracturing technique of well stimulation is one of the major developments in petroleum engineering of the last decade. The technique was introduced to the petroleum industry in a paper by J. B. Clark,&apos; of the Stanolind Oil and Gas Co. in 1948, and since then its use has progressively expanded so that by the end of 1955 more than 100,000 individual treatments had been performed. The technique itself is mechanically related to three other phenomena concerning which ah extensive literature had previously developed. These are: (1) pressure parting in water injection wells in secondary-recovery operations, (2) lost circulation during drilling, and (3) the breakdown of formations during squeeze-cementing operations, all of which appear to involve the formation of open fractures by pressure applied in a wellbore. The most popular interpretation of this mechanism has been that the pressure had parted the formation along a bedding plane and lifted the overburden, notwithstanding the fact that in the great majority of cases where pressures were known they were significantly less than those due to the total weight of the overburden as determined from its density. Prior to 1948, this prevalent opinion had already been queried by Dickey and Andresen,2 in a study of pressure parting; and by Walker,"&apos; who, in studies of squeeze cementing, pointed out that the pressures required were mostly less than those of the overburden, and inferred that the fractures should be vertical. J. B. Clark,&apos; in his paper introducing hydraulic fracturing, and later Howard and Fast," and Scott, Bearden, and Howard," also of Stanolind, postulated that the entire weight of the overburden need not be lifted in producing horizontal fractures, but that it was only necessary to- lift an "effective overburden," requiring a correspondingly lower pressure. Hubbert,7 in discussing the paper by Scott and associates, pointed out that the normal state of stress underground is one of unequal principal stresses; and in tectonically relaxed areas, characterized by normal faults, the least stress should be horizontal. Therefore, in most cases, fracturing should be possible with pressure less than that of the overburden and, moreover, such fractures should be vertical. Harrison, Kieschnick, and McGuire,8 also on the expectation that the least principal stress should be horizontal, argued strongly in favor of vertical fracturing. Scott, Bearden, and Howard6 observed that, when

Jan 1, 1958

By Robert J. White

Strong&apos;s equation for calculating bottom-hole pressure reduction due to gas cutting of drilling mud is corrected, resulting in a simpler equation which is easier to use. Use of the equation is illustrated by examples. It is shown that reduction of weight of a 10 lb/gal mud by 50 per cent at the surface due to gas cutting produces only n 3 per cent decrease in hydrostatic head at the bottom of a 5,000-ft well. It is also shown that at the surface 5 volumes of air per volume of mud are required to reduce the apparent mud weight from 8.5 to 6.5 lh/gal in a 3,000-ft well. Illustrative graphs are included to show the effect of various degrees of gas cutting on apparent mud weight at depths of 2,000, .5,000 and 10,000 ft and to show variation of apparent mud weight, with depth for a 100 lb/cu ft (13.37 lb/gal) mud gas cut to 90 lb/cu ft (12.03 lb/gal) at the surface. INTRODUCTION In 1938, Strong&apos; published the following equation for calculating reduction in bottom-hole pressure due to gas cutting of drilling mud h = 1/D [p + pn/100 In p + p(1-n/100)/p(1- n-100)] where: h = depth in feet P = pressure in atmospheres at depth h due to mud column only D = hydrostatic pressure in atmospheres of a col-umn of uncut mud 1 ft high p = back pressure at well-head (in atmospheres) n/100 = fraction by volume of gas in mud at well-head at back pres-sure p This equation was recently found to contain an error due to an incor-rect expression for variation of vol-ume fraction of gas with pressure. Values calculated by the original equation are close to the correct values for small percentages of gas for which it was derived but become increasingly erroneous as the gas content increases and are completely unusable at the very high gas con-tents employed in aerated mud. DERIVATION The nomenclature used is the same as in Strong&apos;s paper. The pres-sure dP exerted by a lamina of thickness dh at a depth h ft is ex-pressed by Eq. 1 where x = volume fraction of gas at depth h dP = D (l - x)dh . . . (1) When we express x in terms of the wellhead per cent gas n, the well-head pressure p and the pressure P at depth h, the result is Eq. 2. x = (np/p+p)/(100-n) + (np/P + p) When this is substituted in Eq. 1 and rearranged to separate the var-iables the result is dh = 1/D [dp + np/100-n dp/(P + p]. Integration between appropriate lim-its results in or if the wellhead is open so that p = 1 atm and we rearrange Eq. 4, the result is loss in head = hD — P = n/100-n In (P + 1) ...(5) Eq. 5 can be solved easily by successive approximation using hD as a first approximation for P on the right-hand side and calculating a first value for loss in head. This can then be used to get a new value of P using the left-hand side of the equation. Values converge very rap-idly even for quite large values of per cent by volume gas n. EXAMPLES The effect on bottom-hole pressure of 50 per cent gas cutting of the mud in a 5,000-ft well using 10 Ib/ gal mud will be calculated, we have: h = 5,000 ft 10 X 7.48 144 X 14.7 = 0.0353 atm/ft hD = 176.5 atm n = 50 Substituting these values in Eq. 5 and using hD as a first approxima-tion of P on the right, we have, approximate loss in head = 50/50 In (176.5 + 1) = 2.30 log 177.5 = 5.17 atm. We calculate a new value of P as follows:

Jan 1, 1958