Spe-9993-pa Steamflooding.pdf

Distinguished Author Series

Steamflooding by C.S. Matthews, SPE C.S. Matthews is senior petroleum engineering consultant with Shell Oil Co. in Houston. After receiving a as degree in chemical engineering and a PhD degree in chemistry from Rice u., Matthews joined Shell Development Co. in 1944 as an engineer in San Francisco. He transferred to Houston 4 years later to work on production research and became chief reservoir engineer for Shell's Technical Services Div. in 1956. In 1965 he became manager of exploitation engineering in New York City. Later he served as director of production research with Shell Development and as manager of engineering in the Production Dept. Matthews, with D.G. Russell, wrote SPE's first Monograph, Pressure Buildup and Flow Tests in Wells. They were 1967-68 Distinguished Lecturers on that topic. Matthews also chaired the Reservoir Engineering Technical Program Committee for the 1969 Annual Meeting and the Lucas Gold Medal Committee during 1972-73. He received the Lester C. Uren Award in 1975 for distinguished achievement in petroleum engineering technology. His current interests include enhanced recovery of oil and gas, tar sands, and geothermal energy.

Introduction Steamflooding has become an established recovery technique within the last 20 years. This overview discusses its evolution, methods for selecting and designing steamfloods, constraints, and possible improvements. The tenn steamflooding is used here in a general sense. The discussion includes steam soak (cyclic steam injection) and steam drive. For additional infonnation the reader should refer to Farouq Ali and Meldau. I Early Developments: Steam Soak There are records of steam injection into a Texas oil reservoir as far back as 1931 .2 Steam drive on a sustained basis, however, did not begin until 1959-60 when Shell affiliates undertook steam flood pilots in Schoonebeek, the Netherlands,3 Mene Grande , Venezuela,4 and Yorba Linda, CA.5 Steaming operations still under way at Schoonebeek and Yorba Linda are discussed later. Venezuela. The steamdrive pilot in Mene Grande led to development of the steamsoak process. 4 While attempting to relieve the fonnation pressure by opening a steam injector to production, oil was produced surprisingly at a rate of 100 to 200 BID. This was the first steamsoak well. This process underwent considerable development in both 0149·213618310003·9993$00.25 Copyright 1983 Society of Petroleum Engineers of AIME

MARCH 1983

Venezuela and the U.S. Most of the oil produced by steam from 1960 to 1970 was by this process. Typically , steam is injected for several weeks and then the well is closed in (soaked) until the steam has condensed. A pump then is run and the well is placed on production. When oil production falls to a low level, the cycle is repeated. Thennal production in Venezuela is still almost entirely (95 %) from steam soak. Compaction of these thick, unconsolidated sands together with solution-gas drive is leading to satisfactory oil recovery at very high ratios of oil recovered per barrel of steam used. The Netherlands. One of the first large-scale steam drives began at the Schoonebeek field in The Netherlands in 1960. 3 Recovery of the moderately viscous oil (180 cp) was quite successful. This encouraged additional field trials of steam drive in the U.S. and in Venezuela. Coring showed that steam was achieving low residual oil saturations of the order of 8 %. The residual oil in the zone swept only by hot water was about 35 %. Later tests at Schoonebeek were carried out at high pressure and high temperature. Under these conditions C02 and H 2S are generated in situ . Consequently, new alloys had to be developed. Thus, Schoonebeek led to pioneering along both reservoir and metallurgical lines. u.S. Commercial steamflooding began in the U .S. in 1960 in the Yorba Linda field, CA. 6 Prior production in this field had been achieved only through use of 465

TABLE 1-SUMMARY OF TYPICAL WELL PERFORMANCE FOR STEAMSOAK OPERATIONS IN VARIOUS CALIFORNIA FIELDS

Before Steam

After Steam'

Rate cold

Net Sand Open (ft)

15 25 3 10 14 3 5 5 30 7

160 360 140 110 65 52 56 35 85 20

11 14 47 11 4.6 17.3 11 7 2.4 3

40 220 22 250 220 107 240 250 75 80

Production (BID) Field

Zone

Huntington Beach San Ardo Kern River Midway-Sunset Kern River Coalinga Midway-Sunset Midway-Sunset White Wolf Poso Creek

