Search Documents

By John T. Awald

The efficiency of search for porphyry copper and molybdenum deposits can be greatly increased. A model exploration system is here presented to show how geologic literature data and field surveys can be aided by modern statistical and computer methods. Assuming that major ore deposits occur in clusters, it is possible to evaluate rapidly large areas at modest cost and with minimum investment risk. An example of the success of these methods is given. Clusters of mine density and metal production covering the State of Nevada were plotted and machine contoured. Selection of a 47,300 square-mile area on the Nevada-California border was made. Within this area, detailed analyses were performed, and a 3000 square-mile site was picked for subsequent field examination. To initiate efficiently the field work, 81 stream sediment geochemical samples were collected and analyzed for copper, zinc, and molybdenum. These data established local variance and necessary statistical controls. Following this, a total of 151 geochemical samples were used to survey the 3000 square-mile area. Computer contouring of results directed attention to specific target areas for detailed geologic analysis, property acquisition and drilling. A total of 22 copper-zinc and molybdenum anomalies were identified. Four of these were recommended for further exploration. To date, two have been acquired and drilled and preliminary ore reserve and financial analysis have been made on one. The estimated cost and time for the regional exploration phases leading to these targets were thirty man-weeks and $16,000.

Jan 1, 1973

By Sirish L. Shah

It is now common to have archival history of thousands of sensors sampled every second over long time. Yet we frequently have process engineers complain:??.We are drowning in data but starving for information??. How can these rich data sets be put to use? This paper addresses the issue of information and knowledge extraction from data with emphasis on process and performance monitoring including fault detection and isolation. Most of the major plant, factory, process, equipment and tool disruptions are avoidable, and yet preventable fault detection and diagnosis strategies are not the norm in most industries. It is not uncommon to see simple and preventable faults disrupt the operation of an entire integrated manufacturing facility. For example, faults such as malfunctioning sensors or actuators, inoperative alarm systems, poor controller tuning or configuration can render the most sophisticated control systems useless. Such disruptions can cost in the excess of $1 million per day and on the average they rob the plant of 7% of its annual capacity. Over the last decade the fields of multivariate statistics and data visualization and analysis methods have merged to develop powerful sensing and condition-based monitoring systems for predictive fault detection and diagnosis.These methods rely on the notion of sensor fusion whereby data from many sensors or units are combined with process information, such as physical connectivity of process units, to give a holistic picture of health of an integrated plant. Such methods are at a stage where these strategies are being implemented for off-line and on-line deployment. This article attempts to outline the field of sensor fusion - the application of signal processing methods, in the temporal as well as spectral domains, on a multitude and NOT singular sensor signals to detect incipient process abnormality before a catastrophic breakdown is likely to occur Keywords: Process analytics, process and performance monitoring, multivariate statistical analysis, sensor-fusion

Sep 1, 2012

By J. Andrew McGlenn, Brian McInnes, Bruce Downing, Murray Gant

The Second Narrows Water Supply Tunnel crossing beneath Burrard Inlet is an important component of the Greater Vancouver Water District’s plans to increase seismic resilience and meet increasing water demands. The project included the 110-meterdeep (350 ft) South TBM Receival Shaft, one of the deepest shafts to date in the sedimentary rock unit underlying much of Metro Vancouver. The shaft encountered varied sedimentary rock types, including fractured rock, with accompanying high groundwater inflows. This paper discusses the design and construction challenges associated with this shaft and the necessary means to complete the shaft in time for TBM arrival.

Jun 13, 2023

By Gregg W. Davidson, J. Andrew McGlenn, Frank Huber

The Second Narrows Water Supply Tunnel will ensure system reliability in the event of a major earthquake and increase the Greater Vancouver Water District’s capacity to meet the long-term needs of a growing population. The tunnel will be 1,100 m long and 6.5 m in diameter (3,610 LF × 21.3 ft), constructed under Burrard Inlet between North Vancouver and Burnaby in British Columbia. It will be excavated through mixed ground conditions under high groundwater pressure, using a slurry TBM, between a 72 m (236 ft) deep shaft in soil and a 110 m (361 ft) deep shaft in rock. The completed tunnel will accommodate three large-diameter water mains that will be installed as part of the tunnel contract. This paper presents the project from design and contractor procurement, into the current construction stage describing some of the unique challenges encountered during the project development to date.

