Search Documents

By J. M. McDowell, Morris Muskat

In both production and refining operations of the oil industry many processes are controlled by the gas-liquid phase relationships of the hydrocarbon mixtures of interest. The quantitative behavior of fractionating or distillation columns and stills depends on the changes in the equilibrium distributions of the hydrocarbon components between the gas and liquid phases as the pressure or temperature conditions may vary along the length of the column or as the gross operating parameters and compositions are varied. In oil producing operations the nature and amount of gas phase developed within an underground reservoir as it is being depleted and its pressure declines also involve the basic equilibrium gas-liquid phase interactions. The influence of surface conditions of temperature and pressure and various separation procedures on the nature and amount of stock tank oil and natural gas recovered from a given well stream is likewise determined by the same basic phase equilibrium characteristics of the hydrocarbon systems. Such equilibrium separation of the gas and liquid phases in a well stream is of importance both in establishing optimum separator conditions for obtaining maximum stock tank oil yields and in the general problem of crude stabilization. The phenomena involved in these problems and their quantitative aspects can, of course, be established in each individual instance by appropriate laboratory experimentation. It has been found, however, that by associating with the individual hydrocarbon components functional characteristics de- scribing their individual gas - liquid phase equilibrium behavior, the phase properties of the composite mixtures can be predicted by mathematical computation. These characteristics have been termed "equilibrium constants", or "equilibrium ratios"," and are defined by the general equation: where K, is the equilibrium ratio for the jth component, y, is the rnol fraction of that component in the gas phase, and xj is the corresponding mol fraction in the coexisting liquid phase. Considering the Kj as representing available and measurable properties of the individual components of a mixture, and as defined by Eq. (l), it is readily shown that in a coexisting gas-liquid system the rnol fraction concentrations of the individual components in the gas phase, y,, and those in the liquid phase, x,, are given by the expressions: where the nj are the corresponding rnol fraction concentrations in the composite system, and ng is the mol fraction of the whole which is in the gas phase. The basic unknown in Eq. (2) is n,: which gives the gross separation between the gas and liquid phases. This may be determined in principle by imposing the requirement that the sums of the mol fractions in both the gas and liquid phases are unity, i.e. by the equations: The solution of the equivalent Eqs. (3) lying between 0 and 1, when inserted into Eqs. (2), provides a complete description of the compositions of the coexisting gas and liquid phases. While the Eqs. (3) are essentially equivalent to polynomial equations, such transformations are of little value when dealing with multicomponent sys-

Jan 1, 1949

By Harry Gard Knox, F. R. Harris

This paper covers some of the problems presented to the oil industry in the drilling of oil wells in the open waters of the Gulf of Mexico. The hazards, delays and relative costs of offshore drilling are reviewed and various forms of drilling platforms discussed. The requirements of the offshore drilling rig are summarized together with the principal items affecting the cost of construction. Sea waves and their size and characteristics as affecting offshore platforms are reviewed. Drilling platforms are classified as fixed, or fixed and floating. Fixed platforms of the conventional cofferdam and pile-supported type are discussed together with their good points and their limitations. The desirability is pointed out of eliminating as much offshore work as possible where operations are subjected to constant and expensive delay and substituting therefor the prefabrication of structures ashore or in sheltered waters. Marked economy will come with the standardization of offshore rigs aid special means of transporting them to the site. Several types of structure which are partly pile-supported and partly buoyancy-supported are discussed and their possibilities of future usefulness commented upon. General Considerations The predominant attribute of offshore drilling equipment is that it must be equal to the job in hand. We therefore give first thought to the matter of adequacy. In view of the probable depth of drilling, the investment involved and the per diem labor costs, it is assumed heavy modern drilling equipment will be employed. In order to usc such equipment efficiently, the working deck must be ample in size and strength and supported so as to be relatively free from motion and from green seas even in heavy storms. These requirements necessitate a high and rigid land-based support for the drilling rig and machinery with plenty of space for supplies, such as pipe, mud, water and fuel, with reserves for days when transportation from shore is impracticable. In offshore rigs the crew will be housed onboard in comfortable quarters, probably air conditioned and equipped with a galley and refrigerated food storage. An important consideration will be the ability to transfer stores and crew to and from the rig in all but the heaviest weather and the platforms must be protected from the impact of supply vessels when alongside. It will be essential to maintain continuous communication with shore by radio or cable, not forgetting the possibility of a helicopter landing space for the transfer of crew and light stores as well as a means of general communication. Whatever structure is constructed must be secure from storms of hurricane force and a safe place for the operating crew to work. If the well comes in, the drilling rig will presumably be removed and a small but rugged pile-supported platform will remain to carry the Christmas tree with safe working space around it. It will be protected from damage on all sides by buffer piles or dolphins. It is assumed that the flowing oil will be piped to storage tanks in the vicinity of the well. An individual tank for each well or a

Jan 1, 1949

By Alfred H. Bell, Paul F. Simpson

The year 1929 was one of continued activity in the petroleum industry of Indiana and Illinois but the new production obtained has not been sufficient to offset the decline in the production of the old wells. (Table 1). Drilling costs have not changed materially during the year and the prices of crude oil have increased slightly so that the prospects for further testing are good. The total number of completions in Illinois in 1929 (Table 2) was greater than for any year since 1917. More than half of these were in the Dupo field, St. Clair County, and the development of this field was largely responsible for the increased ratio of producers to dry holes in the state as a whole. In the southeastern Illinois field which has produced about 97 per cent. of the state's petroleum the number of completions was greater than in 1928 and the percentage of producers was a little higher. Although a number of scattered wildcat tests were drilled most of them were not located on favorable structure and the only noteworthy discovery was that of a shale-gas horizon at shallow depth in Morgan County in western Illinois where gas has been produced for a number of years from sandstone and limestone. In Indiana the number of completions (Table 2) was only slightly less than in 1928. Most of the new wells were drilled in the southwestern part of the state, which now produces more than 90 per cent. of Indiana's petroleum.