TM Lombardy China Potter A Kern River Temblor Tulare Potter B Reef Ridge Etchegoin

--

Rate hot

Steam Injected (bbl)

Cycle Period (months)

Total (bbl)

Per Barrel of Steam

Additional Oil Recovered per Barrel of Steam

4,500 14,000 4,400 6,000 6,500 9,000 12,000 7,700 14,000 6,700

15 18 6 5 5 5 6 4 4 6

29,000 50,000 11,600 9,240 4,730 4,300 4,640 3,000 6,750 2,660

6.5 3.6 2.62 1.54 0.73 0.48 0.38 0.39 0.48 0.40

5.0 2.8 2.5 1.29 0.43 0.40 0.31 0.29 0.23 0.21

Oil Recovered

• Average of first 30 days.

downhole heaters. Primary production had been estimated to be 5 % of the 100 million bbl in place. After the success of steam soak in this field became known, the process spread rapidly throughout the state. By 1963 there were 29 steamsoak projects in the state,5 and by 1965 there were 267. Steam injection of 15,000 to 25,000 bbl per well led to peak oil rates of 100 to 200 BID per well and to declining but still commercial production for a period of about I year. The injection cycle then would be repeated. It was relatively inexpensive to try steam soak, and a number of operators tried. By 1967 the estimated additional oil production in California from steam soak had reached 120,000 BID. Some 408 steam generators were in use at that time. 5 Steam soaking also was tried widely outside California. Although success was obtained in a few cases, the additional production obtained was very small in comparison with California. In the U.S. outside California, oils are normally of low viscosity. Waterflooding proved the most economical method of supplemental recovery, and steam could not compete. For the thick, viscous oils in California, waterflooding gave poor recoveries. This opened the door to steam. Early attempts to apply steamflooding to U.S. tar sands were singularly unsuccessful. This was largely due to unfavorable reservoir characteristics such as low oil saturation, thief zones, fractures, or extremely high oil viscosity that kept the steam injectivity rate very low. Canada. One of the first sustained steamsoak projects in Canada began in the Cold Lake tar sand in 1964. After many years of experimentation, production had reached a level of 7,000 BID in 1980. Steam soak is not as successful in the more viscous (1 million cp) Athabasca tar sands. The viscosity at Cold Lake is some ten times lower than at Athabasca. Cold Lake tar also apparently has enough gas in solution to drive cold oil into steam-stimulated wells. 466

Other. Steam soaks are also under way in France and the Congo. Soaks probably will be used in connection with steam drives under way or planned for West Germany, Trinidad, Sumatra, Argentina, and Brazil. Methods for Analyzing Steam Soak Steam soak was a process for which field application outstripped theory and laboratory research. It was more economical in many cases to apply steam soak in a particular case than to conduct research to determine whether it would be applicable. The main expense was that of moving in and connecting a portable steam generator. Empirical observations were the first guides to applicability. Some of these are: 1. Thick homogeneous sand reservoirs give the best response, particularly when gravity drainage is effective. 2. Thin (20- to 40-ft) permeable reservoirs also respond well for a few cycles until the pressure is depleted; they also respond well if gravity can provide cold oil inflow. 3. Thin reservoirs or thick reservoirs composed of several sand members respond poorly if pressures are low. 4. Reservoirs producing at high water rate and high water cut respond poorly. 5. Reservoirs with low permeability or containing oils of high viscosity (> 10,000 cp) respond poorly. These observations led to several methods for mathematically analyzing a steam soak. Boberg 7 related the improvement in oil production to the radial distance heated from the wellbore. His analytical method allowed the heated zone to cool by conduction and by removal of hot fluids. For a typical heavy-oil field his method predicts a maximum rate enhancement of 3 to 4 times. Boberg's analysis generally underestimated the rate achieved in a steam soak, probably because his analysis did not account for the increased solution gas JOURNAL OF PETROLEUM TECHNOLOGY

driven out of the crude by the steam and the effect of low-level heating beyond the very hot zone. Boberg's method seems to work best when the reservoir pressure is still relatively high. In such cases the increased drive due to gas release will be small. Owens and Suter 8 presented a very simple model for calculating the response to steam soak. They proposed: Rate hot