Jun 13, 2021

During the second quarter of 1983, the economic recovery showed further strength abetted by the performances of consumer spending, residential investment, and the slowing down of inventory reductions. Real gross national product (GNP) increased from the first to second quarter at an annual rate of 9.2 percent, compared with 2.6 percent from the fourth quarter of 1982 to the first quarter of 1983 and an overall 1.9- percent decrease in 1982. Total industrial production increased 1.8 percent in July, and productivity as measured by output per man-hour increased 1.3 percent between 1982 and the first quarter of 1983 (table 1). Especially noteworthy .was the performance of real consumer durable goods expenditures, which increased at an annual rate of 32.0 percent from the first to the second quarter (about half of the increase consisted of automobile sales) and real consumer spending, which increased between these periods at an annual rate of 9.7 percent. Residential investment increased at an annual rate of 76.1 percent between the first and the second quarters,1/ and private housing starts by about 38,000 units in both March and May.

Jan 1, 1983

By Keith Harris

The U.S. economy grew moderately in the second quarter of 1985 with real gross national product (GNP) recording a seasonally adjusted annual growth rate of 2.0%, up from the 0.3% growth of the previous quarter but significantly below the 7.1% growth of the second quarter of 1984 (table I). Personal consumption expenditures grew 5.3%, with purchases of durable and nondurable goods rising 7.4% and 5.1 %, respectively; nonresidential and residential fixed investment--the two components of fixed investment--grew 11.8% and 5.6%, respectively; and Government spending rose 3.6%. However, these gains were tempered by a major decline in the growth of business inventories (from $19.1 billion to $8.3 billion) and the continuing growth of the net export deficit (from -$28.4 billion to -$33.3 billion). While the high value of the dollar compared with other currencies has been a major cause of the expanding trade deficit, its value, as measured by the monthly Federal Reserve Board (FRB) multilateral trade-weighted dollar index, has declined over 17 index points from its February peak to 140.94 in July. The July dollar value was still high historically. The rate of inflation, as measured by the fixed-weighted price index for GNP, remained low at 4.1%, down from 4.3% in the first quarter, and employment continued to grow at a rate sufficient to keep up with new additions to the labor force. Civilian unemployment has remained steady at 7.3% over the past 6 months. Industry Statistics The FRB recently revised its index of industrial production and changed the base year from 1967 to 1977,1/ The revised total index shows a 0.1 % higher annual growth rate between 1967 and 1984 than the previous index. In the slower growing or declining industries, such as metal mining, primary metals, and fabricated metals, however, downward revisions in the indices and rates of growth have occurred. The previous and revised annual rates of growth between 1977 and 1984 for several of the sectors of interest to the mining industry are shown in table 2.

Jan 1, 1985

By Keith L. Harris

The economy continued to expand during the second quarter, but at a much slower pace than expected. According to latest estimates released by the Department of Commerce, the real gross national product (GNP) grew at a seasonally adjusted annual rate of 0.6% compared with a revised growth rate of 3.8%2 for the first quarter of 1986 and 2.3% for the second quarter of 1985 (table 1). The strong growth of personal consumption expenditures, Government purchases of goods and services, and residential investment was moderated by the growing trade deficit, declining investment in business inventories, and the continuing lag in nonresidential investment. Much of the decreased capital spending for the nonresidential sector for the first and second quarters resulted from reduced petroleum exploration and drilling in response to low petroleum prices. Capital spending and trade continue to be areas of weakness for the economy. Real new plant and equipment expenditures for 1986 are estimated to be 1.3% below those of 1985 (table 2). Real net exports of goods and services decreased $24.6 billion, as exports declined 5.6% and imports grew 16.5%. About two-thirds of the $19.3 billion increase in imports was in petroleum and petroleum products, which, because of their lower prices, should help the economy in the future. In response to the sluggishness of the economy, the Federal Reserve System reduced its discount rate from 6.5% to 6% in early July. Inflation, as measured by the fixed-weighted GNP price index, declined from a revised 2.5% in the first quarter to 1.9% in the second quarter because of significantly lower energy prices.