Jan 1, 1930

By W. M. Pierce

THE present work has been done in an endeavor to correlate and complete the data on the constitution of alloys of zinc with other common, metals, dealing exclusively, however, with the zinc-rich alloys in which the zinc content is between 90 and 100 per cent. This investigation has included the systems lead-zinc, cadmium-zinc, iron-zinc, copper-zinc, aluminum-zinc, nickel-zinc, manganese-zinc, cobalt-zinc, tin-zinc, and magnesium-zinc. The important ternary system zinc-lead-cadmium and the zinc corners of other ternary systems have been investigated in connection with this work but will not be dealt with here. Zinc does not form a single homogenous zinc-rich solid solution over any considerable range of concentration with any of the metals mentioned. A limited range of solid solubility is said,; by previous investigators, to occur in several cases and certain effects on the properties of zinc are attributed (with varying degrees of consistency among different writers) to the presence of certain other metals dissolved in zinc. The increasing use of rolled zinc and the possibilities offered by zinc as a casting alloy in certain fields have it made essential to determine precisely these equilibrium relationships. In carrying out this .investigation reliance was placed chiefly on two methods, conductivity measurements and microscopic analysis. The first method proved valuable in determining the existence and extent of solid solubility. The second method stands alone as a means of determining structural characteristics and supplements the first method in the measurement of solid solubility. The scleroscopic hardness of all conductivity specimens was measured partly for its value from the point of view of mechanical properties and partly; as a check on the conductivity curves in indicating the limits of solid solubility. No hardness curves are included in this paper, for in no case were they important in arriving at the constitution of any system.

Jan 2, 1922

By K. Tangri, B. Swaroop

Simultaneous measurements of lattice (elastic) strain by X-ray line shift method and total strain with an electrical strain gage have been carried out on polycrystalli,ne nickel with the help of a specially designed tensometer attachment for the X-ray dif-fractometer. During the initial stages of deformation, the rate of increase in lattice strain closely follows the total strain until the plastic strain sets in. From then onwards, the two strains deviate from each other and with further increase in afiplied stress the rate of increase of lattice strain eventually decreases. Depending upon the mode of unloading, both compressive and tensile strains have been observed in nickel deformed up to 0.29 pct strain. These results have been explained on the basis of the effects of clustering of dislocations and also the production and behavior of point defects during loading and unloading, respectively. POLYCRYSTALS deformed plastically in a uniaxial tension test show residual lattice strains (hereafter referred to as RLS), which broaden the X-ray line profiles and shift their peak positions.''4 It is known that uniform straining of a crystal lattice (macrostrain) produced movement of ddfraction line peaks, whereas nonuniform straining (microstrain) causes line broadening.= Though the nature and origin of RLS in deformed metals is not yet clearly understood, these are believed to be due to the presence of some form of a locked-up stress system. Several possible detailed interpretations6'" of the stress system have been proposed by various workers and common to all these interpretations is the assumption that different parts of the aggregate have different tensile yield stresses; e.g., that a part A yields under a lower applied stress than a part B. Therefore, during the deformation process, the elastic strain in A will be less than that in B, and, after completion of deformation, B will tend to contract further than A but will be prevented from doing so by the restraining influence of A. Thus, when equilibrium is reached, A will be in compression while B is in tension. Although there is general agreement on the correctness of this argument, controversy still exists as to the exact nature of the parts A and B in a deformed metal. In an alternative approach to the creation of parts A and B, many other investigators have assumed that the RLS observed in unloaded specimens are in some way connected with the lack of proportionality between lattice strain and applied stress in the region above the yield stress." cullity13 has prepared a schematic summary of the results of previous workers, mainly Smith and Wood,2'11 which suggests that, above the elastic limit, the lattice strain may increase less rapidly with respect to strain, or may even decrease with increase in applied stress. If the applied stress is then decreased, the lattice strain would decrease along a line parallel to the loading line in the elastic region and thus produce a compressive residual lattice strain after unloading. At equilibrium, to balance these compressive stresses? there would have to be regions under tensile stress which may well be the grain boundaries or the substructure walls formed during deformation. The present investigation was undertaken to gain a better understanding of the nature and origin of RLS and also to experimentally verify the salient features of this hypothesis. EXPERIMENTAL Stress and strain determinations were made with a specially constructed tensometer attachment for the X-ray diffractometer, Fig. 1: which permits the following measurements simultaneously on a sheet specimen under uniaxial tension at various stress levels during loading and unloading: a) lattice strains, E=. in a direction perpendicular to the direction of pulling (x direction) from shifts in X-ray line peak positions; b) the total strain, ex, in the direction of pulling. with the help of an electrical strain gage affixed to the back of the specimen directly below the area irradiated by X-rays: and c) the applied stress, with the help of a load cell which consisted of a calibrated stainless-steel sample of dimensions identical to those of the specimen under investigation, and to which it was coupled. Because of its high stacking fault energy.I4 nickel

Jan 1, 1970

By T. A. El-Bassyouni, Paul Gordon

Grain-gvowth data in zone-refined aluminum and in dilute alloys of copper with this aluminum have been obtained. The amounts of copper added were from 4 to 400 at. ppm. It was found that in both the unalloyed aluminum and the alloys grain growth was characterized by a high value of n in the isothermal growth equation 6n = Kt The inadvertent presence of 1 at. ppm Fe in the alloys was found to he more potent in slowing growth than the purposely added copper. Possible rationalizations of these observations and their general relevance to the theoretical use of gvain-grozvth data have been discussed. Grain growth data of Winegard and his co-workers' on zone-refined tin and dilute tin alloys were analyzed with some success2 in terms of approximate theoretical equations for boundary migration.2-4 In view of this success, experiments were carried out in an effort to obtain similar data on zone-refined aluminum and on dilute copper-in-aluminum alloys. The results obtained were anomalous in several respects and could not be used to test theory. A discussion of these anomalies is presented here; they have been attributed to impurity effects and serve to emphasize the sensitivity of grain growth in general—and the particularly high sensitivity in aluminum—to impurities. The resulting difficulties in the theoretical use of grain-growth data are also discussed. EXPERIMENTAL DETAILS An aluminum bar was zone-refined and leveled by methods similar to those described previously.' The degree of purity obtained is indicated in Table The data in Table I were determined by mass-spectrographic methods through the courtesy of Mr. Thomas A. Johnson of AIAG Metals, Inc.* Three dilute copper alloys with this aluminum were prepared also by methods described previously.' The alloys contained nominally 4, 40, and 400 at. ppm of Cu, respectively (0.001, 0.01, and 0.1 wt pct). Each material was fabricated by swaging and rolling into a round bar about 1/4 in. in diameter with an annealed penultimate grain size of approximately 0.003 to 0.007 cm. In order to minimize preferred orientation the fabrication was carried out in several steps, each consisting of a reduction of about 40 pct followed by a recrystallization anneal. Annealing temperatures for the grain-growth studies were attained by the use of boiling liquids up to 213°C and molten metal baths at higher temperatures. Temperatures were controlled to within 1°C. Samples for measurements were cylinders about 3/4 in. long and 1/4 in. in diameter cut from the fabricated rods. In the interests of sample economy a single sample was used for the grain-growth measurements at each temperature. Each sample was, thus, annealed for several successive periods of time, the grain size being measured at each annealing period on a transverse surface exposed by