Viscosity hot

Rate cold

Viscosity cold

In general, this method tends to overestimate the effect of steam stimulation. In a typical case an enhancement of 20 to 50 times will be predicted. Oil production rate enhancement in partially depleted fields ranges from about 6 times in some thick Venezuela reservoirs to 10 to 20 times in many California reservoirs (Table 1). A number of additional models for steam soak have been presented. These are generally similar to the Boberg model. All allow an estimate to be made of oil recovered per barrel of steam. Some early field results are shown in Table 1.5 Generally speaking, steam soak does not lead to high recovery. Recoveries in the range of 10 to 20% of OOIP are normal. However, for certain reservoir conditions, particularly where gravity drainage can provide economic production rates, recovery can be high. In the Yorba Linda Upper Conglomerate zone, a recovery of 35 % of OOlP was reported. This excellent recovery was a result of gravity drainage in a thick (325-ft) shallow zone where wells could be drilled on very close spacing (0.8 acre). Thus, the wells were much closer laterally than the pay thickness. Reservoir compaction has given good recovery at main Tia Juana, Venezuela. Total recovery by steam soak through reservoir compaction and solution-gas drive in this reservoir is 25 to 30 % . Steam soak continues to be important, accounting for almost half of current steam-induced production throughout the world. It will remain important as an initiating mechanism for drive projects where initial injectivity is low, and as the major mechanism in reservoirs with poor continuity. It also will remain important in reservoirs with good gravity drainage or reservoir compaction drive.

Methods of Predicting Steam Drives Steam drive began to gain importance in the U.S. in about 1970. (Outside the U.S., it is still relatively unimportant.) At that time, steamdrive production in California was about 30,000 BID. Some 12 years later, production by steam drive had increased to more than 150,000 BID. This section traces the rise of that production, beginning with a discussion of predictive methods. In contrast with the lesser value of steamsoak performance prediction, steamdrive performance prediction is very important. Much time and money are required to conduct a drive pilot. Furthermore, MARCH 1983

scaling up results of a pilot to full scale is difficult without some sort of mathematical or physical reservoir model. Thus, development of predictive methods for steam drive went hand-in-hand with field development.

Early Prediction Methods. Laboratory studies such as those of Willman et al. 9 or Volek and Pryor lO were useful in showing the potentially high displacement obtainable with steam. Distillation by steam of the crude left as residual oil at the steam condensation front results in very low residual oil saturation in the steam-swept zone. For most viscous crudes of interest, the residual oil saturation is 5 to 15 % of PV, well below the residual saturation attainable by waterflood. These early laboratory studies also showed that steam displacement was a highly stable process. Although steam is of low viscosity, it cannot "finger" through an oil bank as can a noncondensible gas. Fingers of steam lose heat rapidly and condense, heating the surrounding rock and fluids. This stability of the steam front is an important advantage of steamflooding. Laboratory measurements of residual oil left by steam are one element in the design of a field project. Prediction methods in the early 1960's made use of these measurements, together with analytical solutions of heat flow and heat loss, to estimate temperature distribution, thermal efficiency, and recovery. The work of Lauwerier ll and Marx and Langenheim 12 were among the earliest analytical heat flow solutions to be used. To use these methods, the rate of steam injection had to be determined from an injection test. The amount of oil displaced by steam then could be calculated for the case of a homogeneous reservoir with a vertical displacement front (no gravity override of steam). Myhill and Stegemeier 13 extended these earlier analytical methods and presented a simple method for estimating the oil/steam ratio for a field project. Scaled Physical Models. These models came into use for steamfloods in the late 1960's. They accounted for reservoir heterogeneity and gravity override of steam, and thus were superior to the previous analytical methods. In addition, they showed where additional wells were needed or where changes in operating policy could improve results. Early scaled models required use of heavy pressure vessels around the model to allow an overburden pressure to be applied to the flow chamber. Consequently, the models were bulky. Much time was required to pack them with sand, to saturate them with water and oil, and to deplete them to conditions prevailing at the beginning of steam drive. In these early models, steam temperatures equivalent to field conditions often were used. A significant breakthrough in modeling steamfloods in the laboratory occurred when it was shown that properly scaled results could be obtained at low 467

