Search Documents

By P. Mageau-Béland, È. -M. Proulx, M. -C. Patoine

Over the past six decades, Hatch has helped several industrial processing facilities in debottlenecking their operations, allowing them to produce more with the same equipment, to reduce and even eliminate expenditure investments, or to increase the return value of their expansion projects. These successes result from the application of a proven analysis approach that can be applied to any industrial facility. A five-stage methodology is discussed and demonstrated with a specific example of a nickel concentrator optimization study. The first stage is an observation stage which allows to meet key actors at the plant, develop mutual understanding of problems, understand operating philosophy, and efficiently collect large amounts of operating and maintenance data. Following is a data compilation stage during which data is analyzed over a representative time period to identify parameters that significantly affect the production, to evaluate data reliability and to define the focus of the upcoming analysis. Quantitative metrics are defined using statistical tools such as pivot tables, histograms, percentiles and MDR (Maximum demonstrated production rate during 3 to 7 consecutive days). Then, the production bottlenecks are identified by a thorough analysis of the facility’s control philosophy, process throughput, equipment capacity, availability, downtime causes, and interaction between components of each circuit. Understanding the maintenance practices and frequency of planned shutdown is also critical to this analysis. Opportunities for improvement are identified in collaboration with the operation, maintenance and engineering teams, ranked as a function of their impact on the bottlenecks and prioritized. The final and most critical stage involves engagement with the plant stakeholders, namely the operators and the maintenance staff members, to present the results of the analysis and obtain their participation in the implementation of the solutions. The benefits of the case study translated into: 10% production capacity gain, improved maintenance planning and avoidance of several million dollars of investment.

Jan 1, 2016

"Chairman: Professor C. O. Frush Rapporteur: Mr A. C. SchmidtPapers:Planning of underground copper mining by J. K. Williams, L. Smith and M. P. WellsLinear programming models of copper leach and solvent extraction plants by W. P. C. Duyvesteyn, J. R. Cutland and J. R. L. RusselPlanning and forecasting stope production by D. W. Kappes and P. F. X. Moussett-JonesA proposed method of dynamic ore reserve assessment for a caving system of mining by A. F. Spinks and S. P. NicholsThe initial point made by Mr L. Smith in presenting his paper was that the planning aid outlined in the preprint was neither sophisticated nor complicated. However, it was a system which worked and was a very real way in which the computer was being used to assist miners, in particular, who at the present moment have very little experience in the use of computers.Mr Smith apologised for the fact that the editors of the paper had eliminated all reference to 'time'. This was very unfortunate because the essence of the system was that it saved time, which was one of the main points made in the paper.The operations being planned by this means cover the Rokana orebody which is a stratiform body mined by open stoping.The planner has to decide how much ore to mine from each base in each time period of his plan, that is, from three to four months hence to a year ahead. Normal procedure is for the planner to have discussions with the underground manager on some of the information he requires. The sort of information needed is (i) tramming limits for each face, (ii) extraction and grade factors, (iii) the present status of mining operations so that the planner knows where to start the new plan, and, (iv) details of any other mining consider¬ations that need to be incorporated in the plan.The objective at this point is to produce an initial plan for discussion purposes. All the information required comes from one sheet per stops and shows, inter alia, the stope identifier, the length along strike, the ore tonnage, the copper and cobalt grades and, finally, the extraction and grade factors.The program itself, among many other things, checks the data for validity, compatibility, completeness and credibility and prints error messages accordingly. Output from this system is in the form of a summary of production from each face, as well as giving the figures of ore reserve depletion. This summary indicates which slopes are to be mined in this time period, the original survey grades (copper and cobalt), the estimated production grades, how much is to be mined from each stope and the copper and cobalt production accordingly."

Jan 1, 2014

By James A. Decker

Hanna has developed three production planning systems to help the mine engineers determine the sequence of mining from available ore supples. The Production Scheduling System was developed to give the staff a tool for long range planning by sequencing the removal of blocks from an open pit property. It was needed to determine the year of removal for each block so that the optimum present-valued pit limits could be established by our pit design program. Since in this type of planning short term grade control is less of a factor than it is in daily or monthly planning, the system gives the program greater control and more latitude in block selection. It permits only full blocks to be taken.