Jan 1, 1986

The Henderson Mine is a large underground panel cave molybdenum mine located near Empire, Colorado. Secondary blasting requirements have varied from one shot per 27 tonnes (30 tons) to one shot per 900 tonnes (1,000 tons) of ore. Geology, structure, cave abutment loads, and location of the ore in the column all affect the amount of oversize boulders (greater than three-fourths cubic meter or one cubic yard) that require secondary breakage. The primary method of breakage utilizes cone charges (plastic funnels filled with an emulsion) placed on accessible oversize rocks. Hangups in the brow area of the drawpoint are drilled with a self contained diesel hydraulic drill and loaded with gelatin dynamite. High hang-ups located beyond the reach of the secondary drill are "bombed" in with strategically placed emulsion bundles or ANFO bags loaded in place though a sectioned aluminum pipe.

Jan 1, 1994

By AIME AIME

LAST month we published (p. 440) the first half of the L discussion by O. E. Kiessling of the paper on copper by Mr. Vogelstein that appeared in the same-issue, but lack of space made it necessary to hold the remainder for this number. Mr. Kiessling now asks to have the last " sentence of the part already published revised to the form given in the first sentence which follows, and also asks that the two parts, though necessarily published separately, be read together, as a mistaken impression may be derived from the first Dart unless it is read in connection with the statement of the Bureau of Mines policy respecting secondary metals which is here published. .

Jan 1, 1931

By J. W. Furness

THE utilization and collection of waste materials have gone on for centuries, and have become a habit of the human race. The degree to which the salvaging of waste plays a part in a nation's industry depends largely on the availability of raw materials within the political confines of the country itself, or its ability to purchase such material from outside. When purchase is necessary careful economical use of the raw materials results. These economies, as a rule, are represented by lengthening the life of the given article as well as much re-use. For many years the steel industry of Italy and Japan has depended largely on importations of scrap, yet both are important. As long as the expenses incident to the conditioning and marketing of materials manufactured from salvaged waste are less than the cost of virgin material, the salvaged material will be re-fabricated and in many cases will compete with primary materials. One of the most obvious methods for furthering the efficient utilization of our limited mineral resources is the salvaging of mineral wastes. This method, moreover, in¬creases the available supply of a given material. If the price depends on a balance of supply and demand and the recovery from wastes enlarges. the supply, there can be no question but that it influences the sales price of primary.

Jan 1, 1931

By LUDWIG VOGELSTEIN

WE are indebted to Mr. Barbour for his valuable contribution to the literature on copper statistics; it is to my knowledge the only intelligent attempt to throw light on a much misunderstood subject. If I should appear to criticize some of his findings and arguments, I do not want to leave the impression that I do not appreciate the article; I think Mr. Barbour has been painstaking, and even where one might not agree with his conclusions, he has opened our eyes and stimulated us in our search for the truth.' It might be well to define the term "secondary copper" and to make a few remarks about the origin and the reasons for the variation of the supply. Mr. Barbour has touched on these points, but not sufficiently to make them clear to those not in the business.

Jan 1, 1931

By M Magee, D Way

The Mount Isa Copper Concentrator was commissioned in 1973. In 1991, two autogenous grinding mills were installed replacing two stages of crushing, one stage of rod milling and one stage of primary ball milling. The cyclone overflow P80 of the secondary ball milling circuit was reduced to 120 microns due to the increased operating power in the AG/SAG mill circuits. A study of the mineralogy of the milled products and the plant flotation recovery indicated that the installed secondary grinding energy was more than what was required for adequate liberation of the chalcopyrite and half of it had potential for more effective use elsewhere. Halving the secondary grinding energy meant only one of the two secondary ball mills would be needed. The predicted saving in operating cost of one less ball mill was more than two million dollars per annum. This paper summarises various plant trials that were conducted in the Copper Concentrator to decommission one of the two secondary ball mills.

Jan 1, 2004

By Ullrich Kammer

Metaleurop's secondary lead smelter at Oker produces 60.000 mtpy of lead and lead alloys. More than 8O2 % of the raw materials are spent lead acid batteries. The other 2O2 % consists of drosses, ashes, scrap lead, type metal, sludges, etc., and is processed in a short rotary furnace (SRF). Also, industrial batteries not fitting in the battery scrap treatment plant are processed in the SRF provided with gas/oxygen-burner. Till Sept. 199O2, a soda ash slag was used in the SRF-smelting similar to the rotary kiln-smelting at Oker. Facing the future problems of handling and dumping soda ash slag, Harz-Metall GmbH has improved SRF-smelting with a non¬hazardous slag, insoluble in water, easy to dump and cheap in fluxes. This slag is based on a CaO-SiO2 -Fe0-type. Aim of the slag smelting is to be within the temperature range of 2 1100° and 1200°C - the eutectic line of Olivine in the corresponding ternary system. Since Sept. 199O2 Harz-Metall GmbH has operated the SRF with an "Olivine-type slag" with good results. Although the material is a "leady" slag. HMG is optimistic of using it as a filling material for construction purposes.

