Search Documents

By Gerald C. Dohm

INTRODUCTION After a mining company has discovered a mineral deposit, the problem is then how to mine and process that deposit the best way. The principal problem facing managers or engineers who must decide on mine plant size, equipment selection, and long-range scheduling is how one can optimize a property not only in terms of efficiency but also as to project duration. Mine opti¬mization is dependent upon the interaction of con¬tributing parameters which leads to maximizing the net present value (NPV) Realistic mine planning is the key, to the analysis. The process is to be used when evaluating undeveloped properties or when a complete revamping of existing facilities is scheduled The proc¬ess is not intended for detailed mine design but rather for property evaluation and feasibility studies. The derivation of the various input parameters will only be discussed to the point where a meaningful correlation can be made with the overall analysis. The interrelationship of the parameters will be the central point because most company guidelines and financial analysts have their own standards when deriving the various inputs. Assume that the data is totally acceptable and pro¬vides at least the following categories of information and guidelines- (1) mineral and waste inventory; (2) geologic, hydrologic, and geotechnical criteria; (3) topographic layout including property boundaries; (4) metallurgical flowsheet, recovery and design criteria; (5) access, water, power, and environmental base-line study; and (6) financial constraints (minimum rate of return, payback period, etc).

Jan 1, 1979

By R. Laird Auchmuty

The A. I. M. E. Subcommittee on Bituminous Mining has been trying for several years to secure the information that was collected by the Marquette Cement Manufacturing &. on the subsidence of its property caused by mining coal beneath it. The Marquette company was willing to have these data used, but it was impossible to obtain them because the records had been filed away. About two years ago these records became available and the engineering data presented herewith were compiled from them by R. L. Auchmuty, assisted by L. E. Young, who was connected with the collection of the information. On account of the cost, it is impossible to reproduce all of the data here, although the original information is on file with the records of the Committee on Ground Movement and Subsidence. The accompanying charts show representative and average data and give only the engineering information with enough description to tell how it was obtained. It is hoped that all of those connected with the original case will participate in the discussion that this will bring forth. Howard N. Eavenson, Chairman, Subcommittee on Bituminous Coal Mining. PRobably the most complete records of subsidence due to coal mining in the United States are those compiled by the Marquette Cement Manufacturing Co. at LaSalle, 111. These records, covering a period of more than three years, were compiled and used as evidence during the suit of the Marquette Cement Manufacturing Co. vs. the Oglesby Coal Co. at Oglesby, LaSalle County, Ill., in which the cement company successfully endeavored to stop the coal company from mining under its property. Fig. 1 shows a typical section of the geological formation. The cement company was mining the LaSalle limestone which lies about 125 ft. under the surface. The limestone is 30 to 35 ft. thick and underlain by about 15 ft. of shale. The cement company was mining about 20 ft. of the limestone, and in some places 10 to 12 ft. of the underlying shale, by the room and pillar method. The mining of the limestone always preceded the mining of the shale. The coal company was mining by longwall advance methods the NO. 2 seam of coal 42 in. thick underlying the limestone bed by 435 to 470 ft. The coal mine, a very old one, had largely mined out its original property

