Search Documents

By Bruce Jago, Joe Zhou, Chris Martin

"From the perspective of metallurgical processing, gold ores can be classified into free-milling and refractory ores. Their extractive metallurgy are largely driven by mineralogical factors such as: gold particle size, association with other minerals, coatings and rimmings, presence of cyanicides, oxygen consumers and preg-robbers, presence of refractory gold minerals and locking of submicroscopic gold in sulfide mineral structure etc. Gold process mineralogy helps address issues and problems related to gold ore processing. It provides useful information on process selection, flowsheet development, recovery improvement and reagent consumption optimization. By combining classic mineralogical techniques, modern instrumental analysis and diagnostic metallurgy, the mineralogist is able to balance the various types of gold occurrence in an ore. This paper will provide a broad review of gold process mineralogy and the techniques available, and use case studies to describe how they can be used together, to provide the complete picture of gold process mineralogy in an ore. Examples will be used where applicable.INTRODUCTIONSince the early 1980’s, a great deal of work has been done in the characterization of gold ores, particularly refractory ores (Gasparrini, 1983; Wang, 1984; Hausen, 1985; Haines, 1986; Henley, 1989; Chryssoulis et al., 1990; Wang et al., 1992, 1994; Kojonen, 1999; Simon et al., 1999; Hong et al., 2000; Pals et al, 2003). Process mineralogy has come into its own as a respected inter-discipline in the fields of mineralogy and metallurgy – a subject that occupies an important place in both research and industry. As the free-milling gold ores are being depleted and more refractory ores are being discovered and processed, gold process mineralogy receives more and more attention from both the mineralogist and metallurgist as it helps to solve various problems encountered during gold ore processing. Steadily rising gold prices are also stimulating the need for gold process mineralogy due to the increasing demand for and production of the yellow metal. At SGS Lakefield Research, gold process mineralogy studies are conducted routinely on all types of gold ores and mill products to assist in a wide variety of metallurgical projects. This paper will briefly discuss the gold ore types and gold mineralogy, review the common mineralogical factors affecting gold extractive metallurgy and the techniques available to gold process mineralogy, introduce the procedure employed at SGS Lakefield Researc"

Jan 1, 2004

By Alexander Malahoff

The summit of L_ihi submarine volcano changed dramatically after seismic activity beneath L_ihi during the month of August 1996. Bathymetric re-surveys of the L_ihi edifice using SeaBeam, Simrad and deep-towed Reson-Focus multibeam systems showed the presence of new north-south striking fissures, normal faults and a new pit crater located at the site of a former 300-meter high volcanic cone. Comparisons between the pre-1996 SeaBeam data and the new bathymetry show a net loss of about 100 million cubic meters of rock volume on the summit. This loss is localized at a new pit crater -- the 300-meter deep by 1,300-meter wide Pele's Pit. Submersible observations with Pisces V show the northern edge of Pele's Pit to be bounded by a normal fault located north of the site of the former Pele's Vents cone that has disappeared into the new pit crater. A new hydrothermal field, the Ikaika Vents, extending from a water depth of 1,311 meters to 1,287 meters has formed at the northeastern edge of the floor of Pele's Pit. Submersible observations of these new Ikaika Vents during September 1996 showed vigorous venting at the site with extensive white-colored bacterial mats hanging off the rock face above the vents. Follow-up mapping and sampling of the vent sites during August 1997 showed vent temperatures of up to 198°C, dense populations of vent shrimps around the vents and a centimeters-thick veneer of fine-grained polymetallic sulfide mineralization along the rock walls above individual vents. The mineralogy of the sulfides shows a barite, chalcopyrite, sphalerite, and pyrite composition.