Jan 1, 1965

By J. Mote, P. Ward Flynn, J. E. Dorn

The effect of strain and strain rate on the critical resolved shear stress for prismatic slip in specially oriented single magnesium crystals was determined over a range of temperatures in an effort to identify the thermally activated strain-rate controlling dislocation process. Below about 450°K prismatic slip was preceded by twinning and fracturing. Above 450°K extensive prismatic slip was obtained followed by fracturing. Neither the Peierls&apos; mechanism nor the dislocation intersection mechanism can account for the observations. The data, however, are in good agreement with the dictates of Friedel&apos;s theory for cross slip to the prismatic plane. PREVIOUS investigations on single and polycrys-talline magnesium have resulted in the identification of the mechanisms of plastic deformation summarized in Table I. The-only well-confirmed mechanisms are {0001}, (1010}, (1011) slip plus (1012) twinning. The other twinning modes were only tentatively established and additional investigations are required to further confirm their operation. As noted in Table I, slip in this hexagonal close-packed metal invariably occurs in the direction of greatest atomic density, revealing that the Burger&apos;s vector for-active slip dislocations is always given by b = a/3<1210>. Slip is easily induced on the basal plane, and recent investigations have shown that the strain-rate controlling mechanism at low temperatures is the thermally activated intersection of the glide dislocations with dislocations threading the basal plane. In contrast, very little is known about prismatic slip. It has been reported to be active at points of stress concentration in polycrystals below

Jan 1, 1962

By H. I. Aaronson, C. Wells

Configurations of ferrite crystals have been found in a plain carbon steel which appear to have resulted from the nucleation of new ferrite crystals at the interphase boundaries of previously formed crystals despite the high carbon concentrations which necessarily develop at these boundaries. This phenomenon has been termed sympathetic nucleation. An attempt has been made to reconcile the occurrence of sympathetic nu-cleation with current nucleation theory. THIS investigation is one of a series on the formation of proeutectoid ferrite from austenite. From the viewpoint of chemical composition, this reaction consists of the nucleation and diffusional growth of crystals of carbon-poor ferrite within a matrix of carbon-rich austenite. The austenite adjacent to the austenite-ferrite boundaries will be greatly enriched in carbon, approximately to the value of the y/(a + y) equilibrium curve or its metastable extrapolation at the temperature of transformation. Those areas of austenite appreciably farther removed from the growing ferrite, on the other hand, will be relatively unaltered in composition, especially at the earlier stages of transformation. Since rates of nucleation are considered to decrease exponentially with decreasing supersaturation,&apos; the frequency with which ferrite nuclei appear at austenite-ferrite boundaries should be negligible in relation to that at which they form in other regions of the austenite. During this investigation, however, many groupings of ferrite crystals have been found which appear to have resulted from the nucleation of ferrite at austenite-ferrite boundaries. This phenomenon has been given the name of sympathetic 71.1tcleation. A number of micrographs of morphological configurations caused by sympathetic nucleation will be presented, after which an explanation for this reaction will be proposed in terms of current nucleation theory. Some of the structures to be considered are composed of bainite, an aggregate of ferrite and carbide, rather than of ferrite. Since ferrite and bainite differ only in that bainite forms under conditions which result in the nucleation of carbides behind the advancing austenite-ferrite boundaries,&apos; it will usually be unnecessary, for the purpose of this paper, to distinguish between the two reaction products. All studies were performed on an electric furnace steel (obtained from the Vanadium Alloy Steel Co.) containing 0.29 pct C, 0.76 pct Mn, 0.25 pct Si, 0.005 pct P, and 0.007 pct S. The alloy was cast as a 150 Ib, 7x7 in. cross section ingot and forged into bars 2x2 in. in cross section. These bars were homogenized for 48 hr at 1250°C in an Endo-Gas atmosphere. The depth to which decarburization penetrated during this heat treatment was determined by chemical and microscopic analyses and the affected metal was removed by machining. Specimens for isothermal transformation studies were cut from the remaining material; most of these specimens were 1/2x1/4X1/16 in., though some with a thickness of 1/32 in. were prepared for use at the shorter reaction times and lower reaction temperatures. Specimens were austenitized for 30 min at 1300°C, isothermally reacted for various times at temperatures ranging from 775" to 475 "C, and then quenched in iced water. The austenite grain sizes within individual specimens ranged from ASTM Nos. 1 through —4. A commercial heat-treating salt which was continuously deoxidized by an immersed graphite crucible served to minimize the loss of carbon during austenitizing; thick covers of powdered graphite and immersed graphite rods effectively prevented decarburization in the lead pots employed for the isothermal reaction treatments. The heat-treated specimens were sectioned and mounted in Bakelite. Following the completion of standard grinding and mechanical polishing procedures, the specimens were electrolytically polished with a Buehler-Waisman apparatus and etched in 2 pct nital. Experimental Results Rules of Evidence for Sympathetic Nucleation—On the basis of observations made on a single plane of polish, one precipitate crystal may be considered to have been sympathetically nucleated at the inter-phase boundary of another precipitate crystal when the following conditions are fulfilled: 1) The sympathetically nucleated crystal is not in contact with a grain boundary or a subboundary in the matrix phase. 2) The shape, size, and location of the crystal at whose boundary sympathetic nucleation occurred (hereafter termed the base crystal) and the crystal formed by sympathetic nucleation substantially pre-clude the possibility that the plane of polish em-ployed may have concealed the fact that both crys-tals actually nucleated at a grain boundary or a sub-boundary in the matrix phase.