Jan 1, 1931

By Robert J. De Angelis, James H. Barker

An improved method for solving dynawzical dislocation problems using a digital computer is described in this paper. Interactions between two distinct types of dislocations were studied: attractive screw dislocations; and Lomer lock forming dislocations. One dislocation is positioned in the lattice and is initially at rest, while the other dislocation is moved through the lattice on an intersecting slip plane at a constant velocity in the range 0.5 to 0.999C. (C is the transverse velocity of sound.) The results obtained from these computations indicate that screw dislocations account for a small fraction of the total strain over a wide portion of the range of velocities studied. They further indicate that mixed dislocations mainly repel other dislocations in the neighborhood of the active glide plane. From this a possible explanation for cell formation is put forth. The density of Lomer locks expected to exist after a strain of 0.2 was found to be 1.4 x 106 cm-2 which is in good agreement with indirect experimental estimates. IN the past, predictions of favorable or nonfavorable dislocation reactions were based on the associated changes in elastic strain energy. Such considerations take no account of the probability of the two dislocations coming into contact to react. Venables1 was the first to approach these probabilities by considering the interactions between two moving screw dislocations on perpendicular glide planes. Because of the restrictive types of dislocations and glide plane geometry employed, his results have limited application to metallic crystals. The work to be presented here develops a general approach to solving dynamical dislocation problems; either dislocation-dislocation interactions, presented here in detail, or dislocation interactions with any other suitably defined stress field. Two types of dislocation-dislocation interactions common to face centered cubic (fee) materials are considered: those between pure screw dislocations of opposite sign on intersecting slip planes and those between mixed dislocations on intersecting slip planes, that can react to form a perfect dislocation. This latter reaction, referred to as the Lomer reaction, produces a locked product dislocation that finds it energitical favorable to disassociate into two Shockley partials and a stair-rod dislocation. This partial configuration known as a Lomer-Cottrell (L-C) lock plays a major role in work hardening of fee crystals. seeger2 names the L-C lock as the prime contributor to Stage II hardening while Kuhlmann-wilsdorf3 and Meakin and Wils- dorf4 also state that it is a significant contributor to work hardening. However, with a few notable exceptions,5-7 direct observations of the Lomer lock and the L-C lock by electron transmission microscopy are scanty, and even these are subject to other interpretations.5,6 In a study of partial dislocations present in austenitic stainless steel, whelan8 did not observe any L-C locks at the head of pile-up groups. This result contradicted existing work hardening theories and led him to postulate an alternate theory based on the stress required to break away dislocations intersecting a pile-up group, from their stacking fault nodes. Due to the importance of the Lomer reaction in producing L-C locks which are an essential feature in current work hardening theories and because there exist no data giving direct quantitative values for the density of locks, and because there has even been some doubt expressed as to whether this important reaction occurs at all, a study of the dynamic behavior of the mixed dislocations which form the Lomer lock was undertaken. Due to their ability to cross-slip with relative ease, screw dislocations play an important role in the deformation of fee crystals. For this reason, the second type of reaction considered here is between screw dislocations of opposite sign. In addition, computations in volving screw dislocation interactions are relatively simple, thus providing a convenient check on the cornputational scheme employed. DEFINITION OF PROBLEM The force exerted on a dislocation due to a generalized stress field is given by the Peach and Koehler9 equation: Here t2 and b2 are respectively the tangent and Burgers vectors of the dislocation, and T1 is the stress dyadic defining the local stress field. The stress field may be externally applied or generated internally by the presence of a lattice defect, such as a second dislocation, as is the case in this work. Frank10 has shown that an equivalent momentum, P, of a screw dislocation can be defined by: Here, EST is the total energy of a screw dislocation and ESo is its rest energy. The left side of Eq. [2] is the time derivative of momentum and the right side is the position derivative of the energy due to the dynamical nature of the dislocation. The total energy of a dislocation is the sum of the potential and kinetic energies. Weertman11 has developed the expressions which were used here; these give the potential and kinetic energies of uniformly moving edge and screw dislocations in an isotropic medium.