TABLE 2-SUCCESSFUL STEAM FLOODS

Depth Field

(tt)

Kern River, CA Inglewood, CA Brea B, CA Coalinga, CA Yorba Linda, CA San Ardo Auginac, CA Mt. Poso, CA Yorba Linda, CA South Belridge, CA Midway-Sunset, CA Schoonebeek, The Netherlands Slocum, TX Smackover, AR Tia Juana, Venezuela Winkleman Dome, WY • k ~ permeability; h ~ net pay;

~, ~

900 1,000 4,600 1,500 2,100 2,350 1,800 650 1,100 1,600 2,600 535 2,000 1,600 1,200

Reservoir Pressure (psig) 35 120 110 300 200 250 100 180 50 120 110 5 300 210

Oil Viscosity

Net Pay _(_ft)_ ~ 4,000 60 1,200 43 6 189 100 50 32 85 2,000 150 280 60 6,400 325 1,600 91 4,000 350 180 83 40 1,300 75 20 5,000 125 900 73

(md-tt/cp)"

4,000 6,000 70 1,000 500 3,000 15,000 600 3,000 4,000 5,000 3,500 5,000 2,800 600

60 220 2,200 500 188 225 3,210 30 170 350 2,300 1,080 1,330 70 50

khll'o

Oil Saturation, % 50 64 49 57 49 58 63 75 60 85 60 80 85 75

Porosity, %

--35 39 24 31 30 39 33 30 33 32 30 38 36 38 25

Oill Steam Ratio (bbl/bbl)

Recovery (% OIP)

0.25 0.50 0.21 0.36 0.21

68 50

0.21 0.49 0.28 0.60 0.37 0.18 0.33 0.83 0.20

65 55 60 65 50 57

39 62

45 50

oil viscosity.

TABLE 3-FACTORS UNFAVORABLE FOR STEAM FLOODING 1. 2. 3. 4. 5. 6.

Oil saturation less than 40%. Porosity less than 20%. Oil-zone thickness less than 30 ft. Permeability less than 100 md. Ratio of net to gross pay less than 50%. Layers of very low oil saturation and high permeability in the oil zone that act as thief zones. 7. Extremely high oil viscosity. 8. Fractures. 9. Large permeability variations in the oil zone. 10. Poor reservoir continuity between injectors and producers. 11. Deep, high-pressure reservoirs and shallow reservoirs with insufficient overburden to permit steam injection without fracturing.

temperatures with subatmospheric-pressure steam. 14 This enabled the packed-bed models to be enclosed by thin plastic sheets, which, upon imposition of a vacuum, became rigid containers. This technique made the heavy pressure vessels unnecessary. Furthermore, it allowed researchers to see where displacement was and was not occurring. Although scaled physical models can be used to portray most mechanisms of a steamflood accurately, they are still time-consuming to pack and operate and often are limited by availability of materials and fluids to achieve proper scaling for particular oils and sands. Considerable time and care are required to pack the model with the proper sizes of sand or glass beads, to saturate it properly with oil and water, and to carry out the steam flood simulation. More recently, development of efficient computational techniques and increases in speed and storage of computers have made mathematical simulation of steamflooding efficient, reliable, and fast. They also have enabled consideration of the effect of additional mechanisms, such as the effect of dissolved gas on a steam drive. Recently, thermal numerical simulators have been able 468