Jan 1, 1984

By D. D. Lei, H. Sevim

Long-term production planning in open-pit mines requires the determination of cut-off grade, mining and milling production rates, a mining sequence, mine life and ultimate pit limits. Since these variables interact in a circular fashion, a genuine solution to the problem necessitates their simultaneous determination. A methodology with this capability is described. This comprises a group of heuristic algorithms and an unconventional dynamic programming algorithm. The method is demonstrated using data for a gold deposit in the western USA

Jul 1, 1996

By Donald L. Katz

FIFTEEN years ago Dr. Manning published a paper in "Petroleum Development and Technology" on '.Fundamental Research Relating to Petroleum." He tabulated for several industries "the proportion of their present success which can be fairly attributed to the application of information and processes obtained by scientific research," giving petroleum a score of 7 compared to 100 for the electrical and coal-tar industries. Today the petroleum industry as a whole and the producing groups both have raised their score beyond the majority and are climbing toward the top. Present practices are being developed both by fundamental research and the application of the research method in field operation.

Jan 1, 1941

By Allen D. Gorrison

EFFORT to develop fundamental quantitative information and improved technique in the production of petroleum has long been faced with difficulties of a particularly evasive nature, owing to a combination of circumstances. In the first place, petroleum comes from far below the ground, out of sight, so that information derived from indirect observations must be extrapolated to reservoir conditions. If one does not force the analogy too far, the petroleum engineer may be said to be somewhat in the position of the astronomer. The actual phenomena which he wishes to observe in detail appear at a place so deep in the earth?s crust that personal observations and direct experiments are impossible.

Jan 1, 1942

By P. E. Fitzgerald

SOME readjustments in the research programs of most of the oil companics and petroleum engineering schools have been made necessary by the war. The most obvious change has been the conversion from production engineering to problems more closely related to the nation's was effort. Many laboratories now devote much of their time to studies on products to be derived from petroleum. such as synthetic rubber, toluene, and substitutes for critical materials. Others are working to improve motor fuels including high-octane gasoline and the special lubricants required by the wide geographical range of the present conflict. Of great importance is the experience of the geophysicists, who are now devoting a great percentage of their time on instrument research and production or special apparatus required by the armed forces.

Jan 1, 1943

By C. W. Baxter

Eveleth taconite company was formed in 1963 to mine and beneficiate taconite on the Mesabi range of Minnesota. Its thunderbird mine is located midway between the cities of Eveleth and Virginia, Minnesota, and the crude ore is transported by rail to its Fairlane plant approximately ten miles to the south of the mine, "figure 1". As is implied by the names of the mine and plant, the ford motor company is the principal owner of Eveleth taconite company, with an 85% interest, Oglebay Norton company holds the remaining 15% interest and also is the managing agent.

Jan 1, 1972

By M Spleit

A long-term production schedule for the LabMag iron ore deposit in northern Quebec, Canada is derived using stochastic integer programming. The optimisation formulation maximises the schedule’s net present value while simultaneously managing the risk of deviations in production tonnages and qualities by considering stochastic simulations of the orebody instead of a single deterministic model. The formulation also smooths and minimises haul truck requirements and ensures that as mining progresses, space is created within the mined out pit for the return of waste material.CITATION:Spleit, M, 2014. Production schedule optimisation – meeting targets by hedging against geological risk while addressing environmental and equipment concerns, in Proceedings Orebody Modelling and Strategic Mine Planning Symposium 2014, pp 397–404 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Nov 24, 2014

By M. Kuchta, A. Newman

LKAB’s Kiruna Mine, located in northern Sweden, produces about 24 Mton of iron ore yearly using sublevel caving. For efficiency, the mine must deliver planned quantities of three ore products. We use mixed integer programming to schedule Kiruna’s operations, specifically, which production blocks should be mined, and when they should be mined to minimize deviations from planned production quantities while adhering to mining restrictions (e.g., sequencing). A new production block database and a well-formulated mixed integer programming model dramatically reduce the size of the original problem; in less than ten minutes we generate five-year schedules with time fidelity of months.