Jan 1, 1998

By David Stephenson, Hugo Granados, Steven Klohn

"HudBay’s Constancia project is a copper molybdenum concentrator located in southern Peru. In 2010 Ausenco was contracted to perform a feasibility study optimization review which resulted in an improved design with significantly increased throughput capacity at a minimal increase in capital cost. Construction began in 2012 with commercial operation commencing in May 2015. This paper outlines the process plant design as well as describing the plant’s major equipment selections, start-up and early operations experiences.IntroductionHudBay Mineral’s Constancia project is located at 4200 meters above sea level (masl) in southern Peru, approximately 630 km south-east of Lima, Peru. The property can be accessed by road from Cusco, Peru, (approximately 110 km to the north), or from Arequipa, Peru (approximately 215 km to the south). Matarani, accessed via Arequipa, is the closest port.Constancia is located within the Yauri-Andahuaylas metallogenic belt. The region contains several large copper-gold-molybdenum porphyry deposits such as Antapaccay (by Glencore) and Los Chancas (by Grupo Mexico), and copper skarn deposits such as Tintaya (by Glencore) and Las Bambas (by MMG). Constancia’s ore reserves contain copper and molybdenum with trace amounts of gold and silver.Ore is mined by an open-pit operation. The process plant recovers copper and molybdenum concentrates, storing tailings in a conventional tailings management facility. Infrastructure at the site includes a grid power supply (via a 70 km overhead transmission line from Tintaya), water supply, workshops and an accommodation camp. Mine life is 22 years.Site construction works began in 2012 with first copper concentrate produced in late 2014. The project was executed for approximately 1.75B$USD.Ausenco’s design approach focussed on developing a capital efficient copper molybdenum concentrator. This paper outlines key features incorporated in the concentrator design, plant layout and equipment selection, as well as start-up experiences and early operational results."

Jan 1, 2016

By R. Grandori, L. Pellegrini

The construction of the La Joya hydropower plant (Costa Rica) was awarded by the Spanish power company, Union Fenosa, to the Italian Joint Venture formed by the companies SELI SpA and Ghella SpA. The project, located in a mountainous region of central Costa Rica, includes a 7.9-km-long head race tunnel having an excavation diameter of 6.18 m. A double-shielded TBM was selected to perform the tunnel excavation, which involved mainly volcanic rocks in variable but generally very weak conditions. The tunnel crossed long and continuous sections in completely weathered rock masses with no residual cohesion and with high water inflows. The paper describes and analyzes: The special methods adopted to overcome these sections with the TBM includ-ing face stabilization by chemical grout and foams as well small by-pass tunnels in the crown combined with TBM advance/support The grouting design and the grouting method that was implemented to meet the waterproof design specifications and restore the original groundwater level in the mountain The additional geological investigations performed to improve the knowledge of the rock mass in front of the TBM and the outcomes of these investigations The performance of the TBM in different ground conditions and on average along the entire tunnel Finally, the author’s comment about the possibility of implementing and improving the technologies experimented with in this project in future similar applications.

Jan 1, 2007

By James W. Wade

This paper describing the mining practices at the Tintic Standard mine is one of a series being prepared by the United States Bureau of Mines. In the preparation of this paper each subject has been presented by the engineer in charge. L. R. Dobbs wrote the part pertaining to tae technique of square-set mining, J. M. Snow the part on geology, and M. D. Paine, the part on ventilation. All of the drawings are the work of H. W. Waterfield of the engineering department. The statistics of the Tintic district ware prepared by A. G. MacKenzie. Each writer has attempted to show the difficulties that arise in operating and the practices used to overcome them.

Jan 1, 1930

By H. K. Lin

Deepwater discharges of sulfide flotation tailings have been practiced for several decades. Oxygen penetration and organic matter content in such environments are the two major factors that determine their reduction-oxidation condition and hence the rate of sulfide mineral dissolution under such discharge environments. In this study, the effects of solution redox potential and pH on the dissolution of some heavy metals from sulfide flotation tailings were examined. Zn, Pb and Cu concentrations in solution were all higher at a higher redox potential and lower pH. Dissolved Zn concentrations reached 100 ppm even at a low redox potential of 105 mV and near neutral pH after 33 days. Pb concentrations were below 2 ppm at 105 mV and near neutral pH after 165 days. At 105 mV, the Cu concentrations were below 0.2 ppm at near neutral pH after 165 days. At pH 2.3, the Fe concentrations were higher at 420 mV than at 105 mV. But at pH 6.8, the Fe concentrations were lower at 420 mV than at 105 mV, due to the oxidation of Fe2+ to Fe3+ and the subsequent precipitation of ferric hydroxide. Arsenic concentrations in solution were all less than 0.002 ppm. For Zn, Cu and Fe, no significant differences were observed between the tests in solutions prepared using seawater and those using distilled water. However, Pb concentrations reached higher levels in solutions prepared using seawater.