Jan 1, 1969

By D. E. Menzie, D. L. Dauben

This paper discusses the physical parameters involved in the slow flow of high molecular weight polymer solutions in porous media. The interacting effects of polymer properties and porous media properties on flow performance are considered. Experiments were conducted with the polyethylene oxides, with molecular weights ranging from 200,000 to over 5,000,000. Frontal advance velocities ranged from I to 30 ft/day. The porous matrix consisted of a flow cell packed with glass beads. Polymer solutions were characterized by viscosity and normal stress measurements. Under certain conditions, unexpectedly high flow resistance was observed. This behavior was observed to be a function of flow rate, pore size, polymer molecular weight and concentration. The polymer solutions exhibit "dilatant" flow behavior in porour media in contrast with the pseudo-plastic behavior in simple flow systems. A theoretical ex-planation of such behavior is presented. INTRODUCTION The use of polymers in the injected water of a waterflood increases oil displacement efficiency by reducing the mobility (k,/pu) of the driving phase. A reduced driving phase mobility results in improvements in the areal sweep efficiency and in the vertical coverage in stratified reservoirs. This mobility reduction may be achieved by a permeability reduction, a viscosity increase or by a combination of the two. Early attempts at increasing the injected water viscosity were not successful because of the poor economics involved. The use of such materials as glycerin, sugar or glycols to increase water viscosity was not economically feasible. Attempts in using certain naturally occurring polymers were not too successful because of the high polymer losses to the rock. A high molecular weight, partially hydrolyzed polyacry-lamide was introduced 3 years ago as a waterflood additive.',' Initial work by Pye' indicated that the presence of these polymers in dilute concentrations decreases the water mobility 5 to 20 times more than would be expected from measurements of the solution viscosity. Such an effect would be of obvious economic value since only a small polymer concentration would be required to accomplish a large reduction in water mobility. This research was designed to study the basic flow mech anisms of polymer solutions in porous media. The interacting effects of polymer properties and porous media properties on flow performance are considered. This study indicates some of the conditions under which these interactions can lead to significant mobility effects. is valid only for Newtonian fluids. For polymer solutions and other non-Newtonian fluids the equation must be modified to consider that viscosity is a variable quantity. Modification of the Darcy equation to include non-Newtonian effects has been the subject of several recent investigations."" The modifications generally adopt a rheological model. such as an Ellis or Dower law model. to Dorous media by defining some characteristic channel radius. Most of these studies showed the porous media flow behavior to be predictable from viscometric data. Investigations involving the flow of high molecular weight polyacrylamide solutions through cores have generally encountered the high flow resistances reported previously by Pye. This high flow resistance has generally been attributed to an in-depth permeability reduction, as evidenced by a reduced permeability to water which has displaced polymer solution from the porous media. Marshall and Metzner" report high flow resistances, which they attribute to viscoelastic effects. In this paper, the effects of polymer molecular weight, pore size, flow rate and concentration are considered. The polymers used are the polyethylene oxides known as Poly-ox. This class of polymers was used because of its availability in a wide range of molecular weights and because of its known ability to propagate well through porous media. THEORY PHYSICAL PROPERTIES OF POLYMER SOLUTIONS The polymer molecule dissolves in water by means of hydrogen bonding, but retains some of its own structural identity while in solution. Nonionic polymers, such as polyethylene oxide, are generally considered to have a random coiling configration. This type of molecule has the ability to "sequester" or hold a large volume of solvent within its coils in a manner similar to that of a sponge. The random coil is easily deformed and under an applied stress

By H. C. LIVINGSTON

At the Hanna No. 4-A mine of The Union Pacific Coal Co. a new system of trackless mining is being utilized to extract a 26-ft coal seam. By using shuttle cars and a conveyor belt for haulage in the rooms and panel entries instead of pit cars and panel hoist a saving of 12 cents per ton is expected. Mining is conducted in a 26-ft coal seam laid down on a synclinal structure with active workings on a pitch of approximately ten to twelve degrees. The coal seam is interbedded between a shale bottom and sand- stone top. The coal is friable, high in moisture, and subject to spontaneous combustion. The mine is rated gassy by the State Coal Mine Inspection Department. Mining is conducted at the present time by unbalanced rope haulage an main slopes spaced on approximate 4300-ft centers with the three-entry system driven to boundary lines, strike entries being turned on approximate 1300-ft centers. Panels are driven to the rise between entries at approximately 750-ft intervals with panel rooms driven on the strike 500 to 600 ft long, leaving 150 ft of barrier pillar between each block of panel rooms.

Jan 1, 1949

By S. Norton, S. LeFevre

In the West, capitalists have expended many millions of dollars developing the low-grade porphyry ores of copper. Half a dozen of these great enterprises have proved to be wonderful commercial successes. They have demanded improved crushing and concentrating machinery and consequently it has been developed. Many improved methods, cheap power, superior business organization, all these have contributed to this success, but the main feature is the handling of the material in enormous quantities, on a manufacturing scale. The mining chance of "striking it rich" has been eliminated by the manufacturing certainty of handling large quantities of material of known value, which while of relatively low grade, is available in large tonnages, assuring a supply for many years' run of the mill. Then the returns on the money invested are sure. The concentration of low-grade magnetic iron ores, separating the magnetite crystals from the gangue by the use of magnets, is a field of work in which the lessons taught by the development of the porphyry coppers can be studied to advantage. Large-scale operations, and the liberal expenditure of enough money at the start to insure the most economical operations, are the means of securing the desired results. The problem is to utilize millions of tons, and we may safely say billions of tons, of now worthless iron-bearing rock and to produce from it 10,000,000 to 20,000,000 tons per year of high-grade ore carrying 60 per cent. iron or higher; to take the lean material as found in nature, varying widely in iron content, and bring it up to a uniform standard of shipping ore. At present these ores are mined carrying from 25 to 50 per cent, iron, and the shipping product is brought up to 60 or 65 per cent. Fe. If future economies of operation make it possible to extend this process so that 15 per cent. iron in the crude ore can be treated as a commercial success, the additional tonnage available will be enormous. A 15 per cent. Fe crude ore raised to 60 per cent. Fe concentrate with 5 per cent.