Jan 1, 2002

By T GRAHAM-TAYLOR

A trend towards mining more complex and economically marginal orebodies, often combined with increasingly tight product specifications, increases the number of variables which need to be considered in production scheduling. Linear programming incorporating integer constraints can be used to create production schedules which are practical as well as mathematically optimum. Factors which can be taken into account include variability in product stockpiles, targets for head or product grades, production of different material types, production from different areas, production targets for loaders, and dependencies between mining blocks. For each schedule an appropriate objective for optimisation must be chosen. Examples of objectives are: to minimise or maximise grade variation; to minimise waste movement; and to minimise movement by front end loaders. A staged approach to the implementation of computer systems based on linear programming optimisation is Proposed.

Jan 1, 1992

By Lyon F. Terry

SPEED-UP of production of crude oil and its products, accompanied by rising prices and the organization of the industry on a war basis, featured the economic aspects of petroleum in 1941. Early in the year American tankers were urgently requested by the British to replace losses and offset slower transit under convoy. Some forty American owned tankers were first assigned to Britain and an additional forty-odd ships were transferred during the summer. This threatened to reduce our transportation facilities from Gulf ports to the Eastern seaboard below requirements, especially should sinkings have increased and the British needed more ships. Co-ordinator Ickes began preparations for restricting gasoline consumption in the Eastern district and plans were formulated for the construction of an emergency trunk pipe line from Texas to New York and Philadelphia refining centers, and for a gigantic tanker-building program. Sales of new oil-heating units in the East practically ceased. The companies were urged to transport as much oil as possible by tank car and by October 140,000 barrels of oil per day were reaching the East Coast district by rail, sufficient to meet 8 per cent of total requirements. But British sinkings from July through October decreased 63 per cent compared with the four months ended June 30, Britain began the return of tankers, the construction plans were dropped, and the threatened shortage was thus averted.

Jan 1, 1942

By AIME AIME

THE fall meeting of the Petroleum Division, held at Tulsa, Okla., on Oct. 11 to 14, devoted two days to technical sessions and two to field excursions. A representative attendance of 250 to 300 engineers and operating officials engaged in oil production marked the sessions and paid close attention throughout to a notable series of highly technical papers. Restricted as the program was to consideration of production problems only, the time proved too short for discussion or even the presentation of the 20 papers submitted. Much informing and pertinent discussion none the less took place, and the sentiment was general to the effect that the meeting was one of the best ever held by the Division. The meetings were held in the auditorium of the Municipal building and local arrangements were well looked after by the Mid-Continent Section of the Institute and the Oil and Gas Association. The combination of the three organizations served to bring together engineers and operators most usefully. F. Julius Fohs, chairman of the Division, presided at the sessions on Monday. Tuesday morning John M. Lovejoy of the local section took over this responsibility in the absence of T. E. Swigart, who was detained in California. The afternoon session was presided over by A. G. Heggam of the Oil and Gas Association and the Tuesday evening banquet by W. C. Franklin of the same organization.