Jan 1, 1993

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

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

Jan 1, 1958

By W. T. Cardwell

Physically absurd, triple-valued saturations appear in the straight-forward solution of the Buckley-Leverell equations for the displacement of oil by water or gas. From an engineering viewpoint, the triple value causen no difficulty. It is well known how it may be compensnted in order to obtain physically meaningful, numerical results. From a scientific viewpoint, the question still arises: What did the triple value mean? This paper explains how and why the triple value arose in non-capillary Buckley-Leverett theory. The discussion should serve as a background for the understanding and use of the modern method of characteristics in di.~pkzc(>ment theory. INTRODUCTION Displacement theory was introduced to petroleum technologists in 1 941 by Buckley and Leverett.' Their first equation, which they wrote down without derivation, was equivalent to the following: ?s/?t= - qr/F ?f/?r (1) where S = saturation of displacing fluid (a fraction) r = time q,. = total volumetric velocity of both the displacing phase and the phase being displaced (usually, but not necessarily, assumed to be a constant) F = porosity (a fraction) f - fraction of the total volumetric velocity that is the volumetric velocity of the displacing phase (assumed to he a function of S only x = distance. Buckley and Leverett called Eq. I "a material balance equation". Actually it is derived from both a continuity equation, or material balance equation, and Darcy's law. Buckley and Leverett jumped from their Eq. 1 immediately to their Eq. 2, which was the equivalent of the following: (?x/?t)x = qr/F df/ds = qr/F f'(s) (2) The jump from Eq. 1 to Eq. 2 involves apparently simple mathematics, but in that apparently simple mathematics lies a subtle point that is part of the key to the meaning of the triple value to be discussed here. Eq. 2 may be readily interpreted as saying that a point of constant saturation (on a saturation-vs-dis-tance curve) moves with a constant velocity that is proportional to the total volumetric rate, inversely proportional to the porosity, and is otherwise a function of the saturation itself. So that if one knows the derivativc with respect to saturation of the fractional flow function, f(S), for each saturation, one knows the velocity of each point of the moving saturation-vs-distance curve. The function f(S) can he determined experimentally and its derivative f'(S) can be calculated. It turns out that the experimentally derived f'(S) is not a monotonic function of S, but instead it has a definite maximum and declines away at both high and low saturations. This in turn means, in accordance with Eq. 2, that on a moving saturation-vs-distance curve, ccrtain intermediate saturations will travel faster than the saturations either higher or lower. The result is indicated in Fig. 1. In Fig. 1, the abscissas represent distances along a column of porous medium, which column is assumed to be uniform in cross-section perpendicular to the ab-scissal direction. The ordinates represent fractions of pore space occupied by the displacing phase, which may be either water or gas. It is assumed that the saturation is uniform over all planes perpendicular to the abscissal direction. The saturation is a function only of time and one space variable. Time variation of the saturation is represented by change in shape of a curve such as curve AGJ, which represents the initial saturation-vs-