Jan 1, 1917

By Archiblad MacMartin

The Kunigin-Marien-Hiitte is the only works in Germany where the Bessemer process is carried on by the direct method. The Besserner plant there, is arranged after the true English type, and the only resemblance to the Swedish mode of 'procedure is the dispensing with the use of spiegeleisen at the end of the "blow." In a new department of the establishment, started within three years, each of the converters is turned by means of a very neat and compact reversible engine, the steel shaft of which is an endless screw, which turns against the oblique cogs of a large wheel attached to the shaft of the vessel. An advantage which this arrangement possesses over the ordinary English hydraulic arrangement, is the fact that the endless screw suffices to turn the vessel, in either direction, any number of complete revolutions; while even the latest American improvements, so clearly explained to us by Mr. Holley at the opening session of this meeting, do not secure even one complete revolution without changing the angle of inclination of the hydraulic piston. In the endless screw arrangement, there being no limit to the working of the motor in either direction (no return-stroke necessary), the vessel can always, unless outside reasons demand the contrary, be turned around to any desired position by the shortest cut, whether backwards or forwards. Also, the diameter of the cogwheel attached to the vessel can be made sufficiently great to avoid all unevenness of motion. In the old department the motive-power still continues to be hydraulic. If reports be true, the new department produces spiegelized steel, for the manufacture of all-steel rails. But the old department is still, as from the beginning, devoted to the production by the direct method of steel for steel-headed rails. The most remarkable fact connected with this direct steel is the ease with which it welds to the iron 'of the rail-packets, although no borax or other fluxing agent is used to facilitate the welding. It is very rarely that an exception occurs, and an ingot or a charge is discarded by the rail-mill. The mixture of pig-iron used for the production of this steel is melted in cupolas of very interesting construction (not to be de-

By William E. Ford, Edward Salisbury Dana

Normal phosphoric acid is H3P04, and consequently normal phosphates have the formulas R3PO4, R3(P04)2 and RPO4, and similarly for the arsenates, etc. Only a comparatively small number of species conform to this simple formula. Most species contain more than one metallic element, and in the prominent Apatite Group the radical (CaF), (CaCI) or (PbCI) enters; in in the Wagnerite Group we have similarly (RF) or (ROH).

Jan 1, 1922

By B. H. Caudle, W. J. Bernard

Factors which influence the success or failure of a waterflood can seldom be determined in the laboratory. For this reason pilot waterfloods are initiated in a repreventative portion of the oil reservoir in question. For a pilot flood to predict quantitatively the recovery to be expected in a field-wide waterflood operation, the pilot area must behave as though it were confined (surrounded by similar areas). In this study, laboratory fluid-flow models were used to determine the simplest pilot pattern, for particular conditions of mobility ratio and initial gas saturation, that would behave as though it were confined. Pilot patterns studied ranged in complexity from a single inverted five-spot to a grouping of nine regular five-spots. Only the innermost producing well in each pattern was studied. Model results showed that the optimum number of wells in the pattern depends upon the oil-water mobility ratio and the expected oil-bank size. Unfavorable mobility ratios will, in general, require more wells in the pilot pattern than will favorable mobility ratios. Pilot patterns in reservoirs which contain a dispersed, flowable, free gas saturation will require fewer wells than for the under-saturated case. The single inverted five-spot pattern was found to be unsatisfactory for predicting behavior of fully developed waterfloods. In particular, it is possible that, in reservoirs which contain a flowable, dispersed gas phase, the oil bank will never be observed at the producers due to the large amounts of free gas which continue to be produced with the oil. INTRODUCTION One method which has been used to predict the performance of a waterflood is the pilot flood. The pilot waterflood is a flood which involves only a small cluster of the reservoir wells and is located in a small, representative portion of the reservoir. The object is that oil produced from the pilot can, in some way, be related to the oil recovery to be expected from a field-wide expansion of the waterflood. However, these field pilot waterfloods have often been unreliable in the prediction of oil recovery in a fully developed waterflood. This unreliability has also been demonstrated in several laboratory studies of pilot floods. Some of the investigators have shown that there are situations in which the pilot flood oil production is far too optimistic with respect to the oil recovery in the fully de- veloped flood. Others4-G have shown that the pilot results can also be pessimistic, especially if the pilot waterflood is initiated in an oil reservoir which has been depleted by primary recovery and is at very low pressure. The major reason for this unreliability of pilot water-floods is the migration of fluids into or out of the pilot area. By the well-known method of images, if straight lines can be drawn to represent vertical planes of symmetry in a porous medium which contains pressure sources and sinks (injectors and producers), then these lines are invariant streamlines, or lines across which there is no potential gradient, and therefore no flow. In an actual reservoir, these lines of symmetry can never be established exactly because of reservoir inhomogeneities and irregular reservoir boundaries. However, if the reservoir is relatively large and contains wells in repetitive patterns, these lines of symmetry are commonly assumed to exist for the pattern units sufficiently far removed from the reservoir boundary. Lines of symmetry for the five-spot injection pattern are shown in Fig. 1. Each five-spot unit in this figure can be considered confined with respect to flow across its boundary. In pilot floods this is not the case. The lines of symmetry for the pilot patterns investigated in this study are shown in Fig. 2. It is obvious that the fluid within these pilots is not confined and is therefore able to migrate into or out of the pilot area. Intuitively, one can see that, if more wells are added to the pilot, the innermost unit tends to behave more and more like the confined pattern. However, there is a practical limit to the number of wells which should be placed in the pilot. This limit is usually determined by economic factors. It was the purpose of this study to use laboratory fluid-flow models to determine which of the previously mentioned pilot patterns will force the innermost producing well to behave as it would in a fully developed waterflood. Since fluid migration is influenced by initial saturation conditions and the mobility ratio, these factors were included in the study. The ultimate objective of this study was to develop data which would allow the operator to choose a pilot pattern and operating conditions that will yield a production history which can be applied directly as an estimate of the performance of each production well of the fully developed waterflood. BASIS FOR THE STUDY The basic problem of field pilot floods is the migration of the reservoir fluids into or out of the pilot area. This problem has been the subject of previously reported model studies on pilot floods. These studies have been concerned mainly with the development of arbitrary "correction factors" to be applied to the simple, unconfined pilot systems such as the single five-spot. The correction factors were intended to adjust the production history of the un-

