Search Documents

Australia is a highly significant hub in the global minerals business. This paper will present the results of global and national studies looking at the potential 20 to 30 year future evolution of the mining and metals industry.   Globally, the World Economic Forum launched a Mining and Metals Scenarios project in January 2009 to explore the question: 'How will the environment for the global mining and metals industry look in 2030?'. As part of a structured strategic foresight process, eight workshops were held across the world over the course of a year which brought together over 200 leaders from the private sector, government, academia and non-governmental organisations. The workshops were supplemented by personal interviews and guided by an international advisory board.   Critical uncertainties for the future of the industry were identified and compare extremely well with the outcomes of a national survey of AusIMM members identifying the key issues and drivers shaping the future of their operations. Global context was one of the most significant issues identified by The AusIMM membership, which is expanded in the World Economic Forum study in terms of: geopolitics; global governance; action on climate change; and economic outlook. More specifically, this global study developed three divergent yet plausible scenarios describing potential environments for the global metals and mining industry in 2030. These scenarios are compared with ground-breaking global foresight work coordinated by the Stockholm Environment Institute in the 1990s. The scenarios will help decision makers rehearse alternative futures, anticipate potential hazards and identify business opportunities - new solutions, new markets and new connections. The presentation will present these scenarios as a way of challenging thinking, testing established wisdom, and stimulating imagination about the business case for the mine of the future. As Marcel Proust put it, 'The real voyage of discovery consists not in seeking new landscapes but in having new eyes'. This is an ABSTRACT ONLY no paper was prepared for this presentation.

Jun 1, 2010

By G. Y. Chin, M. T. Dolan, W. L. Mammel

The least work analysis of constrained deformation, first proposed by Taylor and recently modified by Chin, Hosford, and Mendorf to include twinning, has been applied to the problem of axisynmetric flow. Calculations were made, for the case of (111}(110) slip and {111}(112) twinning applicable to fcc deformation. The solutions were obtained by linear programming techniques and with the aid of computer methods. The results show that in axisymnzetric tension the onset of twinning tends to decrease the plastic strength anisotropy as compared with the case of defortnation by slip alone. In compression, however, the strength anisotropy is increased as a result of twinning. The [100] and 11111 orientations and their vicinity are unfavorable for twinning in tension and compression, respecticely. With slight modifications, the present results may be applied to bcc metals deform ing by {110}(111j slip and (112} (111) twinning. Implications of present results to wire texture development of low stacking fault energy fcc materials arc discussed. PREVIOUSLY we' have applied linear programming techniques to obtain solutions to Taylor's least work analysis2 of constrained deformation of single crystals undergoing axisymmetric flow, such as occurs in an axially loaded polycrystalline aggregate. These solutions yielded values of the parameter M = o/.r = Y/w which relates the external stress-strain (u-c) curve to the shear stress-shear strain (7-y) curve for the crystal. The value of M is dependent on crystal orientation and hence enters significantly in texture strengthening considerations.3'4 Recently the Taylor hypothesis was modified by Chin, Hosford, and Mendorf' to include mechanical twinning. The latter has now been applied to the problem of axisymmetric flow and the results are reported in this paper. Calculations have been made for {111 )( 110) slip and {111 )(112) twinning applicable to fcc crystals. With minor changes in the indices of the slip and twin systems, the present results are also applicable to bcc crystals which slip on (110) (111) and twin on {112)(111) systems.

Jan 1, 1970

By Z. D. Hora

Sand and gravel resources of the southern coastal region of British Columbia may be linked to various episodes of Wisconsin glaciation. During Wisconsin and probably earlier times, the Fraser Lowland and coastal areas were subjected to repeated glaciations separated by non-glacial intervals. Each major glaciation was accompanied by isostatic and eustatic changes in sea level up to 200 m or more. As a result, the Lowland was covered by the sea during much of Quaternary time. Because the Fraser Lowland is surrounded on two sides by high mountain ranges, glaciers terminated in the sea during their maximum advance. Therefore, meltwaters released during deglaciation, together wit h retreating glacier ice, could produce widespread and extensive sand and gravel deposits both along the coast and throughout the Fraser Lowland and adjacent areas.