Permeability (md)

to duplicate physical model results, 15 lending even more confidence to the use of mathematical simulation. Mathematical Simulation. The partial differential equations describing the flow of oil, water, and steam together with the flow of heat were formulated for petroleum reservoirs in the 1950's. At that time, however, computers had insufficient memory and were much too slow to allow these equations to be solved simultaneously as a steamflood progressed. Over the ensuing 20 years, not only did computer speed and storage increase phenomenally, but efficient computation techniques were developed. Coats, Chu, and Marcum 16 and Coats 17 have described some of these techniques. At the current stage of development, it is possible to simulate the behavior of a representative portion of a heterogeneous reservoir under steam drive. The modeled portion can include several injection wells and their surrounding producers. The reservoir can be heterogeneous both vertically and laterally. The effects of dip and gravity can be included. Mathematical reservoir simulation is now the fastest, most accurate, and most efficient method for predicting performance of steamfloods. One problem that arises in such simulation is the difficulty in visualizing results. The computer output is usually a discouragingly large stack of paper showing a large number of computed values of oil production, temperature, saturation distribution, etc. In some cases a series of 2D plots of these data also are provided. The voluminous nature of the output is a deterrent to its analysis. Typical output from one run is about half a million numbers. One recent improvement in this regard is the Dynamic Visual Display method developed by Shell Development Co. 18 By this technique the numerical output from a large computer is processed by a smaller one to develop 2D arrays of the process variables. These digital results are converted to TV signals that JOURNAL OF PETROLEUM TECHNOLOGY

are fed to a color monitor. Typically, the intensity of pressure and temperature is represented by spectra of color, and saturations of oil, steam, and water are in three separate colors. Temperatures, pressures, and individual fluid saturations may be shown on a cross section or aerial view of the reservoir. As the flood progresses, the output on the color monitor is photographed by a movie camera. Results of the calculations may be reviewed by either videotape or movIe. Many possible injection and production schemes may be studied economically in this manner. The pictorial display of results speeds interpretation and allows engineering improvements to be visualized. Field Development of Steam Drive Early Drives. The industry still was learning how to conduct a steam drive in 1970. It was observed in several fields with good vertical permeability that injected steam rose rapidly to the sealing silt or shale at the top of the injected interval and then traveled in a thin layer to a production well. After steam breakthrough occurred at the producer, only a very small pressure differential could be maintained between injector and producer. Thus, the oil was not being produced by a drive but rather by a "drag." Oil directly beneath the overlying steam zone was heated by the steam and flowed cocurrently with the steam to the production well. With the very low pressure gradient between wells, the required injection pressures were very low. Thus, many of the early steam drives could be characterized as continuous steam injection at very low pressure. The steam injection rate was also low, usually just high enough to keep the producers hot and with little steam production. High-injection-rate steam drives appeared somewhat later. High-Rate Steam Drives. In reservoirs with significant dip, laboratory model studies showed that it was possible to drive the oil downdip to producers by injecting steam updip. This usually required initial injection of steam into some of the downdip producers to preheat the oil. Drives of this type began to appear during the 1970's. Examples are found at Mt. Poso l9 and at Midway-Sunset. 20 Partially as a result of these high-rate drives, steamdrive production increased in the U.S. from 30,000 BID in 1970 to 270,000 BID in 1982. Steam drives may lead to dramatic increases in production rate. At Mt. Poso, a 280-cp oil was in the final stages of a strong water drive. Steam was introduced when the field was producing 1,500 BID oil at 99 % water cut. Oil production finally rose to about 25,000 BID. It is estimated that recovery from the steam drive at Mt. Poso may exceed 65 % of OOIP vs. 38 % from the natural water drive. The largest steam drive in the U.S. is at Kern River (107,000 BID). Other large drives are at South Belridge (63,000 BID), San Ardo (26,000 BID), and MARCH 1983