Jan 1, 1926

By Howard E. McKinney

A high-pressure wellhead lubricator has been developed to facilitate telemetering electrical measurements from subsurface reservoirs to the surface with the well under normal flowing conditions. The field unit in use at the present time employs a 5/16-in diameter, single-conductor armored cable and is to be used in wells with surface pressures up to 5,000 psi. It has a pressure sealing element that maintains a continuous and absolute seal around the cable as it enters or leaves the tubing at speeds up to 175 ft per minute. A cable injecting device is used which obviates sinker bars or weight sections on the instrument. INTRODUCTION The development of instruments for securing physical measurements in pressure wells has been confined in the past to those instrument which would operate on a wire line and record the necessary measurements in the downhole instrument itself, or those which would operate on an electric cable in a well with the surface pressure killed by the use of weighted fluids or by the use of tubing plugs. In one case the accuracy of the measurements is questionable, and in the other the range of application is very limited. Considerable work has been done by other interests to provide effective means to accomplish the necessary telemetering. The U. S. Bureau of Mines developed a wellhead lubricator prior to 1947 which permitted a 5/16-in diameter single-conductor cable to be lowered into a well with a surface pressure of approximately 500 psi. This unit was used successfully in a number of tests in Oklahoma. A company which tests wells has recently offered on a service basis a wellhead lubricator with a 3/16-in diameter cable that reportedly operates up to 2,500 psi surface pressures. Each of the above operates on a staging principle to effect a seal. In both of these units some leakage of well fluid occurs when the cable is moved. SPECIFICATIONS FOR DESIGN In order to provide the maximum flexibility in its use, the following parameters were established for the design of our high-pressure wellhead lubricator: L An absolute well fluid seal up to 5,000 psi surface pressures. 2. Means for raising and lowering the instrument at speeds of at least 100 ft per minute. 3. Provision for the instrument to enter and leave the well without disturbing the well flow or the reservoir conditions. 4. A cable large enough to provide suitable electrical characteristics and suitable mechanical properties. 5. A method of operation that will not damage the well in any way. The present high-pressure wellhead lubricator field unit fulfills the above requirements. DESCRIPTION OF THE UNIT AND ITS OPERATION The cable selected for this use is the standard Amergraph 1H2 as manufactured by the American Steel and Wire Co. This particular cable is normally used by the Halliburton Oil Well Cementing Co. in their conventional well logging operations. Although the cable presents a noncircular cross-section (Fig. l), its diameter is very consistent; lor example, maximum diameter is 0.325 in., minimum diameter 0.314 in. This [actor permits the use of the sealing element, as later described. Early in the development of the high-pressure wellhead lubricator it was apparent that we could not effect a direct gas seal around a moving 1H2 cable. This becomes evident when an examination of the cable cross-section is made. The voids between the strands of the outer armor provide natural channels lor the flow and escape in large quantities of a low-viscosity fluid such as gas. However, it is relatively simple to

Jan 1, 1952

By Henry A. Stiff, and Lawrence E. Davis

The authors previously presented a method for predicting the tendency of oil field waters to deposit calcium sulfate. The present paper gives a similar method for calcium carbonate. Methods for predicting calcium carbonate scaling tendencies in fresh waters have been available for some time, but these could not be used for brines. By experimentally deriving the value of the K term in the Langelier equation, a method has been developed which applies to waters of high salt content. A statistical study is included which shows that the experimentally derived values of K are in good agreement with actual conditions. Several applications of the final equation to production practice are given. INTRODUCTION The authors previously presented a method for predicting the formation of calcium sulfate scale in oil field waters.' Although calcium sulfate deposition is important in production operations, the majority of scale problems involve calcium carbonate. The present communication discusses a method by which the formation of this type of scale in oil field waters can be predicted. Calcium carbonate precipitation is caused by a shift toward carbonate in the carbonate-bicarbonate-carbon dioxide equilibrium. When equilibrium shifts in the other direction, the precipitate goes back into solution. Since there is usually considerable delay between the establishment of an equilibrium and the precipitation or solution of calcium carbonate, unstable conditions exist in which a water will precipitate or dissolve calcium carbonate on standing. Tillmans? who did a major part of the early work on carbonate scaling, pointed out that the condition of equilibrium not only indicates the tendency of a water to scale but also is an indication of its corrosive properties. Previously precipitated calcium carbonate combines with iron to form a dense crust which inhibits corrosion. If the water tends to dissolve carbonate, the scale becomes porous, and electrolytic corrosion takes place. Tillmans1 work was extended by several investigators until in 1934 Langeliers developed an equation setting forth the conditions of carbonate equilibrium. By the use of this equation the pH of a water at equilibrium can be calculated. If the actual pH is higher than the calculated pH, the water has a tendency to form scale. If it is lower. the water has a tendency to be corrosive. Langelier's equation can be expressed in a simple form as follows: SI = pH-pCa-pA lk-K where: SI is the stability index. A positive index indicates scale formation. A negative index indicates corrosion. pH is the pH of the water sample, as actually determined. pCa is the negative logarithm of the calcium concentration. pAlk is the negative logarithm of the total alkalinity. K is a constant, the value of which depends on the total salt concentration and the temperature. This equation has been shown to apply to waters with totaI solid concentrations as high as 4.000 ppm. In fact, control of fresh water treatment by means of this equation has been standard text-book practice for a number of years. Nomographs have been worked out so that the stability index of a fresh water sample can be determined in a matter of minutes. Most oil field waters, however, contain well over 4,000 ppm of salts and for this reason the usual application of Langelier's equation can not be made. By an empirical method we have extended the application of the Langelier equation to waters of high salt concentration. By the use of this equation the tendency of oil field waters to deposit calcium carbonate can be predicted.