By F. F. Craig

A series of both welter and gas puttern Hoods was made in the laboratory to study the oil recovery performances of .such operations. These tests were contlrrctt~tl on consolidated sandstone models, using oil. water, and gas. The model floods were scaled to reprotlrrce field performance under gas arid water five-spot injection. X-ray shadowgraphs permitted observation of the gross fluid movement within the models. A method was developed for applying the mobility ratio concept to water flooding and dispersed gas drives in a five-spot well pattern. The area1 sweep efficiency. tit breakthrough for dispersed gas drives is much higher than previorsly expected, Iying in the range of 50 to 100 per cent. A method is presented for predictitig the water-oil ratio performance of five-spot pattern water flood., in uniform sands. This method is verified experimentally for the cotidition of 120 free gas initially present and for va1ues of gas saturation normally encountered in fields following depletion operations. Prodrrction perfort?zut~cc for pattern gas injection is also predictable by this method. INTRODUCTION In field pattern hater flooding or dispersed gas injection operations, the displacing fluid (water or gas) is injected and oil is produced through wells which are, area-wise. small openings in a large container or rescrwir of oil. As a result, not all of the area between the injection and producing wells is necessarily contacted by the injected fluid by the time it first reaches the producing wells. The question arises in predicting oil recovery. as to what traction of the pattern area involved is contacted by either the injected water or gas due to the relative position of the injection and producing wells. This is the areal sweep efficiency. It is also desirable to know how well arrangement affects the oil production performance. A solution to this problem has been attempted by many means. Mathematical analysis¹,²,³, and electrolytic models', have been used to establish a breakthrough areal sweep efficiency of 73 per cent for a five-spot pattern in which the ability of the oil to flow ahead of the invading fluid is equal to the ability of the invading fluid to flow in the invaded zone. All of these studies simulate only one of many field conditions. Improved studies of areal sweep efficiency at breakthrough were made using the fluid mapper5, poten-tiometric models" and X-ray shadowgraph techniques. In these it was possible to study the effect of variations in mobility of the injected and displaced fluids, and so to cover a wide range of field conditions. The limitation of these studies is that, with these models, it was not possible to simulate the saturation gradient behind the flood front when oil is displaced by gas or water. In the most recent published technical paperS on the subject, a method was presented for predicting the oil recovery performance after breakthrough in a five-spot pattern water flood. The method involved correlations which were obtained experimentally using miscible fluids. One limitation of this work was that a saturation gradient, usually found in immiscible fluid displacement, was absent in these studies. Also the laboratory studies did not cover the situation of a water flood following oil recovery by depletion drive. as is the situation in most field water floods. The tests reported in this paper were made using oil. water, and gas, thus eliminating the simplifications of former studies. Also reported are laboratory studies on

Jan 1, 1956

By R. L. Slobod, B. H. Caudle, R. A. Erickson

In predicting the sweepout pattern efficiency to be expected in a .secondary recovery operation, the reservoir engineer is often confronted with a situation in which part of the producing formation lies between the last row of wells and the reservoir boundary. To date, there has been little or no reference in the literalure to sweepout pattern efficiencies in such areas. This paper presents the results, of a laboratory study in which the area contacted around the edge of a reservoir injected in a five-spot pattern was determined. The X-ray shadowgraph technique was used in this study. It was found that surprisingly large portions of the reservoir area lying outside the well network are conracted by the injected fluid before, abndonment conditions are reached. In fact, it can he said that at least 00 per cent of the circa lying outside the last row of wells and within a distance of one well spacing of these wells will ultimately be contacted by the injected fluid in the case of most water floods. Frorn the results given in this paper, it is possible to estitilate not only the extent to which encroachment into these areas will occur but also the injection volume necessary to produce the wells to abnndonment conditions. INTRODUCTION While considerable work has hern done in recent years1.?.'.''.' evaluating the fraction of the reservoir contacted as a function of the mobility ratio for several different well patterns, there has been practically no mention of what happens to the oil in that portion of the reservoir which lies between the last row of wells and the reservoir boundary. In fact, the assumption is generally made that little or no recovery can he expected From this area. This assumption is incorrect and in many cases will lead to seriouq errors in calculating reservoir performance. The fraction of the area outside the regular well pattern contacted in a five-spot injection system has been studied using the X-ray technique5 to follow the progress of actual floods in porous models of reservoir elemental areas. In this paper, the method for carrying out such a study and the results showing the fraction of the area in question swept as a function of mobility ratio and the produced fluids are presented. THE REPRESENTATION OF A RESERVOIR BY ITS ELEMENTAL AREAS She portion of the reservoir lying between the edge of a reservoir and a line connecting the outermost wells may be roughly designated as the part of the reservoir outside the regular well pattern. An example of such an area is shown as the shaded portion in Fig. 1. The case of a square reservoir utilizing a five-spot injection program and with one full well spacing between the outermost wells and the boundary (Fig. 1) was selected as the single example which would provide the most insight into