Jan 1, 1957

By J. F. Abel

It can be concluded that rock mechanics instrumentation, geologic mapping, and operations research in combination will produce an accurate estimate of tunnel support requirements for establishing a steel-set load prediction model. This estimate can be used to increase safety and to reduce tunneling costs. Unnecessary supports can be eliminated; overloaded or insufficient supports can be replaced on a more scientific basis than previously possible. Only a limited number of measurements are required to develop a steel set load prediction model. The results of this study can be applied generally; but, every tunnel must be considered as a separate problem. The construction and geologic variables which were found to be significant in the prediction of rock loads in this tunnel were: steel section modulus, set blocking points, percentage of alteration, relative degree of faulting, relative water condition, distance to the nearest fault, and average joint spacing. INTRODUCTION A combined engineering geology and rock mechanics program was performed throughout the Straight Creek tunnel pilot bore in Colorado in order to develop a technique for estimation of steel tunnel support loads. The technique involves the construction of a statistical model from measured steel set loads, geologic and construction factors. The aim of the investigation was to produce a set of charts, based on these factors, which could be used by a trained man equipped with a hand lens and measuring tape, to specify a safe and economical steel size and set spacing. In order to be of practical value the measured parameters had to be rapidly measurable and capable of quantification. It was apparent from the beginning that quantitative values for many of the geologic factors would have to be arbitrary. It was impossible to ascertain in advance which of the readily determi-nable geologic and construction factors significantly affect steel set loading. As the investigation progressed the number of observations was systematically reduced in order to determine the minimum amount of data which would be necessary for a significant mathematical model. This paper presents an analysis of all the variables measured at each instrumented steel set. The statistical significance of the measured variables was determined. On the basis of this determination, a mathematical model was constructed and charts were prepared to select and space the steel tunnel supports at any given point in the tunnel. Once the effect of geologic and construction factors on set loading is sufficiently established, it is possible to exert close control over the selection of steel tunnel support size and spacing. The magnitude of this rock mechanics program can be appreciated when it is known that there were 44 instrumentation stations for which steel set loads were measured. For each of these sets there were three construction variables recorded, and nine geologic parameters determined (Table I). A rough approximation of the cost for both the collection and reduction of the data would be $50,000. This is in addition to the cost of geologic mapping.

Jan 1, 1967

By Thomas M. Drown

THE means relied on to attain uniformity in Bessemer steel may be enumerated as follows : I. The appearance of the flame. II. The appearance of the slag. III. The spectrum of the flame. IV. Examination of the steel itself. At present, I wish briefly to bring forward a few considerations with reference to the last two mentioned, namely, the conditions favoring the employment of the spectroscope and examination of the steel. From the amount of careful investigation which has been devoted to the examination of the Bessemer flame by the spectroscope, both on the part of scientists and practical metallurgists, one might reasonably expect some results definite and satisfactory in their nature, first, as regards the nature of the flame, and second, as to the practical value of the spectroscope as a guide in conducting the process. We find, in fact, however, that uncertainty and confusion still exist. In Germany, where this subject has received the greatest attention, it is now very generally conceded that manganese is the cause of the well-known groups of dark bands that characterize the Bessemer spectrum. Nor is it easy to see how we can avoid accepting this conclusion, when Wedding and Von Lichtenfels have produced the manganese spectrum and found it to correspond closely with that shown by the Bessemer flame. But it is evident, on a moment&apos;s reflection, as Roscoe has recently brought forward, that manganese cannot be the only cause of the Bessemer spectrum, for we notice the characteristic bands quite as

By Edwin C. Eckel

The title of this study may fairly lead to misunderstanding unless its basal viewpoint is explained at the start. There is of course no chance of misunderstanding the term strategic minerals-for some years past it has been rather generally used to cover those mineral supplies, important for war use, which do not occur in adequate amount on the territory of the Power under consideration. Even here, however, I might add a note of warning. It is commonly understood that the list of strategic minerals will vary, according to what country we are considering; but it is rarely understood, even in Staff studies, that it varies equally according to the particular adversary we happen to be considering. In the present study I have stated the definition in the broadest possible terms, and have at this stage purposely avoided limiting it to any one country or war situation, with the express purpose of being free to bring out certain implications which are commonly, if not always, overlooked. The more important opportunity for misunderstanding is, of course, that my views on the phenomena war and peace do not coincide with those currently fashionable, particularly in advanced intellectual circles. For we must face the fact that for a decade past the natural reaction from war conditions has shattered the public mentality so completely that our public discussions of such questions have very little relation to fact; they are excursions into ideal rather than Realpolitik. And as such they serve merely to prevent public discussion or accept-

Jan 1, 1932

By G. H. Billings

There exists an unconfirmed opinion among many ironmasters that the combination of a small quantity of manganese, chromium, titanium, tungsten, aluminium, nickel, and some of the metalloids with iron has a beneficial effect upon the quality. And the impression prevails in some localities that the excellence of steel greatly depends upon the influence of some of these elements. Brit as the recorded experiments are so meagrely described, and made under such various conditions, the student in search of information upon the subject is somewhat bewildered by the contradiction of the opinions expressed. Observing some interesting phenomena while experimenting with an alloy of iron, copper, and nickel, I was led to determine the effect of some of the metals upon iron as free from contamination by other elements as it is possible to procure in practice, in order that the result of an alloy of an individual metal with iron might be more closely studied; for it is almost an impossibility to determine the influence of a small amount of one element upon a metal combined with a large amount of another, and the mere presence of another sometimes makes this difficult. And as most of the experiments recorded upon this subject have been made with iron containing sufficient carbon and other elements to interfere with the effect of that metal the influence of which it was desired to observe, I have endeavored to avoid these sources of error as far as practicable. In determining the specific gravity of the alloys in these ex-