Jan 1, 1952

By Henry A. Stiff, Lawrence E. Davis

A graphic method was developed which can be used to predict the tendency of oil field waters to precipitate calcium Sulfate under a variety of conditions. Application of this method is made to the prediction of sulfate scale formation in heater treaters. boilers. oil wells. cooling systems and water injection wells. INTRODUCTION A solution may be said to be in equilibrium with respect to a given salt when the concentration of that silt in equal to its solubility. If for any reason the concentration of the salt is increased or its solubility decreased. the equilibrium is upset and precipitation of the salt takes place. Thus, if the concentration of a salt in a particular solution is known and its solubility can be calculated for a given set of conditions, the tendency of this salt to form a precipitate under these conditions can be predicted. Based on these principles a method has been developed for determining the tendency of oil field waters to form calcium Sulfate precipitates under various conditions. THE SOLUBILITY DIAGRAM Calcium sulfate precipitation most frequently occurs when two waters are mixed: one containing calcium ions. and the other sulfate ions. As one of these waters is added to the other a point may be reached where the concentration of calcium Sulfate is greater than its solubility. thus causing the formation of a precipitate. The solubility diagram is a graphic device for determining the location of such points, thereby making possible a prediction of the tendency of such a system to deposit sulfate scale. A similar diagram can be used to solve many other problems involving the precipitation of calcium sulfate. Fig. 1 shows a solubility diagram representing the mixing of two waters. Water "A" contains a high concentration of odium, magnesium, and sulfate ions together with a small amount of calcium. Water "B" contains a high concentration of calcium ions together with sodium, magnesium and a small amount of sulfate ions. The percentage mixture of these waters is plotted along the abscissa of a system of rectangular coordi-

Jan 1, 1952

By Henry J. Welge

The approximate methods which are now in use for calculating oil displacement from reservoirs by gas-cycling or gravity-drainage at constant gas pressure, or by water flooding, make use of fundamental relationships derived by Leverett' and Buckley and Leverett.2 The mathematical equations needed are derived by applying Darcy's law to the flowing phases, and by material balance considerations. ln general, any treatment of this type gives. for any particular exploitation time considered, a plot of oil saturation against distance in the reser-voir. The oil recovery must then be obtained by integrating in some manner the area under the plot. A useful analytical method has been derived for computing the average saturation, and hence the oil recovery. Use of this method simplifies the calculations because it makes unnecessary any numerical integrations. and even the saturation distribution plots are not needed. A further advantage of the method is that knowledge of the relative permeabilities is required only for a limited and intermediate saturation range. In both the Buckley and Leverett method and the method discussed here, a linear sand section is assumed, and in the case of gas drive the gas pressure is assumed sufficiently constant both with respect to reservoir position and time so that changes in gas density. solubility, or reservoir volume factor are negligiklle. Thus, the exploitation contemplates oil displacement as by an immiscible phase. Examples are given to illustrate how the new method can be used. INTRODUCTION AND THEORETICAL BACKGROUND This paper treats a simplified method for computing oil recovery when the oil is displaced, from the reservoir sand by a fluid which, within the limits of accuracy desired, can be assumed to be incompressible and immiscible with the oil. The method makes use of two basic relations originally developed for the case of water displacing oil. However, the case of gas displacing oil saturated with gas at a constant (or nearly constant) pressure may also be considered a displacement by an immiscible fluid. This is possible for the reason that the concentration of the gas in the oil never changes if tile pressure is fixed. consequently, any additional free gas must remain undissolved in the oil, and so must act essentially as an immiscible phase. For convenience, the cape of gas drive will be considered first. Both of the two basic relations needed refer to a linear reservoir, or one which may be somewhat idealized so that a constant cross-section is exposed to fluid flow. Further, the displacing phase is injected at one end, or face; of the reservoir, while the produced fluids are understood to emerge at one point only. which is the opposite face of the reservoir. As indicated in Fig. 1, the reservoir may he inclined by some angle 8. The first of the required equations may be derived by applying Darcy's law to both displacing and displaced phases, and subtracting the two equations. The result for water drive has been given previously, as Equation (13) of the reference by