Jan 1, 1956

By C. W. Arnold, H. L. Stone, D. L. Luffel

The purpose of this research is to determine (I) the efficiency of small banks of enriched gar driven by methane in displacing oil from a porous medium and (2) the effects of variation in bank size and composition of that efficiency. Most of the experiments were conducted in a sand-packed tube 20-ft long and 1/2-in. in diameter. The hydrocarbon system generally used was methane, butane and decane at 2,500 psia and 160°F. The results of these experiments indicate that, in the regions contacted by the gas, a small bank of an oil-miscible gas driven by methane can displace all of the oil in a piston-like manner. If the enriched gas is of such composition as to remain immiscible with the oil, displacement of oil is less efficient than for the miscible case, and the gas bank travels through the sand with a velocity less than that of the driving gas. These data along with theories discussed imply that smaller banks and less total gas are required when the enriched gas and oil are miscible. INTRODUCTION Widespread application of enriched-gas drive to the recovery of oil rests upon a key factor — the use of limited quantities, or "banks", of enriched gas. At the present time, the value of liquefied petroleum gas or other enriching agents discourages their use in a continuous injection technique, or even in a large bank, except in a few isolated reservoirs. If small banks of enriched gas driven by methane were as effective in displacing oil as is continuous injection, the enriched-gas drive process might be applied to a larger number of reservoirs. Previous research on the mechanics of the enriched-gas drive process reported by Stone and Crurnpl and by Kehn, Pyndus and Gaskell has utilized continuous injection of enriched gas. This work has shown that two types of displacements occur. With gases containing sufficient intermediates. the oil is displaced misciblv and complete recovery is obtained from the regions swept. When gases are used which contain insufficient intermediate hydrocarbon for miscible displacement, oil is displaced immiscibly. In the latter type, selective solution of the intermediate hydrocarbons causes a swelling and reduction in viscosity of the oil and leads to an increased recovery over that obtained by dry-gas (methane) drive. The size of the enriched-gas bank necessary for efficient displacement of oil is determined by those factors which cause deterioration of the bank. A differentiation may be made between those factors which operate on a microscopic scale and those which act on a macroscopic scale. On the smaller scale, the enriched gas mixes in the direction of flow by diffusion and convection with the fluids immediately preceding and following it. On the larger scale, the gas may by-pass the oil by flowing through permeable streaks, by overriding the oil because of density difference, or by fingering because of unfavorable viscosity ratios. In such cases, the enriching material tends to mix with the oil both laterally and in the direction of flow. The increase in effective area available for diffusion and dispersion of the enriching components leads to a faster degradation of the bank and a need for a larger bank than is necessary for those cases in which no by-passing occurs. The effects of such macroscopic factors in the deterioration of enriched-gas banks have been reported in a separate paper by Blackwell, Terry and Rayne. The present study was confined to the factors which operate on the smaller scale, in particular to the behavior of banks of enriched gas in sands uniformly swept by the gas. Experiments were designed to answer the following questions. 1. Can small banks of enriched gas driven by methane be used to secure oil recoveries comparable to those obtained by continuous injection of enriched gas? 2. What is the optimum bank size (the minimum bank size necessary to obtain a recovery comparable to that obtained by continuous injection of the same enriched gas)? 3. How many total pore volumes of gas must be injected to obtain the maximum recovery when the optimum bank size is used? 4. What is the effect of varying the number of enriching components in a gas bank? This report describes the experimental investigations and discusses the results in terms of their significance to reservoir behavior.

By B. H. Caudle, L. G. Loncarie

Pilot flooding is one method of evaluating a proposed secondary recovery project. However, the amount and rate of oil recovery from an unconfined pilot area is not usually the same as from an equal area in a large-scale flood. This is true because fluids are free to move across the boundary of a pilot area. The result is that some of the oil produced in a pilot flood may come from outside the designated area and some of the displaced oil may leave the pilot area completely. This paper presents the results of fluid-flow model studies on a five-spot pilot flood. Mobility ratios between 0.1 and to have been studied. The effects of changing the ratio of injection to withdrawal rates are shown. INTRODUCTION A small-scale field test, or pilot flood, is an accepted method for evaluating the applicability of a secondary recovery operation and the economic potential of that operation for a reservoir. The pilot flood is carried on in the same manner, with the same injected materials and at the same pressures as would be used were the secondary recovery program to be expanded to a larger area of the field or to the entire field. If the reservoir fluid and reservoir rock properties are the same in the pilot area as in the rest of the reservoir, the production, injection and other pertinent data associated with the pilot test will represent and measure what should be expected from a full-scale development of this secondary recovery program. There is one major exception which, unfortunately, has been overlooked by many and which has been awaiting quantitative definition. This is the sweep-out pattern the reservoir engineer must use in interpreting the pilot flood and in extrapolating it to the full-scale development. The sweep-out area in a pilot flood is not contained within the five-spot pattern. Further, there is oil flow into and/or out of the five-spot during a pilot test. Thus, production data from pilots must be interpreted with this different flow system and sweep pattern in mind. We have studied the sweep-out pattern that will exist for some conditions of operation of pilot flooding a single five-spot (four injection wells and one producer). In our study, reported here, we have assumed that the reservoir is relatively homogeneous and of uniform thickness in the pilot area. The results of this study en-able the engineer to predict the recovery, injection requirements and project life for a full-scale operation from the pilot flooding data. The effects of various injection to withdrawal-rate ratios and mobility ratios in a relatively gas-free (10 per cent pore volume or less) system are shown. Experimental Method The technique used in this study was the same as that for the X-ray shadowgraph studies of sweep-out pattern efficiencies.' The model of the pilot and surrounding area was made from a uniform packing of round-grain sand consolidated with an epoxy resin. Fig. 1 shows the relation between the model and the pilot area. Only one-eighth of the pilot area had to be represented in the model because we assumed equal injec-