By J. F. Kemp

IN submitting an additional contribution to the discussion on ore-deposits in the recent volumes of the Transactions, it is my desire to adhere closely to matters of material importance as affecting the actual processes. Judging from the discussion by Professor Van Hise, and from one or two reviews or other articles by associates of his in geological work, the impression seems to prevail that, in so far as my paper* referred to ,his masterly essay,†I was guilty of misconceptions regarding both the part attributed by him in it to the igneous rocks and the ubiquitous presence of the ground-water. I should regret extremely to be wanting in these particulars or to have misunderstood one for whose great services in this and other branches of geology no one has a higher esteem than myself. I have, however, in the most careful way, and with these points in mind, read his essay again, anti, taking it as a whole, I cannot gain any different impression of relative magnitudes than the one first received. The importance of the igneous phenomena and the restrictive influence of the processes described by " cementation " in their effect upon the ground-water, as set forth in Professor Van Hise's " Discussion " (Trans., xxxi., 292), impress me as being in very marked contrast with the same things in his first essay. If the point seems to anyone of sufficient importance, it can easily be decided by reading the originals. My paper was the result of some years of observation, reading and reflection, and was meant to be an independent contribution, controversial only in some subordinate particulars. In depicting the drama of the deposition of ores, I cannot but feel that Professor Van Rise assigned to the chorus and the scene-shifters some characters which, it seems to me, should

Jan 1, 1913

By Dae S. Young

A new probabilistic approach is introduced for slope stability analysis including adverse impact of blasting vibrations. The material strength within a slope area is randomized in this bi-variate joint probability analysis. To be a completely general engineering method, the new probabilistic approach employs the random variable transformation technique: the Hermite model of the Gaussian transformation function, which transforms the experimental histogram of strength values to the standard Gaussian variables. Because bi- normal joint probability is analyzed on the true probability region projected on the Gaussian transformed variable plane, it is an exact solution of slope stability based on the available sample data. Details of the probabilistic approach are given through a case study on an iron ore mine. The case study demonstrated that the probabilistic approach is effective and simple enough to be a routine engineering practice for slope analysis.