By B. M. Larsen, T. E. Brower, J. W. Bain

A mass of circumstantial evidence is presented to indicate that the main source of alloy losses in open-hearth tapping is oxidation by air, with the steel apparently reacting with an amount of oxygen equivalent to about 30 times its own volume of air. The effect is erratic from heat to heat, depending largely on turbulence and distance of free fall of the stream of liquid metal. THIS paper reports another phase in a general study of oxidation in the open-hearth, various aspects of which have already been discussed in previous papers.1,2,3 It is based on experiments conducted on commercial, basic open-hearth furnaces at various times when opportunity offered, beginning about 1938, in an attempt to determine the mechanism and measure the amount of air oxidation during the flow of the tapping stream through the air into a ladle. This effect is closely connected, not only with alloy losses and control of chemical analysis, but possibly with various questions regarding inclusion content and steel quality and with a true picture of the deoxidation of steel in general. Thus it is believed that discussion of a large part of the results to date may be of assistance to others who are studying problems related to deoxidation or to steel cleanliness and quality. The literature reveals that strangely little attention has been devoted specifically to the effect of pouring steel through open air. Bardenheuer and Henke4 in 1939, in a paper devoted largely to overall losses of manganese in the basic open-hearth, reported that for a number of heats tapped into a tilted ladle held just beneath the runner so that the maximum distance of drop of the stream through the air was held to less than about 30 in., there was in most cases almost no loss of manganese from furnace to ladle. Hultgren,5 in 1945, showed that in high-carbon basic electric steel deoxidized in the furnace and containing no large inclusions as it left the furnace, large silicate inclusions, often rich in manganese oxide, were present in samples dipped from the pool in the ladle; also that the rise in content of large inclusions (above 0.01 mm dim) was greater when the metal stream was made to spray or flow in a turbulent manner. Our interest in this phase of oxidation in the open-hearth was first aroused by a study of alloy efficiencies on ladle additions. The frequency curves of manganese efficiency (net recovery of metal added in ladle) given in fig. 1 are of interest in this connection. Lowest, and also most variable (47 to 95 pct), recovery was in low-carbon heats having no coal additions to the ladle, progressing upward to a maximum average and least variable (82 to 97 pct) recovery in high-carbon heats killed with aluminum and silicon. Both coal additions and higher carbon contents in the tap stream, as well as additions of silicon and aluminum, help to lower the amount of manganese loss. But the most interesting point is the amount of manganese oxidized in many heats tapped at 0.50 to 1.06 pct carbon. It has been shown in a previous paper3 that the amount of oxygen in the tap stream in such heats is very small, usually below 0.01 pct, too small to oxidize any appreciable amount of manganese. Also, the fact that manganese should be oxidized at all in a metal solution containing such more stable oxide-forming elements as aluminum and silicon is an indication of some rather intense oxidizing effect such as could be caused by reaction with the air during ladle filling. Oxygen Balances from Tap Stream to Teeming Stream: Data in the form of alloy efficiencies, such as those shown in fig. 1, are not a proper measure of such a general oxidation effect. Having available

Jan 1, 1951

By Francis Collins, E. R. Brownscombe

A study has been made of the material balance-fluid flow method of estimating reserves and degree of water drive from pressure and production history data. By considering the effect of random pressure errors it is shown that in a particular example a standard deviation of three and one-half pounds in each of ten pressure survey? permits the determination of the reserves with a standard deviation of 8 per cent and the water drive with a standard deviation of 15 per cent, assuming that certain basic geologic data are correct. It is believed that this method of estimating reserves and water drive is useful and reliable in a number of cases. The method is particularly valuable when reservoir pressure data are accurate within a very few pounds, but may also be applied with less accurate pressure data if a relatively large reservoir pressure decline occurs early in the life of the field, as for example in an under-saturated oil field. INTRODUCTION A knowledge of the magnitude of reserves and degree of water drive present in any newly discovered petroleum reservoir is necessary to early application of proper production practices. A number of investigators have contributed to methods of relating reserves, degree of water drive, and production and pressure history. 1-8 Three types of problems of increasing complexity may be mentioned. If a reservoir is known to have no water drive. and if the ratio of the volume of the reservoir occupied by gas to the volume of the reservoir occupied by oil (which ratio permits fixing the overall compressibility of the reservoir) is known, then only one further extensive reservoir property remains to be determined, namely the magnitude of the reserves. A straightforward application of material balance considerations will permit this determination. The problem becomes very much more difficult if we wish to determine not only the magnitude of the reserves but also the magnitude of water drive, if any, which is present. In principle, a combination of material balance and fluid flow considerations will permit this evaluation. Finally, if neither the magnitude of reserves, the degree of water drive, nor the ratio of oil to gas present in the reservoir is known and it is desired to determine all three of these variables, the problem could in principle be solved by a fluid flow-material balance analysis which determines the overall compressibility of the reservoir at various points in its history. The change in compressibility with pressure would provide a means of determining the ratio of gas to liquid present, since the compressibilities of gas and liquid vary differently with pressure variation. However, in practice this problem is probably so difficult as to defy solution in terms of basic data precision apt to be available.&apos; It is the purpose of this discussion to illustrate the second case, which involves the determination of two unknown variables, single phase reserves and degree of water drive, from pressure and production history and fluid property data, and to study the precision with which these unknowns can be determined in this manner in a particular case. Although an electric analyzer developed by Bruce as used in making the calculations to be described, numerical methods necessary in carrying out the process have been devised and have been applied for this purpose. Schilthuis,&apos; for example, developed a comprehensive equation for the material balance in a reservoir. He combined this with a simplified water drive equation, assuming that the ratio of free gas to oil was fixed by geological data and that a period of constant pressure operation at constant rate of production was available to determine the constant for his water drive equation. On this basis he was able to compute the reserves and predict the future pressure history of the reservoir. Hurst developed a generalized equation permitting the calculation of the water drive by unsteady state expansion from a finite aquifer. He showed in a specific case how the water influx calculated by his equation, using basic geologic and reservoir data to fix the constants, matched the water influx required by material balance considerations. Old3 illustrated the simultaneous use of Schilthuis&apos; material balance equation and Hurst&apos;s fluid flow equation for the determination of the magnitude of reserves and a water drive parameter from pressure and production history. He used this method to calculate the future pressure history of the reservoir under assumed operating conditions. As a basis for determining reserves, Old assumed a value for his water drive parameter and calculated a set of values for the reserves, using the initial reservoir pressure and each successive measured pressure. The sum of the absolute values of the deviations of the resulting reserve numbers from their mean value was taken as a criterion of the closeness of fit to the experimental data possible with the water drive parameter assumed. New values of the water drive parameter were then assumed and new sets of the reserves calculated until a set of reserves numbers having a minimum deviation from the average was established. The average value of- the re-