Jan 1, 2009

This is the seventh in a series of reports on properties of commercial motor fuels, made in accordance with a cooperative agreement between the Cooperative Fuel Research (C.F.R.) Committee and the Bureau of Mines of the United States Department of the Interior. The first report gave data on motor gasolines sold during the winter of 1935-36.3/ The present report covering 1,875 samples represents the products of 83 companies plus those from the Pacific coast, which were averaged by the Pacific Coast Technical Group before being submitted and are published exactly as they were submitted. For the regular grade 985 samples consisting of 73 different brands are averaged. The premium grade consists of 489 samples of 40 different brands. The third grade has 401 samples averaged from 51 different brands. In this report the districts have been reduced to 21 by combining North and South Texas into one district. Fl Paso samples are reported with the South Rocky Mountain district, and Amarillo, Tex., has been included with the Middle Rocker Mountain district. In the present survey samples from Fort Smith, Ark., have been included in the Tulsa area, Indianapolis has been included in the St. Louis district. Cleveland represents the State of Ohio. Samples from Little Rock, Ark., and North Carolina have been included in the Memphis district. The Pittsburgh district includes samples only from this State and Wheeling, W. Va. The actual sampling points in each area are indicated an table 1.

Jan 1, 1939

By Timo Kankaanpää

Keywords: Solvent extraction, dispersion, phase separation, CFD The liquid-liquid dispersion phases are separated by gravity in a solvent extraction settler. This separation step is made more effective by picket fences. The picket fence geometry and its physical placement have a significant effect on the separation process, because the purpose of a fence is to smoothen and control organic-aqueous dispersion flow in the first part of the settler and achieve a deep and dense dispersion layer in the front end of the settler resulting in a clean phase separation. In this study, physical phenomena in the solvent extraction process and phase separation in a solvent extraction settler with and without picket fences have been examined using a commercial CFX software package. It can be concluded that the used CFD-approach proved to be a complementary tool for optimizing and designing the solvent extraction settler.

Jan 1, 2005

This is the seventh in a series of reports on properties of commercial motor fuels, made in accordance with a cooperative agreement between the Cooperative Fuel Research (C.F.R.) Committee and the Bureau of Mines of the United States Department of the Interior. The first report gave data on motor gasolines sold during the winter of 1.935-36.3/ The present report covering 1,87 samples represents the products of 83 companies plus those from the Pacific coast, which were averaged by the Pacific Coast Technical Group before being submitted and are published exactly as they were submitted. For the regular grade 985 samples consisting of 73 different brands are averaged. The premium grade consists of 489 samples of 40 different brands. The third grade has 401 samples averaged from 51 different brands. In this report the districts have been reduced to 21 by combining North and South Texas into one district. El Paso samples are reported with the South Rocky Mountain district, and. Amarillo, Tex., has teen included with the Middle Rocker Mountain district. In the present survey samples from Fort Smith, Ark., have been included in the Tulsa area, Indianapolis has been included in the St. Louis district. Cleveland represents the State of Ohio. Samples from Little Rock, Ark., and North Carolina have been included in the Memphis district. The Pittsburgh district includes samples only from this State and Wheeling, W. Va. The actual sampling points in each area are indicated on table 1.

Jan 1, 1939

The oxidation of ferrous sulphate to ferric sulphate with air and oxygen in a gas lift percolator was investigated with initial Fe2+ concentrations in the range 0·01 M to 1·0 M,temperatures in the range 20-80°C, and pH in the range oto 2.The rate of formation of Fe3+ was found to be well described by the rate expressionEquation 1and a consistent mechanism is proposed.The presence of Cu2+ was found to cause an increased rate of oxidation, the effect being greater with Cu2+ concentrations increasing in the range 0·001 M - 0·01 M in 0·2 M Fe2+. The effect of other solutes containing Ni, Zn, Mn, Co, Hg, Cl, Mo, NH4, As, Cr, or Na at a concentration of 0·001 M in 0·2 M Fe2+ was negligible.

Jan 1, 1972

By S. Sarac

In this study, graphical and mathematical methods used to determine operating points in single or multiple fan ventilation systems are summarized. The problem is mathematically formulized for parallel connected fans in which analytical solutions are not possible. For n number of parallel connected fans, a non-linear equation system that consists of n number of non-linear equation sis derived. The Newton-Raphson method, which is a numerical analysis method, is adapted to solve this non-linear equation system and the solution steps are given as an algorithm. A computer program based on this algorithm is written and its applicability is proved.