Jan 1, 1984

E. R. Rice, Wickenburg, Ariz. (communication to the Secretary*):—While this paper is primarily intended as a discussion of Mr. Riter's, I think it will be best to indicate my criticism by describing my own field-methods. It has been my experience that, for ordinary work, the regular transit-book is to be preferred to the card-system for recording notes. This is specially true when computations in the field are required, for it is then necessary to have at hand the total latitudes, departures, ancl elevations. Moreover, the transit-book is easier to carry and manipulate in wet or cramped places, and is not as liable to damage as the loose leaf or card. I use a regular transit-book in the field, and then copy my notes, sketches, etc., in an o5ce-book, entering also the latitudes, departures, and other reductions. These values are then copied into the field-book, securing a duplicate record, in case either book should be lost or mislaid. In the field-book, the notes are entered on the left-hand, and the sketches and remarks on the right-hand page. The next two pages are left blank for the latitudes, departures, bearings, reduced distances, etc. Of course, the notes for the different parts of the mine are entered in the field-book in the order in which they are surveyed; but in the they are entered systematically. It has been my experience that the system of keeping notes used by a surveyor, is particularly adapted to the needs and temperament of the individual. Otherwise, he would not be

Jan 1, 1912

By G. A. Vissac

IN the presentation which follows, washability curves, such as are commonly used in making studies preliminary to the cleaning of any coal or to the concentration of any mineral, have been reduced in each instance to a single straight line; thus all the most important characteristics of the coal or mineral are represented in a form more advantageous than that in which such data usually are shown. To obtain consistent results, float-and-sink data must be based on a single definite narrow size range, which may be designated "a coal unit." It will be shown that such units, each embodying a different size range, can be grouped to form what may be termed "a composite coal." A three-dimensional presentation will emerge, which will closely identify the coal, illustrate its characteristics in a new manner and explain many obscure facts. FLOAT-AND-SINK TESTS For the preliminary study of any cleaning or concentration problem, or in the control of any operation, float-and-sink tests must be made. The results of these tests are grouped in tabular form and then illustrated in various systems of curves; but if such a graphic representation is to have the utmost practical value, it must offer at least the following advantages: (I) It must be easier to read than the tabular results, and (2) it must be easy to plot, check and interpolate. The first requirement will be attained by reducing the number of curves to a minimum, and the second by some system of representation that will use only straight lines. Of the many washing curves now in use, only the "recovery curve," often known as the "composite curve "-a somewhat misleading term-will be considered. Any point on the recovery curve will show what percentage of weight of clean coal can be recovered when the mineral is washed at a certain specific gravity. As an example, the tests listed below Fig. I are illustrated in the drawing by a curve obtained by plotting (I) as abscissas the ash content of the coal recovered and (2) as ordinates the corresponding percentage of refuse discarded; but, in general practice, instead of writing down that weight, it is customary to set down its complement, which is the corresponding weight of coal recovered. From this curve can be deduced the information necessary for the solution of most practical problems. NEW FORM OF GRAPHIC PRESENTATION The resemblance of the recovery curve in Fig. I, or, for that matter, the recovery curve obtained in common practice, to a power curve is apparent. The general equation of such a curve is y = axn where n is negative. Taking logarithms: log y = a + n log x. Using a logarithmic scale, the equation of the recovery curve will be Y = A + bX. On this logarithmic basis, the recovery graph will be a straight line, using the same variables as in Fig. I ; namely, the ash con-

Jan 1, 1942

By F. Lombard

We consider in this paper the use of a modified version of Hotelling?s statistic in the analysis of particle size distributions. The statistic can be adversely affected by the presence of outliers among the data. We propose a competitor to the statistic that is based on ranks, and hence is less sensitive to outlier effects. The results of a Monte Carlo study suggest that the rank test is highly competitive with the Hotelling test in its ability to detect differences between two mass size distributions. The calculation of the rank statistic is explained in detail and its application is illustrated on two sets of data.