Jan 1, 1952

By Glenn A. Marsh, George Bernard

It is often assumed that aerated oil field brines which are to be injected underground contain dissolved oxygen in amounts which will cause appreciable corrosion. Through the use of a new portable dissolved oxygen meter, accurate determinations were made at three brine conditioning plants. At two older plants, dissolved oxygen concentration in the final treated brine was found to he almost zero In spite of aeration during the conditioning process. At a new plant, dissolved oxygen was initially present but disappeared soon after the brine was "inoculated" with oxygen-free brine from one of the other plants. Corrosion tests at the two older plants have substantiated predictions that the brines would lint be corrosive. These results show that mechanical deaeration may not always be required in brine treating plants. INTRODUCTION The present work was undertaken to add to our knowledge of aeration and deaeration of oil field brines in brine conditioning plants. It is commonly assumed that brines which are exposed to the air soon become more or less saturated with dissolved oxygen. Since oxygen-containing brines are known to be corrosive, precautions are often taken to keep air out of brine through the use of closed systems or mechanical deaeration. The study of oxygen in brine has been seriously hampered in the past because of the lack of a suitable analytical tool. The standard Winkler method for the determination of dis-solved oxygen is satisfactory for use in fresh water, but not in many oil field brines. Corrections have to be made for a number of reducing substances such as ferrous iron and hydrogen sulfide which are commonly present in natural brines. Very often the correction factors are much greater in magnitude than the amounts of oxygen which are to be determined, and consequently erroneous results are obtained. Also. erratic Winkler data are obtained when appreciable concentrations of calcium and magnesium salts are present. The first step in carrying out a systematic study involved designing a portable dissolved oxygen apparatus which would operate in natural brines. Subsequently, this apparatus was used in three brine conditioning plants. Dissolved oxygen data obtained here, and certain corrosion rate data, illustrate the interesting behavior of oxygen in these oil field brines. APPARATUS A portable dissolved oxygen meter described in detail elsewhere,' was developed specifically for use with oil field brines. It is a polarographic apparatus employing a small rotating platinum electrode (cathode) and calomel reference electrode (anode). A potential of 0.7 volts is applied across the electrodes while they are immersed in the brine to be tested. 4References given at end of paper.

Jan 1, 1952

By Harvey T. Kennedy, Charles R. Olson

In many investigations of the performance of petroleum reservoirs the assumption is made that the liquid, if below its bubble-point pressure, is at all times in equilibrium with gas. On the other hand, observations by numerous investigators have indicated that gas-liquid systems including hydrocarbon systems, may exhibit supersaturation to the extent of many hundred psi in the laboratory. Up to the present, there has been no reliable data on which ti judge the actual extent of supersaturation under conditions approaching those existing in petroleum reservoirs. The work reported here deals with observations and measurements on mixtures of methane and kerosene in the presence of silica and calcite crystals. Bubbles were observed to form on crystal-hydrocarbon surfaces in preference to the glass-hydrocarbon interface or to the body of the liquid. Statistically, it was found that the number of bubbles formed per second per square centimeter of crystal surface was a function of the supersaturation only. and the function was evaluated graphically. Supersaturations were observed up to 770 psi, under which condition bubbles formed quickly and with considerable violence. With decreasing degrees of supersaturation, the frequency of bubble formation became less, until at 30 psi supersaturation and lower, no bubbles were observed to form, even though the observation at 30 psi was continued for 138 hours. It was found that silica and calcite crystals had identical effects, within experimental error, in accelerating the formation of bubbles, and that small amounts of water and crude oil had no effect on the results. It is shown that the maximum supersaturation that can exist in a reservoir may be calculated from the data presented and from the area of the rock surface. It is also shown that the number of bubbles formed in the reservoir, in order of magni- tude, may be calculated for any rate of pressure decline imposed on the reservoir by production. The bearing of the number and distribution of bubbles on reservoir performance is discussed. INTRODUCTION A liquid system is supersaturated with gas when the amount of gas dissolved exceeds that corresponding to equilibrium at the existing pressure and temperature. The degree of super-saturation may be conveniently expressed as the difference between the bubble-point of the mixture and the prevailing pressure. Thus, if a mixture having a bubble-point of 1,000 psi at a given temperature exists in single liquid phase at 700 psi at the same temperature. it is supersaturated to the extent of 300 psi. There are many examples of high supersaturations, mostly in aqueous solutions, reported in the literature. Thus, Kenrick, Wismer and Wyatt1 showed that water may be saturated with oxygen, nitrogen or carbon dioxide at 100 atmospheres, and the pressure reduced to one atmosphere without producing bubbles immediately. When liquids are in a state of tension, they may be considered as supersaturated at least to the extent of the tension. The tensile strength of water has been reported as 30 atmospheres by Meyer,1 60 atmospheres bv Budgett. 30 to 50 atmospheres by Temperley and Chambers,4,5 200 atmospheres by Dixon6. and 223 atmospheres by Briggs.7 Vincent8,9 determined the tensile strength of a mineral oil as 45 psi. Gardescu1 maintained pressures for short times in a model reservoir at 115 psi below the bubble-point. It should be noted that the high supersaturations observed were obtained on systems carefully purified to remove particles or surfaces which might promote the formation of bubble; These "nuclei" were considered as contaminants which interfered with the determination of a property of the liquid. In