Midway-Sunset (28,000 BID). In Venezuela, steamdrive production is about 20,000 BID in Lagunillas. At Schoonebeek, steamdrive production is expected to reach 9,000 BID in a few years. Steam drives are under way in Canada and in France, but production is still small. Drives are planned in Trinidad, Argentina, and Indonesia. The latter is planned to be the world's largest steam drive, with an oil production rate of 300,000 BID. Selecting and Designing a Steamflood Criteria. Key properties of some of the most successful steamfloods are given in Table 2.1 Reservoir properties of successful projects cover a wide range of conditions. Table 3 contains a list of factors unfavorable for steamflooding. This table is useful for rapid screening of projects. However, a project may violate one or two of these factors and still be successful, if all other factors are highly favorable. Thus, every potential project should be evaluated on its own merit. Estimating an OillSteam Ratio. After a preliminary screening, the next step in evaluating a potential steam flood is to estimate the average oil/steam ratio. The method of Myhill and Stegemeier 13 is convenient for this purpose. Simulation, Injection Testing, and Economic Analysis. Before a substantial investment is made in a steamflood, a laboratory model or computer simulation study should be run. Refs. 14, 16, and 17 discuss both approaches. The simulations also will allow studies of the effect of well spacing and well arrangement, completion interval, injection and production schedule, pressure level, steamsoak size, and many other variables. An example of the use of a reservoir simulator to improve field performance is given in Ref. 21. With simulation results in hand, an economic evaluation can be made. This evaluation usually will be a strong function of the steam injection rate. Methods for predicting this rate are only approximate, and injection tests in several key wells in the reservoir of interest are a "must." A careful geological and reservoir study is also necessary. There are numerous examples of steamfloods that failed when a careful preflood study would have pointed out the pitfalls. Some of the reasons for failure were: (1) less net pay than originally estimated, especially in sands with shaly intervals; (2) lower oil saturation than originally estimated, especially when the connate water is relatively fresh; (3) poor reservoir continuity; and (4) nearness to an updip gas cap or to bottom water. The final test of a steam drive is a pilot. In contrast with waterflood pilots, steamdrive pilots are often at least semiquantitative. The reason for this is that the pilot area will be surrounded by formation containing cold, viscous oil, usually at a high oil saturation since 469

primary production from heavy-oil reservoirs is low. Heated oil normally will flow into a steamsoaked (and heated) producer because of this surrounding "wall" of viscous oil. Mathematical reservoir simulation is of considerable value during a pilot, particularly if there are observation wells. Simulation may show, for example, that much larger steam soaks are needed in certain wells, or that updip production should cease, to force steam downdip. Continuous interaction between research, engineering, and operations is required at this stage for success. The Potential of Heavy Oil and Tar Sands Resources. The tar sand and heavy oil resources of the world are very great. Canada and Venezuela each have resources (heavy oil and tar in place) of the order of 1 trillion bbl. These tar sands range in depth from the surface-minable deposits to those at several thousand feet. Smaller tar-sand deposits are known to exist in many other countries, including the U.S. The U.S. Natl. Petroleum Council (NPC) made a detailed study in 1976 of the potential of all U.S. heavy oils. They found that the potential recovery by steam is quite low. Less than 4% of the entire U.S. heavy oil resource is estimated to be recoverable in this manner. Even in California, where the world's largest thermal oil production is now occurring, only about 7 % of the heavy oil in place is estimated to be recoverable. Reasons for poor recovery, in addition to those noted previously, include the presence of fractures, low porosity or permeability, and thief zones. Where steam is applicable, oil recovery can be as high as 60%. Where steam is not applicable, oil recovery is very low. The 7 % overall recovery for heavy oil in California indicates that steam injection is not widely applicable, for the reasons discussed. Undoubtedly, price and new technology will change some of these resources to reserves. Many years and much effort will be required. Recovery of heavy oils and tar sands that are unrecoverable with current economics will pose a challenge for future generations. Recovery in Canada. As mentioned, Canada has extremely large resources of heavy oils and tar sands. The very large resources are in the Cold Lake, Peace River, and Athabasca areas. The Cold Lake deposit currently is yielding about 14,000 BID by steam soak. At Peace River a large steamdrive pilot is currently in progress. The giant Athabasca tar-sand reservoirs have not succumbed yet to commercial in-situ exploitation. The shallow Athabasca tar sands now are being mined commercially, but the deeper deposits, in spite of extensive research and field testing, have not yielded their tar commercially. This remains one of the great challenges for the petroleum scientist and engineer. Recovery in Venezuela. Venezuela, like Canada, is blessed with a very large heavy-oil resource. Estimates of 1 trillion bbl in place have been made. Much 470