Jan 1, 2000

By Giuliana A. Zelada

This paper reports on a study of 43 compression load tests (five fully instrumented) and ten pullout tests of auger cast-in-place (ACIP) piles in sand. Eight design methods reported in the literature were used to predict the measured pile capacities. Ultimate load capacities of the ACIP load tests were determined using two different definitions of ultimate capacity and compared with predictions. Load transfer behavior was determined from a study of the fully instrumented pile tests and compared to predicted values. Recommendations are made regarding the design approach most appropriate for these foundations and the best procedure for estimating ultimate capacity from pile load tests.

Jan 1, 2000

By Kristen E. Coose

Concrete is a mixture of a fine aggregate (sand), a coarse aggregate (gravel), cement, and water. Cement itself is the most expensive of the ingredients but generally amounts to only about 10% of the concrete mixture. Cement raw materials include calcium, alumina, iron oxide, silica, and gypsum. They are heated and chemically changed into clinker "minerals," which are then ground and shipped as cement. Each of the four main clinker minerals-tricalcium silicate, dicalcium silicate, tricalcium aluminate, and tetracalcium aluminoferrite-impart specific strength and compositional properties to the final concrete product. During hydration of the cement, chemical changes produce new minerals that include calcium silicate hydrate gel (C-S-H, the main constituent of hydrated cement paste), calcium hydroxide, and calcium, sodium, potassium, and sulfate ions, among others. Two main reactions can occur that may be deleterious to concrete. Normal cement pore solutions, in a hydrated cement, contain sodium hydroxide and potassium hydroxide. If these alkali concentrations are too high, they can react with the siliceous aggregate (alkali-silica reaction, or ASR) to form an alkali-rich calcium silicate hydrate gel (C-N+K-S-H), which, upon prolonged contact with water, can expand to crack the concrete. The other deleterious reaction occurs when the mineral ettringite (Ca6A12(S04)3OH12. 26H20) forms after the concrete is hardened (delayed ettringite formation) and cracks the concrete. Many of the concrete bridges and structures built in New Mexico in the past 20 years are in need of replacement due to deterioration. Many others constructed very early on in the century are still sound. By collecting and analyzing samples of both old and new concrete and good and bad concrete, we compared their differences on both a microscopic and macroscopic scale. Samples taken from concrete sidewalks, parking lots, drainage canals, spillways, and a railroad-trestle foundation have construction dates that range from 1923 to 1997. They were analyzed by electron microprobe and by petrographic microscopy. Cracks, air voids, and aggregate alteration were primary locations for investigation and analysis. Using the electron microprobe, ettringite was found in voids of most of the samples, both in good and bad concrete. This leads us to believe that ettringite formation is not causing distress in the concrete. There was one exception in which ettringite was found lining a crack. In this instance, ettringite may have the ability to further the cracking and cause distress. No alkali-silica gel was identified with the microprobe. Using the petrographic microscope in cross-polarized light, a light-brown material was observed that lined some of the air voids and cracks; however, exact mineralogic identifications were not made. Although many instances of chalcedony and other deleterious quartz phases were found, no clear evidence of ASR-gel deterioration was present in our samples using these methods. However, positive identification of the ASR gel was made on an in-place concrete near one of our sample sites and in several sample cores using a newly developed chemical test for the soluble potassium ions associated with the gel. The chemical used in this test is sodium cobaltinitrite, Na3Co(NO2)6.

Jan 1, 1999

By D. Lusinga

Hydrocyclones are a common feature in almost all mining operations in the world, serving mainly as classifiers. Some of their advantages include low capital costs, low space requirements and their ability to reduce residence time in closed circuit grinding processes. Although an extensive body of literature exists for hydrocyclones, these devices are still inherently inefficient, and more research is currently being undertaken, particularly in the field of modelling. In the vast body of hydrocyclone literature published so far, there has been little or no effort devoted to analysing the effect of the cone force ratio on the performance of hydrocyclones. The cone force ratio is defined as the ratio of the spigot to the vortex finder diameter (Shah, 2005). In this study a total of 44 tests was carried out in a custom-built rig at the University of Cape Town. These tests were aimed at evaluating the effect of the cone force ratio on the performance of as mall diameter hydrocyclone. The cut size and water split were used as the criterion for evaluating the performance of the hydrocyclone. Results from the tests showed that the cut size decreased with an increase in the cone force ratio. The cut size also appeared to decrease as the calculated locus of zero vertical velocity (LZVV) shifted inwards. The water recovery to the underflow appeared to increase with an increase in the cone force ratio. The effect on the cut size of adjusting the cone force ratio was found to be higher for a coarser feed than it was for a finer feed.