Jan 1, 1949

By E. R. Brownscombe, Francis Collins

A study has been made of the material balance-fluid flow method of estimating reserves and degree of water drive from pressure and production history data. By considering the effect of random pressure errors it is shown that in a particular example a standard deviation of three and one-half pounds in each of ten pressure survey? permits the determination of the reserves with a standard deviation of 8 per cent and the water drive with a standard deviation of 15 per cent, assuming that certain basic geologic data are correct. It is believed that this method of estimating reserves and water drive is useful and reliable in a number of cases. The method is particularly valuable when reservoir pressure data are accurate within a very few pounds, but may also be applied with less accurate pressure data if a relatively large reservoir pressure decline occurs early in the life of the field, as for example in an under-saturated oil field. INTRODUCTION A knowledge of the magnitude of reserves and degree of water drive present in any newly discovered petroleum reservoir is necessary to early application of proper production practices. A number of investigators have contributed to methods of relating reserves, degree of water drive, and production and pressure history. 1-8 Three types of problems of increasing complexity may be mentioned. If a reservoir is known to have no water drive. and if the ratio of the volume of the reservoir occupied by gas to the volume of the reservoir occupied by oil (which ratio permits fixing the overall compressibility of the reservoir) is known, then only one further extensive reservoir property remains to be determined, namely the magnitude of the reserves. A straightforward application of material balance considerations will permit this determination. The problem becomes very much more difficult if we wish to determine not only the magnitude of the reserves but also the magnitude of water drive, if any, which is present. In principle, a combination of material balance and fluid flow considerations will permit this evaluation. Finally, if neither the magnitude of reserves, the degree of water drive, nor the ratio of oil to gas present in the reservoir is known and it is desired to determine all three of these variables, the problem could in principle be solved by a fluid flow-material balance analysis which determines the overall compressibility of the reservoir at various points in its history. The change in compressibility with pressure would provide a means of determining the ratio of gas to liquid present, since the compressibilities of gas and liquid vary differently with pressure variation. However, in practice this problem is probably so difficult as to defy solution in terms of basic data precision apt to be available.&apos; It is the purpose of this discussion to illustrate the second case, which involves the determination of two unknown variables, single phase reserves and degree of water drive, from pressure and production history and fluid property data, and to study the precision with which these unknowns can be determined in this manner in a particular case. Although an electric analyzer developed by Bruce as used in making the calculations to be described, numerical methods necessary in carrying out the process have been devised and have been applied for this purpose. Schilthuis,&apos; for example, developed a comprehensive equation for the material balance in a reservoir. He combined this with a simplified water drive equation, assuming that the ratio of free gas to oil was fixed by geological data and that a period of constant pressure operation at constant rate of production was available to determine the constant for his water drive equation. On this basis he was able to compute the reserves and predict the future pressure history of the reservoir. Hurst developed a generalized equation permitting the calculation of the water drive by unsteady state expansion from a finite aquifer. He showed in a specific case how the water influx calculated by his equation, using basic geologic and reservoir data to fix the constants, matched the water influx required by material balance considerations. Old3 illustrated the simultaneous use of Schilthuis&apos; material balance equation and Hurst&apos;s fluid flow equation for the determination of the magnitude of reserves and a water drive parameter from pressure and production history. He used this method to calculate the future pressure history of the reservoir under assumed operating conditions. As a basis for determining reserves, Old assumed a value for his water drive parameter and calculated a set of values for the reserves, using the initial reservoir pressure and each successive measured pressure. The sum of the absolute values of the deviations of the resulting reserve numbers from their mean value was taken as a criterion of the closeness of fit to the experimental data possible with the water drive parameter assumed. New values of the water drive parameter were then assumed and new sets of the reserves calculated until a set of reserves numbers having a minimum deviation from the average was established. The average value of- the re-

Jan 1, 1949

By R. L. Ash, R. R. Rollins, C. J. Konya

An investigation was performed to determine conditions for optimizing the spacing of simultaneously initiated multiple explosive columns. This was done by using models of mortar, dolomite, and Plexiglas with 10-grain mild detonating fuse as the explosive charge. It was desired to simulate blastholes with multiple primers initiated by detonating fuse or when high-velocity explosives are used in low-velocity materials. It was found that optimum spacing between multiple charges was strongly influenced by charge length. At less than optimum charge length, the spacing at which complete shearing was possible between adjacent charges decreased exponentially with a subsequent loss of broken material volume. For charges fired simultaneously, larger burdens and spacings were possible as compared to those necessary for single-crater charges. For each material studied, there was a characteristic optimum charge length and a maximum attainable spacing at any given burden. Proper selection of the spacing distance between charges is fundamental to successful blasting. Its value directly affects the cost of drilling and explosives used per unit of broken material. In addition, the choice of a spacing that is Compatible with a given set of blasting conditions aids in the control of fragmentation sizing, ground vibrations, overbreak, and throw which in turn, influence other production costs. For example, normally loaded blastholes that are spaced too closely invariably promote overbreak and usually give coarse fragmentation. Unless care is taken, airblast and violent flyrock will occur and under certain conditions cutoffs and misfires may result. Too large a spacing, on the other hand, frequently leads to conditions that form bootlegs or toes. The choice of a particular spacicg to use, however, is largely a matter of individual experience and judgment, usually based on trial and error. Very little is known or can be found in the literature with regard to how the spacing between charges is related to field conditions and charge geometry. As a general rule, the firing time sequence of adjacent charges and properties of a material are thought to have the most significant influence on the spacing distance best suited for any given field condition. For example, delayed initiation of adjacent charges usually always requires a closer spacing than when charges are fired at the same time. This should be expected if one considers that the energy normally dissipated and lost in the surrounding ground from charges fired independently would be captured and utilized for breaking material between charges when they are initiated together. Spacing can be extended also when charges are aligned with structural planes of a material, such as jointing, along which shearing is relatively easy. It is customary to relate the spacing (S) between charges to their common burden (B) in the form of a spacing ratio, or SIB. The burden normally is considered as the optimum depth or distance from any single charge perpendicular to the nearest free or open face at which the desired fragmentation and maximum crater yield are obtained. For production blasting, value of the ratio is generally considered to vary from 1 to 2, depending on conditions.1-6 When adjacent charges are fired independent of one another, the value varies from 1 to about 1.4, the closer amount being employed to square corners or produce craters having the ideal 90" apex angle. The larger ratio is the geometric balance value for craters having an apex angle of 135". The basic ideal crater forms in the plane of the charge diameter for charges fired independently are shown in Fig. 1. In the event charges are fired simultaneously, geometric balance in the plane of their charge diameters suggests that a spacing ratio near 2 would be appropriate, as illustrated by Fig. 2. In practice, however, some compromise ratio value must be selected to conform with the specific ground conditions. An example would be where the jointing planes tend to produce 60° or 120° crater angles, the appropriate geometrically balanced charge arrangement being given by Fig. 3. In this condition, the spacing ratio is 1.15, not 1 or 1.4 as suggested for the 90° cratering of independently fired adjacent charges. In view of the foregoing, it would seem logical to assume that whenever charges all having the same burden are fired at the same time, spacing distances always can be greater than those permitted by charges fired independently. In practice this is not the case, however.