Jan 1, 1984

By C. S. Matthews, J. L. Mahaffey, W. M. Rutherford

This paper gives results of an experimental study of the sweep efficiency of a miscible displacement in a five-spot. The study was carried out in a parallel-plate glass model in which effects of diffusion were scaled at or near the molecular diffusion level. The experiments show that very early breakthrough (25 to 35 per cent of pore volume (PV) injected) may be expected in miscible floods because of the unfavorable viscosity ratio. However after 1 PV of displacing fluid is injected, the sweep rises to a reasonable value (50 to 60 per cent). Photographs show that small slugs of less than 10 per cent of PV tend to dissipate before breakthrough. A minimum slug size of 15 per cent of PV would appear to be necessary even in a relatively homogeneous formation. Presence of a slug whose viscosity is inter-mediate between that of oil and gas increases the sweep efficiency of the oil-gas system. In a typical system the sweep at breakthrough rose from 26 to 37 per cent of PV for a 25 per cent slug. The increase in sweep brought about by use of a large slug could well pay for the extra deferment cost of the additional slug material. INTRODUCTION Most miscible displacement processes involve the displacement of oil with fluids of much lower viscosity and density. The displacement process at these adverse viscosity and density ratios is dominated by instability phenomena, i.e., viscous fingers and gravity tongues. These phenomena have highly adverse effects on oil recovery. Although a number of laboratory studies have been made to determine the effect of adverse viscosity ratios on five-spot sweep patterns,l,2 the scaling of diffusion effects is uncertain. In the series of scaled model studies reported herein, an attempt was made to scale diffusion. Model studies of miscible displacements in which molecular diffusion predominates are per-mitted by controlling the parallel plate spacing which reduces convective mixing to arbitrarily small levels. To decide how this scaling relates to any particular Geld displacement necessitates an estimation of diffusion effects for the natural rock being considered and conditions under which displacement will be conducted. The approach normally taken is to extrapolate data obtained from stable miscible displacements performed in the laboratory, such as those presented by Brigham et al.3 The validity with which such an extrapolation can be applied to an unstable flow system has yet to be established. If this approach is accepted, a family of oil recovery curves can be generated for a single viscosity ratio based on Brigham's observation that the magnitude of the dispersion coefficient is dependent, among other things, upon specific rock properties. Objective of our test was to define the lower limit of this range by presenting the case where dispersion effects were reduced to the molecular diffusion level in both the transverse and longitudinal directions.

Jan 1, 1967

By Irwin Roman

IN a previous paper,1 a method was suggested for determining the depth to a bed in the simple case of a uniform overlying layer of constant thickness. The main purpose of the present paper is to show how a slight modification of that method has been used to interpret specific data obtained in the upper peninsula of Michigan by the staff of the Depart- FIG. 1.-THE TECTONIC CONFIGURATION. ment of Mathematics and Physics of the Michigan College of Mining and Technology. The method of Tagg2 has also been used on the same data, thus furnishing a comparison of the e two methods, as well as a method of using the two methods jointly. The method of measuring the apparent resistivity is that suggested by Wenner,3 and illustrated in Fig. 1. A current I enters the earth at the positive pole C and leaves it at the negative pole D. The potential V of the high electrode A, with respect to the low electrode B is measured by suitable means. The separation of the collinear points in the order shown, DBAC, is 1, the same. for each interval.

Jan 1, 1934

By Lester Uren

WITH the depletion of our older, and relatively shallow, oilfields and the necessity for securing new production from deeper horizons, much attention is being given to the improvement of oil-well pumps in order that they may function satisfactorily under the more difficult conditions imposed by increased depth. Until recently, practically all of the world's petroleum was secured from wells less than 3000 ft. deep, whereas horizons are now being explored to depths of from 5000 to 7000 ft. in some of the more prolific fields, and it seems probable that the greater part of the future supply must come from depths in excess of 3000 ft. While much of the early production in a new field is obtained by natural-flow methods, when the stimulus of high gas pressure fails, recourse must be had to some method of pumping. The perfection of a pump capable of economically and efficiently lifting oil from depths ranging from 3000 to 7000 ft. is therefore a matter of prime importance to the oil-producing industry. This paper is offered as an estimate of the present state of development of the oil-well plunger pump and as a review of the general principles involved in its operation in deep-well pumping. A design for a new type of oil-well pump embodying several novel features, is also described.