Jan 1, 2009

By N. Damiano, C. Zhou, B. Whisner, M. Reyes

"The industrial Internet of Things (IIoT)—a concept that combines sensor networks and control systems—has been employed in several industries to improve productivity and safety. NIOSH researchers are investigating IIoT applications to identify challenges of and potential solutions for transferability of IIoT from other industries to the mining industry. Specifically, NIOSH has reviewed existing sensors and communications network systems used in U.S. underground coal mines to determine whether they are capable of supporting IIoT systems. The results show that about 40% of the installed post-accident communication systems as of 2014 require minimal or no modification to support IIoT applications. NIOSH researchers also developed an IIoT monitoring and control prototype system based on low-cost microcontroller Wi-Fi boards, to detect a door opening on a refuge alternative, activate fans located inside the Pittsburgh Experimental Mine, and actuate an alarm beacon on the surface. The results of this feasibility study can be used to explore IIoT applications in underground coal mines based on existing communications and tracking infrastructure. INTRODUCTION The nation’s approach to health and safety in mining has historically been reactive in the sense that health and safety regulations and mandates have traditionally been passed as a result of tragic mining accidents. An example of this is the Mine Improvement and New Emergency Response Act (MINER Act) of 2006, which was passed as a result of several mining accidents where tracking of and communications with underground personnel were severely limited. While this reactive approach has led to the development and implementation of highly impactful safety interventions and safety practices, there currently exists a technological revolution that can shift the paradigm toward a proactive approach based on preventative methods: Internet of Things (IoT). IoT is the network of physical objects or ""things"" embedded with electronics, software, sensors, and connectivity to enable that network to achieve greater value and service by exchanging data with the manufacturer, operator, and/or other connected devices [1, 2]. It should be noted that IoT has been defined from various perspectives, and hence numerous definitions exist in the literature. The apparent ambiguity of the IoT definition might be due to the fact that IoT is syntactically composed of two terms: internet and things. The first one pushes towards a network-oriented vision, while the second tends to put the focus on generic objects. The traditional internet can be viewed as an internet of computers —i.e., “things” connected to the internet are computers only. IoT “extends” the internet by allowing other generic things that are commonly found in our daily life, such as refrigerators and cars, to be interconnected via the internet, and thus creates a bridge between the virtual word and the physical world."

Jan 1, 2017

By W. G. Davenport

Chapter 1 showed that flash smelting is a key step in extracting copper from sulfide ores. It also showed that there , are two versions of flash smelting: (a) Outokumpu flash smelting, which uses oxygen-enriched air blast and injects dry feed downwards into a hot furnace (b) lncoflash? smelting, which uses industrial oxygen blast and injects dry feed horizontally into a hot furnace. This chapter describes Outokumpu flash smelting.

Jan 1, 2001

By J. -P. Chilès

The discrete Gaussian model is a very popular change-of-support model for estimating block grade distributions. It is designed for a stationary random function Z(x) that is regarded as the transform of a stationary random function Y(x) with Gaussian marginal. It is based on an assumption that is not fully true and thus constitutes an approximation to the exact solution. We examine the effect of this approximation on the modelling of the marginal distribution of block values of a lognormal random function. The initial discrete Gaussian model and a variant of that model are considered.