Jan 1, 1970

By A. Romero-Juarez

This paper is concerned with the effect of propping-agent concentration on flow capacity of a fracture in the case in which there is embedment of the propping agent. Previous published studies have shown definitely that there is a relationship between fracture flow capacity and prop ping-agent concentration, and it has been shown that the theoretical results are confirmed by laboratory experiments. The problem of directly finding sand concentration for maximum flow capacity of a sand-propped fracture is solved, and formulas and charts are given to obtain this concentration under various conditions of effective overburden pressure, medium sand-grain diameter and rock properties. It is shown that, for the same effective overburden pressure and the same rock characteristics, optimum sand concentration in pounds per gallon does not depend on medium sand diameter. For conditions met in bydraulic fracturing operations, it is found that sand concentration in grains per square inch for maximum flow capacity varies within a wide range of values; this indicates the convenience of using data of fracturing pressure and rock characteristics for calculating sand concentration in order to achieve the best results in fracture treatments. INTRODUCTION It has been pointed out in the published literature1 that one important factor controls the success of a hydraulic fracturing operation — the propping of the fracture. The main effect of the propping agent is to hold the fracture open by means of the reaction forces that oppose the pressure due to the overburden. It is assumed in this paper that the propping agent is sand, that the grains are spherical and uniform in size, and that this sand is distributed in a monolayer in the fracture. The existence of a sand concentration value for maximum fracture flow capacity has been recognized since the publication of one of the first studies on the subject.2 In the reference cited, flow capacity was assumed to be proportional to the cube of the fracture volume per square inch not filled by sand. The obtained experimental curves relating to flow capacity and sand concentration had the same appearance as the curves obtained with the afore-mentioned assumption. Recent work3 has shown that fracture permeability, in the case of a monolayer, can be calculated accurately by means of a modified Kozeny-Cacman equation; thus, it can be shown that fracture pqmeability is a function of free volume in the fracture and of the rock and sand wet surfaces. The flow capacity of a fracture is dependent on the permeability of the fracture and on the fracture width. If all excessive quantity of sand is present in the fracture, its width may be large but its permeability i:; low due to the greater flow resistance; too few grains of sand per square inch are present in the fracture, sand embedment in the rock is greater and the width becomes small enough to appreciably decrease the rate of flow. To determine optimum conditions of flow, therefore, it is necessary to know the dependence of fracture width on density or concentration of propping agent. This relation is given in a study4 in which the approach to embedment is similar to the one made in a penetration hardness test of metal. With the equacions presented by Darin and Huitt,3 the number of sand grains per square inch that results in maxirnum fracture capacity can be determined graphically by plotting fracture capacity vs sand concentration. Since the relationship between fracture capacity and sand concentration may be considered known, one is led to formulate the analytical problem of directly determining sand concentration for maximum fracture flow capacity, in order to study the effect of the various quantities that come into the process. The solution of this problem appears to be interesting both from the viewpoint of fluid mechanics and from the practical viewpoint of oil production in hydraulic fracturing treatments. EQUATION FOR FRACTURE FLOW CAPACITY It can be shown3 that the fracture flow capacity of a monolayer is given by the following expression,

The Library of the above-named Societies is open from 9 A.M. to 9 P.M. on all week-days, except holidays, from September 1 to June 30, and from 9 A.M. to 6 P.M. during July and August. The Library contains about 55,000 volumes, including sets of technical periodicals and the publications of scientific and technical societies. Members of the Institute, with few exceptions, are forced to spend a portion of their time in localities isolated from sources of information. To these the Library can render valuable service through correspondence; letters requesting information will receive special attention. The Library is prepared to furbish references and copies of articles on mining and metallurgical subjects; to determine the existence of mining-maps, and to furnish general information as to the geology and mineral resources of all countries. All communications should be made as definite as possible so that the information received may be what is desired and not include collateral matter which may not be of interest. The time spent in searching for such collateral matter will be saved, and the information will be sent more promptly and in more usable shape. The members of the Institute can be of service to the Library by forwarding copies of mining-reports, maps privately issued, and similar material, which will be classified, indexed, and made available to other members. Suggestions for additions to the Library, either by purchase or personal solicitation as gifts, will be welcomed. It is hoped that members while in the city will use the Library freely, and assurance is given that most careful service will be rendered to them.