By R. J. Blackwell, J. R. Rayne, J. R. Henderson, W. M. Terry, D. C. Lindley

This paper presents the results of a laboratory investigation of the efficiency of water-solvent mixtures in recovery of oil. These mixtures may have the high displacement efficiencies characteristic of solvent floods and the high sweep efficiencies characteristic of water floods. Thus, the water-solvent process may increase the number of reservoirs in which a miscible-type displacement can be used profitably. The experiments on the use of water-solvent mixtures for recovery of oil were conducted to find the general applicability of the process. These studies demonstrated that, in flowing through sands, water and solvent segregated into a solvent layer on the top and a water layer on the bottom rather than flowing through the sands as a uniform mixture. Calculations based on the simultaneous flow of the water and solvent in layers were used to predict the effective mobility of the mixtures and the optimum operation of the process in steeply dipping, homogeneous reservoirs. As most reservoirs are not suited for the operation of the process under ideal conditions, experimental studies were conducted with sand-packed models scaled to represent more realistic reservoirs. These studies included the effects on recovery of oil of rate of injection, viscosity of oil, variations of permeablity within a formation and variations in water-solvent ratio. For the range of con-ditions studied, higher recoveries of oil were obtained with water-solvent mixtures than with water or practical volumes of solvent alone. INTRODUCTION A group of intriguing—because of their great possibilities—new oil recovery methods at the disposal of the petroleum engineer are the miscible displacement processes. These processes (high-pressure gas drive, en-riched-gas drive and LPG bank driven by methane) displace all of the oil from the portions of the reservoir swept by the injection fluid. The key question confronting the engineer applying one of these techniques to a particular reservoir is, "What fraction of the reservoir can be swept by injection of a practical volume of solvent?"." Intensive laboratory studies have been made during the past several years in seeking answers to the question of the sweep efficiencies which can be expected in solvent floods.' - These studies provided the answer that low-viscosity, low-density solvents channel and by-pass oil in sands with no dip. In horizontal sands, solvent flooding becomes less efficient as the viscosity of the oil increases, the recovery of oil at solvent breakthrough decreases and larger volumes of solvent are required to achieve a given recovery. More efficient displacement of oil by solvent is observed under certain conditions in sands with dip.".' If the permeability and dip of the sand are sufficiently high, gravity segregation of low-density solvent injected updip can prevent channeling. At rates of depletion below a critical rate,"." no channeling occurs. Unfortunately, in many reservoirs, the critical rate is so low that production of oil at rates below this rate is not economically attractive. And at rates over four times the critical rate, channeling is almost as severe in sands with dip as in horizontal sands. These findings pointed out that, for solvent floods to be generally applicable in recovering oil from all types of reservoirs, new methods of improving their sweep efficiencies are needed. Simultaneous injection of water with the miscible fluid was suggested by Caudle and Dyes7 as a method for improving sweep efficiencies. They theorize that water flowing with the solvent would decrease the effective mobility of the solvent and cause it to contact more oil sand. If, indeed, the water and solvent flowed as a uniform mixture, the process should have the advantages of the high displacement efficiencies characteristic of miscible floods and the high sweep efficiencies of water floods. Thus, the method (at least in theory) would greatly increase the number of reservoirs in which miscible-type displacements would be feasible. A research program was conducted to probe the technical feasibility of the water-solvent process. It