Jan 1, 2012

By F. Lombard

We consider in this paper the use of a modified version of Hotelling?s statistic in the analysis of particle size distributions. The statistic can be adversely affected by the presence of outliers among the data. We propose a competitor to the statistic that is based on ranks, and hence is less sensitive to outlier effects. The results of a Monte Carlo study suggest that the rank test is highly competitive with the Hotelling test in its ability to detect differences between two mass size distributions. The calculation of the rank statistic is explained in detail and its application is illustrated on two sets of data.

Jan 1, 2012

By Michael J. Hannan, Thomas F. Torries

The nature of industry structure is examined and the structure of the aluminum, copper, lead and zinc industries are described by the use of factor analysis. Capacity is defined and the relationship between industry structure and capacity utilization is examined. It is demonstrated that excess capacity is possible in industries with competitive or monopolistic characteristics, but the effects of the overcapacity are vastly different in the two cases. In a competitive industry, cycles of overcapacity with attendant fluctuating prices are expected. A change in industry structure from monopolistic to competitive characteristics over time causes profits and economic rents to be transferred from the producer to consumers of the mineral.

Jan 1, 1989

The education system today is turning out potential Mine Managers and Superintendents for the coal industry who are lacking in adequate management knowledge. They are required by the Mining Act to be highly skilled operators who are operations and production oriented. This puts them in much the same category as airline pilots and ships' captains. The real need is for them to be management and profit orientated. Changes over the years in Managers' and Superintendents' roles are highlighted by identifying those professions and vocations with whom they must now work. Comparing classifications in the 1916 award and the current award, and making a comparison between management and mining degree courses, leads to the conclusion that separate degree courses are necessary for future Mine Managers and Mining Engineers. Mine management must become a profession in its own right if optimum profits are to be made from mining in the future.

Jan 1, 1980

By Wendell E. Dunn

INTRODUCTION The early history of the chloride process has been recorded in a chapter of a Bureau of Mines bulletin (9 by one of the inventors of a chloride process, Stewart Croasdale, who is famous for the Pohle- Croasdale Process. The earliest use of chlorine in metallurgy was in 1557 by Bartholome Medina. But apparently even earlier there was knowledge that salt could be used in gold and silver extraction because a treatise, de la Pyrotechnica, published in 1540 in Venice, asserted that silver could be obtained from its ores by a wet salt process. In 1609 a priest, Alonzo Barba, invented a process using a copper vessel and boiling a solution of salt to extract silver chloride ores. The copper vessel was, of course, one of the reactants. This led to the Kronke process introduced in Chile in 1860 which substituted cuprous chloride for the kettle. A major step was taken in Europe in the late 1700's to chloridize roasted silver ores in combination with amalgamation using iron to reduce the silver chloride. By 1843 the Augustan process was introduced. Ores were roasted with salt and leached with a saturated salt solution. The silver chlorides passed into solution from which silver was precipitated by copper. In 1848 sodium or calcium thiosulfate solution of silver chloride was suggested by Dr. John Percy, and the idea was applied in 1856 by Von Patera in Bohemia.

Jan 1, 1983

By Lloyd A. Morley, Alan M. Christman

General Six open-pit metal mines, five copper and one molybdenum, were visited during the course of the research. Four were investigated twice, in summer and in winter, in order to determine the effects of seasonal variations In temperature and moisture upon the measured parameters, Permission to enter the fifth mine was received too late to make an examination during the summer months. The sixth mine was visited to confirm the results obtained from the other five and specifically, to observe how the project recommendations (as will be presented in this chapter) would apply to open-pit metal mines in general. Information gathered during the ten plant visits is sufficient to form an adequate data base for the report. A complete presentation of all field trip data is given in Appendix 1. Further, a detailed description of the electrical distribution and grounding practices at each mine is also contained there. Typical System. At a typical mine as shown schematically in Figure 62, one or more substations with resistance-grounded secondaries are employed to transform the incoming utility-voltage to the lower potential used by the mining machines, generally three-phase 4160 V. At this level, pit distribution is carried on overhead lines, supported by wooden poles, to supply switchhouses located in benches near the particular piece of equipment. A switchhouse is a small portable (often skid-mounted) structure at the termination of an overhead line which contains disconnect