additional drilling will have to be done, however, to determine the characteristics of this resource. Very likely, drilling will indicate, as at Athabasca, that this tar belt is composed of many different reservoirs, each with different sand and oil characteristics. A start has been made at process development. The Jobo steam pilot was begun in 1981. Planning for upgrading the heavy tar is also under way. The tar belt could prove to be a major sustainable source of energy for many years. Potential. The current world production rate from steam is about 550,000 BID. From the number of projects planned it seems likely that production in the U.S. will rise to 500,000 BID by 1990. A production rate of 350,000 BID in the rest of the world appears reasonable, for an overall world production of 850,000 BID in 1990. In the 1990's, commercial projects at Cold Lake, Peace River, and in the Venezuela tar belt easily could double this figure. The tar and heavy-oil resource is so large that it could sustain production of a few million barrels per day many years into the next century. Constraints to Development In the U.S. there are a number of factors that probably will prevent the rate of oil production from steam injection from reaching its maximum potential. Some of these are: 1. Delays in obtaining air emission permits. Three years or more have been required in the past to obtain some permits to install and operate steam generators in the U.S. 2. Evolving regulations. As new regulations were issued, new engineering studies were required. These in turn led to alteration of proposed projects, which meant additional studies and revisions. 3. Scrubber waste disposal. Most steam generators in California burn crude oil containing several percent sulfur. Caustic scrubbers currently are used to remove S02 from exhaust gases. Constructive efforts of both industry and government are required to assure that scrubber waste disposal is both environmentally acceptable and cost effective. 4. Water supply. Approximately 4 bbl of water are required for each barrel of oil produced. Oilfield steam generators can use nonpotable water with a high content of dissolved solids. It is also common practice to recycle produced water. Thus, water for steam generation is not usually a problem. However, S02 scrubbers require fresh water. Although the requirements are small, they are critical. This may be a problem in some areas. 5. Taxes. The U.S. federal government provides a disincentive in the form of the "Windfall Profit Tax. " According to the 1976 NPC study, this will lower the potential production rate of U.S. thermal oil in 1990 by up to 300,000 BID. JOURNAL OF PETROLEUM TECHNOLOGY

New Possibilities Other Fuels for Steam. In steam drives, the equivalent of approximately one-third of the produced oil is required to generate steam. The ratio is lower in steam soak. Substitution of petroleum coke or coal for part or all of the oil used as fuel may be possible in large installations. Fluidized-bed combustion of coal is one of the promising technologies for such large-scale steam generation. Burning natural gas, when available, under a field steam generator is an efficient use of energy. For every Btu burned, about 3 to 5 Btu will be produced in the form of heavy oil. Use of gas also eliminates most emission problems in the field. Cogeneration. By generating steam at higher temperature and pressure than needed for steam injection, electrical power first could be generated. After expansion through an electrical power generator turbine, the lower-pressure steam could be injected for heavy-oil recovery. Burning the fuel in a turbine before raising steam offers an additional possibility of further increasing the efficiency of the entire cycle. Downhole Steam Generators. For deep steaminjection projects, for injection into low-permeability reservoirs, and for use offshore, a downhole steam generator would be attractive. Several types are currently under development. In one type the products of combustion are injected along with the steam. In another the combustion products are vented to the surface. If development is successful, these downhole generators undoubtedly will make their contribution to increased thermal oil production. Conclusions Steam now has proved an important agent for oil production. The oil production rate for steam projects continues to increase and the worldwide rate may reach 850,000 BID by 1990. Regulatory and economic constraints may prevent the oil production rate from increasing to its maximum potential in the short term, especially in the U. S. Changing environmental regulations, delays in issuing permits, and increased taxes are among the major constraints to production. On the brighter side, there is still considerable potential for new technology to bring on currently marginal steam projects. References I. Farouq Ali, S.M. and Meldau, R.F.: "Current Steamtlood Technology," 1. Pet. Tech. (Oct. 1979) 1332-42. 2. Stoval, S.L.: "Recovery of Oil from Depleted Sands by Means of Dry Steam," Oil Weekly (1934) 74, 9. 3. Van Dijk, e.: "Steam-Drive Project in the Schoonebeek Field, The Netherlands," Trans., AIME (1968) 243, 295-302.