Jan 1, 2014

By R. G. Cawthorn

The vertical distribution of platinum-group elements (PGE) in the Merensky and UG2 chromitite Reefs is highly variable, but does follow certain general patterns that allow the testing of geological models. In the Merensky reef, historically popular geological representations and interpretations have suggested that the mineralization occurs in a pegmatitic pyroxenite bounded by two chromitite layers. However, there are now enough published exceptions to this relationship to expose its fallacy. The vertical distribution of the PGE depends largely upon the separation between these two chromitite layers or the absence of one chromitite layer. Where separation is small or only one chromitite layer is present, there is considerable PGE mineralization in the footwall below the pyroxenite and lower chromitite layer. With increasing separation between the two chromitites, the mineral¬ization occurs progressively higher in the succession, and tends to track the upper chromitite layer. The view presented here is that the entire Merensky package is comprised of three events of magma emplacement, in which each has produced variable reaction with, and erosion of, the footwall, causing recrystallization of pyroxenite into pegmatitic pyroxenite. The magma then produced a chromitite layer followed by pyroxenite. Each chromitite layer is associated with some mineralization, which was extensively complemented by the subsequent formation of a sulphide liquid carrying PGE that trickled down through about 1 m of permeable crystal pile. The sulphide liquid formed after, not coincident with, the pyroxenite layers. The distribution of PGE in the UG2 chromitite shows up to three upward depletion sequences of PGE, also suggesting three events. The PGE are associated only with the chromite. The contribution from a sulphide liquid is considered to have been minimal for two reasons. Cu and Ni are extremely low in the UG2 which would not be expected if sulphides had been present. Also, by analogy with the Merensky, downward trickling of sulphides would have caused some PGE to be present in the footwall. There are some PGE in the footwall, but only where remnants of chromitite occur, and the hanging wall is lacking in PGE. In the northeastern Bushveld the UG2 chromitite is thinner than elsewhere, and there is a UG3 chromitite layer, not present elsewhere, that has a high Pt/Pd ratio. It is surmized that in the rest of the Bushveld the UG3 only appears to be absent because it has merged with the UG2 to produce the high Pt concentration at the top of the composite UG2. The non-accumulation of the package of pyroxenite-norite-anorthosite between the UG2 and UG3 in the northeastern Bushveld, which is not seen elsewhere, creates an interesting challenge to the formation of silicate layering in the complex. The mineralization in the Platreef is not addressed here. Keywords Bushveld complex, Merensky Reef, UG2 chromitite, geological evolution.

Jan 1, 2011

Tumbling mills and some theoretical and practical aspects of milling are discussed. Attention is directed towards the difficulty of relating the dynamical analyses of the motions of particles within mills to the existing treatments of breakage processes. This synthesis may be advanced by a study of milling that utilizes the matrix analysis of breakage processes. Application of the matrix analysis to results obtained on (i) grinding in open-circuit a zinc ore in a conical pebble mill and in a modified conical mill, (ii) "grinding of cement clinker in a compound mill, and (iii) to some drop-weight grindability tests is discussed. It is shown that the 'IT-breakage model adequately describes the breakage processes of each of the tests. (The proportion broken in each breakage cycle is given by 'IT and is independent of particle size.)The size distributions of the products are accurately predicted in the cases considered. Matrix analyses of the breakage of coals, various ores and a cement clinker have been made. Appendices give an elementary account of the matrix methods and give the equations fully describing reverse order closed circuit operations. The matrix analyses of breakage processes and the survey are used to examine briefly closed circuit grinding.INTRODUCTIONTumbling mills are slow-speed crushing and grinding machines which have been developed from some fairly simple types into many widely differing forms. Many of these versatile mills have proved extremely useful in ore dressing and mineral beneficiation, whereas others have been used in the cement industry and less extensively in the coal industry.The performance of even the simplest tumbling mill is very imperfectly understood, whilst the more difficult problems associated with the balancing of complex milling circuits and with the establishment of the optimum operating conditions for the whole of a complex circuit are even less well apprehended.Our knowledge of the performance of tumbling mills is rather sharply divided into two categories: the first consists...

Jan 1, 1957

By W. M. Merritts

A joint investigation by the Bureau of Mines and the Christopher Coal Co. at the Humphrey No. 7 mine, operating in the Pittsburgh coalbed in northern West Virginia, was undertaken to gain more insight concerning the hazardous emission of methane gas in mining, This report describes the equipment, procedure, and results. Horizontal bleeder holes drilled in the faces drained large quantities of methane, eliminating downtime of mining equipment. Also, larger quantities of methane were removed by forcing 14 to 25 g.p.m. water at 25 to 200 p.s.i. 5 pressure into the bleeder holes, In one infusion test a fluorescein dye was added to the infusion water; in later mining operations, the infusion water was traced for a distance of 1,500 feet. With vacuum pump applied to the horizontal hole the gas-flow rate increased, and the methane content decreased.

Jan 1, 1962