Jan 1, 1976

By Robert Ignouf

MY REMARKS, which I feel highly honored in being invited to make, shall be limited to a consideration of -the mining and metallurgical problems involved in this question; in fact, these problems alone might well be considered as suffi¬cient to explain the existence of the Ruhr situation and to justify the steps taken by France in this affair. Comparing the France of pre-war days with the France of the period following the Versailles treaty indicates clearly how the Ruhr question is a direct consequence of the new economic conditions which have prevailed in Europe since peace was signed between the Allies and Germany. Before the war, in 1913 for instance, the French out¬put of iron ore was 22,000,000 tons. This output, which decreased considerably during and immediately after the war, has about regained its previous level. It should be noted that 96 per cent. of the production of iron in France is obtained from the eastern part of the country, in the districts of Nancy, Briey,. Longwy, Metz, and Thionville. This concentration of practi¬cally all the iron mines into one district creates a very serious problem for France, involving strategic and economic consequences. The strategic problem arises from the fact that the whole iron resources of France lie close to the frontier of Germany, at a distance of only a few miles; hence it is of the' utmost importance for France to be assured that her eastern frontier shall not be at the mercy of a sudden invasion by an army which could in a few hours deprive France of her iron resources and render her powerless to defend herself. Invaded by her Eastern neighbor three times during the past 100 years, France has had dire experience with what such invasions mean for her. Wherever the German armies have passed, they have left an inefface¬able trail of desolation. France, therefore, intends to protect herself by sufficient and efficient guarantees against renewed invasions of her territory.

Jan 5, 1923

The Osborne copper-gold deposit (within the Trough Tank project area) is located in the Cloncurry Shire about 195 kilometres south-east of Mount Isa. It is hosted by Proterozoic metasediments of the Eastern succession of the Mount Isa Inlier. The deposit is overlain by 25-40 metres of Mesozoic cover, so it represents a true "blind target". Placer Exploration has 100% outright ownership of the property, having purchased the remaining 10% interest in the property from Shell in early 1993.Intensive exploration, both surface and via drilling at Osborne, has taken place since 1985, when CSR Limited Mineral Exploration and Development Group (acquired by Placer in 1988) entered into a joint venture with Billiton Minerals. Exploration effort from 1985 to June 1988 outlined a 20 kilometre strike length of magnetic horizons, in what was termed the North-East, Central and Southern anomalies, which were drilled with weak indications of copper-gold mineralization in the North-East anomaly. During the later half of 1989, four core holes were drilled into the shallower northern partofthe North-East anomalywhere previous drilling had outlined economic intersections of mineralization. These holes encountered high-grade copper-gold mineralization confirming the potential for an economic deposit to be hosted within the ironstone sequence. Further drilling involving over 400 holes (100 kilometres of drilling) has brought the deposit up to its present resource status.The current resource inventory is 36 million tonnes grading 2.0% copper and 1.0 grams pertonne gold and includes a high-grade zone of 13 million tonnes grading 3.6% copper and 1.5 grams per tonne gold. The discovery of this potential economic mineralization resulted from a persistent exploration effort based on indications that a large mineralizing system existed within a sequence of ironstones.