MARCH 1983

4. deHaan, H.J. and Schenk, L.: "Performance and Analysis of a Major Steam Drive Project in the Tia Juana Field, Western Venezuela," Trans., AIME (1969) 246, II I - 19. 5. Bums, J.: "A Review of Steam Soak Operations in California," Trans., AIME (1969) 246, 25-34. 6. Stokes, D.D. and Doscher, T.M.: "Shell Makes a Success of Steam Flood at Yorba Linda," Oil and Gas 1. (Sept. 2, 1974) 71-76. 7. Boberg, T.C.: "What's the Score on Thermal Recovery and Thermal Stimulation?", Oil and Gas 1. (Aug. 23, 1965) 78. 8. Owens, W.D. and Suter, V.E.: "Steam Stimulation-Newest Form of Secondary Petroleum Recovery," Oil and Gas 1. (April 26, 1965) 82. 9. Willman, B.T., et al.: "Laboratory Studies of Oil Recovery by Steam Injection," 1. Pet. Tech. (July 1961)681-90. 10. Volek, e.W. and Pryor, J.A.: "Steam Distillation Drive, Brea Field, California," 1. Pet. Tech. (Aug. 1972) 899-906. II. Lauwerier, H.A.: "The Transport of Heat in an Oil Layer Caused by the Injection of Hot Fluid," Applied Science Research (1955) A-5,145. 12. Marx, J.W. and Langenheim, K.H.: "Reservoir Heating by Hot Fluid Injection," Trans., AIME (1959) 216, 312-15. 13. Myhill, N.A. and Stegemeier, G.L.: "Steam Drive Correlation and Prediction," 1. Pet. Tech. (Feb. 1978) 173-82. 14. Stegemeier, G.L., Laumbach, D.D., and Volek, e.W.: "Representing Steam Processes with Vacuum Models," Soc. Pet. Eng. 1. (June 1980) 151-74. 15. Myhill, N.A.: "A Check on Numerical Thermal Simulation," paper SPE 8822 presented at the 1980 Enhanced Oil Recovery Symposium, Tulsa, April 20-23. 16. Coats, K.H., Chu, e., and Marcum, B.D.: "Three-Dimensional Simulation of Steamtlooding," Soc. Pet. Eng. 1. (Dec. 1974) 573-92. 17. Coats, K.H.: "A Highly Implicit Steamtlood Model," Soc. Pet. Eng. 1. (Oct. 1978) 369-83. 18. Good, P.A. et al.: "Use of Color Movies for Interpretation and Presentation of Reservoir Simulation Results," 1. Pet. Tech. (Aug. 1980) 1331-38. 19. Stokes, D.D. et al.: "Steam Drive as a Supplemental Recovery Process in an Intermediate Viscosity Reservoir, Mt. Poso Field, California," 1. Pet. Tech. (Jan. 1978) 125-31. 20. Duerksen, J.H., Webb, M.G., and Gomaa, E.E.: "Status of the Section 26C Steamtlood, Midway-Sunset Field, California," paper SPE 6748 presented at the 1977 SPE Annual Technical Conference and Exhibition, Denver, Oct. 9-12. 21. O'Dell, P.M. and Rogers, W.e.: "Use of Numerical Simulation to Improve Thermal Recovery Performance in the Mt. Poso Field, California," paper SPE 7078 presented at the Fifth Symposium on Improved Methods for Oil Recovery, Tulsa, April 16-19,1978. 22. Ramey, H.J.: "How to Calculate Heat Transmission in Hot Fluid," Pet. Eng. (Nov. 1964) 110.

SI Metric Conversion Factors acre X 4.046 873 bbl X 1.589 873 Btu X 1.055 056 cp X 1.0* ft X 3.048* psi X 6.894 757 "Conversion factor is exact.

E+03 = m 2 E-Ol = m 3 E+03 =1 E-03 Pa's E-Ol m E-03 MPa JPT

Distinguished Author Series articles are general, descriptive presentations that summarize the state of the art in an area of technology by describing recent developments for readers who are not specialists in the topics discussed. Written by individuals recognized as experts in the areas, these articles provide key references to more definitive work and present specific details only to illustrate the technology. Purpose: To inform the general readership of recent advances in various areas of petroleum engineering. The series is a project of the Technical Coverage Committee.

471