Jan 9, 1925

By W. R. Clark, W. W. Cotter

Jan 1, 1946

By W. W. Cotter, W. R. Clark

Jan 1, 1946

By Earl Oliver, J. B. Umpleby

It is believed that in connection with the study of unit operation by the Petroleum Division of the A. I. M. E. a review of the simple principles of unit operation would be helpful. To that end, the following discussion is submitted. It is not to be considered in any sense as the conclusion of the committee, but rather as material submitted for its consideration and represents at this time only the viewpoints of its authors. Growth of the Oil Industry In 70 years the oil industry has grown from a single well on Oil Creek, Venango County, Pennsylvania, producing 25 bbl, per day, to more than 400,000 wells scattered over five continents with a daily output of 4,000,000 bbl. In that time petroleum has expanded from its simple use as a medicine in a few localities to a multitude of diverse uses in homes of every country the world over. During those years the oil fields of the United States have produced two-thirds of the world's output and have shipped their excess production into almost every country. In those years several hundred thousand good wells in the United States declined into small ones and depleted shallow deposits gave way to deep zones; so that the United States crude oil industry gradually changed from one of low cost production to one of higher cost. Throughout its entire life it has labored under two serious economic disadvantages, namely, constantly recurring periods of demoralization through overproduction, and competitive development methods that are wasteful. Development of Foreign Competition During the transition from low to higher cost production in the United States, events even more significant were taking place in foreign lands. Jungles were tamed and deserts conquered. Oil deposits were found and developed until, in the words of Sir John Cadman, "the globe is now drenched with a commodity which, although necessary to life, involves but an irregular and sometimes meager expenditure of life in its

Jan 1, 1930

By Herty, C. H.

No attempt will be made here to review the previous work done by investigators on the general subject of inclusions, because it was discussed sufficiently in an early cooperative bulletin of this series (No. 36, 1928), and reference was then made to general bibliographies. The work described in this present bulletin deals directly with the study of the mechanism of deoxidation of steel with aluminum and the formation of inclusions containing alumina. Unquestion- ably, such inclusions are encountered in most steel-making practices and little has been done to determine their characteristics. Aluminum is used to ensure rapid deoxidation during the solidification stages of the steel. For this reason the possibilities of inclusion elimination are slight and alumina-containing particles will undoubtedly be present in the ingot. Besides this common practice, aluminum is often used in the ladle with other deoxidizing agents such as silicon and manganese. Ladle additions of aluminum are made to rimming steels to obtain the correct partial deoxidation desired in this type of steel. In killed steels the aluminum is added to ensure soundness, as far as blowholes are concerned, in the ingot. In either rimmed or killed steels, alumina or alumina-containing inclusions are found therein, the amount and type depending on the oxidation of the metal, the amount of aluminum added, and the addition of other deoxidizers.