Jan 1, 1993

A significant resource of uranium has been discovered within a buried palaeochannel located in the south western corner of the Officer Basin of Western Australia by PNC Exploration (Australia) Pty. Ltd. The mineral- isation is redox controlled and is hosted by Eocene organic rich paludal sediments. Prob- able insitu reserves are 13,000 t of contained uranium with an average ore grade of 0.12% at a cutoff grade of 0.03% U. The Mulga Rock palaeochannel has been traced for at least 100 km beneath the Quaternary surficial cover. The main channel is horseshoe shaped with the channel arms vary- ing from 5 to 15 km in width. The western arm connects to the Lake Raeside-Ponton Creek drainage system, whilst the eastern arm'is open to the south. The headwaters of the palaeo- channel are thought to have been the Lake Minigwal drainage system, prior to its capture by the Lake Rason system. Differential.erosion of the basement rocks during the Late Cretac- eous to Early Tertiary strongly influenced channel shape and resulted in the main channel being confined to a Permian,mudstone filled valley. Upwards of 100 m of fluviatile sediments accumulated within the palaeovalley during a period of subsidence in the Early Tertiary. Three broad stratigraphic units have been re- cognised within the channel: 1. An upper unit of fluviatile sands with interbedded lacustrine sediments, which have subsequently been wholly oxidised and variably silicified and ferruginised. 2. A middle unit of lacustrine to paludal sediments, which host the main mineral- isation. 3. A basal unit of fluviatile gravels and sand. The presence of excellently preserved Nothofagidites in organic rich paludal sediments from the middle unit, suggest that unit was laid down in a non-marine, fluviatile, humid environ- ment during the Middle Eocene.

Jan 1, 1988

By T. B. Whitmoyer, J. E. Sneckenberger, M. A. Jerabek, C. F. Stanley

The measurement of acoustic impedance at frequencies up to 6,400 Hz has been proposed as a diagnostic tool to yield more detailed information about lung condition. A three generation branching system of plastic Y-tubes has been constructed. The impedance of the branched system is measured by an impedance tube using the two-microphone technique. The impedance spectrum is used to calculate the reflection coefficient spectrum, which is converted to an impulse response by an inverse FFT. The area function is extracted from the impulse response by the Ware-Aki algorithm. Measurements of acoustic impedance of the branched system were made for several different end conditions of the eight terminal outlets. The outlets were either blocked with modeling clay, plastic packing material, or aluminum blocks or were unblocked. Six measurements were made of each configuration. In general, the more outlets blocked, the larger the shift to the right of the impedance magnitude peaks for frequencies above 3000 Hz. The impedance magnitude peaks for the plastic packing material were shifted in frequency to the right relative to the peaks of the other system terminations. The impedance magnitude peaks for the clay and aluminum system terminations were significantly different in magnitude but not shifted in frequency. The area functions at the end of the branched system were significantly different (a > 99%) for most of the end conditions. Area functions for the modeled branched lung systems between the end of the impedance tube and the termination outlets of the branched system were similar but distinctively different for the various end conditions. These results indicate that the two-microphone technique can sense small changes in the terminating end conditions of a rigid system of branched tubes and thus may sense impedance changes due to lung disease in the conducting airways. Keywords: acoustic impedance, lung model, area function.

Jan 1, 1991

By W. A. Bruce

This paper covers the theory and present state of development of an apparatus for the nonmathematical analysis of complex problems of reservoir and well behavior. At the present stage of development of this apparatus, it is capable of solving the water-drive problem of any reservoir performance for which a mathematical solution has been presented in the literature, either on the basis of steady or unsteady-state flow analysis. The theory is presented for the extension of the scope of this analyzer along the lines of the present development. The resulting apparatus should be able to solve reservoir-performance problems previously thought too difficult for practical solution by mathematical means. As an illustration of its operation and present scope, the analysis of a reservoir problem and synthesis of expected performance is presented. Introduction In this method† of analyzing problems of petroleum-reservoir and well behavior, the porous continuum of the reservoir is assumed to be divided into small blocks, or units. Each block is assumed to be small enough so that for a material balance on the fluids in the block an average pressure for the block can be used. In analyzing reservoir behavior from this point of view by mathematical means, a set of simultaneous difference equations would be obtained (two for each block), which must be solved or combined into a partial differential equation. In the method to be described here, instead of solving these sets of equations analytically, electrical units are constructed, which will behave in respect to the flow of electricity exactly as the reservoir units behave under various conditions of fluid flow. The electrical units are then wired together in the same way that the reservoir units are connected naturally by virtue of their geometrical locations and shapes. Fluid-Flow Equations Consider a block in a reservoir. Assume that reservoir fluid (or fluids) is flowing in at one face and out at the opposite face, and that the block is small enough so that at all points the stream lines are within one degree of perpendicular to the face. Furthermore, assume that the pressure, pi, at the upstream face (Fig, I) is not more than one per cent greater than P2 the pressure at the downstream face. A material balance will show that the mass of fluid flowing out less the mass of fluid flowing in will equal the change in density times the fluid volume. This is expressed in the equation:

Jan 1, 1943

By W. A. Bruce

This paper covers the theory and present state of development of an apparatus for the nonmathematical analysis of complex problems of reservoir and well behavior. At the present stage of development of this apparatus, it is capable of solving the water-drive problem of any reservoir performance for which a mathematical solution has been presented in the literature, either on the basis of steady or unsteady-state flow analysis. The theory is presented for the extension of the scope of this analyzer along the lines of the present development. The resulting apparatus should be able to solve reservoir-performance problems previously thought too difficult for practical solution by mathematical means. As an illustration of its operation and present scope, the analysis of a reservoir problem and synthesis of expected performance is presented. Introduction In this method† of analyzing problems of petroleum-reservoir and well behavior, the porous continuum of the reservoir is assumed to be divided into small blocks, or units. Each block is assumed to be small enough so that for a material balance on the fluids in the block an average pressure for the block can be used. In analyzing reservoir behavior from this point of view by mathematical means, a set of simultaneous difference equations would be obtained (two for each block), which must be solved or combined into a partial differential equation. In the method to be described here, instead of solving these sets of equations analytically, electrical units are constructed, which will behave in respect to the flow of electricity exactly as the reservoir units behave under various conditions of fluid flow. The electrical units are then wired together in the same way that the reservoir units are connected naturally by virtue of their geometrical locations and shapes. Fluid-Flow Equations Consider a block in a reservoir. Assume that reservoir fluid (or fluids) is flowing in at one face and out at the opposite face, and that the block is small enough so that at all points the stream lines are within one degree of perpendicular to the face. Furthermore, assume that the pressure, pi, at the upstream face (Fig, I) is not more than one per cent greater than P2 the pressure at the downstream face. A material balance will show that the mass of fluid flowing out less the mass of fluid flowing in will equal the change in density times the fluid volume. This is expressed in the equation:

Jan 1, 1943

By James E. Lawver

A SURVEY of Minerals Beneficiation Div. membership indicated genuine interest in automatic process control, but revealed that the average mill is operating with a minimum of self-regulating devices. Apparently, the mill operators are content to continue without controls until the practicality of self-regulating devices is more generally accepted, or, in the opinion of the writer, until automatic controls become a necessity to meet economic demands of product quality and cost. Both to stimulate interest in control, and to solicit publication of technical know-how by members and equipment manufacturers, this paper presents a cross section of control problems in two general categories: (a) Control problems that can usually be solved by reference to the literature, i.e. controls that have proven successful in the mineral or chemical industries; (b) Problems that are troublesome and frequently unsolved due to unavoidable adverse conditions in the mill circuit to be controlled.