Jan 1, 1952

By R. H. Jacoby, M. J. Rzasa

Experimental equilibrium vaporization ratios (K values) were obtained for nitrogen, methane, carbon dioxide, ethane and hydrogen sulfide in two natural gas-absorber oil mixtures and in two natural gas-Elk Basin crude oil mixtures. For each mixture of constant over-all composition, data were obtained at 100°, 150° and 200°F and at various pressures in the range 200 to 5,000 psia. Some effects of composition on the K values were obtained to serve as a guide in choosing K's for engineering calculations on other mixtures. The pressure cell used to obtain the data is a new type and is described here for the first time. INTRODUCTION The phase equilibria of complex hydrocarbon mixtures such as natural gases, crude oils and their mixtures have been studied previously for the purpose of finding equilibrium vaporization ratios for the hydrocarbon constitutents. In a few cases, such non-hydrocarbon.: as nitrogen, carbon dioxide, and hydrogen sulfide were included in the mixture. studied because they occur. in the fluids obtained from petroleum reservoirs. The increasing occurrerce. of these non-liydrocarbon.; antl their growing economic importance. make it necessary to account for them more accurately in engineering calculation.: than is now possible using the meager published data. The available data for non-hydrocarbons may he classified into two groups: namely, binary mixtures of a hydrocarbon and a non-hydrocarbon. and mixtures containing three or more components, only one of which is a non-hydrocarbon. Amon: the former. phase analyses are available for two nitrogen-HC mixtures, tve carbon dioxide-HC mixtures and three hydrogen sulfide-HC mixtures. Data for more complex mixtures are available as follows: Nitrogen-Methane-Pentane'; Nitrogen-Methane-Hexane,'; Nitrogen-Methane-Heptane'; Carbon dioxide-Natural Gas-Natural Gasorline": Carbon dioxide-Natural Gas-Crude Oil". These data have been used generally lor engineering calculations, often with little regard for their precise applicability because data for the non-hydrocarbons in the complex mixtures being dealt with were not available. The use of K values from binary data for calculation:; involving complex mixtures is open to criticism because such K's are not a function of composition and very often show large differences from the K's for the same components in a complex mixture. Thus the effects of composition must be evaluated when obtaining K's for components in complex mixtures. Poettmann and Katz " varied the carbon dioxide content of their mixture.; from about five to ten mol per cent and found substantially no effect on the K values due to this change. This was not surprising since one might expect that changing a minor com-ponent by * three mol per cent would have little effect. It is more important to find the composition change which have the largest effect on the K values. For example. varying the methane or heptanes — plus compositions over a wide rangr may have a large effect on the K'. of all components in the mixture. Interesting data to this effect were obtained by Eilerts and Smith." It was toward these problems that the experimental data presented here were directed. EQUIPMENT The pressure cell used to obtain the data i:. somewhat dif-ferent from those which have been described in the past. Photographs of the equipment are shown in Figs. 1 and 2. Fig. 1 Shows. the front panel: from a .rated position at the right the

Jan 1, 1952