Jan 1, 1957

By J. N. Maycock, P. W. M. Jacobs

Previous measurements of the thermoelectric power of ionic crystals are reviewed briefly. It is concluded that, while extensive measurerements are available on systems in which the electrode M has a cation in cornon with the sall MX, there is as yet little modern work on systems in which the metal electrode M' does not haue a calion in common with the salt. We haue now measured the thermoelectric power of- single cryslals of KC1 and of KC1 + SrCl2 between plati.nu?n electrodes over a much wider range of temperature than before. individual ionic heats of transport, which, while probably not very accurate, are of a reasonable order of magnitude, are obtained from this data on the basis of an assumption that charge carriers do not cross the electrode/salt interface. RELATIVELY little work has been done on thermal diffusion in ionic crystals. Reinhold and coworkers1-I measured the thermoelectric power and Soret effect in silver and cuprous salts and their mixed crystals. One electrode was kept at constant temperature and electromotive force was measured as a function of the temperature gradient. This is not a satisfactory technique if the results are to be compared with theory; rather should the electromotive force be measured for small temperature differences as a function of the mean temperature of the two electrodes. They used electrodes having the same cation as the salt. Patrick and Lawsons measured the thermoelectric power of AgBr and of AgBr containing 0.25 mole pct CdBr2 using silver electrodes and temperature gradients of 10°C in the temperature range 100° to 400°C. The results were analyzed using the Teltow9 values of the ratio of the mobility of the two types of charge carrier (cation vacancies and interstitials) found from conductance measurements and the wagnerlo estimate of the heterogeneous contribution to the thermopower, Ohet They obtained values of qi = 0.017 ev and q$ = -0.385 ev for the heats of transport of interstitial Ag ions and cation vacancies, respectively. Their data has been reanalyzed by Haga using the corrections to the Wagner formula for ehet given by Howard andLidiard; Haga found ev. Christy et a1.14 have measured the thermopower of AgBr containing 0.01 to 1.0 mole pct CdBr,. They analyzed their results in terms of the theory of Howard and Lidiard and found that excellent agreement with the theory was obtained if q5 + : + hF was set equal to 0.70 ev at 175°C and equal to 0.58 ev at 250°C. Since Teltow's value for hF, the enthalpy of formation of a Frenkel defect pair, is 1.27 ev, this means that qt + q: increases in magnitude from -0.57 ev at 175°C to -0.69 ev at 250°C. This result is in poor agreement with Patrick and Lawson's value of -0.37 ev for pure AgBr (for which the corrections to the Wagner estimate for ehet are very small). christy1' has also reported thermopower measurements for pure silver halides and for the systems he finds that qr + qt for AgBr increases in magnitude from about -0.55 ev at 150°C to about -1.1 ev at 400°C. The temperature dependence of the sum of the heats of transport in AgCl is very similar in magnitude. The sum qi + q cannot be decomposed into the individual interstitial and vacancy heats of transport without a knowledge of the entropy contribution from the change in the lattice vibrational frequencies around the interstitial and vacancy. Despite this it seems an inescapable conclusion from Christy's analysis that and -6 = Q Ag are quite strongly temperature-de pendent and that -q: is much greater in magnitude than the corresponding activation energy for vacancy migration, +0.37 ev. It remains a mystery as to why the data for pure AgBr are fitted so well by the assumption of temperature-independent heats of transport if association of impurity ions and vacancies is neglected.l The values for q$ and q: so obtained are also much more in line with those expected from a simple kinetic interpretation of the heats of transport due to wirtz.17 Christy and Hsueh have also measured the thermopower of CuCl between copper electrodes; 0 for pure CuCl is comparable with that for the silver halides but, owing to the limited solubility of divalent cation impurities it was not found possible to deduce values for the heats of transport. In contrast to these extensive measurements on cells of the type M MX 1 M, there have been relatively few attempts to measure the thermoelectric power of cells of the form M' 1 MX I M'. (We specifically exclude from discussion cells of this type in which the metal electrodes M' are in contact with gas containing a controlled partial pressure of either M or X,, since in these cells the electrochemical reactions are analogous to those of the first type.) The thermoelectric power of Pb12, PbC12, TlC1, TlBr, TlI, NaNO3, and KNO 3 between platinum electrodes was measured by Thie1e.l3 In discussing these measurements Holtan20 asserts that the thermopotential initially fluctuated wildly and even changed sign before approaching a steady value. A high degree of irreproducibility is also evident in the work of Nikitinskaya and Murin21 who first studied the thermopower of NaCl and KC1 between platinum electrodes. Much better reproduci-bility was achieved in our earlier measurements on KC1;22 these were, however, confined to a relatively narrow temperature range and one of the objects of the work reported in this paper was to extend the temperature range of the measurements of the thermoelectric power 8 of KC1. The only other published

Jan 1, 1967

By Zay, Jeffries

THE effect of thoria (ThO2) on grain growth in tungsten was discussed in some detail in a paper presented before this Institute by one, of the authors in 1918.1 In that paper it " was assumed that the, thoria was present in the tungsten entirely as a mechanical mixture. There was no indication in the experiments described that even a trace of the thoria was in solid solution in the tungsten. Langmuir,2 however, observed that thoriated-tungsten filaments under certain conditions showed electron emissions many thousand times greater than unthoriated tungsten at the same temperature. He concluded from his studies that a slight amount of the thoria had dissociated on heating the filament to a temperature near 220°. C. At this temperature and higher, part of the thorium diffuses to the surface of the filament where it evaporates at. a much faster rate relatively, than the tungsten, and also at a rate, nearly as fast as it is brought to the surface by diffusion: If the temperature is lowered to say 1800°,. C., diffusion of the thorium atoms to the surface is more rapid than evaporation, sown adsorbed film of thorium atoms is produced on the surface of the tungsten filament. The electron emission is governed by the outside layer of atoms so the emission from such a filament might be expected to approxi¬mate that of a thorium filament at the same temperature. PRESENCE OF METALLIC THORIUM IN THORIATED TUNGSTEN If the filament containing the adsorbed film of thorium. is heated to a sufficiently high temperature, the thorium atoms evaporate: rapidly and the electron emission becomes more nearly that of tungsten. On cooling again to 1800° C., or thereabouts, the emission 'builds up gradually to a maximum. The filament containing the adsorbed layer of thorium atoms is said to be "activated." By studying the rates of activation under different conditions of prior thermal treatment and at different temperatures, Langmuir was able to arrive at an estimate of the

Jan 1, 1927