Jan 10, 1953

By R. Schuhmann

PROBLEMS in expressing, interpreting, and using size-distribution data recur in many phases of mineral dressing; therefore it is desirable that size analyses be expressed in such a form, either numerically or graphically, that comparisons can be made readily. The size distribution of comminuted material is a result of a specific physical operation on the material and therefore should be governed by some definite physical principle, or cause-effect relation. Although this principle has not been discovered, it has been found that the size distribution of comminuted, homogeneous solids may be expressed by relatively simple mathematical equations. Two equations in particular, proposed by Gaudin1 and by Rosin and Rammler,2 respectively, have stood out, both having wide demonstrated ranges of application. The form and the method of application of the two appear almost irreconcilably different on the surface, but, as will be shown, they are asymptotic in the fine sizes when expressed in the same units. This paper gives the results of a further study of the Gaudin and Rosin-Rammler relations, made with the object of developing a form of size-distribution equation that will be more usable and more significant-particularly in comminution studies, both for interpreting the nature of size distribution and for evaluating surface. The work is summarized in the following: I. An equation relating cumulative per cent finer to particle size in the fine sizes is derived from Gaudin's original equation. That equation relates weight per cent retained on one size and through the next larger size in a geometric series of sizes to size (of the retaining screen, for example). 2. Comparisons with Gaudin's equation and with the Rosin-Rammler equation show that the cumulative equation applies to the fine sizes of all size distributions that fit either of these earlier relations, and, furthermore, has some advantages over each. 3. A simple graphical method of surface calculation based on the cumulative-equation plot is evolved and illustrated. 4. Additional confirmation of the cumulative equation was obtained in the more severe test of sizing down to the limit of a sedimentation balance a sample of jaw-crushed quartz. The equation was closely followed over a 20o-fold range in size, from approximately 0.4 mm. (35 mesh) down to 0.0023 mm. (2.3 microns), the lower limit of the size analysis made. There is no indication that the equation will not hold further to well below 2.3 microns. 5. Calculations of the surface produced in the jaw-crushing test mentioned above showed that at least 8o per cent of the total new surface was on particles finer than 35 mesh (about ½ 5 of the crusher setting). The minus 35-mesh portion was only

Jan 1, 1940

By R. B. Basham, C. W. Macune

Very small anomalies in oil well temperatures are detected and measured by recording the difference in temperature existing between two thermally sensitive elements which are Spaced several feet apart on a small diameter carrier and lowered into the bore hole. The elements operate in a balanced electrical circuit, sending only a reference signal to the surface as long as normal gradient temperatures are encountered. When either element enters a temperature disturbance, the circuit is unbalanced and produces large recorder deflections for minute anomalies. The system is highly sensitive to small changes in thermal gradients caused by gas or fluid movements and to the boundaries between beds of different thermal conductivities. Typical logs show successful applications of the process to locating gas and water leaks in casing, gas entry. and gas-oil contact. INTRODUCTION Temperature measurements in oil wells have been made for many years in the search for additional sub-surface data. It has been generally agreed that many conditions in the bore hole cause temperature anomalies which, if located and measured, could yield useful information. But it has been recognized also that many of the irregularities, such as those produced by small water and gas leaks in casing, might be very small and therefore difficult to detect with certainty.' Although tem-perature measurements have been made which show anomalies of less than one degree deviation from normal gradient temperatures, the results are usually uncertain and unconvincing. The general practice, therefore. has been to apply temperature surveys only to well conditions known to produce appreciable disturbances, such as large gas movements and cement hydration. The recent demand for additional data has focused attention on the problem of making dependable temperature measurements in the many instances when anomalies are small but none the less significant. It was for the purpose of detecting these minute irregularities in bore hole temperature that the program described in the following text was undertaken. It is emphasized that this presentation is not offered as a complete study, for the process is still partially experimental. But it is believed that the results obtained thus far will be of interest to the industry and will indicate some of the system's potential usefulness. DIFFERENTIAL MEASUREMENT The chief value of differential measurement lies in the fact that a difference between two factors can be measured with a scale of values best suited to that difference, a scale that might be cumbersome or impractical if applied to the factors themselves. When it is known that all factors will be nearly equal in value, and therefore the differences consistently small, it is possible to select a scale which will permit very accurate measurement of the differences. This advantage is in contrast to the difficulty of measuring small increments of a single variable, where the increment must be measured by a large scale which is suitable for measuring the entire factor. Gradient temperatures are usually measured under this latter condition where temperature measuring apparatus must have a scale extending over the range of expected temperatures, say 50' to 25O°F, and it becomes almost impossible to detect small irregularities of less than one degree. The use of large strip chart recorders and divided multiple ranges tend to amplify the small anomaly and make it more pronounced. But the instability of instruments, both

Jan 1, 1952