Jan 10, 1913

By T. C. Atchison, W. I. Duvall

Over the past ten years a series of investigations have been conducted to determine some of the pnysical processes involved in breaking rock with confined concentrated charges. Detailed discussions of many of these investigations have been published elsewhere.1-6 Laboratory experiments made by other investigators using Hopkinson pressure-bar techniques have shown that solid materials are fractured in tension by the reflection of an incident compressive stress pulse at a free surface.7-12 In these tests a small charge of explosive is placed in contact with one end of a bar. Detonation produces some plastic flow and crushing of the bar near the charge and generates a compressive stress pulse that travels along the length of the bar. At the free end of the bar the compressive stress pulse is reflected back into the bar as a tensile stress pulse. If the tensile strength of the bar is exceeded during this reflection process, a tensile fracture normal to the length of the bar is produced, and the broken end of the bar moves forward with a constant velocity equal to the average particle velocity trapped in the broken fragment. The new surface formed by the fracture becomes the new free end of the bar that reflects the remaining portion of the incident compressive stress pulse. This process is repeated any number of times until all of the incident stress pulse is reflected. Hino has demonstrated this kind of breakage for three rock types—marble, granite, and sandstone." He has defined a blastibility coefficient, B, as the ratio of compressive strength, C, to tensile strength, T, thus: C The quantity B is the maximum number of slabs that can be produced by reflection breakage. Normally fewer slabs are produced because of loss of energy as the stress pulse travels through the rock. Fig. 1 illustrates reflection-type fracture for a triangular compressive stress pulse. The number of slabs produced by the reflection breakage process is the first whole number less than the ratio of the peak stress of the incident pulse to the tensile breaking stress of the solid. Thus the number of slabs is given by the thickness of each slab is given by and the total length of rock broken is given by N = number of slabs S - peak stress in incident pulse T = tensile strength of the rock F = fall length of incident stress pulse h = thickness of each slab D = total length of rock broken During the reflection process the particle velocity at the free surface is twice the particle velocity in the incident stress pulse. Thus the velocity with which the broken fragments move forward is given by where v, = velocity of broken fragment, and v = average particle velocity contained in that portion of the incident pulse trapped in the broken fragment. Results of these laboratory experiments cannot be applied directly to rock blasting where the explosive charge is placed in a drillhole. In laboratory tests the charge is unconfined and in contact with the rock in only one direction. In a drillhole additional confinement is offered by the rock surrounding the charge and by the stemming placed above it. This additional confinement may be enough to allow the explosive gases to do additional work on the rock during their expansion. Other writers have discussed possible effects of gas expansion&apos;on rock breakage.13-10 However, very few experimental data are available to determine to what extent expansion of the gases is responsible for rock fragmentation. The USBM has studied the physical processes involved in breaking rock with confined concentrated charges by using simple crater tests breaking to one free surface. Crater tests have been performed in four rock types: granite, sandstone, marlstone, and chalk. Table I gives some physical properties of these rocks. Fig. 2 shows plan and section drawings of two typical crater tests and illustrates some of the test variables measured. For these tests the charge was placed at the bottom of the drillhole and primed with an electric cap. The hole was stemmed to the collar with sand and the charge detonated. Size and shape of the crater were measured after it was cleared of broken rock. As a given charge size was buried deeper in a drillhole, the crater depth usually was equal to or

Jan 1, 1960

By J Price, M. R. Geer, H. F. Yancey

COAL mine operators recognize coal as an abrasive material, because the wear of drilling, cutting, and conveying equipment is reflected as a cost item for replacement of parts. Similarly, industrial consumers of coal experience abrasive wear on all coal-handling equipment. Operators of pulverized fuel plants are doubtless most keenly aware of the abrasiveness of coal, because under the high contact pressures developed between coal and metal in pulverizers, abrasive wear is increased many fold. Moreover, experience in operating pulverized fuel plants has demonstrated that some coals are much more abrasive than others. Hardgrove&apos; stated that maintenance costs entailed by the wear of grinding elements is often a more important variable than the cost of the power required to pulverize different coals. Craig2 also reports that one coal may cause pulverizer parts to wear several times faster than another. It is apparent, therefore, that those concerned with pulverizing coal could profitably employ a method for estimating the abrasiveness of different coals, just as they utilize standard tests for thermal value, grindability, and ash-fusion temperature to assist in selecting the most suitable and economical coal to use in a particular plant. The objective of this investigation was to develop a test procedure that would be suitable for general use in estimating the abrasiveness of coals. However, few, if any, of the standard tests now used for evaluating the properties of coal are the product of a single investigation or the result of a single investigator&apos;s efforts. Rather, in each case, a testing procedure was devised by one investigator, used by others on a wider variety of coals, and finally refined completely as the result of the joint efforts of a number of interested people. Thus, the test procedure for estimating abrasiveness developed in the course of this work may not be refined sufficiently in its present form for general use, but it may serve as the starting point from which an acceptable test procedure can be developed. The method has been used thus far on only about a dozen coals, and there has been no opportunity to attempt a correlation between experimental results and actual plant experience. Only wider use of the procedure by other investigators and correlation with plant experience can determine to what extent the method will have to be modified to render it suitable for general application. Test Method Although the literature contains no record of an attempt to devise a method for estimating the abrasiveness of coal that could be used industrially, several investigators have tested properties of coal that are closely related to its abrasiveness. The abrasiveness of a material generally is considered to be related to its hardness, and hardness tests for coal have been employed by Heywood,&apos; O&apos;Neill," and Mathes. Also, the resistance of coal to abrasion, a property that presumably is related to the abrasiveness of coal, was measured by Heywooda and by Simek, Pulkrabek, and Coufalik.2 11 these investigators tested only individual pieces of coal. Since coal is a heterogeneous material having components of varying properties, tests of this type can yield results having little more than academic interest. Only a test method that utilizes a representative sample of coal can give results that are useful industrially. The abrasion tests used for various other materials have been considered for adaptation to testing the abrasiveness of coal. The tests used for metals,7-9 paving and flooring,&apos;" and rubber," cannot be used because coal is not sufficiently abrasive.~ The present experimental work was begun before World War II and was conducted by three research fellows"&apos;" working under a joint agreement between the University of Washington and the Bureau of Mines. After a great deal of preliminary work with a variety of apparatus and materials, a test procedure was developed which consisted of rotating a test disk 2Yz in. diam in a steel mortar containing the coal sample. The shaft carrying the test disk at the lower end and a 100-lb load on the upper end was free to move vertically. The bed of coal in the mortar was kept fluid by low-pressure air admitted through a port near the bottom of the mortar. Measurable wear on an Armco iron disk could be obtained in this test procedure, but, despite extensive efforts to eliminate them, several major disadvantages remained in this test method. First, with most coals the amount of wear on the iron disk did not exceed a few milligrams. Second, a single type of disk was not applicable for all coals. A smooth iron disk gave satisfactory results with both bituminous and sub-bituminous coals, but hardly any wear with anthracite or coke. A disk having studs or projections gave more satisfactory abrasion losses with anthracite and coke and presented no operating difficulties with free-burning bituminous and sub-bituminous coals. It could not, however, be used with caking coals because these coals formed a

Jan 1, 1952